diff options
Diffstat (limited to 'drivers/gpu/drm/radeon')
63 files changed, 8981 insertions, 1993 deletions
diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile index 3c91312dea9a..84b1f2729d43 100644 --- a/drivers/gpu/drm/radeon/Makefile +++ b/drivers/gpu/drm/radeon/Makefile @@ -33,6 +33,9 @@ $(obj)/rs600_reg_safe.h: $(src)/reg_srcs/rs600 $(obj)/mkregtable $(obj)/r600_reg_safe.h: $(src)/reg_srcs/r600 $(obj)/mkregtable $(call if_changed,mkregtable) +$(obj)/evergreen_reg_safe.h: $(src)/reg_srcs/evergreen $(obj)/mkregtable + $(call if_changed,mkregtable) + $(obj)/r100.o: $(obj)/r100_reg_safe.h $(obj)/rn50_reg_safe.h $(obj)/r200.o: $(obj)/r200_reg_safe.h @@ -47,6 +50,8 @@ $(obj)/rs600.o: $(obj)/rs600_reg_safe.h $(obj)/r600_cs.o: $(obj)/r600_reg_safe.h +$(obj)/evergreen_cs.o: $(obj)/evergreen_reg_safe.h + radeon-y := radeon_drv.o radeon_cp.o radeon_state.o radeon_mem.o \ radeon_irq.o r300_cmdbuf.o r600_cp.o # add KMS driver @@ -60,7 +65,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \ rs400.o rs600.o rs690.o rv515.o r520.o r600.o rv770.o radeon_test.o \ r200.o radeon_legacy_tv.o r600_cs.o r600_blit.o r600_blit_shaders.o \ r600_blit_kms.o radeon_pm.o atombios_dp.o r600_audio.o r600_hdmi.o \ - evergreen.o + evergreen.o evergreen_cs.o radeon-$(CONFIG_COMPAT) += radeon_ioc32.o radeon-$(CONFIG_VGA_SWITCHEROO) += radeon_atpx_handler.o diff --git a/drivers/gpu/drm/radeon/atombios.h b/drivers/gpu/drm/radeon/atombios.h index 27e2c715be11..1bc72c3190a9 100644 --- a/drivers/gpu/drm/radeon/atombios.h +++ b/drivers/gpu/drm/radeon/atombios.h @@ -3780,7 +3780,7 @@ typedef struct _ATOM_ASIC_SS_ASSIGNMENT UCHAR ucReserved[2]; }ATOM_ASIC_SS_ASSIGNMENT; -//Define ucClockIndication, SW uses the IDs below to search if the SS is requried/enabled on a clock branch/signal type. +//Define ucClockIndication, SW uses the IDs below to search if the SS is required/enabled on a clock branch/signal type. //SS is not required or enabled if a match is not found. #define ASIC_INTERNAL_MEMORY_SS 1 #define ASIC_INTERNAL_ENGINE_SS 2 @@ -5742,6 +5742,9 @@ typedef struct _ATOM_PPLIB_THERMALCONTROLLER #define ATOM_PP_THERMALCONTROLLER_RV6xx 7 #define ATOM_PP_THERMALCONTROLLER_RV770 8 #define ATOM_PP_THERMALCONTROLLER_ADT7473 9 +#define ATOM_PP_THERMALCONTROLLER_EXTERNAL_GPIO 11 +#define ATOM_PP_THERMALCONTROLLER_EVERGREEN 12 +#define ATOM_PP_THERMALCONTROLLER_ADT7473_WITH_INTERNAL 0x89 // ADT7473 Fan Control + Internal Thermal Controller typedef struct _ATOM_PPLIB_STATE { @@ -5749,6 +5752,26 @@ typedef struct _ATOM_PPLIB_STATE UCHAR ucClockStateIndices[1]; // variable-sized } ATOM_PPLIB_STATE; +typedef struct _ATOM_PPLIB_FANTABLE +{ + UCHAR ucFanTableFormat; // Change this if the table format changes or version changes so that the other fields are not the same. + UCHAR ucTHyst; // Temperature hysteresis. Integer. + USHORT usTMin; // The temperature, in 0.01 centigrades, below which we just run at a minimal PWM. + USHORT usTMed; // The middle temperature where we change slopes. + USHORT usTHigh; // The high point above TMed for adjusting the second slope. + USHORT usPWMMin; // The minimum PWM value in percent (0.01% increments). + USHORT usPWMMed; // The PWM value (in percent) at TMed. + USHORT usPWMHigh; // The PWM value at THigh. +} ATOM_PPLIB_FANTABLE; + +typedef struct _ATOM_PPLIB_EXTENDEDHEADER +{ + USHORT usSize; + ULONG ulMaxEngineClock; // For Overdrive. + ULONG ulMaxMemoryClock; // For Overdrive. + // Add extra system parameters here, always adjust size to include all fields. +} ATOM_PPLIB_EXTENDEDHEADER; + //// ATOM_PPLIB_POWERPLAYTABLE::ulPlatformCaps #define ATOM_PP_PLATFORM_CAP_BACKBIAS 1 #define ATOM_PP_PLATFORM_CAP_POWERPLAY 2 @@ -5762,6 +5785,12 @@ typedef struct _ATOM_PPLIB_STATE #define ATOM_PP_PLATFORM_CAP_SIDEPORTCONTROL 512 #define ATOM_PP_PLATFORM_CAP_TURNOFFPLL_ASPML1 1024 #define ATOM_PP_PLATFORM_CAP_HTLINKCONTROL 2048 +#define ATOM_PP_PLATFORM_CAP_MVDDCONTROL 4096 +#define ATOM_PP_PLATFORM_CAP_GOTO_BOOT_ON_ALERT 0x2000 // Go to boot state on alerts, e.g. on an AC->DC transition. +#define ATOM_PP_PLATFORM_CAP_DONT_WAIT_FOR_VBLANK_ON_ALERT 0x4000 // Do NOT wait for VBLANK during an alert (e.g. AC->DC transition). +#define ATOM_PP_PLATFORM_CAP_VDDCI_CONTROL 0x8000 // Does the driver control VDDCI independently from VDDC. +#define ATOM_PP_PLATFORM_CAP_REGULATOR_HOT 0x00010000 // Enable the 'regulator hot' feature. +#define ATOM_PP_PLATFORM_CAP_BACO 0x00020000 // Does the driver supports BACO state. typedef struct _ATOM_PPLIB_POWERPLAYTABLE { @@ -5797,6 +5826,21 @@ typedef struct _ATOM_PPLIB_POWERPLAYTABLE } ATOM_PPLIB_POWERPLAYTABLE; +typedef struct _ATOM_PPLIB_POWERPLAYTABLE2 +{ + ATOM_PPLIB_POWERPLAYTABLE basicTable; + UCHAR ucNumCustomThermalPolicy; + USHORT usCustomThermalPolicyArrayOffset; +}ATOM_PPLIB_POWERPLAYTABLE2, *LPATOM_PPLIB_POWERPLAYTABLE2; + +typedef struct _ATOM_PPLIB_POWERPLAYTABLE3 +{ + ATOM_PPLIB_POWERPLAYTABLE2 basicTable2; + USHORT usFormatID; // To be used ONLY by PPGen. + USHORT usFanTableOffset; + USHORT usExtendendedHeaderOffset; +} ATOM_PPLIB_POWERPLAYTABLE3, *LPATOM_PPLIB_POWERPLAYTABLE3; + //// ATOM_PPLIB_NONCLOCK_INFO::usClassification #define ATOM_PPLIB_CLASSIFICATION_UI_MASK 0x0007 #define ATOM_PPLIB_CLASSIFICATION_UI_SHIFT 0 @@ -5816,7 +5860,9 @@ typedef struct _ATOM_PPLIB_POWERPLAYTABLE #define ATOM_PPLIB_CLASSIFICATION_UVDSTATE 0x0400 #define ATOM_PPLIB_CLASSIFICATION_3DLOW 0x0800 #define ATOM_PPLIB_CLASSIFICATION_ACPI 0x1000 -// remaining 3 bits are reserved +#define ATOM_PPLIB_CLASSIFICATION_HD2STATE 0x2000 +#define ATOM_PPLIB_CLASSIFICATION_HDSTATE 0x4000 +#define ATOM_PPLIB_CLASSIFICATION_SDSTATE 0x8000 //// ATOM_PPLIB_NONCLOCK_INFO::ulCapsAndSettings #define ATOM_PPLIB_SINGLE_DISPLAY_ONLY 0x00000001 @@ -5840,9 +5886,15 @@ typedef struct _ATOM_PPLIB_POWERPLAYTABLE #define ATOM_PPLIB_SOFTWARE_DISABLE_LOADBALANCING 0x00001000 #define ATOM_PPLIB_SOFTWARE_ENABLE_SLEEP_FOR_TIMESTAMPS 0x00002000 +#define ATOM_PPLIB_DISALLOW_ON_DC 0x00004000 #define ATOM_PPLIB_ENABLE_VARIBRIGHT 0x00008000 -#define ATOM_PPLIB_DISALLOW_ON_DC 0x00004000 +//memory related flags +#define ATOM_PPLIB_SWSTATE_MEMORY_DLL_OFF 0x000010000 + +//M3 Arb //2bits, current 3 sets of parameters in total +#define ATOM_PPLIB_M3ARB_MASK 0x00060000 +#define ATOM_PPLIB_M3ARB_SHIFT 17 // Contained in an array starting at the offset // in ATOM_PPLIB_POWERPLAYTABLE::usNonClockInfoArrayOffset. @@ -5860,6 +5912,9 @@ typedef struct _ATOM_PPLIB_NONCLOCK_INFO // Contained in an array starting at the offset // in ATOM_PPLIB_POWERPLAYTABLE::usClockInfoArrayOffset. // referenced from ATOM_PPLIB_STATE::ucClockStateIndices +#define ATOM_PPLIB_NONCLOCKINFO_VER1 12 +#define ATOM_PPLIB_NONCLOCKINFO_VER2 24 + typedef struct _ATOM_PPLIB_R600_CLOCK_INFO { USHORT usEngineClockLow; @@ -5882,6 +5937,23 @@ typedef struct _ATOM_PPLIB_R600_CLOCK_INFO #define ATOM_PPLIB_R600_FLAGS_BACKBIASENABLE 4 #define ATOM_PPLIB_R600_FLAGS_MEMORY_ODT_OFF 8 #define ATOM_PPLIB_R600_FLAGS_MEMORY_DLL_OFF 16 +#define ATOM_PPLIB_R600_FLAGS_LOWPOWER 32 // On the RV770 use 'low power' setting (sequencer S0). + +typedef struct _ATOM_PPLIB_EVERGREEN_CLOCK_INFO +{ + USHORT usEngineClockLow; + UCHAR ucEngineClockHigh; + + USHORT usMemoryClockLow; + UCHAR ucMemoryClockHigh; + + USHORT usVDDC; + USHORT usVDDCI; + USHORT usUnused; + + ULONG ulFlags; // ATOM_PPLIB_R600_FLAGS_* + +} ATOM_PPLIB_EVERGREEN_CLOCK_INFO; typedef struct _ATOM_PPLIB_RS780_CLOCK_INFO @@ -5895,7 +5967,7 @@ typedef struct _ATOM_PPLIB_RS780_CLOCK_INFO UCHAR ucPadding; // For proper alignment and size. USHORT usVDDC; // For the 780, use: None, Low, High, Variable UCHAR ucMaxHTLinkWidth; // From SBIOS - {2, 4, 8, 16} - UCHAR ucMinHTLinkWidth; // From SBIOS - {2, 4, 8, 16}. Effective only if CDLW enabled. Minimum down stream width could be bigger as display BW requriement. + UCHAR ucMinHTLinkWidth; // From SBIOS - {2, 4, 8, 16}. Effective only if CDLW enabled. Minimum down stream width could be bigger as display BW requirement. USHORT usHTLinkFreq; // See definition ATOM_PPLIB_RS780_HTLINKFREQ_xxx or in MHz(>=200). ULONG ulFlags; } ATOM_PPLIB_RS780_CLOCK_INFO; diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index a87990b3ae84..8c2d6478a221 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -26,7 +26,7 @@ #include <drm/drmP.h> #include <drm/drm_crtc_helper.h> #include <drm/radeon_drm.h> -#include "radeon_fixed.h" +#include <drm/drm_fixed.h> #include "radeon.h" #include "atom.h" #include "atom-bits.h" @@ -245,25 +245,27 @@ void atombios_crtc_dpms(struct drm_crtc *crtc, int mode) switch (mode) { case DRM_MODE_DPMS_ON: + radeon_crtc->enabled = true; + /* adjust pm to dpms changes BEFORE enabling crtcs */ + radeon_pm_compute_clocks(rdev); atombios_enable_crtc(crtc, ATOM_ENABLE); if (ASIC_IS_DCE3(rdev)) atombios_enable_crtc_memreq(crtc, ATOM_ENABLE); atombios_blank_crtc(crtc, ATOM_DISABLE); - /* XXX re-enable when interrupt support is added */ - if (!ASIC_IS_DCE4(rdev)) - drm_vblank_post_modeset(dev, radeon_crtc->crtc_id); + drm_vblank_post_modeset(dev, radeon_crtc->crtc_id); radeon_crtc_load_lut(crtc); break; case DRM_MODE_DPMS_STANDBY: case DRM_MODE_DPMS_SUSPEND: case DRM_MODE_DPMS_OFF: - /* XXX re-enable when interrupt support is added */ - if (!ASIC_IS_DCE4(rdev)) - drm_vblank_pre_modeset(dev, radeon_crtc->crtc_id); + drm_vblank_pre_modeset(dev, radeon_crtc->crtc_id); atombios_blank_crtc(crtc, ATOM_ENABLE); if (ASIC_IS_DCE3(rdev)) atombios_enable_crtc_memreq(crtc, ATOM_DISABLE); atombios_enable_crtc(crtc, ATOM_DISABLE); + radeon_crtc->enabled = false; + /* adjust pm to dpms changes AFTER disabling crtcs */ + radeon_pm_compute_clocks(rdev); break; } } @@ -496,7 +498,7 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, if ((rdev->family == CHIP_RS600) || (rdev->family == CHIP_RS690) || (rdev->family == CHIP_RS740)) - pll->flags |= (RADEON_PLL_USE_FRAC_FB_DIV | + pll->flags |= (/*RADEON_PLL_USE_FRAC_FB_DIV |*/ RADEON_PLL_PREFER_CLOSEST_LOWER); if (ASIC_IS_DCE32(rdev) && mode->clock > 200000) /* range limits??? */ @@ -705,6 +707,7 @@ static void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode break; case ATOM_DCPLL: case ATOM_PPLL_INVALID: + default: pll = &rdev->clock.dcpll; break; } @@ -1160,6 +1163,12 @@ static bool atombios_crtc_mode_fixup(struct drm_crtc *crtc, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode) { + struct drm_device *dev = crtc->dev; + struct radeon_device *rdev = dev->dev_private; + + /* adjust pm to upcoming mode change */ + radeon_pm_compute_clocks(rdev); + if (!radeon_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) return false; return true; diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c index 28b31c64f48d..abffb1499e22 100644 --- a/drivers/gpu/drm/radeon/atombios_dp.c +++ b/drivers/gpu/drm/radeon/atombios_dp.c @@ -351,7 +351,7 @@ retry: args.v1.ucChannelID = chan->rec.i2c_id; args.v1.ucDelay = delay / 10; if (ASIC_IS_DCE4(rdev)) - args.v2.ucHPD_ID = chan->rec.hpd_id; + args.v2.ucHPD_ID = chan->rec.hpd; atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index e8f447e20507..1caf625e472b 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -28,39 +28,246 @@ #include "radeon.h" #include "radeon_asic.h" #include "radeon_drm.h" -#include "rv770d.h" +#include "evergreend.h" #include "atom.h" #include "avivod.h" #include "evergreen_reg.h" +#define EVERGREEN_PFP_UCODE_SIZE 1120 +#define EVERGREEN_PM4_UCODE_SIZE 1376 + static void evergreen_gpu_init(struct radeon_device *rdev); void evergreen_fini(struct radeon_device *rdev); +void evergreen_pm_misc(struct radeon_device *rdev) +{ + int req_ps_idx = rdev->pm.requested_power_state_index; + int req_cm_idx = rdev->pm.requested_clock_mode_index; + struct radeon_power_state *ps = &rdev->pm.power_state[req_ps_idx]; + struct radeon_voltage *voltage = &ps->clock_info[req_cm_idx].voltage; + + if ((voltage->type == VOLTAGE_SW) && voltage->voltage) { + if (voltage->voltage != rdev->pm.current_vddc) { + radeon_atom_set_voltage(rdev, voltage->voltage); + rdev->pm.current_vddc = voltage->voltage; + DRM_DEBUG("Setting: v: %d\n", voltage->voltage); + } + } +} + +void evergreen_pm_prepare(struct radeon_device *rdev) +{ + struct drm_device *ddev = rdev->ddev; + struct drm_crtc *crtc; + struct radeon_crtc *radeon_crtc; + u32 tmp; + + /* disable any active CRTCs */ + list_for_each_entry(crtc, &ddev->mode_config.crtc_list, head) { + radeon_crtc = to_radeon_crtc(crtc); + if (radeon_crtc->enabled) { + tmp = RREG32(EVERGREEN_CRTC_CONTROL + radeon_crtc->crtc_offset); + tmp |= EVERGREEN_CRTC_DISP_READ_REQUEST_DISABLE; + WREG32(EVERGREEN_CRTC_CONTROL + radeon_crtc->crtc_offset, tmp); + } + } +} + +void evergreen_pm_finish(struct radeon_device *rdev) +{ + struct drm_device *ddev = rdev->ddev; + struct drm_crtc *crtc; + struct radeon_crtc *radeon_crtc; + u32 tmp; + + /* enable any active CRTCs */ + list_for_each_entry(crtc, &ddev->mode_config.crtc_list, head) { + radeon_crtc = to_radeon_crtc(crtc); + if (radeon_crtc->enabled) { + tmp = RREG32(EVERGREEN_CRTC_CONTROL + radeon_crtc->crtc_offset); + tmp &= ~EVERGREEN_CRTC_DISP_READ_REQUEST_DISABLE; + WREG32(EVERGREEN_CRTC_CONTROL + radeon_crtc->crtc_offset, tmp); + } + } +} + bool evergreen_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd) { bool connected = false; - /* XXX */ + + switch (hpd) { + case RADEON_HPD_1: + if (RREG32(DC_HPD1_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + case RADEON_HPD_2: + if (RREG32(DC_HPD2_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + case RADEON_HPD_3: + if (RREG32(DC_HPD3_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + case RADEON_HPD_4: + if (RREG32(DC_HPD4_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + case RADEON_HPD_5: + if (RREG32(DC_HPD5_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + case RADEON_HPD_6: + if (RREG32(DC_HPD6_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + default: + break; + } + return connected; } void evergreen_hpd_set_polarity(struct radeon_device *rdev, enum radeon_hpd_id hpd) { - /* XXX */ + u32 tmp; + bool connected = evergreen_hpd_sense(rdev, hpd); + + switch (hpd) { + case RADEON_HPD_1: + tmp = RREG32(DC_HPD1_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD1_INT_CONTROL, tmp); + break; + case RADEON_HPD_2: + tmp = RREG32(DC_HPD2_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD2_INT_CONTROL, tmp); + break; + case RADEON_HPD_3: + tmp = RREG32(DC_HPD3_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD3_INT_CONTROL, tmp); + break; + case RADEON_HPD_4: + tmp = RREG32(DC_HPD4_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD4_INT_CONTROL, tmp); + break; + case RADEON_HPD_5: + tmp = RREG32(DC_HPD5_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD5_INT_CONTROL, tmp); + break; + case RADEON_HPD_6: + tmp = RREG32(DC_HPD6_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD6_INT_CONTROL, tmp); + break; + default: + break; + } } void evergreen_hpd_init(struct radeon_device *rdev) { - /* XXX */ + struct drm_device *dev = rdev->ddev; + struct drm_connector *connector; + u32 tmp = DC_HPDx_CONNECTION_TIMER(0x9c4) | + DC_HPDx_RX_INT_TIMER(0xfa) | DC_HPDx_EN; + + list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + struct radeon_connector *radeon_connector = to_radeon_connector(connector); + switch (radeon_connector->hpd.hpd) { + case RADEON_HPD_1: + WREG32(DC_HPD1_CONTROL, tmp); + rdev->irq.hpd[0] = true; + break; + case RADEON_HPD_2: + WREG32(DC_HPD2_CONTROL, tmp); + rdev->irq.hpd[1] = true; + break; + case RADEON_HPD_3: + WREG32(DC_HPD3_CONTROL, tmp); + rdev->irq.hpd[2] = true; + break; + case RADEON_HPD_4: + WREG32(DC_HPD4_CONTROL, tmp); + rdev->irq.hpd[3] = true; + break; + case RADEON_HPD_5: + WREG32(DC_HPD5_CONTROL, tmp); + rdev->irq.hpd[4] = true; + break; + case RADEON_HPD_6: + WREG32(DC_HPD6_CONTROL, tmp); + rdev->irq.hpd[5] = true; + break; + default: + break; + } + } + if (rdev->irq.installed) + evergreen_irq_set(rdev); } - -void evergreen_bandwidth_update(struct radeon_device *rdev) +void evergreen_hpd_fini(struct radeon_device *rdev) { - /* XXX */ + struct drm_device *dev = rdev->ddev; + struct drm_connector *connector; + + list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + struct radeon_connector *radeon_connector = to_radeon_connector(connector); + switch (radeon_connector->hpd.hpd) { + case RADEON_HPD_1: + WREG32(DC_HPD1_CONTROL, 0); + rdev->irq.hpd[0] = false; + break; + case RADEON_HPD_2: + WREG32(DC_HPD2_CONTROL, 0); + rdev->irq.hpd[1] = false; + break; + case RADEON_HPD_3: + WREG32(DC_HPD3_CONTROL, 0); + rdev->irq.hpd[2] = false; + break; + case RADEON_HPD_4: + WREG32(DC_HPD4_CONTROL, 0); + rdev->irq.hpd[3] = false; + break; + case RADEON_HPD_5: + WREG32(DC_HPD5_CONTROL, 0); + rdev->irq.hpd[4] = false; + break; + case RADEON_HPD_6: + WREG32(DC_HPD6_CONTROL, 0); + rdev->irq.hpd[5] = false; + break; + default: + break; + } + } } -void evergreen_hpd_fini(struct radeon_device *rdev) +void evergreen_bandwidth_update(struct radeon_device *rdev) { /* XXX */ } @@ -83,10 +290,31 @@ static int evergreen_mc_wait_for_idle(struct radeon_device *rdev) /* * GART */ +void evergreen_pcie_gart_tlb_flush(struct radeon_device *rdev) +{ + unsigned i; + u32 tmp; + + WREG32(VM_CONTEXT0_REQUEST_RESPONSE, REQUEST_TYPE(1)); + for (i = 0; i < rdev->usec_timeout; i++) { + /* read MC_STATUS */ + tmp = RREG32(VM_CONTEXT0_REQUEST_RESPONSE); + tmp = (tmp & RESPONSE_TYPE_MASK) >> RESPONSE_TYPE_SHIFT; + if (tmp == 2) { + printk(KERN_WARNING "[drm] r600 flush TLB failed\n"); + return; + } + if (tmp) { + return; + } + udelay(1); + } +} + int evergreen_pcie_gart_enable(struct radeon_device *rdev) { u32 tmp; - int r, i; + int r; if (rdev->gart.table.vram.robj == NULL) { dev_err(rdev->dev, "No VRAM object for PCIE GART.\n"); @@ -121,10 +349,9 @@ int evergreen_pcie_gart_enable(struct radeon_device *rdev) RANGE_PROTECTION_FAULT_ENABLE_DEFAULT); WREG32(VM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR, (u32)(rdev->dummy_page.addr >> 12)); - for (i = 1; i < 7; i++) - WREG32(VM_CONTEXT0_CNTL + (i * 4), 0); + WREG32(VM_CONTEXT1_CNTL, 0); - r600_pcie_gart_tlb_flush(rdev); + evergreen_pcie_gart_tlb_flush(rdev); rdev->gart.ready = true; return 0; } @@ -132,11 +359,11 @@ int evergreen_pcie_gart_enable(struct radeon_device *rdev) void evergreen_pcie_gart_disable(struct radeon_device *rdev) { u32 tmp; - int i, r; + int r; /* Disable all tables */ - for (i = 0; i < 7; i++) - WREG32(VM_CONTEXT0_CNTL + (i * 4), 0); + WREG32(VM_CONTEXT0_CNTL, 0); + WREG32(VM_CONTEXT1_CNTL, 0); /* Setup L2 cache */ WREG32(VM_L2_CNTL, ENABLE_L2_FRAGMENT_PROCESSING | @@ -173,7 +400,6 @@ void evergreen_pcie_gart_fini(struct radeon_device *rdev) void evergreen_agp_enable(struct radeon_device *rdev) { u32 tmp; - int i; /* Setup L2 cache */ WREG32(VM_L2_CNTL, ENABLE_L2_CACHE | ENABLE_L2_FRAGMENT_PROCESSING | @@ -193,8 +419,8 @@ void evergreen_agp_enable(struct radeon_device *rdev) WREG32(MC_VM_MB_L1_TLB1_CNTL, tmp); WREG32(MC_VM_MB_L1_TLB2_CNTL, tmp); WREG32(MC_VM_MB_L1_TLB3_CNTL, tmp); - for (i = 0; i < 7; i++) - WREG32(VM_CONTEXT0_CNTL + (i * 4), 0); + WREG32(VM_CONTEXT0_CNTL, 0); + WREG32(VM_CONTEXT1_CNTL, 0); } static void evergreen_mc_stop(struct radeon_device *rdev, struct evergreen_mc_save *save) @@ -381,7 +607,7 @@ static void evergreen_mc_program(struct radeon_device *rdev) WREG32(MC_VM_FB_LOCATION, tmp); WREG32(HDP_NONSURFACE_BASE, (rdev->mc.vram_start >> 8)); WREG32(HDP_NONSURFACE_INFO, (2 << 7)); - WREG32(HDP_NONSURFACE_SIZE, (rdev->mc.mc_vram_size - 1) | 0x3FF); + WREG32(HDP_NONSURFACE_SIZE, 0x3FFFFFFF); if (rdev->flags & RADEON_IS_AGP) { WREG32(MC_VM_AGP_TOP, rdev->mc.gtt_end >> 16); WREG32(MC_VM_AGP_BOT, rdev->mc.gtt_start >> 16); @@ -400,40 +626,679 @@ static void evergreen_mc_program(struct radeon_device *rdev) rv515_vga_render_disable(rdev); } -#if 0 /* * CP. */ -static void evergreen_cp_stop(struct radeon_device *rdev) -{ - /* XXX */ -} - static int evergreen_cp_load_microcode(struct radeon_device *rdev) { - /* XXX */ + const __be32 *fw_data; + int i; + if (!rdev->me_fw || !rdev->pfp_fw) + return -EINVAL; + + r700_cp_stop(rdev); + WREG32(CP_RB_CNTL, RB_NO_UPDATE | (15 << 8) | (3 << 0)); + + fw_data = (const __be32 *)rdev->pfp_fw->data; + WREG32(CP_PFP_UCODE_ADDR, 0); + for (i = 0; i < EVERGREEN_PFP_UCODE_SIZE; i++) + WREG32(CP_PFP_UCODE_DATA, be32_to_cpup(fw_data++)); + WREG32(CP_PFP_UCODE_ADDR, 0); + + fw_data = (const __be32 *)rdev->me_fw->data; + WREG32(CP_ME_RAM_WADDR, 0); + for (i = 0; i < EVERGREEN_PM4_UCODE_SIZE; i++) + WREG32(CP_ME_RAM_DATA, be32_to_cpup(fw_data++)); + + WREG32(CP_PFP_UCODE_ADDR, 0); + WREG32(CP_ME_RAM_WADDR, 0); + WREG32(CP_ME_RAM_RADDR, 0); return 0; } +int evergreen_cp_resume(struct radeon_device *rdev) +{ + u32 tmp; + u32 rb_bufsz; + int r; + + /* Reset cp; if cp is reset, then PA, SH, VGT also need to be reset */ + WREG32(GRBM_SOFT_RESET, (SOFT_RESET_CP | + SOFT_RESET_PA | + SOFT_RESET_SH | + SOFT_RESET_VGT | + SOFT_RESET_SX)); + RREG32(GRBM_SOFT_RESET); + mdelay(15); + WREG32(GRBM_SOFT_RESET, 0); + RREG32(GRBM_SOFT_RESET); + + /* Set ring buffer size */ + rb_bufsz = drm_order(rdev->cp.ring_size / 8); + tmp = RB_NO_UPDATE | (drm_order(RADEON_GPU_PAGE_SIZE/8) << 8) | rb_bufsz; +#ifdef __BIG_ENDIAN + tmp |= BUF_SWAP_32BIT; +#endif + WREG32(CP_RB_CNTL, tmp); + WREG32(CP_SEM_WAIT_TIMER, 0x4); + + /* Set the write pointer delay */ + WREG32(CP_RB_WPTR_DELAY, 0); + + /* Initialize the ring buffer's read and write pointers */ + WREG32(CP_RB_CNTL, tmp | RB_RPTR_WR_ENA); + WREG32(CP_RB_RPTR_WR, 0); + WREG32(CP_RB_WPTR, 0); + WREG32(CP_RB_RPTR_ADDR, rdev->cp.gpu_addr & 0xFFFFFFFF); + WREG32(CP_RB_RPTR_ADDR_HI, upper_32_bits(rdev->cp.gpu_addr)); + mdelay(1); + WREG32(CP_RB_CNTL, tmp); + + WREG32(CP_RB_BASE, rdev->cp.gpu_addr >> 8); + WREG32(CP_DEBUG, (1 << 27) | (1 << 28)); + + rdev->cp.rptr = RREG32(CP_RB_RPTR); + rdev->cp.wptr = RREG32(CP_RB_WPTR); + + r600_cp_start(rdev); + rdev->cp.ready = true; + r = radeon_ring_test(rdev); + if (r) { + rdev->cp.ready = false; + return r; + } + return 0; +} /* * Core functions */ -static u32 evergreen_get_tile_pipe_to_backend_map(u32 num_tile_pipes, +static u32 evergreen_get_tile_pipe_to_backend_map(struct radeon_device *rdev, + u32 num_tile_pipes, u32 num_backends, u32 backend_disable_mask) { u32 backend_map = 0; + u32 enabled_backends_mask = 0; + u32 enabled_backends_count = 0; + u32 cur_pipe; + u32 swizzle_pipe[EVERGREEN_MAX_PIPES]; + u32 cur_backend = 0; + u32 i; + bool force_no_swizzle; + + if (num_tile_pipes > EVERGREEN_MAX_PIPES) + num_tile_pipes = EVERGREEN_MAX_PIPES; + if (num_tile_pipes < 1) + num_tile_pipes = 1; + if (num_backends > EVERGREEN_MAX_BACKENDS) + num_backends = EVERGREEN_MAX_BACKENDS; + if (num_backends < 1) + num_backends = 1; + + for (i = 0; i < EVERGREEN_MAX_BACKENDS; ++i) { + if (((backend_disable_mask >> i) & 1) == 0) { + enabled_backends_mask |= (1 << i); + ++enabled_backends_count; + } + if (enabled_backends_count == num_backends) + break; + } + + if (enabled_backends_count == 0) { + enabled_backends_mask = 1; + enabled_backends_count = 1; + } + + if (enabled_backends_count != num_backends) + num_backends = enabled_backends_count; + + memset((uint8_t *)&swizzle_pipe[0], 0, sizeof(u32) * EVERGREEN_MAX_PIPES); + switch (rdev->family) { + case CHIP_CEDAR: + case CHIP_REDWOOD: + force_no_swizzle = false; + break; + case CHIP_CYPRESS: + case CHIP_HEMLOCK: + case CHIP_JUNIPER: + default: + force_no_swizzle = true; + break; + } + if (force_no_swizzle) { + bool last_backend_enabled = false; + + force_no_swizzle = false; + for (i = 0; i < EVERGREEN_MAX_BACKENDS; ++i) { + if (((enabled_backends_mask >> i) & 1) == 1) { + if (last_backend_enabled) + force_no_swizzle = true; + last_backend_enabled = true; + } else + last_backend_enabled = false; + } + } + + switch (num_tile_pipes) { + case 1: + case 3: + case 5: + case 7: + DRM_ERROR("odd number of pipes!\n"); + break; + case 2: + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 1; + break; + case 4: + if (force_no_swizzle) { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 1; + swizzle_pipe[2] = 2; + swizzle_pipe[3] = 3; + } else { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 2; + swizzle_pipe[2] = 1; + swizzle_pipe[3] = 3; + } + break; + case 6: + if (force_no_swizzle) { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 1; + swizzle_pipe[2] = 2; + swizzle_pipe[3] = 3; + swizzle_pipe[4] = 4; + swizzle_pipe[5] = 5; + } else { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 2; + swizzle_pipe[2] = 4; + swizzle_pipe[3] = 1; + swizzle_pipe[4] = 3; + swizzle_pipe[5] = 5; + } + break; + case 8: + if (force_no_swizzle) { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 1; + swizzle_pipe[2] = 2; + swizzle_pipe[3] = 3; + swizzle_pipe[4] = 4; + swizzle_pipe[5] = 5; + swizzle_pipe[6] = 6; + swizzle_pipe[7] = 7; + } else { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 2; + swizzle_pipe[2] = 4; + swizzle_pipe[3] = 6; + swizzle_pipe[4] = 1; + swizzle_pipe[5] = 3; + swizzle_pipe[6] = 5; + swizzle_pipe[7] = 7; + } + break; + } + + for (cur_pipe = 0; cur_pipe < num_tile_pipes; ++cur_pipe) { + while (((1 << cur_backend) & enabled_backends_mask) == 0) + cur_backend = (cur_backend + 1) % EVERGREEN_MAX_BACKENDS; + + backend_map |= (((cur_backend & 0xf) << (swizzle_pipe[cur_pipe] * 4))); + + cur_backend = (cur_backend + 1) % EVERGREEN_MAX_BACKENDS; + } return backend_map; } -#endif static void evergreen_gpu_init(struct radeon_device *rdev) { - /* XXX */ + u32 cc_rb_backend_disable = 0; + u32 cc_gc_shader_pipe_config; + u32 gb_addr_config = 0; + u32 mc_shared_chmap, mc_arb_ramcfg; + u32 gb_backend_map; + u32 grbm_gfx_index; + u32 sx_debug_1; + u32 smx_dc_ctl0; + u32 sq_config; + u32 sq_lds_resource_mgmt; + u32 sq_gpr_resource_mgmt_1; + u32 sq_gpr_resource_mgmt_2; + u32 sq_gpr_resource_mgmt_3; + u32 sq_thread_resource_mgmt; + u32 sq_thread_resource_mgmt_2; + u32 sq_stack_resource_mgmt_1; + u32 sq_stack_resource_mgmt_2; + u32 sq_stack_resource_mgmt_3; + u32 vgt_cache_invalidation; + u32 hdp_host_path_cntl; + int i, j, num_shader_engines, ps_thread_count; + + switch (rdev->family) { + case CHIP_CYPRESS: + case CHIP_HEMLOCK: + rdev->config.evergreen.num_ses = 2; + rdev->config.evergreen.max_pipes = 4; + rdev->config.evergreen.max_tile_pipes = 8; + rdev->config.evergreen.max_simds = 10; + rdev->config.evergreen.max_backends = 4 * rdev->config.evergreen.num_ses; + rdev->config.evergreen.max_gprs = 256; + rdev->config.evergreen.max_threads = 248; + rdev->config.evergreen.max_gs_threads = 32; + rdev->config.evergreen.max_stack_entries = 512; + rdev->config.evergreen.sx_num_of_sets = 4; + rdev->config.evergreen.sx_max_export_size = 256; + rdev->config.evergreen.sx_max_export_pos_size = 64; + rdev->config.evergreen.sx_max_export_smx_size = 192; + rdev->config.evergreen.max_hw_contexts = 8; + rdev->config.evergreen.sq_num_cf_insts = 2; + + rdev->config.evergreen.sc_prim_fifo_size = 0x100; + rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30; + rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130; + break; + case CHIP_JUNIPER: + rdev->config.evergreen.num_ses = 1; + rdev->config.evergreen.max_pipes = 4; + rdev->config.evergreen.max_tile_pipes = 4; + rdev->config.evergreen.max_simds = 10; + rdev->config.evergreen.max_backends = 4 * rdev->config.evergreen.num_ses; + rdev->config.evergreen.max_gprs = 256; + rdev->config.evergreen.max_threads = 248; + rdev->config.evergreen.max_gs_threads = 32; + rdev->config.evergreen.max_stack_entries = 512; + rdev->config.evergreen.sx_num_of_sets = 4; + rdev->config.evergreen.sx_max_export_size = 256; + rdev->config.evergreen.sx_max_export_pos_size = 64; + rdev->config.evergreen.sx_max_export_smx_size = 192; + rdev->config.evergreen.max_hw_contexts = 8; + rdev->config.evergreen.sq_num_cf_insts = 2; + + rdev->config.evergreen.sc_prim_fifo_size = 0x100; + rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30; + rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130; + break; + case CHIP_REDWOOD: + rdev->config.evergreen.num_ses = 1; + rdev->config.evergreen.max_pipes = 4; + rdev->config.evergreen.max_tile_pipes = 4; + rdev->config.evergreen.max_simds = 5; + rdev->config.evergreen.max_backends = 2 * rdev->config.evergreen.num_ses; + rdev->config.evergreen.max_gprs = 256; + rdev->config.evergreen.max_threads = 248; + rdev->config.evergreen.max_gs_threads = 32; + rdev->config.evergreen.max_stack_entries = 256; + rdev->config.evergreen.sx_num_of_sets = 4; + rdev->config.evergreen.sx_max_export_size = 256; + rdev->config.evergreen.sx_max_export_pos_size = 64; + rdev->config.evergreen.sx_max_export_smx_size = 192; + rdev->config.evergreen.max_hw_contexts = 8; + rdev->config.evergreen.sq_num_cf_insts = 2; + + rdev->config.evergreen.sc_prim_fifo_size = 0x100; + rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30; + rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130; + break; + case CHIP_CEDAR: + default: + rdev->config.evergreen.num_ses = 1; + rdev->config.evergreen.max_pipes = 2; + rdev->config.evergreen.max_tile_pipes = 2; + rdev->config.evergreen.max_simds = 2; + rdev->config.evergreen.max_backends = 1 * rdev->config.evergreen.num_ses; + rdev->config.evergreen.max_gprs = 256; + rdev->config.evergreen.max_threads = 192; + rdev->config.evergreen.max_gs_threads = 16; + rdev->config.evergreen.max_stack_entries = 256; + rdev->config.evergreen.sx_num_of_sets = 4; + rdev->config.evergreen.sx_max_export_size = 128; + rdev->config.evergreen.sx_max_export_pos_size = 32; + rdev->config.evergreen.sx_max_export_smx_size = 96; + rdev->config.evergreen.max_hw_contexts = 4; + rdev->config.evergreen.sq_num_cf_insts = 1; + + rdev->config.evergreen.sc_prim_fifo_size = 0x40; + rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30; + rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130; + break; + } + + /* Initialize HDP */ + for (i = 0, j = 0; i < 32; i++, j += 0x18) { + WREG32((0x2c14 + j), 0x00000000); + WREG32((0x2c18 + j), 0x00000000); + WREG32((0x2c1c + j), 0x00000000); + WREG32((0x2c20 + j), 0x00000000); + WREG32((0x2c24 + j), 0x00000000); + } + + WREG32(GRBM_CNTL, GRBM_READ_TIMEOUT(0xff)); + + cc_gc_shader_pipe_config = RREG32(CC_GC_SHADER_PIPE_CONFIG) & ~2; + + cc_gc_shader_pipe_config |= + INACTIVE_QD_PIPES((EVERGREEN_MAX_PIPES_MASK << rdev->config.evergreen.max_pipes) + & EVERGREEN_MAX_PIPES_MASK); + cc_gc_shader_pipe_config |= + INACTIVE_SIMDS((EVERGREEN_MAX_SIMDS_MASK << rdev->config.evergreen.max_simds) + & EVERGREEN_MAX_SIMDS_MASK); + + cc_rb_backend_disable = + BACKEND_DISABLE((EVERGREEN_MAX_BACKENDS_MASK << rdev->config.evergreen.max_backends) + & EVERGREEN_MAX_BACKENDS_MASK); + + + mc_shared_chmap = RREG32(MC_SHARED_CHMAP); + mc_arb_ramcfg = RREG32(MC_ARB_RAMCFG); + + switch (rdev->config.evergreen.max_tile_pipes) { + case 1: + default: + gb_addr_config |= NUM_PIPES(0); + break; + case 2: + gb_addr_config |= NUM_PIPES(1); + break; + case 4: + gb_addr_config |= NUM_PIPES(2); + break; + case 8: + gb_addr_config |= NUM_PIPES(3); + break; + } + + gb_addr_config |= PIPE_INTERLEAVE_SIZE((mc_arb_ramcfg & BURSTLENGTH_MASK) >> BURSTLENGTH_SHIFT); + gb_addr_config |= BANK_INTERLEAVE_SIZE(0); + gb_addr_config |= NUM_SHADER_ENGINES(rdev->config.evergreen.num_ses - 1); + gb_addr_config |= SHADER_ENGINE_TILE_SIZE(1); + gb_addr_config |= NUM_GPUS(0); /* Hemlock? */ + gb_addr_config |= MULTI_GPU_TILE_SIZE(2); + + if (((mc_arb_ramcfg & NOOFCOLS_MASK) >> NOOFCOLS_SHIFT) > 2) + gb_addr_config |= ROW_SIZE(2); + else + gb_addr_config |= ROW_SIZE((mc_arb_ramcfg & NOOFCOLS_MASK) >> NOOFCOLS_SHIFT); + + if (rdev->ddev->pdev->device == 0x689e) { + u32 efuse_straps_4; + u32 efuse_straps_3; + u8 efuse_box_bit_131_124; + + WREG32(RCU_IND_INDEX, 0x204); + efuse_straps_4 = RREG32(RCU_IND_DATA); + WREG32(RCU_IND_INDEX, 0x203); + efuse_straps_3 = RREG32(RCU_IND_DATA); + efuse_box_bit_131_124 = (u8)(((efuse_straps_4 & 0xf) << 4) | ((efuse_straps_3 & 0xf0000000) >> 28)); + + switch(efuse_box_bit_131_124) { + case 0x00: + gb_backend_map = 0x76543210; + break; + case 0x55: + gb_backend_map = 0x77553311; + break; + case 0x56: + gb_backend_map = 0x77553300; + break; + case 0x59: + gb_backend_map = 0x77552211; + break; + case 0x66: + gb_backend_map = 0x77443300; + break; + case 0x99: + gb_backend_map = 0x66552211; + break; + case 0x5a: + gb_backend_map = 0x77552200; + break; + case 0xaa: + gb_backend_map = 0x66442200; + break; + case 0x95: + gb_backend_map = 0x66553311; + break; + default: + DRM_ERROR("bad backend map, using default\n"); + gb_backend_map = + evergreen_get_tile_pipe_to_backend_map(rdev, + rdev->config.evergreen.max_tile_pipes, + rdev->config.evergreen.max_backends, + ((EVERGREEN_MAX_BACKENDS_MASK << + rdev->config.evergreen.max_backends) & + EVERGREEN_MAX_BACKENDS_MASK)); + break; + } + } else if (rdev->ddev->pdev->device == 0x68b9) { + u32 efuse_straps_3; + u8 efuse_box_bit_127_124; + + WREG32(RCU_IND_INDEX, 0x203); + efuse_straps_3 = RREG32(RCU_IND_DATA); + efuse_box_bit_127_124 = (u8)(efuse_straps_3 & 0xF0000000) >> 28; + + switch(efuse_box_bit_127_124) { + case 0x0: + gb_backend_map = 0x00003210; + break; + case 0x5: + case 0x6: + case 0x9: + case 0xa: + gb_backend_map = 0x00003311; + break; + default: + DRM_ERROR("bad backend map, using default\n"); + gb_backend_map = + evergreen_get_tile_pipe_to_backend_map(rdev, + rdev->config.evergreen.max_tile_pipes, + rdev->config.evergreen.max_backends, + ((EVERGREEN_MAX_BACKENDS_MASK << + rdev->config.evergreen.max_backends) & + EVERGREEN_MAX_BACKENDS_MASK)); + break; + } + } else + gb_backend_map = + evergreen_get_tile_pipe_to_backend_map(rdev, + rdev->config.evergreen.max_tile_pipes, + rdev->config.evergreen.max_backends, + ((EVERGREEN_MAX_BACKENDS_MASK << + rdev->config.evergreen.max_backends) & + EVERGREEN_MAX_BACKENDS_MASK)); + + WREG32(GB_BACKEND_MAP, gb_backend_map); + WREG32(GB_ADDR_CONFIG, gb_addr_config); + WREG32(DMIF_ADDR_CONFIG, gb_addr_config); + WREG32(HDP_ADDR_CONFIG, gb_addr_config); + + num_shader_engines = ((RREG32(GB_ADDR_CONFIG) & NUM_SHADER_ENGINES(3)) >> 12) + 1; + grbm_gfx_index = INSTANCE_BROADCAST_WRITES; + + for (i = 0; i < rdev->config.evergreen.num_ses; i++) { + u32 rb = cc_rb_backend_disable | (0xf0 << 16); + u32 sp = cc_gc_shader_pipe_config; + u32 gfx = grbm_gfx_index | SE_INDEX(i); + + if (i == num_shader_engines) { + rb |= BACKEND_DISABLE(EVERGREEN_MAX_BACKENDS_MASK); + sp |= INACTIVE_SIMDS(EVERGREEN_MAX_SIMDS_MASK); + } + + WREG32(GRBM_GFX_INDEX, gfx); + WREG32(RLC_GFX_INDEX, gfx); + + WREG32(CC_RB_BACKEND_DISABLE, rb); + WREG32(CC_SYS_RB_BACKEND_DISABLE, rb); + WREG32(GC_USER_RB_BACKEND_DISABLE, rb); + WREG32(CC_GC_SHADER_PIPE_CONFIG, sp); + } + + grbm_gfx_index |= SE_BROADCAST_WRITES; + WREG32(GRBM_GFX_INDEX, grbm_gfx_index); + WREG32(RLC_GFX_INDEX, grbm_gfx_index); + + WREG32(CGTS_SYS_TCC_DISABLE, 0); + WREG32(CGTS_TCC_DISABLE, 0); + WREG32(CGTS_USER_SYS_TCC_DISABLE, 0); + WREG32(CGTS_USER_TCC_DISABLE, 0); + + /* set HW defaults for 3D engine */ + WREG32(CP_QUEUE_THRESHOLDS, (ROQ_IB1_START(0x16) | + ROQ_IB2_START(0x2b))); + + WREG32(CP_MEQ_THRESHOLDS, STQ_SPLIT(0x30)); + + WREG32(TA_CNTL_AUX, (DISABLE_CUBE_ANISO | + SYNC_GRADIENT | + SYNC_WALKER | + SYNC_ALIGNER)); + + sx_debug_1 = RREG32(SX_DEBUG_1); + sx_debug_1 |= ENABLE_NEW_SMX_ADDRESS; + WREG32(SX_DEBUG_1, sx_debug_1); + + + smx_dc_ctl0 = RREG32(SMX_DC_CTL0); + smx_dc_ctl0 &= ~NUMBER_OF_SETS(0x1ff); + smx_dc_ctl0 |= NUMBER_OF_SETS(rdev->config.evergreen.sx_num_of_sets); + WREG32(SMX_DC_CTL0, smx_dc_ctl0); + + WREG32(SX_EXPORT_BUFFER_SIZES, (COLOR_BUFFER_SIZE((rdev->config.evergreen.sx_max_export_size / 4) - 1) | + POSITION_BUFFER_SIZE((rdev->config.evergreen.sx_max_export_pos_size / 4) - 1) | + SMX_BUFFER_SIZE((rdev->config.evergreen.sx_max_export_smx_size / 4) - 1))); + + WREG32(PA_SC_FIFO_SIZE, (SC_PRIM_FIFO_SIZE(rdev->config.evergreen.sc_prim_fifo_size) | + SC_HIZ_TILE_FIFO_SIZE(rdev->config.evergreen.sc_hiz_tile_fifo_size) | + SC_EARLYZ_TILE_FIFO_SIZE(rdev->config.evergreen.sc_earlyz_tile_fifo_size))); + + WREG32(VGT_NUM_INSTANCES, 1); + WREG32(SPI_CONFIG_CNTL, 0); + WREG32(SPI_CONFIG_CNTL_1, VTX_DONE_DELAY(4)); + WREG32(CP_PERFMON_CNTL, 0); + + WREG32(SQ_MS_FIFO_SIZES, (CACHE_FIFO_SIZE(16 * rdev->config.evergreen.sq_num_cf_insts) | + FETCH_FIFO_HIWATER(0x4) | + DONE_FIFO_HIWATER(0xe0) | + ALU_UPDATE_FIFO_HIWATER(0x8))); + + sq_config = RREG32(SQ_CONFIG); + sq_config &= ~(PS_PRIO(3) | + VS_PRIO(3) | + GS_PRIO(3) | + ES_PRIO(3)); + sq_config |= (VC_ENABLE | + EXPORT_SRC_C | + PS_PRIO(0) | + VS_PRIO(1) | + GS_PRIO(2) | + ES_PRIO(3)); + + if (rdev->family == CHIP_CEDAR) + /* no vertex cache */ + sq_config &= ~VC_ENABLE; + + sq_lds_resource_mgmt = RREG32(SQ_LDS_RESOURCE_MGMT); + + sq_gpr_resource_mgmt_1 = NUM_PS_GPRS((rdev->config.evergreen.max_gprs - (4 * 2))* 12 / 32); + sq_gpr_resource_mgmt_1 |= NUM_VS_GPRS((rdev->config.evergreen.max_gprs - (4 * 2)) * 6 / 32); + sq_gpr_resource_mgmt_1 |= NUM_CLAUSE_TEMP_GPRS(4); + sq_gpr_resource_mgmt_2 = NUM_GS_GPRS((rdev->config.evergreen.max_gprs - (4 * 2)) * 4 / 32); + sq_gpr_resource_mgmt_2 |= NUM_ES_GPRS((rdev->config.evergreen.max_gprs - (4 * 2)) * 4 / 32); + sq_gpr_resource_mgmt_3 = NUM_HS_GPRS((rdev->config.evergreen.max_gprs - (4 * 2)) * 3 / 32); + sq_gpr_resource_mgmt_3 |= NUM_LS_GPRS((rdev->config.evergreen.max_gprs - (4 * 2)) * 3 / 32); + + if (rdev->family == CHIP_CEDAR) + ps_thread_count = 96; + else + ps_thread_count = 128; + + sq_thread_resource_mgmt = NUM_PS_THREADS(ps_thread_count); + sq_thread_resource_mgmt |= NUM_VS_THREADS((((rdev->config.evergreen.max_threads - ps_thread_count) / 6) / 8) * 8); + sq_thread_resource_mgmt |= NUM_GS_THREADS((((rdev->config.evergreen.max_threads - ps_thread_count) / 6) / 8) * 8); + sq_thread_resource_mgmt |= NUM_ES_THREADS((((rdev->config.evergreen.max_threads - ps_thread_count) / 6) / 8) * 8); + sq_thread_resource_mgmt_2 = NUM_HS_THREADS((((rdev->config.evergreen.max_threads - ps_thread_count) / 6) / 8) * 8); + sq_thread_resource_mgmt_2 |= NUM_LS_THREADS((((rdev->config.evergreen.max_threads - ps_thread_count) / 6) / 8) * 8); + + sq_stack_resource_mgmt_1 = NUM_PS_STACK_ENTRIES((rdev->config.evergreen.max_stack_entries * 1) / 6); + sq_stack_resource_mgmt_1 |= NUM_VS_STACK_ENTRIES((rdev->config.evergreen.max_stack_entries * 1) / 6); + sq_stack_resource_mgmt_2 = NUM_GS_STACK_ENTRIES((rdev->config.evergreen.max_stack_entries * 1) / 6); + sq_stack_resource_mgmt_2 |= NUM_ES_STACK_ENTRIES((rdev->config.evergreen.max_stack_entries * 1) / 6); + sq_stack_resource_mgmt_3 = NUM_HS_STACK_ENTRIES((rdev->config.evergreen.max_stack_entries * 1) / 6); + sq_stack_resource_mgmt_3 |= NUM_LS_STACK_ENTRIES((rdev->config.evergreen.max_stack_entries * 1) / 6); + + WREG32(SQ_CONFIG, sq_config); + WREG32(SQ_GPR_RESOURCE_MGMT_1, sq_gpr_resource_mgmt_1); + WREG32(SQ_GPR_RESOURCE_MGMT_2, sq_gpr_resource_mgmt_2); + WREG32(SQ_GPR_RESOURCE_MGMT_3, sq_gpr_resource_mgmt_3); + WREG32(SQ_THREAD_RESOURCE_MGMT, sq_thread_resource_mgmt); + WREG32(SQ_THREAD_RESOURCE_MGMT_2, sq_thread_resource_mgmt_2); + WREG32(SQ_STACK_RESOURCE_MGMT_1, sq_stack_resource_mgmt_1); + WREG32(SQ_STACK_RESOURCE_MGMT_2, sq_stack_resource_mgmt_2); + WREG32(SQ_STACK_RESOURCE_MGMT_3, sq_stack_resource_mgmt_3); + WREG32(SQ_DYN_GPR_CNTL_PS_FLUSH_REQ, 0); + WREG32(SQ_LDS_RESOURCE_MGMT, sq_lds_resource_mgmt); + + WREG32(PA_SC_FORCE_EOV_MAX_CNTS, (FORCE_EOV_MAX_CLK_CNT(4095) | + FORCE_EOV_MAX_REZ_CNT(255))); + + if (rdev->family == CHIP_CEDAR) + vgt_cache_invalidation = CACHE_INVALIDATION(TC_ONLY); + else + vgt_cache_invalidation = CACHE_INVALIDATION(VC_AND_TC); + vgt_cache_invalidation |= AUTO_INVLD_EN(ES_AND_GS_AUTO); + WREG32(VGT_CACHE_INVALIDATION, vgt_cache_invalidation); + + WREG32(VGT_GS_VERTEX_REUSE, 16); + WREG32(PA_SC_LINE_STIPPLE_STATE, 0); + + WREG32(VGT_VERTEX_REUSE_BLOCK_CNTL, 14); + WREG32(VGT_OUT_DEALLOC_CNTL, 16); + + WREG32(CB_PERF_CTR0_SEL_0, 0); + WREG32(CB_PERF_CTR0_SEL_1, 0); + WREG32(CB_PERF_CTR1_SEL_0, 0); + WREG32(CB_PERF_CTR1_SEL_1, 0); + WREG32(CB_PERF_CTR2_SEL_0, 0); + WREG32(CB_PERF_CTR2_SEL_1, 0); + WREG32(CB_PERF_CTR3_SEL_0, 0); + WREG32(CB_PERF_CTR3_SEL_1, 0); + + /* clear render buffer base addresses */ + WREG32(CB_COLOR0_BASE, 0); + WREG32(CB_COLOR1_BASE, 0); + WREG32(CB_COLOR2_BASE, 0); + WREG32(CB_COLOR3_BASE, 0); + WREG32(CB_COLOR4_BASE, 0); + WREG32(CB_COLOR5_BASE, 0); + WREG32(CB_COLOR6_BASE, 0); + WREG32(CB_COLOR7_BASE, 0); + WREG32(CB_COLOR8_BASE, 0); + WREG32(CB_COLOR9_BASE, 0); + WREG32(CB_COLOR10_BASE, 0); + WREG32(CB_COLOR11_BASE, 0); + + /* set the shader const cache sizes to 0 */ + for (i = SQ_ALU_CONST_BUFFER_SIZE_PS_0; i < 0x28200; i += 4) + WREG32(i, 0); + for (i = SQ_ALU_CONST_BUFFER_SIZE_HS_0; i < 0x29000; i += 4) + WREG32(i, 0); + + hdp_host_path_cntl = RREG32(HDP_HOST_PATH_CNTL); + WREG32(HDP_HOST_PATH_CNTL, hdp_host_path_cntl); + + WREG32(PA_CL_ENHANCE, CLIP_VTX_REORDER_ENA | NUM_CLIP_SEQ(3)); + + udelay(50); + } int evergreen_mc_init(struct radeon_device *rdev) @@ -476,26 +1341,627 @@ int evergreen_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE) * 1024 * 1024; rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE) * 1024 * 1024; rdev->mc.visible_vram_size = rdev->mc.aper_size; - /* FIXME remove this once we support unmappable VRAM */ - if (rdev->mc.mc_vram_size > rdev->mc.aper_size) { - rdev->mc.mc_vram_size = rdev->mc.aper_size; - rdev->mc.real_vram_size = rdev->mc.aper_size; - } r600_vram_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); return 0; } -int evergreen_gpu_reset(struct radeon_device *rdev) +bool evergreen_gpu_is_lockup(struct radeon_device *rdev) { /* FIXME: implement for evergreen */ + return false; +} + +static int evergreen_gpu_soft_reset(struct radeon_device *rdev) +{ + struct evergreen_mc_save save; + u32 srbm_reset = 0; + u32 grbm_reset = 0; + + dev_info(rdev->dev, "GPU softreset \n"); + dev_info(rdev->dev, " GRBM_STATUS=0x%08X\n", + RREG32(GRBM_STATUS)); + dev_info(rdev->dev, " GRBM_STATUS_SE0=0x%08X\n", + RREG32(GRBM_STATUS_SE0)); + dev_info(rdev->dev, " GRBM_STATUS_SE1=0x%08X\n", + RREG32(GRBM_STATUS_SE1)); + dev_info(rdev->dev, " SRBM_STATUS=0x%08X\n", + RREG32(SRBM_STATUS)); + evergreen_mc_stop(rdev, &save); + if (evergreen_mc_wait_for_idle(rdev)) { + dev_warn(rdev->dev, "Wait for MC idle timedout !\n"); + } + /* Disable CP parsing/prefetching */ + WREG32(CP_ME_CNTL, CP_ME_HALT | CP_PFP_HALT); + + /* reset all the gfx blocks */ + grbm_reset = (SOFT_RESET_CP | + SOFT_RESET_CB | + SOFT_RESET_DB | + SOFT_RESET_PA | + SOFT_RESET_SC | + SOFT_RESET_SPI | + SOFT_RESET_SH | + SOFT_RESET_SX | + SOFT_RESET_TC | + SOFT_RESET_TA | + SOFT_RESET_VC | + SOFT_RESET_VGT); + + dev_info(rdev->dev, " GRBM_SOFT_RESET=0x%08X\n", grbm_reset); + WREG32(GRBM_SOFT_RESET, grbm_reset); + (void)RREG32(GRBM_SOFT_RESET); + udelay(50); + WREG32(GRBM_SOFT_RESET, 0); + (void)RREG32(GRBM_SOFT_RESET); + + /* reset all the system blocks */ + srbm_reset = SRBM_SOFT_RESET_ALL_MASK; + + dev_info(rdev->dev, " SRBM_SOFT_RESET=0x%08X\n", srbm_reset); + WREG32(SRBM_SOFT_RESET, srbm_reset); + (void)RREG32(SRBM_SOFT_RESET); + udelay(50); + WREG32(SRBM_SOFT_RESET, 0); + (void)RREG32(SRBM_SOFT_RESET); + /* Wait a little for things to settle down */ + udelay(50); + dev_info(rdev->dev, " GRBM_STATUS=0x%08X\n", + RREG32(GRBM_STATUS)); + dev_info(rdev->dev, " GRBM_STATUS_SE0=0x%08X\n", + RREG32(GRBM_STATUS_SE0)); + dev_info(rdev->dev, " GRBM_STATUS_SE1=0x%08X\n", + RREG32(GRBM_STATUS_SE1)); + dev_info(rdev->dev, " SRBM_STATUS=0x%08X\n", + RREG32(SRBM_STATUS)); + /* After reset we need to reinit the asic as GPU often endup in an + * incoherent state. + */ + atom_asic_init(rdev->mode_info.atom_context); + evergreen_mc_resume(rdev, &save); + return 0; +} + +int evergreen_asic_reset(struct radeon_device *rdev) +{ + return evergreen_gpu_soft_reset(rdev); +} + +/* Interrupts */ + +u32 evergreen_get_vblank_counter(struct radeon_device *rdev, int crtc) +{ + switch (crtc) { + case 0: + return RREG32(CRTC_STATUS_FRAME_COUNT + EVERGREEN_CRTC0_REGISTER_OFFSET); + case 1: + return RREG32(CRTC_STATUS_FRAME_COUNT + EVERGREEN_CRTC1_REGISTER_OFFSET); + case 2: + return RREG32(CRTC_STATUS_FRAME_COUNT + EVERGREEN_CRTC2_REGISTER_OFFSET); + case 3: + return RREG32(CRTC_STATUS_FRAME_COUNT + EVERGREEN_CRTC3_REGISTER_OFFSET); + case 4: + return RREG32(CRTC_STATUS_FRAME_COUNT + EVERGREEN_CRTC4_REGISTER_OFFSET); + case 5: + return RREG32(CRTC_STATUS_FRAME_COUNT + EVERGREEN_CRTC5_REGISTER_OFFSET); + default: + return 0; + } +} + +void evergreen_disable_interrupt_state(struct radeon_device *rdev) +{ + u32 tmp; + + WREG32(CP_INT_CNTL, 0); + WREG32(GRBM_INT_CNTL, 0); + WREG32(INT_MASK + EVERGREEN_CRTC0_REGISTER_OFFSET, 0); + WREG32(INT_MASK + EVERGREEN_CRTC1_REGISTER_OFFSET, 0); + WREG32(INT_MASK + EVERGREEN_CRTC2_REGISTER_OFFSET, 0); + WREG32(INT_MASK + EVERGREEN_CRTC3_REGISTER_OFFSET, 0); + WREG32(INT_MASK + EVERGREEN_CRTC4_REGISTER_OFFSET, 0); + WREG32(INT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, 0); + + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, 0); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, 0); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, 0); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, 0); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, 0); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, 0); + + WREG32(DACA_AUTODETECT_INT_CONTROL, 0); + WREG32(DACB_AUTODETECT_INT_CONTROL, 0); + + tmp = RREG32(DC_HPD1_INT_CONTROL) & DC_HPDx_INT_POLARITY; + WREG32(DC_HPD1_INT_CONTROL, tmp); + tmp = RREG32(DC_HPD2_INT_CONTROL) & DC_HPDx_INT_POLARITY; + WREG32(DC_HPD2_INT_CONTROL, tmp); + tmp = RREG32(DC_HPD3_INT_CONTROL) & DC_HPDx_INT_POLARITY; + WREG32(DC_HPD3_INT_CONTROL, tmp); + tmp = RREG32(DC_HPD4_INT_CONTROL) & DC_HPDx_INT_POLARITY; + WREG32(DC_HPD4_INT_CONTROL, tmp); + tmp = RREG32(DC_HPD5_INT_CONTROL) & DC_HPDx_INT_POLARITY; + WREG32(DC_HPD5_INT_CONTROL, tmp); + tmp = RREG32(DC_HPD6_INT_CONTROL) & DC_HPDx_INT_POLARITY; + WREG32(DC_HPD6_INT_CONTROL, tmp); + +} + +int evergreen_irq_set(struct radeon_device *rdev) +{ + u32 cp_int_cntl = CNTX_BUSY_INT_ENABLE | CNTX_EMPTY_INT_ENABLE; + u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0; + u32 hpd1, hpd2, hpd3, hpd4, hpd5, hpd6; + u32 grbm_int_cntl = 0; + + if (!rdev->irq.installed) { + WARN(1, "Can't enable IRQ/MSI because no handler is installed.\n"); + return -EINVAL; + } + /* don't enable anything if the ih is disabled */ + if (!rdev->ih.enabled) { + r600_disable_interrupts(rdev); + /* force the active interrupt state to all disabled */ + evergreen_disable_interrupt_state(rdev); + return 0; + } + + hpd1 = RREG32(DC_HPD1_INT_CONTROL) & ~DC_HPDx_INT_EN; + hpd2 = RREG32(DC_HPD2_INT_CONTROL) & ~DC_HPDx_INT_EN; + hpd3 = RREG32(DC_HPD3_INT_CONTROL) & ~DC_HPDx_INT_EN; + hpd4 = RREG32(DC_HPD4_INT_CONTROL) & ~DC_HPDx_INT_EN; + hpd5 = RREG32(DC_HPD5_INT_CONTROL) & ~DC_HPDx_INT_EN; + hpd6 = RREG32(DC_HPD6_INT_CONTROL) & ~DC_HPDx_INT_EN; + + if (rdev->irq.sw_int) { + DRM_DEBUG("evergreen_irq_set: sw int\n"); + cp_int_cntl |= RB_INT_ENABLE; + } + if (rdev->irq.crtc_vblank_int[0]) { + DRM_DEBUG("evergreen_irq_set: vblank 0\n"); + crtc1 |= VBLANK_INT_MASK; + } + if (rdev->irq.crtc_vblank_int[1]) { + DRM_DEBUG("evergreen_irq_set: vblank 1\n"); + crtc2 |= VBLANK_INT_MASK; + } + if (rdev->irq.crtc_vblank_int[2]) { + DRM_DEBUG("evergreen_irq_set: vblank 2\n"); + crtc3 |= VBLANK_INT_MASK; + } + if (rdev->irq.crtc_vblank_int[3]) { + DRM_DEBUG("evergreen_irq_set: vblank 3\n"); + crtc4 |= VBLANK_INT_MASK; + } + if (rdev->irq.crtc_vblank_int[4]) { + DRM_DEBUG("evergreen_irq_set: vblank 4\n"); + crtc5 |= VBLANK_INT_MASK; + } + if (rdev->irq.crtc_vblank_int[5]) { + DRM_DEBUG("evergreen_irq_set: vblank 5\n"); + crtc6 |= VBLANK_INT_MASK; + } + if (rdev->irq.hpd[0]) { + DRM_DEBUG("evergreen_irq_set: hpd 1\n"); + hpd1 |= DC_HPDx_INT_EN; + } + if (rdev->irq.hpd[1]) { + DRM_DEBUG("evergreen_irq_set: hpd 2\n"); + hpd2 |= DC_HPDx_INT_EN; + } + if (rdev->irq.hpd[2]) { + DRM_DEBUG("evergreen_irq_set: hpd 3\n"); + hpd3 |= DC_HPDx_INT_EN; + } + if (rdev->irq.hpd[3]) { + DRM_DEBUG("evergreen_irq_set: hpd 4\n"); + hpd4 |= DC_HPDx_INT_EN; + } + if (rdev->irq.hpd[4]) { + DRM_DEBUG("evergreen_irq_set: hpd 5\n"); + hpd5 |= DC_HPDx_INT_EN; + } + if (rdev->irq.hpd[5]) { + DRM_DEBUG("evergreen_irq_set: hpd 6\n"); + hpd6 |= DC_HPDx_INT_EN; + } + if (rdev->irq.gui_idle) { + DRM_DEBUG("gui idle\n"); + grbm_int_cntl |= GUI_IDLE_INT_ENABLE; + } + + WREG32(CP_INT_CNTL, cp_int_cntl); + WREG32(GRBM_INT_CNTL, grbm_int_cntl); + + WREG32(INT_MASK + EVERGREEN_CRTC0_REGISTER_OFFSET, crtc1); + WREG32(INT_MASK + EVERGREEN_CRTC1_REGISTER_OFFSET, crtc2); + WREG32(INT_MASK + EVERGREEN_CRTC2_REGISTER_OFFSET, crtc3); + WREG32(INT_MASK + EVERGREEN_CRTC3_REGISTER_OFFSET, crtc4); + WREG32(INT_MASK + EVERGREEN_CRTC4_REGISTER_OFFSET, crtc5); + WREG32(INT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, crtc6); + + WREG32(DC_HPD1_INT_CONTROL, hpd1); + WREG32(DC_HPD2_INT_CONTROL, hpd2); + WREG32(DC_HPD3_INT_CONTROL, hpd3); + WREG32(DC_HPD4_INT_CONTROL, hpd4); + WREG32(DC_HPD5_INT_CONTROL, hpd5); + WREG32(DC_HPD6_INT_CONTROL, hpd6); + return 0; } +static inline void evergreen_irq_ack(struct radeon_device *rdev, + u32 *disp_int, + u32 *disp_int_cont, + u32 *disp_int_cont2, + u32 *disp_int_cont3, + u32 *disp_int_cont4, + u32 *disp_int_cont5) +{ + u32 tmp; + + *disp_int = RREG32(DISP_INTERRUPT_STATUS); + *disp_int_cont = RREG32(DISP_INTERRUPT_STATUS_CONTINUE); + *disp_int_cont2 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE2); + *disp_int_cont3 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE3); + *disp_int_cont4 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE4); + *disp_int_cont5 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE5); + + if (*disp_int & LB_D1_VBLANK_INTERRUPT) + WREG32(VBLANK_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, VBLANK_ACK); + if (*disp_int & LB_D1_VLINE_INTERRUPT) + WREG32(VLINE_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, VLINE_ACK); + + if (*disp_int_cont & LB_D2_VBLANK_INTERRUPT) + WREG32(VBLANK_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, VBLANK_ACK); + if (*disp_int_cont & LB_D2_VLINE_INTERRUPT) + WREG32(VLINE_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, VLINE_ACK); + + if (*disp_int_cont2 & LB_D3_VBLANK_INTERRUPT) + WREG32(VBLANK_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, VBLANK_ACK); + if (*disp_int_cont2 & LB_D3_VLINE_INTERRUPT) + WREG32(VLINE_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, VLINE_ACK); + + if (*disp_int_cont3 & LB_D4_VBLANK_INTERRUPT) + WREG32(VBLANK_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET, VBLANK_ACK); + if (*disp_int_cont3 & LB_D4_VLINE_INTERRUPT) + WREG32(VLINE_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET, VLINE_ACK); + + if (*disp_int_cont4 & LB_D5_VBLANK_INTERRUPT) + WREG32(VBLANK_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, VBLANK_ACK); + if (*disp_int_cont4 & LB_D5_VLINE_INTERRUPT) + WREG32(VLINE_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, VLINE_ACK); + + if (*disp_int_cont5 & LB_D6_VBLANK_INTERRUPT) + WREG32(VBLANK_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET, VBLANK_ACK); + if (*disp_int_cont5 & LB_D6_VLINE_INTERRUPT) + WREG32(VLINE_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET, VLINE_ACK); + + if (*disp_int & DC_HPD1_INTERRUPT) { + tmp = RREG32(DC_HPD1_INT_CONTROL); + tmp |= DC_HPDx_INT_ACK; + WREG32(DC_HPD1_INT_CONTROL, tmp); + } + if (*disp_int_cont & DC_HPD2_INTERRUPT) { + tmp = RREG32(DC_HPD2_INT_CONTROL); + tmp |= DC_HPDx_INT_ACK; + WREG32(DC_HPD2_INT_CONTROL, tmp); + } + if (*disp_int_cont2 & DC_HPD3_INTERRUPT) { + tmp = RREG32(DC_HPD3_INT_CONTROL); + tmp |= DC_HPDx_INT_ACK; + WREG32(DC_HPD3_INT_CONTROL, tmp); + } + if (*disp_int_cont3 & DC_HPD4_INTERRUPT) { + tmp = RREG32(DC_HPD4_INT_CONTROL); + tmp |= DC_HPDx_INT_ACK; + WREG32(DC_HPD4_INT_CONTROL, tmp); + } + if (*disp_int_cont4 & DC_HPD5_INTERRUPT) { + tmp = RREG32(DC_HPD5_INT_CONTROL); + tmp |= DC_HPDx_INT_ACK; + WREG32(DC_HPD5_INT_CONTROL, tmp); + } + if (*disp_int_cont5 & DC_HPD6_INTERRUPT) { + tmp = RREG32(DC_HPD5_INT_CONTROL); + tmp |= DC_HPDx_INT_ACK; + WREG32(DC_HPD6_INT_CONTROL, tmp); + } +} + +void evergreen_irq_disable(struct radeon_device *rdev) +{ + u32 disp_int, disp_int_cont, disp_int_cont2; + u32 disp_int_cont3, disp_int_cont4, disp_int_cont5; + + r600_disable_interrupts(rdev); + /* Wait and acknowledge irq */ + mdelay(1); + evergreen_irq_ack(rdev, &disp_int, &disp_int_cont, &disp_int_cont2, + &disp_int_cont3, &disp_int_cont4, &disp_int_cont5); + evergreen_disable_interrupt_state(rdev); +} + +static void evergreen_irq_suspend(struct radeon_device *rdev) +{ + evergreen_irq_disable(rdev); + r600_rlc_stop(rdev); +} + +static inline u32 evergreen_get_ih_wptr(struct radeon_device *rdev) +{ + u32 wptr, tmp; + + /* XXX use writeback */ + wptr = RREG32(IH_RB_WPTR); + + if (wptr & RB_OVERFLOW) { + /* When a ring buffer overflow happen start parsing interrupt + * from the last not overwritten vector (wptr + 16). Hopefully + * this should allow us to catchup. + */ + dev_warn(rdev->dev, "IH ring buffer overflow (0x%08X, %d, %d)\n", + wptr, rdev->ih.rptr, (wptr + 16) + rdev->ih.ptr_mask); + rdev->ih.rptr = (wptr + 16) & rdev->ih.ptr_mask; + tmp = RREG32(IH_RB_CNTL); + tmp |= IH_WPTR_OVERFLOW_CLEAR; + WREG32(IH_RB_CNTL, tmp); + } + return (wptr & rdev->ih.ptr_mask); +} + +int evergreen_irq_process(struct radeon_device *rdev) +{ + u32 wptr = evergreen_get_ih_wptr(rdev); + u32 rptr = rdev->ih.rptr; + u32 src_id, src_data; + u32 ring_index; + u32 disp_int, disp_int_cont, disp_int_cont2; + u32 disp_int_cont3, disp_int_cont4, disp_int_cont5; + unsigned long flags; + bool queue_hotplug = false; + + DRM_DEBUG("r600_irq_process start: rptr %d, wptr %d\n", rptr, wptr); + if (!rdev->ih.enabled) + return IRQ_NONE; + + spin_lock_irqsave(&rdev->ih.lock, flags); + + if (rptr == wptr) { + spin_unlock_irqrestore(&rdev->ih.lock, flags); + return IRQ_NONE; + } + if (rdev->shutdown) { + spin_unlock_irqrestore(&rdev->ih.lock, flags); + return IRQ_NONE; + } + +restart_ih: + /* display interrupts */ + evergreen_irq_ack(rdev, &disp_int, &disp_int_cont, &disp_int_cont2, + &disp_int_cont3, &disp_int_cont4, &disp_int_cont5); + + rdev->ih.wptr = wptr; + while (rptr != wptr) { + /* wptr/rptr are in bytes! */ + ring_index = rptr / 4; + src_id = rdev->ih.ring[ring_index] & 0xff; + src_data = rdev->ih.ring[ring_index + 1] & 0xfffffff; + + switch (src_id) { + case 1: /* D1 vblank/vline */ + switch (src_data) { + case 0: /* D1 vblank */ + if (disp_int & LB_D1_VBLANK_INTERRUPT) { + drm_handle_vblank(rdev->ddev, 0); + wake_up(&rdev->irq.vblank_queue); + disp_int &= ~LB_D1_VBLANK_INTERRUPT; + DRM_DEBUG("IH: D1 vblank\n"); + } + break; + case 1: /* D1 vline */ + if (disp_int & LB_D1_VLINE_INTERRUPT) { + disp_int &= ~LB_D1_VLINE_INTERRUPT; + DRM_DEBUG("IH: D1 vline\n"); + } + break; + default: + DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data); + break; + } + break; + case 2: /* D2 vblank/vline */ + switch (src_data) { + case 0: /* D2 vblank */ + if (disp_int_cont & LB_D2_VBLANK_INTERRUPT) { + drm_handle_vblank(rdev->ddev, 1); + wake_up(&rdev->irq.vblank_queue); + disp_int_cont &= ~LB_D2_VBLANK_INTERRUPT; + DRM_DEBUG("IH: D2 vblank\n"); + } + break; + case 1: /* D2 vline */ + if (disp_int_cont & LB_D2_VLINE_INTERRUPT) { + disp_int_cont &= ~LB_D2_VLINE_INTERRUPT; + DRM_DEBUG("IH: D2 vline\n"); + } + break; + default: + DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data); + break; + } + break; + case 3: /* D3 vblank/vline */ + switch (src_data) { + case 0: /* D3 vblank */ + if (disp_int_cont2 & LB_D3_VBLANK_INTERRUPT) { + drm_handle_vblank(rdev->ddev, 2); + wake_up(&rdev->irq.vblank_queue); + disp_int_cont2 &= ~LB_D3_VBLANK_INTERRUPT; + DRM_DEBUG("IH: D3 vblank\n"); + } + break; + case 1: /* D3 vline */ + if (disp_int_cont2 & LB_D3_VLINE_INTERRUPT) { + disp_int_cont2 &= ~LB_D3_VLINE_INTERRUPT; + DRM_DEBUG("IH: D3 vline\n"); + } + break; + default: + DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data); + break; + } + break; + case 4: /* D4 vblank/vline */ + switch (src_data) { + case 0: /* D4 vblank */ + if (disp_int_cont3 & LB_D4_VBLANK_INTERRUPT) { + drm_handle_vblank(rdev->ddev, 3); + wake_up(&rdev->irq.vblank_queue); + disp_int_cont3 &= ~LB_D4_VBLANK_INTERRUPT; + DRM_DEBUG("IH: D4 vblank\n"); + } + break; + case 1: /* D4 vline */ + if (disp_int_cont3 & LB_D4_VLINE_INTERRUPT) { + disp_int_cont3 &= ~LB_D4_VLINE_INTERRUPT; + DRM_DEBUG("IH: D4 vline\n"); + } + break; + default: + DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data); + break; + } + break; + case 5: /* D5 vblank/vline */ + switch (src_data) { + case 0: /* D5 vblank */ + if (disp_int_cont4 & LB_D5_VBLANK_INTERRUPT) { + drm_handle_vblank(rdev->ddev, 4); + wake_up(&rdev->irq.vblank_queue); + disp_int_cont4 &= ~LB_D5_VBLANK_INTERRUPT; + DRM_DEBUG("IH: D5 vblank\n"); + } + break; + case 1: /* D5 vline */ + if (disp_int_cont4 & LB_D5_VLINE_INTERRUPT) { + disp_int_cont4 &= ~LB_D5_VLINE_INTERRUPT; + DRM_DEBUG("IH: D5 vline\n"); + } + break; + default: + DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data); + break; + } + break; + case 6: /* D6 vblank/vline */ + switch (src_data) { + case 0: /* D6 vblank */ + if (disp_int_cont5 & LB_D6_VBLANK_INTERRUPT) { + drm_handle_vblank(rdev->ddev, 5); + wake_up(&rdev->irq.vblank_queue); + disp_int_cont5 &= ~LB_D6_VBLANK_INTERRUPT; + DRM_DEBUG("IH: D6 vblank\n"); + } + break; + case 1: /* D6 vline */ + if (disp_int_cont5 & LB_D6_VLINE_INTERRUPT) { + disp_int_cont5 &= ~LB_D6_VLINE_INTERRUPT; + DRM_DEBUG("IH: D6 vline\n"); + } + break; + default: + DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data); + break; + } + break; + case 42: /* HPD hotplug */ + switch (src_data) { + case 0: + if (disp_int & DC_HPD1_INTERRUPT) { + disp_int &= ~DC_HPD1_INTERRUPT; + queue_hotplug = true; + DRM_DEBUG("IH: HPD1\n"); + } + break; + case 1: + if (disp_int_cont & DC_HPD2_INTERRUPT) { + disp_int_cont &= ~DC_HPD2_INTERRUPT; + queue_hotplug = true; + DRM_DEBUG("IH: HPD2\n"); + } + break; + case 2: + if (disp_int_cont2 & DC_HPD3_INTERRUPT) { + disp_int_cont2 &= ~DC_HPD3_INTERRUPT; + queue_hotplug = true; + DRM_DEBUG("IH: HPD3\n"); + } + break; + case 3: + if (disp_int_cont3 & DC_HPD4_INTERRUPT) { + disp_int_cont3 &= ~DC_HPD4_INTERRUPT; + queue_hotplug = true; + DRM_DEBUG("IH: HPD4\n"); + } + break; + case 4: + if (disp_int_cont4 & DC_HPD5_INTERRUPT) { + disp_int_cont4 &= ~DC_HPD5_INTERRUPT; + queue_hotplug = true; + DRM_DEBUG("IH: HPD5\n"); + } + break; + case 5: + if (disp_int_cont5 & DC_HPD6_INTERRUPT) { + disp_int_cont5 &= ~DC_HPD6_INTERRUPT; + queue_hotplug = true; + DRM_DEBUG("IH: HPD6\n"); + } + break; + default: + DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data); + break; + } + break; + case 176: /* CP_INT in ring buffer */ + case 177: /* CP_INT in IB1 */ + case 178: /* CP_INT in IB2 */ + DRM_DEBUG("IH: CP int: 0x%08x\n", src_data); + radeon_fence_process(rdev); + break; + case 181: /* CP EOP event */ + DRM_DEBUG("IH: CP EOP\n"); + break; + case 233: /* GUI IDLE */ + DRM_DEBUG("IH: CP EOP\n"); + rdev->pm.gui_idle = true; + wake_up(&rdev->irq.idle_queue); + break; + default: + DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data); + break; + } + + /* wptr/rptr are in bytes! */ + rptr += 16; + rptr &= rdev->ih.ptr_mask; + } + /* make sure wptr hasn't changed while processing */ + wptr = evergreen_get_ih_wptr(rdev); + if (wptr != rdev->ih.wptr) + goto restart_ih; + if (queue_hotplug) + queue_work(rdev->wq, &rdev->hotplug_work); + rdev->ih.rptr = rptr; + WREG32(IH_RB_RPTR, rdev->ih.rptr); + spin_unlock_irqrestore(&rdev->ih.lock, flags); + return IRQ_HANDLED; +} + static int evergreen_startup(struct radeon_device *rdev) { -#if 0 int r; if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) { @@ -505,17 +1971,15 @@ static int evergreen_startup(struct radeon_device *rdev) return r; } } -#endif + evergreen_mc_program(rdev); -#if 0 if (rdev->flags & RADEON_IS_AGP) { - evergreem_agp_enable(rdev); + evergreen_agp_enable(rdev); } else { r = evergreen_pcie_gart_enable(rdev); if (r) return r; } -#endif evergreen_gpu_init(rdev); #if 0 if (!rdev->r600_blit.shader_obj) { @@ -536,6 +2000,7 @@ static int evergreen_startup(struct radeon_device *rdev) DRM_ERROR("failed to pin blit object %d\n", r); return r; } +#endif /* Enable IRQ */ r = r600_irq_init(rdev); @@ -544,7 +2009,7 @@ static int evergreen_startup(struct radeon_device *rdev) radeon_irq_kms_fini(rdev); return r; } - r600_irq_set(rdev); + evergreen_irq_set(rdev); r = radeon_ring_init(rdev, rdev->cp.ring_size); if (r) @@ -552,12 +2017,12 @@ static int evergreen_startup(struct radeon_device *rdev) r = evergreen_cp_load_microcode(rdev); if (r) return r; - r = r600_cp_resume(rdev); + r = evergreen_cp_resume(rdev); if (r) return r; /* write back buffer are not vital so don't worry about failure */ r600_wb_enable(rdev); -#endif + return 0; } @@ -582,13 +2047,13 @@ int evergreen_resume(struct radeon_device *rdev) DRM_ERROR("r600 startup failed on resume\n"); return r; } -#if 0 + r = r600_ib_test(rdev); if (r) { DRM_ERROR("radeon: failled testing IB (%d).\n", r); return r; } -#endif + return r; } @@ -597,12 +2062,14 @@ int evergreen_suspend(struct radeon_device *rdev) { #if 0 int r; - +#endif /* FIXME: we should wait for ring to be empty */ r700_cp_stop(rdev); rdev->cp.ready = false; + evergreen_irq_suspend(rdev); r600_wb_disable(rdev); evergreen_pcie_gart_disable(rdev); +#if 0 /* unpin shaders bo */ r = radeon_bo_reserve(rdev->r600_blit.shader_obj, false); if (likely(r == 0)) { @@ -682,8 +2149,6 @@ int evergreen_init(struct radeon_device *rdev) r = radeon_clocks_init(rdev); if (r) return r; - /* Initialize power management */ - radeon_pm_init(rdev); /* Fence driver */ r = radeon_fence_driver_init(rdev); if (r) @@ -702,7 +2167,7 @@ int evergreen_init(struct radeon_device *rdev) r = radeon_bo_init(rdev); if (r) return r; -#if 0 + r = radeon_irq_kms_init(rdev); if (r) return r; @@ -716,14 +2181,16 @@ int evergreen_init(struct radeon_device *rdev) r = r600_pcie_gart_init(rdev); if (r) return r; -#endif - rdev->accel_working = false; + + rdev->accel_working = true; r = evergreen_startup(rdev); if (r) { - evergreen_suspend(rdev); - /*r600_wb_fini(rdev);*/ - /*radeon_ring_fini(rdev);*/ - /*evergreen_pcie_gart_fini(rdev);*/ + dev_err(rdev->dev, "disabling GPU acceleration\n"); + r700_cp_fini(rdev); + r600_wb_fini(rdev); + r600_irq_fini(rdev); + radeon_irq_kms_fini(rdev); + evergreen_pcie_gart_fini(rdev); rdev->accel_working = false; } if (rdev->accel_working) { @@ -743,16 +2210,12 @@ int evergreen_init(struct radeon_device *rdev) void evergreen_fini(struct radeon_device *rdev) { - radeon_pm_fini(rdev); - evergreen_suspend(rdev); -#if 0 - r600_blit_fini(rdev); + /*r600_blit_fini(rdev);*/ + r700_cp_fini(rdev); + r600_wb_fini(rdev); r600_irq_fini(rdev); radeon_irq_kms_fini(rdev); - radeon_ring_fini(rdev); - r600_wb_fini(rdev); evergreen_pcie_gart_fini(rdev); -#endif radeon_gem_fini(rdev); radeon_fence_driver_fini(rdev); radeon_clocks_fini(rdev); diff --git a/drivers/gpu/drm/radeon/evergreen_cs.c b/drivers/gpu/drm/radeon/evergreen_cs.c new file mode 100644 index 000000000000..345a75a03c96 --- /dev/null +++ b/drivers/gpu/drm/radeon/evergreen_cs.c @@ -0,0 +1,1354 @@ +/* + * Copyright 2010 Advanced Micro Devices, Inc. + * Copyright 2008 Red Hat Inc. + * Copyright 2009 Jerome Glisse. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Dave Airlie + * Alex Deucher + * Jerome Glisse + */ +#include "drmP.h" +#include "radeon.h" +#include "evergreend.h" +#include "evergreen_reg_safe.h" + +static int evergreen_cs_packet_next_reloc(struct radeon_cs_parser *p, + struct radeon_cs_reloc **cs_reloc); + +struct evergreen_cs_track { + u32 group_size; + u32 nbanks; + u32 npipes; + /* value we track */ + u32 nsamples; + u32 cb_color_base_last[12]; + struct radeon_bo *cb_color_bo[12]; + u32 cb_color_bo_offset[12]; + struct radeon_bo *cb_color_fmask_bo[8]; + struct radeon_bo *cb_color_cmask_bo[8]; + u32 cb_color_info[12]; + u32 cb_color_view[12]; + u32 cb_color_pitch_idx[12]; + u32 cb_color_slice_idx[12]; + u32 cb_color_dim_idx[12]; + u32 cb_color_dim[12]; + u32 cb_color_pitch[12]; + u32 cb_color_slice[12]; + u32 cb_color_cmask_slice[8]; + u32 cb_color_fmask_slice[8]; + u32 cb_target_mask; + u32 cb_shader_mask; + u32 vgt_strmout_config; + u32 vgt_strmout_buffer_config; + u32 db_depth_control; + u32 db_depth_view; + u32 db_depth_size; + u32 db_depth_size_idx; + u32 db_z_info; + u32 db_z_idx; + u32 db_z_read_offset; + u32 db_z_write_offset; + struct radeon_bo *db_z_read_bo; + struct radeon_bo *db_z_write_bo; + u32 db_s_info; + u32 db_s_idx; + u32 db_s_read_offset; + u32 db_s_write_offset; + struct radeon_bo *db_s_read_bo; + struct radeon_bo *db_s_write_bo; +}; + +static void evergreen_cs_track_init(struct evergreen_cs_track *track) +{ + int i; + + for (i = 0; i < 8; i++) { + track->cb_color_fmask_bo[i] = NULL; + track->cb_color_cmask_bo[i] = NULL; + track->cb_color_cmask_slice[i] = 0; + track->cb_color_fmask_slice[i] = 0; + } + + for (i = 0; i < 12; i++) { + track->cb_color_base_last[i] = 0; + track->cb_color_bo[i] = NULL; + track->cb_color_bo_offset[i] = 0xFFFFFFFF; + track->cb_color_info[i] = 0; + track->cb_color_view[i] = 0; + track->cb_color_pitch_idx[i] = 0; + track->cb_color_slice_idx[i] = 0; + track->cb_color_dim[i] = 0; + track->cb_color_pitch[i] = 0; + track->cb_color_slice[i] = 0; + track->cb_color_dim[i] = 0; + } + track->cb_target_mask = 0xFFFFFFFF; + track->cb_shader_mask = 0xFFFFFFFF; + + track->db_depth_view = 0xFFFFC000; + track->db_depth_size = 0xFFFFFFFF; + track->db_depth_size_idx = 0; + track->db_depth_control = 0xFFFFFFFF; + track->db_z_info = 0xFFFFFFFF; + track->db_z_idx = 0xFFFFFFFF; + track->db_z_read_offset = 0xFFFFFFFF; + track->db_z_write_offset = 0xFFFFFFFF; + track->db_z_read_bo = NULL; + track->db_z_write_bo = NULL; + track->db_s_info = 0xFFFFFFFF; + track->db_s_idx = 0xFFFFFFFF; + track->db_s_read_offset = 0xFFFFFFFF; + track->db_s_write_offset = 0xFFFFFFFF; + track->db_s_read_bo = NULL; + track->db_s_write_bo = NULL; +} + +static inline int evergreen_cs_track_validate_cb(struct radeon_cs_parser *p, int i) +{ + /* XXX fill in */ + return 0; +} + +static int evergreen_cs_track_check(struct radeon_cs_parser *p) +{ + struct evergreen_cs_track *track = p->track; + + /* we don't support stream out buffer yet */ + if (track->vgt_strmout_config || track->vgt_strmout_buffer_config) { + dev_warn(p->dev, "this kernel doesn't support SMX output buffer\n"); + return -EINVAL; + } + + /* XXX fill in */ + return 0; +} + +/** + * evergreen_cs_packet_parse() - parse cp packet and point ib index to next packet + * @parser: parser structure holding parsing context. + * @pkt: where to store packet informations + * + * Assume that chunk_ib_index is properly set. Will return -EINVAL + * if packet is bigger than remaining ib size. or if packets is unknown. + **/ +int evergreen_cs_packet_parse(struct radeon_cs_parser *p, + struct radeon_cs_packet *pkt, + unsigned idx) +{ + struct radeon_cs_chunk *ib_chunk = &p->chunks[p->chunk_ib_idx]; + uint32_t header; + + if (idx >= ib_chunk->length_dw) { + DRM_ERROR("Can not parse packet at %d after CS end %d !\n", + idx, ib_chunk->length_dw); + return -EINVAL; + } + header = radeon_get_ib_value(p, idx); + pkt->idx = idx; + pkt->type = CP_PACKET_GET_TYPE(header); + pkt->count = CP_PACKET_GET_COUNT(header); + pkt->one_reg_wr = 0; + switch (pkt->type) { + case PACKET_TYPE0: + pkt->reg = CP_PACKET0_GET_REG(header); + break; + case PACKET_TYPE3: + pkt->opcode = CP_PACKET3_GET_OPCODE(header); + break; + case PACKET_TYPE2: + pkt->count = -1; + break; + default: + DRM_ERROR("Unknown packet type %d at %d !\n", pkt->type, idx); + return -EINVAL; + } + if ((pkt->count + 1 + pkt->idx) >= ib_chunk->length_dw) { + DRM_ERROR("Packet (%d:%d:%d) end after CS buffer (%d) !\n", + pkt->idx, pkt->type, pkt->count, ib_chunk->length_dw); + return -EINVAL; + } + return 0; +} + +/** + * evergreen_cs_packet_next_reloc() - parse next packet which should be reloc packet3 + * @parser: parser structure holding parsing context. + * @data: pointer to relocation data + * @offset_start: starting offset + * @offset_mask: offset mask (to align start offset on) + * @reloc: reloc informations + * + * Check next packet is relocation packet3, do bo validation and compute + * GPU offset using the provided start. + **/ +static int evergreen_cs_packet_next_reloc(struct radeon_cs_parser *p, + struct radeon_cs_reloc **cs_reloc) +{ + struct radeon_cs_chunk *relocs_chunk; + struct radeon_cs_packet p3reloc; + unsigned idx; + int r; + + if (p->chunk_relocs_idx == -1) { + DRM_ERROR("No relocation chunk !\n"); + return -EINVAL; + } + *cs_reloc = NULL; + relocs_chunk = &p->chunks[p->chunk_relocs_idx]; + r = evergreen_cs_packet_parse(p, &p3reloc, p->idx); + if (r) { + return r; + } + p->idx += p3reloc.count + 2; + if (p3reloc.type != PACKET_TYPE3 || p3reloc.opcode != PACKET3_NOP) { + DRM_ERROR("No packet3 for relocation for packet at %d.\n", + p3reloc.idx); + return -EINVAL; + } + idx = radeon_get_ib_value(p, p3reloc.idx + 1); + if (idx >= relocs_chunk->length_dw) { + DRM_ERROR("Relocs at %d after relocations chunk end %d !\n", + idx, relocs_chunk->length_dw); + return -EINVAL; + } + /* FIXME: we assume reloc size is 4 dwords */ + *cs_reloc = p->relocs_ptr[(idx / 4)]; + return 0; +} + +/** + * evergreen_cs_packet_next_is_pkt3_nop() - test if next packet is packet3 nop for reloc + * @parser: parser structure holding parsing context. + * + * Check next packet is relocation packet3, do bo validation and compute + * GPU offset using the provided start. + **/ +static inline int evergreen_cs_packet_next_is_pkt3_nop(struct radeon_cs_parser *p) +{ + struct radeon_cs_packet p3reloc; + int r; + + r = evergreen_cs_packet_parse(p, &p3reloc, p->idx); + if (r) { + return 0; + } + if (p3reloc.type != PACKET_TYPE3 || p3reloc.opcode != PACKET3_NOP) { + return 0; + } + return 1; +} + +/** + * evergreen_cs_packet_next_vline() - parse userspace VLINE packet + * @parser: parser structure holding parsing context. + * + * Userspace sends a special sequence for VLINE waits. + * PACKET0 - VLINE_START_END + value + * PACKET3 - WAIT_REG_MEM poll vline status reg + * RELOC (P3) - crtc_id in reloc. + * + * This function parses this and relocates the VLINE START END + * and WAIT_REG_MEM packets to the correct crtc. + * It also detects a switched off crtc and nulls out the + * wait in that case. + */ +static int evergreen_cs_packet_parse_vline(struct radeon_cs_parser *p) +{ + struct drm_mode_object *obj; + struct drm_crtc *crtc; + struct radeon_crtc *radeon_crtc; + struct radeon_cs_packet p3reloc, wait_reg_mem; + int crtc_id; + int r; + uint32_t header, h_idx, reg, wait_reg_mem_info; + volatile uint32_t *ib; + + ib = p->ib->ptr; + + /* parse the WAIT_REG_MEM */ + r = evergreen_cs_packet_parse(p, &wait_reg_mem, p->idx); + if (r) + return r; + + /* check its a WAIT_REG_MEM */ + if (wait_reg_mem.type != PACKET_TYPE3 || + wait_reg_mem.opcode != PACKET3_WAIT_REG_MEM) { + DRM_ERROR("vline wait missing WAIT_REG_MEM segment\n"); + r = -EINVAL; + return r; + } + + wait_reg_mem_info = radeon_get_ib_value(p, wait_reg_mem.idx + 1); + /* bit 4 is reg (0) or mem (1) */ + if (wait_reg_mem_info & 0x10) { + DRM_ERROR("vline WAIT_REG_MEM waiting on MEM rather than REG\n"); + r = -EINVAL; + return r; + } + /* waiting for value to be equal */ + if ((wait_reg_mem_info & 0x7) != 0x3) { + DRM_ERROR("vline WAIT_REG_MEM function not equal\n"); + r = -EINVAL; + return r; + } + if ((radeon_get_ib_value(p, wait_reg_mem.idx + 2) << 2) != EVERGREEN_VLINE_STATUS) { + DRM_ERROR("vline WAIT_REG_MEM bad reg\n"); + r = -EINVAL; + return r; + } + + if (radeon_get_ib_value(p, wait_reg_mem.idx + 5) != EVERGREEN_VLINE_STAT) { + DRM_ERROR("vline WAIT_REG_MEM bad bit mask\n"); + r = -EINVAL; + return r; + } + + /* jump over the NOP */ + r = evergreen_cs_packet_parse(p, &p3reloc, p->idx + wait_reg_mem.count + 2); + if (r) + return r; + + h_idx = p->idx - 2; + p->idx += wait_reg_mem.count + 2; + p->idx += p3reloc.count + 2; + + header = radeon_get_ib_value(p, h_idx); + crtc_id = radeon_get_ib_value(p, h_idx + 2 + 7 + 1); + reg = CP_PACKET0_GET_REG(header); + obj = drm_mode_object_find(p->rdev->ddev, crtc_id, DRM_MODE_OBJECT_CRTC); + if (!obj) { + DRM_ERROR("cannot find crtc %d\n", crtc_id); + r = -EINVAL; + goto out; + } + crtc = obj_to_crtc(obj); + radeon_crtc = to_radeon_crtc(crtc); + crtc_id = radeon_crtc->crtc_id; + + if (!crtc->enabled) { + /* if the CRTC isn't enabled - we need to nop out the WAIT_REG_MEM */ + ib[h_idx + 2] = PACKET2(0); + ib[h_idx + 3] = PACKET2(0); + ib[h_idx + 4] = PACKET2(0); + ib[h_idx + 5] = PACKET2(0); + ib[h_idx + 6] = PACKET2(0); + ib[h_idx + 7] = PACKET2(0); + ib[h_idx + 8] = PACKET2(0); + } else { + switch (reg) { + case EVERGREEN_VLINE_START_END: + header &= ~R600_CP_PACKET0_REG_MASK; + header |= (EVERGREEN_VLINE_START_END + radeon_crtc->crtc_offset) >> 2; + ib[h_idx] = header; + ib[h_idx + 4] = (EVERGREEN_VLINE_STATUS + radeon_crtc->crtc_offset) >> 2; + break; + default: + DRM_ERROR("unknown crtc reloc\n"); + r = -EINVAL; + goto out; + } + } +out: + return r; +} + +static int evergreen_packet0_check(struct radeon_cs_parser *p, + struct radeon_cs_packet *pkt, + unsigned idx, unsigned reg) +{ + int r; + + switch (reg) { + case EVERGREEN_VLINE_START_END: + r = evergreen_cs_packet_parse_vline(p); + if (r) { + DRM_ERROR("No reloc for ib[%d]=0x%04X\n", + idx, reg); + return r; + } + break; + default: + printk(KERN_ERR "Forbidden register 0x%04X in cs at %d\n", + reg, idx); + return -EINVAL; + } + return 0; +} + +static int evergreen_cs_parse_packet0(struct radeon_cs_parser *p, + struct radeon_cs_packet *pkt) +{ + unsigned reg, i; + unsigned idx; + int r; + + idx = pkt->idx + 1; + reg = pkt->reg; + for (i = 0; i <= pkt->count; i++, idx++, reg += 4) { + r = evergreen_packet0_check(p, pkt, idx, reg); + if (r) { + return r; + } + } + return 0; +} + +/** + * evergreen_cs_check_reg() - check if register is authorized or not + * @parser: parser structure holding parsing context + * @reg: register we are testing + * @idx: index into the cs buffer + * + * This function will test against evergreen_reg_safe_bm and return 0 + * if register is safe. If register is not flag as safe this function + * will test it against a list of register needind special handling. + */ +static inline int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx) +{ + struct evergreen_cs_track *track = (struct evergreen_cs_track *)p->track; + struct radeon_cs_reloc *reloc; + u32 last_reg = ARRAY_SIZE(evergreen_reg_safe_bm); + u32 m, i, tmp, *ib; + int r; + + i = (reg >> 7); + if (i > last_reg) { + dev_warn(p->dev, "forbidden register 0x%08x at %d\n", reg, idx); + return -EINVAL; + } + m = 1 << ((reg >> 2) & 31); + if (!(evergreen_reg_safe_bm[i] & m)) + return 0; + ib = p->ib->ptr; + switch (reg) { + /* force following reg to 0 in an attemp to disable out buffer + * which will need us to better understand how it works to perform + * security check on it (Jerome) + */ + case SQ_ESGS_RING_SIZE: + case SQ_GSVS_RING_SIZE: + case SQ_ESTMP_RING_SIZE: + case SQ_GSTMP_RING_SIZE: + case SQ_HSTMP_RING_SIZE: + case SQ_LSTMP_RING_SIZE: + case SQ_PSTMP_RING_SIZE: + case SQ_VSTMP_RING_SIZE: + case SQ_ESGS_RING_ITEMSIZE: + case SQ_ESTMP_RING_ITEMSIZE: + case SQ_GSTMP_RING_ITEMSIZE: + case SQ_GSVS_RING_ITEMSIZE: + case SQ_GS_VERT_ITEMSIZE: + case SQ_GS_VERT_ITEMSIZE_1: + case SQ_GS_VERT_ITEMSIZE_2: + case SQ_GS_VERT_ITEMSIZE_3: + case SQ_GSVS_RING_OFFSET_1: + case SQ_GSVS_RING_OFFSET_2: + case SQ_GSVS_RING_OFFSET_3: + case SQ_HSTMP_RING_ITEMSIZE: + case SQ_LSTMP_RING_ITEMSIZE: + case SQ_PSTMP_RING_ITEMSIZE: + case SQ_VSTMP_RING_ITEMSIZE: + case VGT_TF_RING_SIZE: + /* get value to populate the IB don't remove */ + tmp =radeon_get_ib_value(p, idx); + ib[idx] = 0; + break; + case DB_DEPTH_CONTROL: + track->db_depth_control = radeon_get_ib_value(p, idx); + break; + case DB_Z_INFO: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + track->db_z_info = radeon_get_ib_value(p, idx); + ib[idx] &= ~Z_ARRAY_MODE(0xf); + track->db_z_info &= ~Z_ARRAY_MODE(0xf); + if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) { + ib[idx] |= Z_ARRAY_MODE(ARRAY_2D_TILED_THIN1); + track->db_z_info |= Z_ARRAY_MODE(ARRAY_2D_TILED_THIN1); + } else { + ib[idx] |= Z_ARRAY_MODE(ARRAY_1D_TILED_THIN1); + track->db_z_info |= Z_ARRAY_MODE(ARRAY_1D_TILED_THIN1); + } + break; + case DB_STENCIL_INFO: + track->db_s_info = radeon_get_ib_value(p, idx); + break; + case DB_DEPTH_VIEW: + track->db_depth_view = radeon_get_ib_value(p, idx); + break; + case DB_DEPTH_SIZE: + track->db_depth_size = radeon_get_ib_value(p, idx); + track->db_depth_size_idx = idx; + break; + case DB_Z_READ_BASE: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + track->db_z_read_offset = radeon_get_ib_value(p, idx); + ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + track->db_z_read_bo = reloc->robj; + break; + case DB_Z_WRITE_BASE: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + track->db_z_write_offset = radeon_get_ib_value(p, idx); + ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + track->db_z_write_bo = reloc->robj; + break; + case DB_STENCIL_READ_BASE: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + track->db_s_read_offset = radeon_get_ib_value(p, idx); + ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + track->db_s_read_bo = reloc->robj; + break; + case DB_STENCIL_WRITE_BASE: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + track->db_s_write_offset = radeon_get_ib_value(p, idx); + ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + track->db_s_write_bo = reloc->robj; + break; + case VGT_STRMOUT_CONFIG: + track->vgt_strmout_config = radeon_get_ib_value(p, idx); + break; + case VGT_STRMOUT_BUFFER_CONFIG: + track->vgt_strmout_buffer_config = radeon_get_ib_value(p, idx); + break; + case CB_TARGET_MASK: + track->cb_target_mask = radeon_get_ib_value(p, idx); + break; + case CB_SHADER_MASK: + track->cb_shader_mask = radeon_get_ib_value(p, idx); + break; + case PA_SC_AA_CONFIG: + tmp = radeon_get_ib_value(p, idx) & MSAA_NUM_SAMPLES_MASK; + track->nsamples = 1 << tmp; + break; + case CB_COLOR0_VIEW: + case CB_COLOR1_VIEW: + case CB_COLOR2_VIEW: + case CB_COLOR3_VIEW: + case CB_COLOR4_VIEW: + case CB_COLOR5_VIEW: + case CB_COLOR6_VIEW: + case CB_COLOR7_VIEW: + tmp = (reg - CB_COLOR0_VIEW) / 0x3c; + track->cb_color_view[tmp] = radeon_get_ib_value(p, idx); + break; + case CB_COLOR8_VIEW: + case CB_COLOR9_VIEW: + case CB_COLOR10_VIEW: + case CB_COLOR11_VIEW: + tmp = ((reg - CB_COLOR8_VIEW) / 0x1c) + 8; + track->cb_color_view[tmp] = radeon_get_ib_value(p, idx); + break; + case CB_COLOR0_INFO: + case CB_COLOR1_INFO: + case CB_COLOR2_INFO: + case CB_COLOR3_INFO: + case CB_COLOR4_INFO: + case CB_COLOR5_INFO: + case CB_COLOR6_INFO: + case CB_COLOR7_INFO: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + tmp = (reg - CB_COLOR0_INFO) / 0x3c; + track->cb_color_info[tmp] = radeon_get_ib_value(p, idx); + if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) { + ib[idx] |= CB_ARRAY_MODE(ARRAY_2D_TILED_THIN1); + track->cb_color_info[tmp] |= CB_ARRAY_MODE(ARRAY_2D_TILED_THIN1); + } else if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) { + ib[idx] |= CB_ARRAY_MODE(ARRAY_1D_TILED_THIN1); + track->cb_color_info[tmp] |= CB_ARRAY_MODE(ARRAY_1D_TILED_THIN1); + } + break; + case CB_COLOR8_INFO: + case CB_COLOR9_INFO: + case CB_COLOR10_INFO: + case CB_COLOR11_INFO: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + tmp = ((reg - CB_COLOR8_INFO) / 0x1c) + 8; + track->cb_color_info[tmp] = radeon_get_ib_value(p, idx); + if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) { + ib[idx] |= CB_ARRAY_MODE(ARRAY_2D_TILED_THIN1); + track->cb_color_info[tmp] |= CB_ARRAY_MODE(ARRAY_2D_TILED_THIN1); + } else if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) { + ib[idx] |= CB_ARRAY_MODE(ARRAY_1D_TILED_THIN1); + track->cb_color_info[tmp] |= CB_ARRAY_MODE(ARRAY_1D_TILED_THIN1); + } + break; + case CB_COLOR0_PITCH: + case CB_COLOR1_PITCH: + case CB_COLOR2_PITCH: + case CB_COLOR3_PITCH: + case CB_COLOR4_PITCH: + case CB_COLOR5_PITCH: + case CB_COLOR6_PITCH: + case CB_COLOR7_PITCH: + tmp = (reg - CB_COLOR0_PITCH) / 0x3c; + track->cb_color_pitch[tmp] = radeon_get_ib_value(p, idx); + track->cb_color_pitch_idx[tmp] = idx; + break; + case CB_COLOR8_PITCH: + case CB_COLOR9_PITCH: + case CB_COLOR10_PITCH: + case CB_COLOR11_PITCH: + tmp = ((reg - CB_COLOR8_PITCH) / 0x1c) + 8; + track->cb_color_pitch[tmp] = radeon_get_ib_value(p, idx); + track->cb_color_pitch_idx[tmp] = idx; + break; + case CB_COLOR0_SLICE: + case CB_COLOR1_SLICE: + case CB_COLOR2_SLICE: + case CB_COLOR3_SLICE: + case CB_COLOR4_SLICE: + case CB_COLOR5_SLICE: + case CB_COLOR6_SLICE: + case CB_COLOR7_SLICE: + tmp = (reg - CB_COLOR0_SLICE) / 0x3c; + track->cb_color_slice[tmp] = radeon_get_ib_value(p, idx); + track->cb_color_slice_idx[tmp] = idx; + break; + case CB_COLOR8_SLICE: + case CB_COLOR9_SLICE: + case CB_COLOR10_SLICE: + case CB_COLOR11_SLICE: + tmp = ((reg - CB_COLOR8_SLICE) / 0x1c) + 8; + track->cb_color_slice[tmp] = radeon_get_ib_value(p, idx); + track->cb_color_slice_idx[tmp] = idx; + break; + case CB_COLOR0_ATTRIB: + case CB_COLOR1_ATTRIB: + case CB_COLOR2_ATTRIB: + case CB_COLOR3_ATTRIB: + case CB_COLOR4_ATTRIB: + case CB_COLOR5_ATTRIB: + case CB_COLOR6_ATTRIB: + case CB_COLOR7_ATTRIB: + case CB_COLOR8_ATTRIB: + case CB_COLOR9_ATTRIB: + case CB_COLOR10_ATTRIB: + case CB_COLOR11_ATTRIB: + break; + case CB_COLOR0_DIM: + case CB_COLOR1_DIM: + case CB_COLOR2_DIM: + case CB_COLOR3_DIM: + case CB_COLOR4_DIM: + case CB_COLOR5_DIM: + case CB_COLOR6_DIM: + case CB_COLOR7_DIM: + tmp = (reg - CB_COLOR0_DIM) / 0x3c; + track->cb_color_dim[tmp] = radeon_get_ib_value(p, idx); + track->cb_color_dim_idx[tmp] = idx; + break; + case CB_COLOR8_DIM: + case CB_COLOR9_DIM: + case CB_COLOR10_DIM: + case CB_COLOR11_DIM: + tmp = ((reg - CB_COLOR8_DIM) / 0x1c) + 8; + track->cb_color_dim[tmp] = radeon_get_ib_value(p, idx); + track->cb_color_dim_idx[tmp] = idx; + break; + case CB_COLOR0_FMASK: + case CB_COLOR1_FMASK: + case CB_COLOR2_FMASK: + case CB_COLOR3_FMASK: + case CB_COLOR4_FMASK: + case CB_COLOR5_FMASK: + case CB_COLOR6_FMASK: + case CB_COLOR7_FMASK: + tmp = (reg - CB_COLOR0_FMASK) / 0x3c; + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_err(p->dev, "bad SET_CONTEXT_REG 0x%04X\n", reg); + return -EINVAL; + } + ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + track->cb_color_fmask_bo[tmp] = reloc->robj; + break; + case CB_COLOR0_CMASK: + case CB_COLOR1_CMASK: + case CB_COLOR2_CMASK: + case CB_COLOR3_CMASK: + case CB_COLOR4_CMASK: + case CB_COLOR5_CMASK: + case CB_COLOR6_CMASK: + case CB_COLOR7_CMASK: + tmp = (reg - CB_COLOR0_CMASK) / 0x3c; + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_err(p->dev, "bad SET_CONTEXT_REG 0x%04X\n", reg); + return -EINVAL; + } + ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + track->cb_color_cmask_bo[tmp] = reloc->robj; + break; + case CB_COLOR0_FMASK_SLICE: + case CB_COLOR1_FMASK_SLICE: + case CB_COLOR2_FMASK_SLICE: + case CB_COLOR3_FMASK_SLICE: + case CB_COLOR4_FMASK_SLICE: + case CB_COLOR5_FMASK_SLICE: + case CB_COLOR6_FMASK_SLICE: + case CB_COLOR7_FMASK_SLICE: + tmp = (reg - CB_COLOR0_FMASK_SLICE) / 0x3c; + track->cb_color_fmask_slice[tmp] = radeon_get_ib_value(p, idx); + break; + case CB_COLOR0_CMASK_SLICE: + case CB_COLOR1_CMASK_SLICE: + case CB_COLOR2_CMASK_SLICE: + case CB_COLOR3_CMASK_SLICE: + case CB_COLOR4_CMASK_SLICE: + case CB_COLOR5_CMASK_SLICE: + case CB_COLOR6_CMASK_SLICE: + case CB_COLOR7_CMASK_SLICE: + tmp = (reg - CB_COLOR0_CMASK_SLICE) / 0x3c; + track->cb_color_cmask_slice[tmp] = radeon_get_ib_value(p, idx); + break; + case CB_COLOR0_BASE: + case CB_COLOR1_BASE: + case CB_COLOR2_BASE: + case CB_COLOR3_BASE: + case CB_COLOR4_BASE: + case CB_COLOR5_BASE: + case CB_COLOR6_BASE: + case CB_COLOR7_BASE: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + tmp = (reg - CB_COLOR0_BASE) / 0x3c; + track->cb_color_bo_offset[tmp] = radeon_get_ib_value(p, idx); + ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + track->cb_color_base_last[tmp] = ib[idx]; + track->cb_color_bo[tmp] = reloc->robj; + break; + case CB_COLOR8_BASE: + case CB_COLOR9_BASE: + case CB_COLOR10_BASE: + case CB_COLOR11_BASE: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + tmp = ((reg - CB_COLOR8_BASE) / 0x1c) + 8; + track->cb_color_bo_offset[tmp] = radeon_get_ib_value(p, idx); + ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + track->cb_color_base_last[tmp] = ib[idx]; + track->cb_color_bo[tmp] = reloc->robj; + break; + case CB_IMMED0_BASE: + case CB_IMMED1_BASE: + case CB_IMMED2_BASE: + case CB_IMMED3_BASE: + case CB_IMMED4_BASE: + case CB_IMMED5_BASE: + case CB_IMMED6_BASE: + case CB_IMMED7_BASE: + case CB_IMMED8_BASE: + case CB_IMMED9_BASE: + case CB_IMMED10_BASE: + case CB_IMMED11_BASE: + case DB_HTILE_DATA_BASE: + case SQ_PGM_START_FS: + case SQ_PGM_START_ES: + case SQ_PGM_START_VS: + case SQ_PGM_START_GS: + case SQ_PGM_START_PS: + case SQ_PGM_START_HS: + case SQ_PGM_START_LS: + case GDS_ADDR_BASE: + case SQ_CONST_MEM_BASE: + case SQ_ALU_CONST_CACHE_GS_0: + case SQ_ALU_CONST_CACHE_GS_1: + case SQ_ALU_CONST_CACHE_GS_2: + case SQ_ALU_CONST_CACHE_GS_3: + case SQ_ALU_CONST_CACHE_GS_4: + case SQ_ALU_CONST_CACHE_GS_5: + case SQ_ALU_CONST_CACHE_GS_6: + case SQ_ALU_CONST_CACHE_GS_7: + case SQ_ALU_CONST_CACHE_GS_8: + case SQ_ALU_CONST_CACHE_GS_9: + case SQ_ALU_CONST_CACHE_GS_10: + case SQ_ALU_CONST_CACHE_GS_11: + case SQ_ALU_CONST_CACHE_GS_12: + case SQ_ALU_CONST_CACHE_GS_13: + case SQ_ALU_CONST_CACHE_GS_14: + case SQ_ALU_CONST_CACHE_GS_15: + case SQ_ALU_CONST_CACHE_PS_0: + case SQ_ALU_CONST_CACHE_PS_1: + case SQ_ALU_CONST_CACHE_PS_2: + case SQ_ALU_CONST_CACHE_PS_3: + case SQ_ALU_CONST_CACHE_PS_4: + case SQ_ALU_CONST_CACHE_PS_5: + case SQ_ALU_CONST_CACHE_PS_6: + case SQ_ALU_CONST_CACHE_PS_7: + case SQ_ALU_CONST_CACHE_PS_8: + case SQ_ALU_CONST_CACHE_PS_9: + case SQ_ALU_CONST_CACHE_PS_10: + case SQ_ALU_CONST_CACHE_PS_11: + case SQ_ALU_CONST_CACHE_PS_12: + case SQ_ALU_CONST_CACHE_PS_13: + case SQ_ALU_CONST_CACHE_PS_14: + case SQ_ALU_CONST_CACHE_PS_15: + case SQ_ALU_CONST_CACHE_VS_0: + case SQ_ALU_CONST_CACHE_VS_1: + case SQ_ALU_CONST_CACHE_VS_2: + case SQ_ALU_CONST_CACHE_VS_3: + case SQ_ALU_CONST_CACHE_VS_4: + case SQ_ALU_CONST_CACHE_VS_5: + case SQ_ALU_CONST_CACHE_VS_6: + case SQ_ALU_CONST_CACHE_VS_7: + case SQ_ALU_CONST_CACHE_VS_8: + case SQ_ALU_CONST_CACHE_VS_9: + case SQ_ALU_CONST_CACHE_VS_10: + case SQ_ALU_CONST_CACHE_VS_11: + case SQ_ALU_CONST_CACHE_VS_12: + case SQ_ALU_CONST_CACHE_VS_13: + case SQ_ALU_CONST_CACHE_VS_14: + case SQ_ALU_CONST_CACHE_VS_15: + case SQ_ALU_CONST_CACHE_HS_0: + case SQ_ALU_CONST_CACHE_HS_1: + case SQ_ALU_CONST_CACHE_HS_2: + case SQ_ALU_CONST_CACHE_HS_3: + case SQ_ALU_CONST_CACHE_HS_4: + case SQ_ALU_CONST_CACHE_HS_5: + case SQ_ALU_CONST_CACHE_HS_6: + case SQ_ALU_CONST_CACHE_HS_7: + case SQ_ALU_CONST_CACHE_HS_8: + case SQ_ALU_CONST_CACHE_HS_9: + case SQ_ALU_CONST_CACHE_HS_10: + case SQ_ALU_CONST_CACHE_HS_11: + case SQ_ALU_CONST_CACHE_HS_12: + case SQ_ALU_CONST_CACHE_HS_13: + case SQ_ALU_CONST_CACHE_HS_14: + case SQ_ALU_CONST_CACHE_HS_15: + case SQ_ALU_CONST_CACHE_LS_0: + case SQ_ALU_CONST_CACHE_LS_1: + case SQ_ALU_CONST_CACHE_LS_2: + case SQ_ALU_CONST_CACHE_LS_3: + case SQ_ALU_CONST_CACHE_LS_4: + case SQ_ALU_CONST_CACHE_LS_5: + case SQ_ALU_CONST_CACHE_LS_6: + case SQ_ALU_CONST_CACHE_LS_7: + case SQ_ALU_CONST_CACHE_LS_8: + case SQ_ALU_CONST_CACHE_LS_9: + case SQ_ALU_CONST_CACHE_LS_10: + case SQ_ALU_CONST_CACHE_LS_11: + case SQ_ALU_CONST_CACHE_LS_12: + case SQ_ALU_CONST_CACHE_LS_13: + case SQ_ALU_CONST_CACHE_LS_14: + case SQ_ALU_CONST_CACHE_LS_15: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + break; + default: + dev_warn(p->dev, "forbidden register 0x%08x at %d\n", reg, idx); + return -EINVAL; + } + return 0; +} + +/** + * evergreen_check_texture_resource() - check if register is authorized or not + * @p: parser structure holding parsing context + * @idx: index into the cs buffer + * @texture: texture's bo structure + * @mipmap: mipmap's bo structure + * + * This function will check that the resource has valid field and that + * the texture and mipmap bo object are big enough to cover this resource. + */ +static inline int evergreen_check_texture_resource(struct radeon_cs_parser *p, u32 idx, + struct radeon_bo *texture, + struct radeon_bo *mipmap) +{ + /* XXX fill in */ + return 0; +} + +static int evergreen_packet3_check(struct radeon_cs_parser *p, + struct radeon_cs_packet *pkt) +{ + struct radeon_cs_reloc *reloc; + struct evergreen_cs_track *track; + volatile u32 *ib; + unsigned idx; + unsigned i; + unsigned start_reg, end_reg, reg; + int r; + u32 idx_value; + + track = (struct evergreen_cs_track *)p->track; + ib = p->ib->ptr; + idx = pkt->idx + 1; + idx_value = radeon_get_ib_value(p, idx); + + switch (pkt->opcode) { + case PACKET3_CONTEXT_CONTROL: + if (pkt->count != 1) { + DRM_ERROR("bad CONTEXT_CONTROL\n"); + return -EINVAL; + } + break; + case PACKET3_INDEX_TYPE: + case PACKET3_NUM_INSTANCES: + case PACKET3_CLEAR_STATE: + if (pkt->count) { + DRM_ERROR("bad INDEX_TYPE/NUM_INSTANCES/CLEAR_STATE\n"); + return -EINVAL; + } + break; + case PACKET3_INDEX_BASE: + if (pkt->count != 1) { + DRM_ERROR("bad INDEX_BASE\n"); + return -EINVAL; + } + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad INDEX_BASE\n"); + return -EINVAL; + } + ib[idx+0] = idx_value + (u32)(reloc->lobj.gpu_offset & 0xffffffff); + ib[idx+1] += upper_32_bits(reloc->lobj.gpu_offset) & 0xff; + r = evergreen_cs_track_check(p); + if (r) { + dev_warn(p->dev, "%s:%d invalid cmd stream\n", __func__, __LINE__); + return r; + } + break; + case PACKET3_DRAW_INDEX: + if (pkt->count != 3) { + DRM_ERROR("bad DRAW_INDEX\n"); + return -EINVAL; + } + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad DRAW_INDEX\n"); + return -EINVAL; + } + ib[idx+0] = idx_value + (u32)(reloc->lobj.gpu_offset & 0xffffffff); + ib[idx+1] += upper_32_bits(reloc->lobj.gpu_offset) & 0xff; + r = evergreen_cs_track_check(p); + if (r) { + dev_warn(p->dev, "%s:%d invalid cmd stream\n", __func__, __LINE__); + return r; + } + break; + case PACKET3_DRAW_INDEX_2: + if (pkt->count != 4) { + DRM_ERROR("bad DRAW_INDEX_2\n"); + return -EINVAL; + } + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad DRAW_INDEX_2\n"); + return -EINVAL; + } + ib[idx+1] = idx_value + (u32)(reloc->lobj.gpu_offset & 0xffffffff); + ib[idx+2] += upper_32_bits(reloc->lobj.gpu_offset) & 0xff; + r = evergreen_cs_track_check(p); + if (r) { + dev_warn(p->dev, "%s:%d invalid cmd stream\n", __func__, __LINE__); + return r; + } + break; + case PACKET3_DRAW_INDEX_AUTO: + if (pkt->count != 1) { + DRM_ERROR("bad DRAW_INDEX_AUTO\n"); + return -EINVAL; + } + r = evergreen_cs_track_check(p); + if (r) { + dev_warn(p->dev, "%s:%d invalid cmd stream %d\n", __func__, __LINE__, idx); + return r; + } + break; + case PACKET3_DRAW_INDEX_MULTI_AUTO: + if (pkt->count != 2) { + DRM_ERROR("bad DRAW_INDEX_MULTI_AUTO\n"); + return -EINVAL; + } + r = evergreen_cs_track_check(p); + if (r) { + dev_warn(p->dev, "%s:%d invalid cmd stream %d\n", __func__, __LINE__, idx); + return r; + } + break; + case PACKET3_DRAW_INDEX_IMMD: + if (pkt->count < 2) { + DRM_ERROR("bad DRAW_INDEX_IMMD\n"); + return -EINVAL; + } + r = evergreen_cs_track_check(p); + if (r) { + dev_warn(p->dev, "%s:%d invalid cmd stream\n", __func__, __LINE__); + return r; + } + break; + case PACKET3_DRAW_INDEX_OFFSET: + if (pkt->count != 2) { + DRM_ERROR("bad DRAW_INDEX_OFFSET\n"); + return -EINVAL; + } + r = evergreen_cs_track_check(p); + if (r) { + dev_warn(p->dev, "%s:%d invalid cmd stream\n", __func__, __LINE__); + return r; + } + break; + case PACKET3_DRAW_INDEX_OFFSET_2: + if (pkt->count != 3) { + DRM_ERROR("bad DRAW_INDEX_OFFSET_2\n"); + return -EINVAL; + } + r = evergreen_cs_track_check(p); + if (r) { + dev_warn(p->dev, "%s:%d invalid cmd stream\n", __func__, __LINE__); + return r; + } + break; + case PACKET3_WAIT_REG_MEM: + if (pkt->count != 5) { + DRM_ERROR("bad WAIT_REG_MEM\n"); + return -EINVAL; + } + /* bit 4 is reg (0) or mem (1) */ + if (idx_value & 0x10) { + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad WAIT_REG_MEM\n"); + return -EINVAL; + } + ib[idx+1] += (u32)(reloc->lobj.gpu_offset & 0xffffffff); + ib[idx+2] += upper_32_bits(reloc->lobj.gpu_offset) & 0xff; + } + break; + case PACKET3_SURFACE_SYNC: + if (pkt->count != 3) { + DRM_ERROR("bad SURFACE_SYNC\n"); + return -EINVAL; + } + /* 0xffffffff/0x0 is flush all cache flag */ + if (radeon_get_ib_value(p, idx + 1) != 0xffffffff || + radeon_get_ib_value(p, idx + 2) != 0) { + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad SURFACE_SYNC\n"); + return -EINVAL; + } + ib[idx+2] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + } + break; + case PACKET3_EVENT_WRITE: + if (pkt->count != 2 && pkt->count != 0) { + DRM_ERROR("bad EVENT_WRITE\n"); + return -EINVAL; + } + if (pkt->count) { + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad EVENT_WRITE\n"); + return -EINVAL; + } + ib[idx+1] += (u32)(reloc->lobj.gpu_offset & 0xffffffff); + ib[idx+2] += upper_32_bits(reloc->lobj.gpu_offset) & 0xff; + } + break; + case PACKET3_EVENT_WRITE_EOP: + if (pkt->count != 4) { + DRM_ERROR("bad EVENT_WRITE_EOP\n"); + return -EINVAL; + } + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad EVENT_WRITE_EOP\n"); + return -EINVAL; + } + ib[idx+1] += (u32)(reloc->lobj.gpu_offset & 0xffffffff); + ib[idx+2] += upper_32_bits(reloc->lobj.gpu_offset) & 0xff; + break; + case PACKET3_EVENT_WRITE_EOS: + if (pkt->count != 3) { + DRM_ERROR("bad EVENT_WRITE_EOS\n"); + return -EINVAL; + } + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad EVENT_WRITE_EOS\n"); + return -EINVAL; + } + ib[idx+1] += (u32)(reloc->lobj.gpu_offset & 0xffffffff); + ib[idx+2] += upper_32_bits(reloc->lobj.gpu_offset) & 0xff; + break; + case PACKET3_SET_CONFIG_REG: + start_reg = (idx_value << 2) + PACKET3_SET_CONFIG_REG_START; + end_reg = 4 * pkt->count + start_reg - 4; + if ((start_reg < PACKET3_SET_CONFIG_REG_START) || + (start_reg >= PACKET3_SET_CONFIG_REG_END) || + (end_reg >= PACKET3_SET_CONFIG_REG_END)) { + DRM_ERROR("bad PACKET3_SET_CONFIG_REG\n"); + return -EINVAL; + } + for (i = 0; i < pkt->count; i++) { + reg = start_reg + (4 * i); + r = evergreen_cs_check_reg(p, reg, idx+1+i); + if (r) + return r; + } + break; + case PACKET3_SET_CONTEXT_REG: + start_reg = (idx_value << 2) + PACKET3_SET_CONTEXT_REG_START; + end_reg = 4 * pkt->count + start_reg - 4; + if ((start_reg < PACKET3_SET_CONTEXT_REG_START) || + (start_reg >= PACKET3_SET_CONTEXT_REG_END) || + (end_reg >= PACKET3_SET_CONTEXT_REG_END)) { + DRM_ERROR("bad PACKET3_SET_CONTEXT_REG\n"); + return -EINVAL; + } + for (i = 0; i < pkt->count; i++) { + reg = start_reg + (4 * i); + r = evergreen_cs_check_reg(p, reg, idx+1+i); + if (r) + return r; + } + break; + case PACKET3_SET_RESOURCE: + if (pkt->count % 8) { + DRM_ERROR("bad SET_RESOURCE\n"); + return -EINVAL; + } + start_reg = (idx_value << 2) + PACKET3_SET_RESOURCE_START; + end_reg = 4 * pkt->count + start_reg - 4; + if ((start_reg < PACKET3_SET_RESOURCE_START) || + (start_reg >= PACKET3_SET_RESOURCE_END) || + (end_reg >= PACKET3_SET_RESOURCE_END)) { + DRM_ERROR("bad SET_RESOURCE\n"); + return -EINVAL; + } + for (i = 0; i < (pkt->count / 8); i++) { + struct radeon_bo *texture, *mipmap; + u32 size, offset; + + switch (G__SQ_CONSTANT_TYPE(radeon_get_ib_value(p, idx+1+(i*8)+7))) { + case SQ_TEX_VTX_VALID_TEXTURE: + /* tex base */ + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad SET_RESOURCE (tex)\n"); + return -EINVAL; + } + ib[idx+1+(i*8)+2] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) + ib[idx+1+(i*8)+1] |= TEX_ARRAY_MODE(ARRAY_2D_TILED_THIN1); + else if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) + ib[idx+1+(i*8)+1] |= TEX_ARRAY_MODE(ARRAY_1D_TILED_THIN1); + texture = reloc->robj; + /* tex mip base */ + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad SET_RESOURCE (tex)\n"); + return -EINVAL; + } + ib[idx+1+(i*8)+3] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); + mipmap = reloc->robj; + r = evergreen_check_texture_resource(p, idx+1+(i*8), + texture, mipmap); + if (r) + return r; + break; + case SQ_TEX_VTX_VALID_BUFFER: + /* vtx base */ + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad SET_RESOURCE (vtx)\n"); + return -EINVAL; + } + offset = radeon_get_ib_value(p, idx+1+(i*8)+0); + size = radeon_get_ib_value(p, idx+1+(i*8)+1); + if (p->rdev && (size + offset) > radeon_bo_size(reloc->robj)) { + /* force size to size of the buffer */ + dev_warn(p->dev, "vbo resource seems too big for the bo\n"); + ib[idx+1+(i*8)+1] = radeon_bo_size(reloc->robj); + } + ib[idx+1+(i*8)+0] += (u32)((reloc->lobj.gpu_offset) & 0xffffffff); + ib[idx+1+(i*8)+2] += upper_32_bits(reloc->lobj.gpu_offset) & 0xff; + break; + case SQ_TEX_VTX_INVALID_TEXTURE: + case SQ_TEX_VTX_INVALID_BUFFER: + default: + DRM_ERROR("bad SET_RESOURCE\n"); + return -EINVAL; + } + } + break; + case PACKET3_SET_ALU_CONST: + /* XXX fix me ALU const buffers only */ + break; + case PACKET3_SET_BOOL_CONST: + start_reg = (idx_value << 2) + PACKET3_SET_BOOL_CONST_START; + end_reg = 4 * pkt->count + start_reg - 4; + if ((start_reg < PACKET3_SET_BOOL_CONST_START) || + (start_reg >= PACKET3_SET_BOOL_CONST_END) || + (end_reg >= PACKET3_SET_BOOL_CONST_END)) { + DRM_ERROR("bad SET_BOOL_CONST\n"); + return -EINVAL; + } + break; + case PACKET3_SET_LOOP_CONST: + start_reg = (idx_value << 2) + PACKET3_SET_LOOP_CONST_START; + end_reg = 4 * pkt->count + start_reg - 4; + if ((start_reg < PACKET3_SET_LOOP_CONST_START) || + (start_reg >= PACKET3_SET_LOOP_CONST_END) || + (end_reg >= PACKET3_SET_LOOP_CONST_END)) { + DRM_ERROR("bad SET_LOOP_CONST\n"); + return -EINVAL; + } + break; + case PACKET3_SET_CTL_CONST: + start_reg = (idx_value << 2) + PACKET3_SET_CTL_CONST_START; + end_reg = 4 * pkt->count + start_reg - 4; + if ((start_reg < PACKET3_SET_CTL_CONST_START) || + (start_reg >= PACKET3_SET_CTL_CONST_END) || + (end_reg >= PACKET3_SET_CTL_CONST_END)) { + DRM_ERROR("bad SET_CTL_CONST\n"); + return -EINVAL; + } + break; + case PACKET3_SET_SAMPLER: + if (pkt->count % 3) { + DRM_ERROR("bad SET_SAMPLER\n"); + return -EINVAL; + } + start_reg = (idx_value << 2) + PACKET3_SET_SAMPLER_START; + end_reg = 4 * pkt->count + start_reg - 4; + if ((start_reg < PACKET3_SET_SAMPLER_START) || + (start_reg >= PACKET3_SET_SAMPLER_END) || + (end_reg >= PACKET3_SET_SAMPLER_END)) { + DRM_ERROR("bad SET_SAMPLER\n"); + return -EINVAL; + } + break; + case PACKET3_NOP: + break; + default: + DRM_ERROR("Packet3 opcode %x not supported\n", pkt->opcode); + return -EINVAL; + } + return 0; +} + +int evergreen_cs_parse(struct radeon_cs_parser *p) +{ + struct radeon_cs_packet pkt; + struct evergreen_cs_track *track; + int r; + + if (p->track == NULL) { + /* initialize tracker, we are in kms */ + track = kzalloc(sizeof(*track), GFP_KERNEL); + if (track == NULL) + return -ENOMEM; + evergreen_cs_track_init(track); + track->npipes = p->rdev->config.evergreen.tiling_npipes; + track->nbanks = p->rdev->config.evergreen.tiling_nbanks; + track->group_size = p->rdev->config.evergreen.tiling_group_size; + p->track = track; + } + do { + r = evergreen_cs_packet_parse(p, &pkt, p->idx); + if (r) { + kfree(p->track); + p->track = NULL; + return r; + } + p->idx += pkt.count + 2; + switch (pkt.type) { + case PACKET_TYPE0: + r = evergreen_cs_parse_packet0(p, &pkt); + break; + case PACKET_TYPE2: + break; + case PACKET_TYPE3: + r = evergreen_packet3_check(p, &pkt); + break; + default: + DRM_ERROR("Unknown packet type %d !\n", pkt.type); + kfree(p->track); + p->track = NULL; + return -EINVAL; + } + if (r) { + kfree(p->track); + p->track = NULL; + return r; + } + } while (p->idx < p->chunks[p->chunk_ib_idx].length_dw); +#if 0 + for (r = 0; r < p->ib->length_dw; r++) { + printk(KERN_INFO "%05d 0x%08X\n", r, p->ib->ptr[r]); + mdelay(1); + } +#endif + kfree(p->track); + p->track = NULL; + return 0; +} + diff --git a/drivers/gpu/drm/radeon/evergreen_reg.h b/drivers/gpu/drm/radeon/evergreen_reg.h index f7c7c9643433..e028c1cd9d9b 100644 --- a/drivers/gpu/drm/radeon/evergreen_reg.h +++ b/drivers/gpu/drm/radeon/evergreen_reg.h @@ -151,6 +151,9 @@ #define EVERGREEN_DATA_FORMAT 0x6b00 # define EVERGREEN_INTERLEAVE_EN (1 << 0) #define EVERGREEN_DESKTOP_HEIGHT 0x6b04 +#define EVERGREEN_VLINE_START_END 0x6b08 +#define EVERGREEN_VLINE_STATUS 0x6bb8 +# define EVERGREEN_VLINE_STAT (1 << 12) #define EVERGREEN_VIEWPORT_START 0x6d70 #define EVERGREEN_VIEWPORT_SIZE 0x6d74 @@ -164,8 +167,12 @@ #define EVERGREEN_CRTC5_REGISTER_OFFSET (0x129f0 - 0x6df0) /* CRTC blocks at 0x6df0, 0x79f0, 0x105f0, 0x111f0, 0x11df0, 0x129f0 */ +#define EVERGREEN_CRTC_V_BLANK_START_END 0x6e34 #define EVERGREEN_CRTC_CONTROL 0x6e70 # define EVERGREEN_CRTC_MASTER_EN (1 << 0) +# define EVERGREEN_CRTC_DISP_READ_REQUEST_DISABLE (1 << 24) +#define EVERGREEN_CRTC_STATUS 0x6e8c +#define EVERGREEN_CRTC_STATUS_POSITION 0x6e90 #define EVERGREEN_CRTC_UPDATE_LOCK 0x6ed4 #define EVERGREEN_DC_GPIO_HPD_MASK 0x64b0 diff --git a/drivers/gpu/drm/radeon/evergreend.h b/drivers/gpu/drm/radeon/evergreend.h new file mode 100644 index 000000000000..a1cd621780e2 --- /dev/null +++ b/drivers/gpu/drm/radeon/evergreend.h @@ -0,0 +1,1023 @@ +/* + * Copyright 2010 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Alex Deucher + */ +#ifndef EVERGREEND_H +#define EVERGREEND_H + +#define EVERGREEN_MAX_SH_GPRS 256 +#define EVERGREEN_MAX_TEMP_GPRS 16 +#define EVERGREEN_MAX_SH_THREADS 256 +#define EVERGREEN_MAX_SH_STACK_ENTRIES 4096 +#define EVERGREEN_MAX_FRC_EOV_CNT 16384 +#define EVERGREEN_MAX_BACKENDS 8 +#define EVERGREEN_MAX_BACKENDS_MASK 0xFF +#define EVERGREEN_MAX_SIMDS 16 +#define EVERGREEN_MAX_SIMDS_MASK 0xFFFF +#define EVERGREEN_MAX_PIPES 8 +#define EVERGREEN_MAX_PIPES_MASK 0xFF +#define EVERGREEN_MAX_LDS_NUM 0xFFFF + +/* Registers */ + +#define RCU_IND_INDEX 0x100 +#define RCU_IND_DATA 0x104 + +#define GRBM_GFX_INDEX 0x802C +#define INSTANCE_INDEX(x) ((x) << 0) +#define SE_INDEX(x) ((x) << 16) +#define INSTANCE_BROADCAST_WRITES (1 << 30) +#define SE_BROADCAST_WRITES (1 << 31) +#define RLC_GFX_INDEX 0x3fC4 +#define CC_GC_SHADER_PIPE_CONFIG 0x8950 +#define WRITE_DIS (1 << 0) +#define CC_RB_BACKEND_DISABLE 0x98F4 +#define BACKEND_DISABLE(x) ((x) << 16) +#define GB_ADDR_CONFIG 0x98F8 +#define NUM_PIPES(x) ((x) << 0) +#define PIPE_INTERLEAVE_SIZE(x) ((x) << 4) +#define BANK_INTERLEAVE_SIZE(x) ((x) << 8) +#define NUM_SHADER_ENGINES(x) ((x) << 12) +#define SHADER_ENGINE_TILE_SIZE(x) ((x) << 16) +#define NUM_GPUS(x) ((x) << 20) +#define MULTI_GPU_TILE_SIZE(x) ((x) << 24) +#define ROW_SIZE(x) ((x) << 28) +#define GB_BACKEND_MAP 0x98FC +#define DMIF_ADDR_CONFIG 0xBD4 +#define HDP_ADDR_CONFIG 0x2F48 + +#define CC_SYS_RB_BACKEND_DISABLE 0x3F88 +#define GC_USER_RB_BACKEND_DISABLE 0x9B7C + +#define CGTS_SYS_TCC_DISABLE 0x3F90 +#define CGTS_TCC_DISABLE 0x9148 +#define CGTS_USER_SYS_TCC_DISABLE 0x3F94 +#define CGTS_USER_TCC_DISABLE 0x914C + +#define CONFIG_MEMSIZE 0x5428 + +#define CP_ME_CNTL 0x86D8 +#define CP_ME_HALT (1 << 28) +#define CP_PFP_HALT (1 << 26) +#define CP_ME_RAM_DATA 0xC160 +#define CP_ME_RAM_RADDR 0xC158 +#define CP_ME_RAM_WADDR 0xC15C +#define CP_MEQ_THRESHOLDS 0x8764 +#define STQ_SPLIT(x) ((x) << 0) +#define CP_PERFMON_CNTL 0x87FC +#define CP_PFP_UCODE_ADDR 0xC150 +#define CP_PFP_UCODE_DATA 0xC154 +#define CP_QUEUE_THRESHOLDS 0x8760 +#define ROQ_IB1_START(x) ((x) << 0) +#define ROQ_IB2_START(x) ((x) << 8) +#define CP_RB_BASE 0xC100 +#define CP_RB_CNTL 0xC104 +#define RB_BUFSZ(x) ((x) << 0) +#define RB_BLKSZ(x) ((x) << 8) +#define RB_NO_UPDATE (1 << 27) +#define RB_RPTR_WR_ENA (1 << 31) +#define BUF_SWAP_32BIT (2 << 16) +#define CP_RB_RPTR 0x8700 +#define CP_RB_RPTR_ADDR 0xC10C +#define CP_RB_RPTR_ADDR_HI 0xC110 +#define CP_RB_RPTR_WR 0xC108 +#define CP_RB_WPTR 0xC114 +#define CP_RB_WPTR_ADDR 0xC118 +#define CP_RB_WPTR_ADDR_HI 0xC11C +#define CP_RB_WPTR_DELAY 0x8704 +#define CP_SEM_WAIT_TIMER 0x85BC +#define CP_DEBUG 0xC1FC + + +#define GC_USER_SHADER_PIPE_CONFIG 0x8954 +#define INACTIVE_QD_PIPES(x) ((x) << 8) +#define INACTIVE_QD_PIPES_MASK 0x0000FF00 +#define INACTIVE_SIMDS(x) ((x) << 16) +#define INACTIVE_SIMDS_MASK 0x00FF0000 + +#define GRBM_CNTL 0x8000 +#define GRBM_READ_TIMEOUT(x) ((x) << 0) +#define GRBM_SOFT_RESET 0x8020 +#define SOFT_RESET_CP (1 << 0) +#define SOFT_RESET_CB (1 << 1) +#define SOFT_RESET_DB (1 << 3) +#define SOFT_RESET_PA (1 << 5) +#define SOFT_RESET_SC (1 << 6) +#define SOFT_RESET_SPI (1 << 8) +#define SOFT_RESET_SH (1 << 9) +#define SOFT_RESET_SX (1 << 10) +#define SOFT_RESET_TC (1 << 11) +#define SOFT_RESET_TA (1 << 12) +#define SOFT_RESET_VC (1 << 13) +#define SOFT_RESET_VGT (1 << 14) + +#define GRBM_STATUS 0x8010 +#define CMDFIFO_AVAIL_MASK 0x0000000F +#define SRBM_RQ_PENDING (1 << 5) +#define CF_RQ_PENDING (1 << 7) +#define PF_RQ_PENDING (1 << 8) +#define GRBM_EE_BUSY (1 << 10) +#define SX_CLEAN (1 << 11) +#define DB_CLEAN (1 << 12) +#define CB_CLEAN (1 << 13) +#define TA_BUSY (1 << 14) +#define VGT_BUSY_NO_DMA (1 << 16) +#define VGT_BUSY (1 << 17) +#define SX_BUSY (1 << 20) +#define SH_BUSY (1 << 21) +#define SPI_BUSY (1 << 22) +#define SC_BUSY (1 << 24) +#define PA_BUSY (1 << 25) +#define DB_BUSY (1 << 26) +#define CP_COHERENCY_BUSY (1 << 28) +#define CP_BUSY (1 << 29) +#define CB_BUSY (1 << 30) +#define GUI_ACTIVE (1 << 31) +#define GRBM_STATUS_SE0 0x8014 +#define GRBM_STATUS_SE1 0x8018 +#define SE_SX_CLEAN (1 << 0) +#define SE_DB_CLEAN (1 << 1) +#define SE_CB_CLEAN (1 << 2) +#define SE_TA_BUSY (1 << 25) +#define SE_SX_BUSY (1 << 26) +#define SE_SPI_BUSY (1 << 27) +#define SE_SH_BUSY (1 << 28) +#define SE_SC_BUSY (1 << 29) +#define SE_DB_BUSY (1 << 30) +#define SE_CB_BUSY (1 << 31) + +#define HDP_HOST_PATH_CNTL 0x2C00 +#define HDP_NONSURFACE_BASE 0x2C04 +#define HDP_NONSURFACE_INFO 0x2C08 +#define HDP_NONSURFACE_SIZE 0x2C0C +#define HDP_REG_COHERENCY_FLUSH_CNTL 0x54A0 +#define HDP_TILING_CONFIG 0x2F3C + +#define MC_SHARED_CHMAP 0x2004 +#define NOOFCHAN_SHIFT 12 +#define NOOFCHAN_MASK 0x00003000 + +#define MC_ARB_RAMCFG 0x2760 +#define NOOFBANK_SHIFT 0 +#define NOOFBANK_MASK 0x00000003 +#define NOOFRANK_SHIFT 2 +#define NOOFRANK_MASK 0x00000004 +#define NOOFROWS_SHIFT 3 +#define NOOFROWS_MASK 0x00000038 +#define NOOFCOLS_SHIFT 6 +#define NOOFCOLS_MASK 0x000000C0 +#define CHANSIZE_SHIFT 8 +#define CHANSIZE_MASK 0x00000100 +#define BURSTLENGTH_SHIFT 9 +#define BURSTLENGTH_MASK 0x00000200 +#define CHANSIZE_OVERRIDE (1 << 11) +#define MC_VM_AGP_TOP 0x2028 +#define MC_VM_AGP_BOT 0x202C +#define MC_VM_AGP_BASE 0x2030 +#define MC_VM_FB_LOCATION 0x2024 +#define MC_VM_MB_L1_TLB0_CNTL 0x2234 +#define MC_VM_MB_L1_TLB1_CNTL 0x2238 +#define MC_VM_MB_L1_TLB2_CNTL 0x223C +#define MC_VM_MB_L1_TLB3_CNTL 0x2240 +#define ENABLE_L1_TLB (1 << 0) +#define ENABLE_L1_FRAGMENT_PROCESSING (1 << 1) +#define SYSTEM_ACCESS_MODE_PA_ONLY (0 << 3) +#define SYSTEM_ACCESS_MODE_USE_SYS_MAP (1 << 3) +#define SYSTEM_ACCESS_MODE_IN_SYS (2 << 3) +#define SYSTEM_ACCESS_MODE_NOT_IN_SYS (3 << 3) +#define SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU (0 << 5) +#define EFFECTIVE_L1_TLB_SIZE(x) ((x)<<15) +#define EFFECTIVE_L1_QUEUE_SIZE(x) ((x)<<18) +#define MC_VM_MD_L1_TLB0_CNTL 0x2654 +#define MC_VM_MD_L1_TLB1_CNTL 0x2658 +#define MC_VM_MD_L1_TLB2_CNTL 0x265C +#define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR 0x203C +#define MC_VM_SYSTEM_APERTURE_HIGH_ADDR 0x2038 +#define MC_VM_SYSTEM_APERTURE_LOW_ADDR 0x2034 + +#define PA_CL_ENHANCE 0x8A14 +#define CLIP_VTX_REORDER_ENA (1 << 0) +#define NUM_CLIP_SEQ(x) ((x) << 1) +#define PA_SC_AA_CONFIG 0x28C04 +#define MSAA_NUM_SAMPLES_SHIFT 0 +#define MSAA_NUM_SAMPLES_MASK 0x3 +#define PA_SC_CLIPRECT_RULE 0x2820C +#define PA_SC_EDGERULE 0x28230 +#define PA_SC_FIFO_SIZE 0x8BCC +#define SC_PRIM_FIFO_SIZE(x) ((x) << 0) +#define SC_HIZ_TILE_FIFO_SIZE(x) ((x) << 12) +#define SC_EARLYZ_TILE_FIFO_SIZE(x) ((x) << 20) +#define PA_SC_FORCE_EOV_MAX_CNTS 0x8B24 +#define FORCE_EOV_MAX_CLK_CNT(x) ((x) << 0) +#define FORCE_EOV_MAX_REZ_CNT(x) ((x) << 16) +#define PA_SC_LINE_STIPPLE 0x28A0C +#define PA_SC_LINE_STIPPLE_STATE 0x8B10 + +#define SCRATCH_REG0 0x8500 +#define SCRATCH_REG1 0x8504 +#define SCRATCH_REG2 0x8508 +#define SCRATCH_REG3 0x850C +#define SCRATCH_REG4 0x8510 +#define SCRATCH_REG5 0x8514 +#define SCRATCH_REG6 0x8518 +#define SCRATCH_REG7 0x851C +#define SCRATCH_UMSK 0x8540 +#define SCRATCH_ADDR 0x8544 + +#define SMX_DC_CTL0 0xA020 +#define USE_HASH_FUNCTION (1 << 0) +#define NUMBER_OF_SETS(x) ((x) << 1) +#define FLUSH_ALL_ON_EVENT (1 << 10) +#define STALL_ON_EVENT (1 << 11) +#define SMX_EVENT_CTL 0xA02C +#define ES_FLUSH_CTL(x) ((x) << 0) +#define GS_FLUSH_CTL(x) ((x) << 3) +#define ACK_FLUSH_CTL(x) ((x) << 6) +#define SYNC_FLUSH_CTL (1 << 8) + +#define SPI_CONFIG_CNTL 0x9100 +#define GPR_WRITE_PRIORITY(x) ((x) << 0) +#define SPI_CONFIG_CNTL_1 0x913C +#define VTX_DONE_DELAY(x) ((x) << 0) +#define INTERP_ONE_PRIM_PER_ROW (1 << 4) +#define SPI_INPUT_Z 0x286D8 +#define SPI_PS_IN_CONTROL_0 0x286CC +#define NUM_INTERP(x) ((x)<<0) +#define POSITION_ENA (1<<8) +#define POSITION_CENTROID (1<<9) +#define POSITION_ADDR(x) ((x)<<10) +#define PARAM_GEN(x) ((x)<<15) +#define PARAM_GEN_ADDR(x) ((x)<<19) +#define BARYC_SAMPLE_CNTL(x) ((x)<<26) +#define PERSP_GRADIENT_ENA (1<<28) +#define LINEAR_GRADIENT_ENA (1<<29) +#define POSITION_SAMPLE (1<<30) +#define BARYC_AT_SAMPLE_ENA (1<<31) + +#define SQ_CONFIG 0x8C00 +#define VC_ENABLE (1 << 0) +#define EXPORT_SRC_C (1 << 1) +#define CS_PRIO(x) ((x) << 18) +#define LS_PRIO(x) ((x) << 20) +#define HS_PRIO(x) ((x) << 22) +#define PS_PRIO(x) ((x) << 24) +#define VS_PRIO(x) ((x) << 26) +#define GS_PRIO(x) ((x) << 28) +#define ES_PRIO(x) ((x) << 30) +#define SQ_GPR_RESOURCE_MGMT_1 0x8C04 +#define NUM_PS_GPRS(x) ((x) << 0) +#define NUM_VS_GPRS(x) ((x) << 16) +#define NUM_CLAUSE_TEMP_GPRS(x) ((x) << 28) +#define SQ_GPR_RESOURCE_MGMT_2 0x8C08 +#define NUM_GS_GPRS(x) ((x) << 0) +#define NUM_ES_GPRS(x) ((x) << 16) +#define SQ_GPR_RESOURCE_MGMT_3 0x8C0C +#define NUM_HS_GPRS(x) ((x) << 0) +#define NUM_LS_GPRS(x) ((x) << 16) +#define SQ_THREAD_RESOURCE_MGMT 0x8C18 +#define NUM_PS_THREADS(x) ((x) << 0) +#define NUM_VS_THREADS(x) ((x) << 8) +#define NUM_GS_THREADS(x) ((x) << 16) +#define NUM_ES_THREADS(x) ((x) << 24) +#define SQ_THREAD_RESOURCE_MGMT_2 0x8C1C +#define NUM_HS_THREADS(x) ((x) << 0) +#define NUM_LS_THREADS(x) ((x) << 8) +#define SQ_STACK_RESOURCE_MGMT_1 0x8C20 +#define NUM_PS_STACK_ENTRIES(x) ((x) << 0) +#define NUM_VS_STACK_ENTRIES(x) ((x) << 16) +#define SQ_STACK_RESOURCE_MGMT_2 0x8C24 +#define NUM_GS_STACK_ENTRIES(x) ((x) << 0) +#define NUM_ES_STACK_ENTRIES(x) ((x) << 16) +#define SQ_STACK_RESOURCE_MGMT_3 0x8C28 +#define NUM_HS_STACK_ENTRIES(x) ((x) << 0) +#define NUM_LS_STACK_ENTRIES(x) ((x) << 16) +#define SQ_DYN_GPR_CNTL_PS_FLUSH_REQ 0x8D8C +#define SQ_LDS_RESOURCE_MGMT 0x8E2C + +#define SQ_MS_FIFO_SIZES 0x8CF0 +#define CACHE_FIFO_SIZE(x) ((x) << 0) +#define FETCH_FIFO_HIWATER(x) ((x) << 8) +#define DONE_FIFO_HIWATER(x) ((x) << 16) +#define ALU_UPDATE_FIFO_HIWATER(x) ((x) << 24) + +#define SX_DEBUG_1 0x9058 +#define ENABLE_NEW_SMX_ADDRESS (1 << 16) +#define SX_EXPORT_BUFFER_SIZES 0x900C +#define COLOR_BUFFER_SIZE(x) ((x) << 0) +#define POSITION_BUFFER_SIZE(x) ((x) << 8) +#define SMX_BUFFER_SIZE(x) ((x) << 16) +#define SX_MISC 0x28350 + +#define CB_PERF_CTR0_SEL_0 0x9A20 +#define CB_PERF_CTR0_SEL_1 0x9A24 +#define CB_PERF_CTR1_SEL_0 0x9A28 +#define CB_PERF_CTR1_SEL_1 0x9A2C +#define CB_PERF_CTR2_SEL_0 0x9A30 +#define CB_PERF_CTR2_SEL_1 0x9A34 +#define CB_PERF_CTR3_SEL_0 0x9A38 +#define CB_PERF_CTR3_SEL_1 0x9A3C + +#define TA_CNTL_AUX 0x9508 +#define DISABLE_CUBE_WRAP (1 << 0) +#define DISABLE_CUBE_ANISO (1 << 1) +#define SYNC_GRADIENT (1 << 24) +#define SYNC_WALKER (1 << 25) +#define SYNC_ALIGNER (1 << 26) + +#define VGT_CACHE_INVALIDATION 0x88C4 +#define CACHE_INVALIDATION(x) ((x) << 0) +#define VC_ONLY 0 +#define TC_ONLY 1 +#define VC_AND_TC 2 +#define AUTO_INVLD_EN(x) ((x) << 6) +#define NO_AUTO 0 +#define ES_AUTO 1 +#define GS_AUTO 2 +#define ES_AND_GS_AUTO 3 +#define VGT_GS_VERTEX_REUSE 0x88D4 +#define VGT_NUM_INSTANCES 0x8974 +#define VGT_OUT_DEALLOC_CNTL 0x28C5C +#define DEALLOC_DIST_MASK 0x0000007F +#define VGT_VERTEX_REUSE_BLOCK_CNTL 0x28C58 +#define VTX_REUSE_DEPTH_MASK 0x000000FF + +#define VM_CONTEXT0_CNTL 0x1410 +#define ENABLE_CONTEXT (1 << 0) +#define PAGE_TABLE_DEPTH(x) (((x) & 3) << 1) +#define RANGE_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 4) +#define VM_CONTEXT1_CNTL 0x1414 +#define VM_CONTEXT0_PAGE_TABLE_BASE_ADDR 0x153C +#define VM_CONTEXT0_PAGE_TABLE_END_ADDR 0x157C +#define VM_CONTEXT0_PAGE_TABLE_START_ADDR 0x155C +#define VM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR 0x1518 +#define VM_CONTEXT0_REQUEST_RESPONSE 0x1470 +#define REQUEST_TYPE(x) (((x) & 0xf) << 0) +#define RESPONSE_TYPE_MASK 0x000000F0 +#define RESPONSE_TYPE_SHIFT 4 +#define VM_L2_CNTL 0x1400 +#define ENABLE_L2_CACHE (1 << 0) +#define ENABLE_L2_FRAGMENT_PROCESSING (1 << 1) +#define ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE (1 << 9) +#define EFFECTIVE_L2_QUEUE_SIZE(x) (((x) & 7) << 14) +#define VM_L2_CNTL2 0x1404 +#define INVALIDATE_ALL_L1_TLBS (1 << 0) +#define INVALIDATE_L2_CACHE (1 << 1) +#define VM_L2_CNTL3 0x1408 +#define BANK_SELECT(x) ((x) << 0) +#define CACHE_UPDATE_MODE(x) ((x) << 6) +#define VM_L2_STATUS 0x140C +#define L2_BUSY (1 << 0) + +#define WAIT_UNTIL 0x8040 + +#define SRBM_STATUS 0x0E50 +#define SRBM_SOFT_RESET 0x0E60 +#define SRBM_SOFT_RESET_ALL_MASK 0x00FEEFA6 +#define SOFT_RESET_BIF (1 << 1) +#define SOFT_RESET_CG (1 << 2) +#define SOFT_RESET_DC (1 << 5) +#define SOFT_RESET_GRBM (1 << 8) +#define SOFT_RESET_HDP (1 << 9) +#define SOFT_RESET_IH (1 << 10) +#define SOFT_RESET_MC (1 << 11) +#define SOFT_RESET_RLC (1 << 13) +#define SOFT_RESET_ROM (1 << 14) +#define SOFT_RESET_SEM (1 << 15) +#define SOFT_RESET_VMC (1 << 17) +#define SOFT_RESET_TST (1 << 21) +#define SOFT_RESET_REGBB (1 << 22) +#define SOFT_RESET_ORB (1 << 23) + +#define IH_RB_CNTL 0x3e00 +# define IH_RB_ENABLE (1 << 0) +# define IH_IB_SIZE(x) ((x) << 1) /* log2 */ +# define IH_RB_FULL_DRAIN_ENABLE (1 << 6) +# define IH_WPTR_WRITEBACK_ENABLE (1 << 8) +# define IH_WPTR_WRITEBACK_TIMER(x) ((x) << 9) /* log2 */ +# define IH_WPTR_OVERFLOW_ENABLE (1 << 16) +# define IH_WPTR_OVERFLOW_CLEAR (1 << 31) +#define IH_RB_BASE 0x3e04 +#define IH_RB_RPTR 0x3e08 +#define IH_RB_WPTR 0x3e0c +# define RB_OVERFLOW (1 << 0) +# define WPTR_OFFSET_MASK 0x3fffc +#define IH_RB_WPTR_ADDR_HI 0x3e10 +#define IH_RB_WPTR_ADDR_LO 0x3e14 +#define IH_CNTL 0x3e18 +# define ENABLE_INTR (1 << 0) +# define IH_MC_SWAP(x) ((x) << 2) +# define IH_MC_SWAP_NONE 0 +# define IH_MC_SWAP_16BIT 1 +# define IH_MC_SWAP_32BIT 2 +# define IH_MC_SWAP_64BIT 3 +# define RPTR_REARM (1 << 4) +# define MC_WRREQ_CREDIT(x) ((x) << 15) +# define MC_WR_CLEAN_CNT(x) ((x) << 20) + +#define CP_INT_CNTL 0xc124 +# define CNTX_BUSY_INT_ENABLE (1 << 19) +# define CNTX_EMPTY_INT_ENABLE (1 << 20) +# define SCRATCH_INT_ENABLE (1 << 25) +# define TIME_STAMP_INT_ENABLE (1 << 26) +# define IB2_INT_ENABLE (1 << 29) +# define IB1_INT_ENABLE (1 << 30) +# define RB_INT_ENABLE (1 << 31) +#define CP_INT_STATUS 0xc128 +# define SCRATCH_INT_STAT (1 << 25) +# define TIME_STAMP_INT_STAT (1 << 26) +# define IB2_INT_STAT (1 << 29) +# define IB1_INT_STAT (1 << 30) +# define RB_INT_STAT (1 << 31) + +#define GRBM_INT_CNTL 0x8060 +# define RDERR_INT_ENABLE (1 << 0) +# define GUI_IDLE_INT_ENABLE (1 << 19) + +/* 0x6e98, 0x7a98, 0x10698, 0x11298, 0x11e98, 0x12a98 */ +#define CRTC_STATUS_FRAME_COUNT 0x6e98 + +/* 0x6bb8, 0x77b8, 0x103b8, 0x10fb8, 0x11bb8, 0x127b8 */ +#define VLINE_STATUS 0x6bb8 +# define VLINE_OCCURRED (1 << 0) +# define VLINE_ACK (1 << 4) +# define VLINE_STAT (1 << 12) +# define VLINE_INTERRUPT (1 << 16) +# define VLINE_INTERRUPT_TYPE (1 << 17) +/* 0x6bbc, 0x77bc, 0x103bc, 0x10fbc, 0x11bbc, 0x127bc */ +#define VBLANK_STATUS 0x6bbc +# define VBLANK_OCCURRED (1 << 0) +# define VBLANK_ACK (1 << 4) +# define VBLANK_STAT (1 << 12) +# define VBLANK_INTERRUPT (1 << 16) +# define VBLANK_INTERRUPT_TYPE (1 << 17) + +/* 0x6b40, 0x7740, 0x10340, 0x10f40, 0x11b40, 0x12740 */ +#define INT_MASK 0x6b40 +# define VBLANK_INT_MASK (1 << 0) +# define VLINE_INT_MASK (1 << 4) + +#define DISP_INTERRUPT_STATUS 0x60f4 +# define LB_D1_VLINE_INTERRUPT (1 << 2) +# define LB_D1_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD1_INTERRUPT (1 << 17) +# define DC_HPD1_RX_INTERRUPT (1 << 18) +# define DACA_AUTODETECT_INTERRUPT (1 << 22) +# define DACB_AUTODETECT_INTERRUPT (1 << 23) +# define DC_I2C_SW_DONE_INTERRUPT (1 << 24) +# define DC_I2C_HW_DONE_INTERRUPT (1 << 25) +#define DISP_INTERRUPT_STATUS_CONTINUE 0x60f8 +# define LB_D2_VLINE_INTERRUPT (1 << 2) +# define LB_D2_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD2_INTERRUPT (1 << 17) +# define DC_HPD2_RX_INTERRUPT (1 << 18) +# define DISP_TIMER_INTERRUPT (1 << 24) +#define DISP_INTERRUPT_STATUS_CONTINUE2 0x60fc +# define LB_D3_VLINE_INTERRUPT (1 << 2) +# define LB_D3_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD3_INTERRUPT (1 << 17) +# define DC_HPD3_RX_INTERRUPT (1 << 18) +#define DISP_INTERRUPT_STATUS_CONTINUE3 0x6100 +# define LB_D4_VLINE_INTERRUPT (1 << 2) +# define LB_D4_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD4_INTERRUPT (1 << 17) +# define DC_HPD4_RX_INTERRUPT (1 << 18) +#define DISP_INTERRUPT_STATUS_CONTINUE4 0x614c +# define LB_D5_VLINE_INTERRUPT (1 << 2) +# define LB_D5_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD5_INTERRUPT (1 << 17) +# define DC_HPD5_RX_INTERRUPT (1 << 18) +#define DISP_INTERRUPT_STATUS_CONTINUE5 0x6050 +# define LB_D6_VLINE_INTERRUPT (1 << 2) +# define LB_D6_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD6_INTERRUPT (1 << 17) +# define DC_HPD6_RX_INTERRUPT (1 << 18) + +/* 0x6858, 0x7458, 0x10058, 0x10c58, 0x11858, 0x12458 */ +#define GRPH_INT_STATUS 0x6858 +# define GRPH_PFLIP_INT_OCCURRED (1 << 0) +# define GRPH_PFLIP_INT_CLEAR (1 << 8) +/* 0x685c, 0x745c, 0x1005c, 0x10c5c, 0x1185c, 0x1245c */ +#define GRPH_INT_CONTROL 0x685c +# define GRPH_PFLIP_INT_MASK (1 << 0) +# define GRPH_PFLIP_INT_TYPE (1 << 8) + +#define DACA_AUTODETECT_INT_CONTROL 0x66c8 +#define DACB_AUTODETECT_INT_CONTROL 0x67c8 + +#define DC_HPD1_INT_STATUS 0x601c +#define DC_HPD2_INT_STATUS 0x6028 +#define DC_HPD3_INT_STATUS 0x6034 +#define DC_HPD4_INT_STATUS 0x6040 +#define DC_HPD5_INT_STATUS 0x604c +#define DC_HPD6_INT_STATUS 0x6058 +# define DC_HPDx_INT_STATUS (1 << 0) +# define DC_HPDx_SENSE (1 << 1) +# define DC_HPDx_RX_INT_STATUS (1 << 8) + +#define DC_HPD1_INT_CONTROL 0x6020 +#define DC_HPD2_INT_CONTROL 0x602c +#define DC_HPD3_INT_CONTROL 0x6038 +#define DC_HPD4_INT_CONTROL 0x6044 +#define DC_HPD5_INT_CONTROL 0x6050 +#define DC_HPD6_INT_CONTROL 0x605c +# define DC_HPDx_INT_ACK (1 << 0) +# define DC_HPDx_INT_POLARITY (1 << 8) +# define DC_HPDx_INT_EN (1 << 16) +# define DC_HPDx_RX_INT_ACK (1 << 20) +# define DC_HPDx_RX_INT_EN (1 << 24) + +#define DC_HPD1_CONTROL 0x6024 +#define DC_HPD2_CONTROL 0x6030 +#define DC_HPD3_CONTROL 0x603c +#define DC_HPD4_CONTROL 0x6048 +#define DC_HPD5_CONTROL 0x6054 +#define DC_HPD6_CONTROL 0x6060 +# define DC_HPDx_CONNECTION_TIMER(x) ((x) << 0) +# define DC_HPDx_RX_INT_TIMER(x) ((x) << 16) +# define DC_HPDx_EN (1 << 28) + +/* + * PM4 + */ +#define PACKET_TYPE0 0 +#define PACKET_TYPE1 1 +#define PACKET_TYPE2 2 +#define PACKET_TYPE3 3 + +#define CP_PACKET_GET_TYPE(h) (((h) >> 30) & 3) +#define CP_PACKET_GET_COUNT(h) (((h) >> 16) & 0x3FFF) +#define CP_PACKET0_GET_REG(h) (((h) & 0xFFFF) << 2) +#define CP_PACKET3_GET_OPCODE(h) (((h) >> 8) & 0xFF) +#define PACKET0(reg, n) ((PACKET_TYPE0 << 30) | \ + (((reg) >> 2) & 0xFFFF) | \ + ((n) & 0x3FFF) << 16) +#define CP_PACKET2 0x80000000 +#define PACKET2_PAD_SHIFT 0 +#define PACKET2_PAD_MASK (0x3fffffff << 0) + +#define PACKET2(v) (CP_PACKET2 | REG_SET(PACKET2_PAD, (v))) + +#define PACKET3(op, n) ((PACKET_TYPE3 << 30) | \ + (((op) & 0xFF) << 8) | \ + ((n) & 0x3FFF) << 16) + +/* Packet 3 types */ +#define PACKET3_NOP 0x10 +#define PACKET3_SET_BASE 0x11 +#define PACKET3_CLEAR_STATE 0x12 +#define PACKET3_INDIRECT_BUFFER_SIZE 0x13 +#define PACKET3_DISPATCH_DIRECT 0x15 +#define PACKET3_DISPATCH_INDIRECT 0x16 +#define PACKET3_INDIRECT_BUFFER_END 0x17 +#define PACKET3_SET_PREDICATION 0x20 +#define PACKET3_REG_RMW 0x21 +#define PACKET3_COND_EXEC 0x22 +#define PACKET3_PRED_EXEC 0x23 +#define PACKET3_DRAW_INDIRECT 0x24 +#define PACKET3_DRAW_INDEX_INDIRECT 0x25 +#define PACKET3_INDEX_BASE 0x26 +#define PACKET3_DRAW_INDEX_2 0x27 +#define PACKET3_CONTEXT_CONTROL 0x28 +#define PACKET3_DRAW_INDEX_OFFSET 0x29 +#define PACKET3_INDEX_TYPE 0x2A +#define PACKET3_DRAW_INDEX 0x2B +#define PACKET3_DRAW_INDEX_AUTO 0x2D +#define PACKET3_DRAW_INDEX_IMMD 0x2E +#define PACKET3_NUM_INSTANCES 0x2F +#define PACKET3_DRAW_INDEX_MULTI_AUTO 0x30 +#define PACKET3_STRMOUT_BUFFER_UPDATE 0x34 +#define PACKET3_DRAW_INDEX_OFFSET_2 0x35 +#define PACKET3_DRAW_INDEX_MULTI_ELEMENT 0x36 +#define PACKET3_MEM_SEMAPHORE 0x39 +#define PACKET3_MPEG_INDEX 0x3A +#define PACKET3_WAIT_REG_MEM 0x3C +#define PACKET3_MEM_WRITE 0x3D +#define PACKET3_INDIRECT_BUFFER 0x32 +#define PACKET3_SURFACE_SYNC 0x43 +# define PACKET3_CB0_DEST_BASE_ENA (1 << 6) +# define PACKET3_CB1_DEST_BASE_ENA (1 << 7) +# define PACKET3_CB2_DEST_BASE_ENA (1 << 8) +# define PACKET3_CB3_DEST_BASE_ENA (1 << 9) +# define PACKET3_CB4_DEST_BASE_ENA (1 << 10) +# define PACKET3_CB5_DEST_BASE_ENA (1 << 11) +# define PACKET3_CB6_DEST_BASE_ENA (1 << 12) +# define PACKET3_CB7_DEST_BASE_ENA (1 << 13) +# define PACKET3_DB_DEST_BASE_ENA (1 << 14) +# define PACKET3_CB8_DEST_BASE_ENA (1 << 15) +# define PACKET3_CB9_DEST_BASE_ENA (1 << 16) +# define PACKET3_CB10_DEST_BASE_ENA (1 << 17) +# define PACKET3_CB11_DEST_BASE_ENA (1 << 17) +# define PACKET3_FULL_CACHE_ENA (1 << 20) +# define PACKET3_TC_ACTION_ENA (1 << 23) +# define PACKET3_VC_ACTION_ENA (1 << 24) +# define PACKET3_CB_ACTION_ENA (1 << 25) +# define PACKET3_DB_ACTION_ENA (1 << 26) +# define PACKET3_SH_ACTION_ENA (1 << 27) +# define PACKET3_SMX_ACTION_ENA (1 << 28) +#define PACKET3_ME_INITIALIZE 0x44 +#define PACKET3_ME_INITIALIZE_DEVICE_ID(x) ((x) << 16) +#define PACKET3_COND_WRITE 0x45 +#define PACKET3_EVENT_WRITE 0x46 +#define PACKET3_EVENT_WRITE_EOP 0x47 +#define PACKET3_EVENT_WRITE_EOS 0x48 +#define PACKET3_PREAMBLE_CNTL 0x4A +#define PACKET3_RB_OFFSET 0x4B +#define PACKET3_ALU_PS_CONST_BUFFER_COPY 0x4C +#define PACKET3_ALU_VS_CONST_BUFFER_COPY 0x4D +#define PACKET3_ALU_PS_CONST_UPDATE 0x4E +#define PACKET3_ALU_VS_CONST_UPDATE 0x4F +#define PACKET3_ONE_REG_WRITE 0x57 +#define PACKET3_SET_CONFIG_REG 0x68 +#define PACKET3_SET_CONFIG_REG_START 0x00008000 +#define PACKET3_SET_CONFIG_REG_END 0x0000ac00 +#define PACKET3_SET_CONTEXT_REG 0x69 +#define PACKET3_SET_CONTEXT_REG_START 0x00028000 +#define PACKET3_SET_CONTEXT_REG_END 0x00029000 +#define PACKET3_SET_ALU_CONST 0x6A +/* alu const buffers only; no reg file */ +#define PACKET3_SET_BOOL_CONST 0x6B +#define PACKET3_SET_BOOL_CONST_START 0x0003a500 +#define PACKET3_SET_BOOL_CONST_END 0x0003a518 +#define PACKET3_SET_LOOP_CONST 0x6C +#define PACKET3_SET_LOOP_CONST_START 0x0003a200 +#define PACKET3_SET_LOOP_CONST_END 0x0003a500 +#define PACKET3_SET_RESOURCE 0x6D +#define PACKET3_SET_RESOURCE_START 0x00030000 +#define PACKET3_SET_RESOURCE_END 0x00038000 +#define PACKET3_SET_SAMPLER 0x6E +#define PACKET3_SET_SAMPLER_START 0x0003c000 +#define PACKET3_SET_SAMPLER_END 0x0003c600 +#define PACKET3_SET_CTL_CONST 0x6F +#define PACKET3_SET_CTL_CONST_START 0x0003cff0 +#define PACKET3_SET_CTL_CONST_END 0x0003ff0c +#define PACKET3_SET_RESOURCE_OFFSET 0x70 +#define PACKET3_SET_ALU_CONST_VS 0x71 +#define PACKET3_SET_ALU_CONST_DI 0x72 +#define PACKET3_SET_CONTEXT_REG_INDIRECT 0x73 +#define PACKET3_SET_RESOURCE_INDIRECT 0x74 +#define PACKET3_SET_APPEND_CNT 0x75 + +#define SQ_RESOURCE_CONSTANT_WORD7_0 0x3001c +#define S__SQ_CONSTANT_TYPE(x) (((x) & 3) << 30) +#define G__SQ_CONSTANT_TYPE(x) (((x) >> 30) & 3) +#define SQ_TEX_VTX_INVALID_TEXTURE 0x0 +#define SQ_TEX_VTX_INVALID_BUFFER 0x1 +#define SQ_TEX_VTX_VALID_TEXTURE 0x2 +#define SQ_TEX_VTX_VALID_BUFFER 0x3 + +#define SQ_CONST_MEM_BASE 0x8df8 + +#define SQ_ESGS_RING_SIZE 0x8c44 +#define SQ_GSVS_RING_SIZE 0x8c4c +#define SQ_ESTMP_RING_SIZE 0x8c54 +#define SQ_GSTMP_RING_SIZE 0x8c5c +#define SQ_VSTMP_RING_SIZE 0x8c64 +#define SQ_PSTMP_RING_SIZE 0x8c6c +#define SQ_LSTMP_RING_SIZE 0x8e14 +#define SQ_HSTMP_RING_SIZE 0x8e1c +#define VGT_TF_RING_SIZE 0x8988 + +#define SQ_ESGS_RING_ITEMSIZE 0x28900 +#define SQ_GSVS_RING_ITEMSIZE 0x28904 +#define SQ_ESTMP_RING_ITEMSIZE 0x28908 +#define SQ_GSTMP_RING_ITEMSIZE 0x2890c +#define SQ_VSTMP_RING_ITEMSIZE 0x28910 +#define SQ_PSTMP_RING_ITEMSIZE 0x28914 +#define SQ_LSTMP_RING_ITEMSIZE 0x28830 +#define SQ_HSTMP_RING_ITEMSIZE 0x28834 + +#define SQ_GS_VERT_ITEMSIZE 0x2891c +#define SQ_GS_VERT_ITEMSIZE_1 0x28920 +#define SQ_GS_VERT_ITEMSIZE_2 0x28924 +#define SQ_GS_VERT_ITEMSIZE_3 0x28928 +#define SQ_GSVS_RING_OFFSET_1 0x2892c +#define SQ_GSVS_RING_OFFSET_2 0x28930 +#define SQ_GSVS_RING_OFFSET_3 0x28934 + +#define SQ_ALU_CONST_BUFFER_SIZE_PS_0 0x28140 +#define SQ_ALU_CONST_BUFFER_SIZE_HS_0 0x28f80 + +#define SQ_ALU_CONST_CACHE_PS_0 0x28940 +#define SQ_ALU_CONST_CACHE_PS_1 0x28944 +#define SQ_ALU_CONST_CACHE_PS_2 0x28948 +#define SQ_ALU_CONST_CACHE_PS_3 0x2894c +#define SQ_ALU_CONST_CACHE_PS_4 0x28950 +#define SQ_ALU_CONST_CACHE_PS_5 0x28954 +#define SQ_ALU_CONST_CACHE_PS_6 0x28958 +#define SQ_ALU_CONST_CACHE_PS_7 0x2895c +#define SQ_ALU_CONST_CACHE_PS_8 0x28960 +#define SQ_ALU_CONST_CACHE_PS_9 0x28964 +#define SQ_ALU_CONST_CACHE_PS_10 0x28968 +#define SQ_ALU_CONST_CACHE_PS_11 0x2896c +#define SQ_ALU_CONST_CACHE_PS_12 0x28970 +#define SQ_ALU_CONST_CACHE_PS_13 0x28974 +#define SQ_ALU_CONST_CACHE_PS_14 0x28978 +#define SQ_ALU_CONST_CACHE_PS_15 0x2897c +#define SQ_ALU_CONST_CACHE_VS_0 0x28980 +#define SQ_ALU_CONST_CACHE_VS_1 0x28984 +#define SQ_ALU_CONST_CACHE_VS_2 0x28988 +#define SQ_ALU_CONST_CACHE_VS_3 0x2898c +#define SQ_ALU_CONST_CACHE_VS_4 0x28990 +#define SQ_ALU_CONST_CACHE_VS_5 0x28994 +#define SQ_ALU_CONST_CACHE_VS_6 0x28998 +#define SQ_ALU_CONST_CACHE_VS_7 0x2899c +#define SQ_ALU_CONST_CACHE_VS_8 0x289a0 +#define SQ_ALU_CONST_CACHE_VS_9 0x289a4 +#define SQ_ALU_CONST_CACHE_VS_10 0x289a8 +#define SQ_ALU_CONST_CACHE_VS_11 0x289ac +#define SQ_ALU_CONST_CACHE_VS_12 0x289b0 +#define SQ_ALU_CONST_CACHE_VS_13 0x289b4 +#define SQ_ALU_CONST_CACHE_VS_14 0x289b8 +#define SQ_ALU_CONST_CACHE_VS_15 0x289bc +#define SQ_ALU_CONST_CACHE_GS_0 0x289c0 +#define SQ_ALU_CONST_CACHE_GS_1 0x289c4 +#define SQ_ALU_CONST_CACHE_GS_2 0x289c8 +#define SQ_ALU_CONST_CACHE_GS_3 0x289cc +#define SQ_ALU_CONST_CACHE_GS_4 0x289d0 +#define SQ_ALU_CONST_CACHE_GS_5 0x289d4 +#define SQ_ALU_CONST_CACHE_GS_6 0x289d8 +#define SQ_ALU_CONST_CACHE_GS_7 0x289dc +#define SQ_ALU_CONST_CACHE_GS_8 0x289e0 +#define SQ_ALU_CONST_CACHE_GS_9 0x289e4 +#define SQ_ALU_CONST_CACHE_GS_10 0x289e8 +#define SQ_ALU_CONST_CACHE_GS_11 0x289ec +#define SQ_ALU_CONST_CACHE_GS_12 0x289f0 +#define SQ_ALU_CONST_CACHE_GS_13 0x289f4 +#define SQ_ALU_CONST_CACHE_GS_14 0x289f8 +#define SQ_ALU_CONST_CACHE_GS_15 0x289fc +#define SQ_ALU_CONST_CACHE_HS_0 0x28f00 +#define SQ_ALU_CONST_CACHE_HS_1 0x28f04 +#define SQ_ALU_CONST_CACHE_HS_2 0x28f08 +#define SQ_ALU_CONST_CACHE_HS_3 0x28f0c +#define SQ_ALU_CONST_CACHE_HS_4 0x28f10 +#define SQ_ALU_CONST_CACHE_HS_5 0x28f14 +#define SQ_ALU_CONST_CACHE_HS_6 0x28f18 +#define SQ_ALU_CONST_CACHE_HS_7 0x28f1c +#define SQ_ALU_CONST_CACHE_HS_8 0x28f20 +#define SQ_ALU_CONST_CACHE_HS_9 0x28f24 +#define SQ_ALU_CONST_CACHE_HS_10 0x28f28 +#define SQ_ALU_CONST_CACHE_HS_11 0x28f2c +#define SQ_ALU_CONST_CACHE_HS_12 0x28f30 +#define SQ_ALU_CONST_CACHE_HS_13 0x28f34 +#define SQ_ALU_CONST_CACHE_HS_14 0x28f38 +#define SQ_ALU_CONST_CACHE_HS_15 0x28f3c +#define SQ_ALU_CONST_CACHE_LS_0 0x28f40 +#define SQ_ALU_CONST_CACHE_LS_1 0x28f44 +#define SQ_ALU_CONST_CACHE_LS_2 0x28f48 +#define SQ_ALU_CONST_CACHE_LS_3 0x28f4c +#define SQ_ALU_CONST_CACHE_LS_4 0x28f50 +#define SQ_ALU_CONST_CACHE_LS_5 0x28f54 +#define SQ_ALU_CONST_CACHE_LS_6 0x28f58 +#define SQ_ALU_CONST_CACHE_LS_7 0x28f5c +#define SQ_ALU_CONST_CACHE_LS_8 0x28f60 +#define SQ_ALU_CONST_CACHE_LS_9 0x28f64 +#define SQ_ALU_CONST_CACHE_LS_10 0x28f68 +#define SQ_ALU_CONST_CACHE_LS_11 0x28f6c +#define SQ_ALU_CONST_CACHE_LS_12 0x28f70 +#define SQ_ALU_CONST_CACHE_LS_13 0x28f74 +#define SQ_ALU_CONST_CACHE_LS_14 0x28f78 +#define SQ_ALU_CONST_CACHE_LS_15 0x28f7c + +#define DB_DEPTH_CONTROL 0x28800 +#define DB_DEPTH_VIEW 0x28008 +#define DB_HTILE_DATA_BASE 0x28014 +#define DB_Z_INFO 0x28040 +# define Z_ARRAY_MODE(x) ((x) << 4) +#define DB_STENCIL_INFO 0x28044 +#define DB_Z_READ_BASE 0x28048 +#define DB_STENCIL_READ_BASE 0x2804c +#define DB_Z_WRITE_BASE 0x28050 +#define DB_STENCIL_WRITE_BASE 0x28054 +#define DB_DEPTH_SIZE 0x28058 + +#define SQ_PGM_START_PS 0x28840 +#define SQ_PGM_START_VS 0x2885c +#define SQ_PGM_START_GS 0x28874 +#define SQ_PGM_START_ES 0x2888c +#define SQ_PGM_START_FS 0x288a4 +#define SQ_PGM_START_HS 0x288b8 +#define SQ_PGM_START_LS 0x288d0 + +#define VGT_STRMOUT_CONFIG 0x28b94 +#define VGT_STRMOUT_BUFFER_CONFIG 0x28b98 + +#define CB_TARGET_MASK 0x28238 +#define CB_SHADER_MASK 0x2823c + +#define GDS_ADDR_BASE 0x28720 + +#define CB_IMMED0_BASE 0x28b9c +#define CB_IMMED1_BASE 0x28ba0 +#define CB_IMMED2_BASE 0x28ba4 +#define CB_IMMED3_BASE 0x28ba8 +#define CB_IMMED4_BASE 0x28bac +#define CB_IMMED5_BASE 0x28bb0 +#define CB_IMMED6_BASE 0x28bb4 +#define CB_IMMED7_BASE 0x28bb8 +#define CB_IMMED8_BASE 0x28bbc +#define CB_IMMED9_BASE 0x28bc0 +#define CB_IMMED10_BASE 0x28bc4 +#define CB_IMMED11_BASE 0x28bc8 + +/* all 12 CB blocks have these regs */ +#define CB_COLOR0_BASE 0x28c60 +#define CB_COLOR0_PITCH 0x28c64 +#define CB_COLOR0_SLICE 0x28c68 +#define CB_COLOR0_VIEW 0x28c6c +#define CB_COLOR0_INFO 0x28c70 +# define CB_ARRAY_MODE(x) ((x) << 8) +# define ARRAY_LINEAR_GENERAL 0 +# define ARRAY_LINEAR_ALIGNED 1 +# define ARRAY_1D_TILED_THIN1 2 +# define ARRAY_2D_TILED_THIN1 4 +#define CB_COLOR0_ATTRIB 0x28c74 +#define CB_COLOR0_DIM 0x28c78 +/* only CB0-7 blocks have these regs */ +#define CB_COLOR0_CMASK 0x28c7c +#define CB_COLOR0_CMASK_SLICE 0x28c80 +#define CB_COLOR0_FMASK 0x28c84 +#define CB_COLOR0_FMASK_SLICE 0x28c88 +#define CB_COLOR0_CLEAR_WORD0 0x28c8c +#define CB_COLOR0_CLEAR_WORD1 0x28c90 +#define CB_COLOR0_CLEAR_WORD2 0x28c94 +#define CB_COLOR0_CLEAR_WORD3 0x28c98 + +#define CB_COLOR1_BASE 0x28c9c +#define CB_COLOR2_BASE 0x28cd8 +#define CB_COLOR3_BASE 0x28d14 +#define CB_COLOR4_BASE 0x28d50 +#define CB_COLOR5_BASE 0x28d8c +#define CB_COLOR6_BASE 0x28dc8 +#define CB_COLOR7_BASE 0x28e04 +#define CB_COLOR8_BASE 0x28e40 +#define CB_COLOR9_BASE 0x28e5c +#define CB_COLOR10_BASE 0x28e78 +#define CB_COLOR11_BASE 0x28e94 + +#define CB_COLOR1_PITCH 0x28ca0 +#define CB_COLOR2_PITCH 0x28cdc +#define CB_COLOR3_PITCH 0x28d18 +#define CB_COLOR4_PITCH 0x28d54 +#define CB_COLOR5_PITCH 0x28d90 +#define CB_COLOR6_PITCH 0x28dcc +#define CB_COLOR7_PITCH 0x28e08 +#define CB_COLOR8_PITCH 0x28e44 +#define CB_COLOR9_PITCH 0x28e60 +#define CB_COLOR10_PITCH 0x28e7c +#define CB_COLOR11_PITCH 0x28e98 + +#define CB_COLOR1_SLICE 0x28ca4 +#define CB_COLOR2_SLICE 0x28ce0 +#define CB_COLOR3_SLICE 0x28d1c +#define CB_COLOR4_SLICE 0x28d58 +#define CB_COLOR5_SLICE 0x28d94 +#define CB_COLOR6_SLICE 0x28dd0 +#define CB_COLOR7_SLICE 0x28e0c +#define CB_COLOR8_SLICE 0x28e48 +#define CB_COLOR9_SLICE 0x28e64 +#define CB_COLOR10_SLICE 0x28e80 +#define CB_COLOR11_SLICE 0x28e9c + +#define CB_COLOR1_VIEW 0x28ca8 +#define CB_COLOR2_VIEW 0x28ce4 +#define CB_COLOR3_VIEW 0x28d20 +#define CB_COLOR4_VIEW 0x28d5c +#define CB_COLOR5_VIEW 0x28d98 +#define CB_COLOR6_VIEW 0x28dd4 +#define CB_COLOR7_VIEW 0x28e10 +#define CB_COLOR8_VIEW 0x28e4c +#define CB_COLOR9_VIEW 0x28e68 +#define CB_COLOR10_VIEW 0x28e84 +#define CB_COLOR11_VIEW 0x28ea0 + +#define CB_COLOR1_INFO 0x28cac +#define CB_COLOR2_INFO 0x28ce8 +#define CB_COLOR3_INFO 0x28d24 +#define CB_COLOR4_INFO 0x28d60 +#define CB_COLOR5_INFO 0x28d9c +#define CB_COLOR6_INFO 0x28dd8 +#define CB_COLOR7_INFO 0x28e14 +#define CB_COLOR8_INFO 0x28e50 +#define CB_COLOR9_INFO 0x28e6c +#define CB_COLOR10_INFO 0x28e88 +#define CB_COLOR11_INFO 0x28ea4 + +#define CB_COLOR1_ATTRIB 0x28cb0 +#define CB_COLOR2_ATTRIB 0x28cec +#define CB_COLOR3_ATTRIB 0x28d28 +#define CB_COLOR4_ATTRIB 0x28d64 +#define CB_COLOR5_ATTRIB 0x28da0 +#define CB_COLOR6_ATTRIB 0x28ddc +#define CB_COLOR7_ATTRIB 0x28e18 +#define CB_COLOR8_ATTRIB 0x28e54 +#define CB_COLOR9_ATTRIB 0x28e70 +#define CB_COLOR10_ATTRIB 0x28e8c +#define CB_COLOR11_ATTRIB 0x28ea8 + +#define CB_COLOR1_DIM 0x28cb4 +#define CB_COLOR2_DIM 0x28cf0 +#define CB_COLOR3_DIM 0x28d2c +#define CB_COLOR4_DIM 0x28d68 +#define CB_COLOR5_DIM 0x28da4 +#define CB_COLOR6_DIM 0x28de0 +#define CB_COLOR7_DIM 0x28e1c +#define CB_COLOR8_DIM 0x28e58 +#define CB_COLOR9_DIM 0x28e74 +#define CB_COLOR10_DIM 0x28e90 +#define CB_COLOR11_DIM 0x28eac + +#define CB_COLOR1_CMASK 0x28cb8 +#define CB_COLOR2_CMASK 0x28cf4 +#define CB_COLOR3_CMASK 0x28d30 +#define CB_COLOR4_CMASK 0x28d6c +#define CB_COLOR5_CMASK 0x28da8 +#define CB_COLOR6_CMASK 0x28de4 +#define CB_COLOR7_CMASK 0x28e20 + +#define CB_COLOR1_CMASK_SLICE 0x28cbc +#define CB_COLOR2_CMASK_SLICE 0x28cf8 +#define CB_COLOR3_CMASK_SLICE 0x28d34 +#define CB_COLOR4_CMASK_SLICE 0x28d70 +#define CB_COLOR5_CMASK_SLICE 0x28dac +#define CB_COLOR6_CMASK_SLICE 0x28de8 +#define CB_COLOR7_CMASK_SLICE 0x28e24 + +#define CB_COLOR1_FMASK 0x28cc0 +#define CB_COLOR2_FMASK 0x28cfc +#define CB_COLOR3_FMASK 0x28d38 +#define CB_COLOR4_FMASK 0x28d74 +#define CB_COLOR5_FMASK 0x28db0 +#define CB_COLOR6_FMASK 0x28dec +#define CB_COLOR7_FMASK 0x28e28 + +#define CB_COLOR1_FMASK_SLICE 0x28cc4 +#define CB_COLOR2_FMASK_SLICE 0x28d00 +#define CB_COLOR3_FMASK_SLICE 0x28d3c +#define CB_COLOR4_FMASK_SLICE 0x28d78 +#define CB_COLOR5_FMASK_SLICE 0x28db4 +#define CB_COLOR6_FMASK_SLICE 0x28df0 +#define CB_COLOR7_FMASK_SLICE 0x28e2c + +#define CB_COLOR1_CLEAR_WORD0 0x28cc8 +#define CB_COLOR2_CLEAR_WORD0 0x28d04 +#define CB_COLOR3_CLEAR_WORD0 0x28d40 +#define CB_COLOR4_CLEAR_WORD0 0x28d7c +#define CB_COLOR5_CLEAR_WORD0 0x28db8 +#define CB_COLOR6_CLEAR_WORD0 0x28df4 +#define CB_COLOR7_CLEAR_WORD0 0x28e30 + +#define CB_COLOR1_CLEAR_WORD1 0x28ccc +#define CB_COLOR2_CLEAR_WORD1 0x28d08 +#define CB_COLOR3_CLEAR_WORD1 0x28d44 +#define CB_COLOR4_CLEAR_WORD1 0x28d80 +#define CB_COLOR5_CLEAR_WORD1 0x28dbc +#define CB_COLOR6_CLEAR_WORD1 0x28df8 +#define CB_COLOR7_CLEAR_WORD1 0x28e34 + +#define CB_COLOR1_CLEAR_WORD2 0x28cd0 +#define CB_COLOR2_CLEAR_WORD2 0x28d0c +#define CB_COLOR3_CLEAR_WORD2 0x28d48 +#define CB_COLOR4_CLEAR_WORD2 0x28d84 +#define CB_COLOR5_CLEAR_WORD2 0x28dc0 +#define CB_COLOR6_CLEAR_WORD2 0x28dfc +#define CB_COLOR7_CLEAR_WORD2 0x28e38 + +#define CB_COLOR1_CLEAR_WORD3 0x28cd4 +#define CB_COLOR2_CLEAR_WORD3 0x28d10 +#define CB_COLOR3_CLEAR_WORD3 0x28d4c +#define CB_COLOR4_CLEAR_WORD3 0x28d88 +#define CB_COLOR5_CLEAR_WORD3 0x28dc4 +#define CB_COLOR6_CLEAR_WORD3 0x28e00 +#define CB_COLOR7_CLEAR_WORD3 0x28e3c + +#define SQ_TEX_RESOURCE_WORD0_0 0x30000 +#define SQ_TEX_RESOURCE_WORD1_0 0x30004 +# define TEX_ARRAY_MODE(x) ((x) << 28) +#define SQ_TEX_RESOURCE_WORD2_0 0x30008 +#define SQ_TEX_RESOURCE_WORD3_0 0x3000C +#define SQ_TEX_RESOURCE_WORD4_0 0x30010 +#define SQ_TEX_RESOURCE_WORD5_0 0x30014 +#define SQ_TEX_RESOURCE_WORD6_0 0x30018 +#define SQ_TEX_RESOURCE_WORD7_0 0x3001c + + +#endif diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index cf60c0b3ef15..a89a15ab524d 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -37,6 +37,7 @@ #include "rs100d.h" #include "rv200d.h" #include "rv250d.h" +#include "atom.h" #include <linux/firmware.h> #include <linux/platform_device.h> @@ -67,6 +68,274 @@ MODULE_FIRMWARE(FIRMWARE_R520); * r100,rv100,rs100,rv200,rs200,r200,rv250,rs300,rv280 */ +void r100_pm_get_dynpm_state(struct radeon_device *rdev) +{ + int i; + rdev->pm.dynpm_can_upclock = true; + rdev->pm.dynpm_can_downclock = true; + + switch (rdev->pm.dynpm_planned_action) { + case DYNPM_ACTION_MINIMUM: + rdev->pm.requested_power_state_index = 0; + rdev->pm.dynpm_can_downclock = false; + break; + case DYNPM_ACTION_DOWNCLOCK: + if (rdev->pm.current_power_state_index == 0) { + rdev->pm.requested_power_state_index = rdev->pm.current_power_state_index; + rdev->pm.dynpm_can_downclock = false; + } else { + if (rdev->pm.active_crtc_count > 1) { + for (i = 0; i < rdev->pm.num_power_states; i++) { + if (rdev->pm.power_state[i].flags & RADEON_PM_STATE_SINGLE_DISPLAY_ONLY) + continue; + else if (i >= rdev->pm.current_power_state_index) { + rdev->pm.requested_power_state_index = rdev->pm.current_power_state_index; + break; + } else { + rdev->pm.requested_power_state_index = i; + break; + } + } + } else + rdev->pm.requested_power_state_index = + rdev->pm.current_power_state_index - 1; + } + /* don't use the power state if crtcs are active and no display flag is set */ + if ((rdev->pm.active_crtc_count > 0) && + (rdev->pm.power_state[rdev->pm.requested_power_state_index].clock_info[0].flags & + RADEON_PM_MODE_NO_DISPLAY)) { + rdev->pm.requested_power_state_index++; + } + break; + case DYNPM_ACTION_UPCLOCK: + if (rdev->pm.current_power_state_index == (rdev->pm.num_power_states - 1)) { + rdev->pm.requested_power_state_index = rdev->pm.current_power_state_index; + rdev->pm.dynpm_can_upclock = false; + } else { + if (rdev->pm.active_crtc_count > 1) { + for (i = (rdev->pm.num_power_states - 1); i >= 0; i--) { + if (rdev->pm.power_state[i].flags & RADEON_PM_STATE_SINGLE_DISPLAY_ONLY) + continue; + else if (i <= rdev->pm.current_power_state_index) { + rdev->pm.requested_power_state_index = rdev->pm.current_power_state_index; + break; + } else { + rdev->pm.requested_power_state_index = i; + break; + } + } + } else + rdev->pm.requested_power_state_index = + rdev->pm.current_power_state_index + 1; + } + break; + case DYNPM_ACTION_DEFAULT: + rdev->pm.requested_power_state_index = rdev->pm.default_power_state_index; + rdev->pm.dynpm_can_upclock = false; + break; + case DYNPM_ACTION_NONE: + default: + DRM_ERROR("Requested mode for not defined action\n"); + return; + } + /* only one clock mode per power state */ + rdev->pm.requested_clock_mode_index = 0; + + DRM_DEBUG("Requested: e: %d m: %d p: %d\n", + rdev->pm.power_state[rdev->pm.requested_power_state_index]. + clock_info[rdev->pm.requested_clock_mode_index].sclk, + rdev->pm.power_state[rdev->pm.requested_power_state_index]. + clock_info[rdev->pm.requested_clock_mode_index].mclk, + rdev->pm.power_state[rdev->pm.requested_power_state_index]. + pcie_lanes); +} + +void r100_pm_init_profile(struct radeon_device *rdev) +{ + /* default */ + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_cm_idx = 0; + /* low sh */ + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_cm_idx = 0; + /* mid sh */ + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_cm_idx = 0; + /* high sh */ + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_cm_idx = 0; + /* low mh */ + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_cm_idx = 0; + /* mid mh */ + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_cm_idx = 0; + /* high mh */ + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_cm_idx = 0; +} + +void r100_pm_misc(struct radeon_device *rdev) +{ + int requested_index = rdev->pm.requested_power_state_index; + struct radeon_power_state *ps = &rdev->pm.power_state[requested_index]; + struct radeon_voltage *voltage = &ps->clock_info[0].voltage; + u32 tmp, sclk_cntl, sclk_cntl2, sclk_more_cntl; + + if ((voltage->type == VOLTAGE_GPIO) && (voltage->gpio.valid)) { + if (ps->misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_SUPPORT) { + tmp = RREG32(voltage->gpio.reg); + if (voltage->active_high) + tmp |= voltage->gpio.mask; + else + tmp &= ~(voltage->gpio.mask); + WREG32(voltage->gpio.reg, tmp); + if (voltage->delay) + udelay(voltage->delay); + } else { + tmp = RREG32(voltage->gpio.reg); + if (voltage->active_high) + tmp &= ~voltage->gpio.mask; + else + tmp |= voltage->gpio.mask; + WREG32(voltage->gpio.reg, tmp); + if (voltage->delay) + udelay(voltage->delay); + } + } + + sclk_cntl = RREG32_PLL(SCLK_CNTL); + sclk_cntl2 = RREG32_PLL(SCLK_CNTL2); + sclk_cntl2 &= ~REDUCED_SPEED_SCLK_SEL(3); + sclk_more_cntl = RREG32_PLL(SCLK_MORE_CNTL); + sclk_more_cntl &= ~VOLTAGE_DELAY_SEL(3); + if (ps->misc & ATOM_PM_MISCINFO_ASIC_REDUCED_SPEED_SCLK_EN) { + sclk_more_cntl |= REDUCED_SPEED_SCLK_EN; + if (ps->misc & ATOM_PM_MISCINFO_DYN_CLK_3D_IDLE) + sclk_cntl2 |= REDUCED_SPEED_SCLK_MODE; + else + sclk_cntl2 &= ~REDUCED_SPEED_SCLK_MODE; + if (ps->misc & ATOM_PM_MISCINFO_DYNAMIC_CLOCK_DIVIDER_BY_2) + sclk_cntl2 |= REDUCED_SPEED_SCLK_SEL(0); + else if (ps->misc & ATOM_PM_MISCINFO_DYNAMIC_CLOCK_DIVIDER_BY_4) + sclk_cntl2 |= REDUCED_SPEED_SCLK_SEL(2); + } else + sclk_more_cntl &= ~REDUCED_SPEED_SCLK_EN; + + if (ps->misc & ATOM_PM_MISCINFO_ASIC_DYNAMIC_VOLTAGE_EN) { + sclk_more_cntl |= IO_CG_VOLTAGE_DROP; + if (voltage->delay) { + sclk_more_cntl |= VOLTAGE_DROP_SYNC; + switch (voltage->delay) { + case 33: + sclk_more_cntl |= VOLTAGE_DELAY_SEL(0); + break; + case 66: + sclk_more_cntl |= VOLTAGE_DELAY_SEL(1); + break; + case 99: + sclk_more_cntl |= VOLTAGE_DELAY_SEL(2); + break; + case 132: + sclk_more_cntl |= VOLTAGE_DELAY_SEL(3); + break; + } + } else + sclk_more_cntl &= ~VOLTAGE_DROP_SYNC; + } else + sclk_more_cntl &= ~IO_CG_VOLTAGE_DROP; + + if (ps->misc & ATOM_PM_MISCINFO_DYNAMIC_HDP_BLOCK_EN) + sclk_cntl &= ~FORCE_HDP; + else + sclk_cntl |= FORCE_HDP; + + WREG32_PLL(SCLK_CNTL, sclk_cntl); + WREG32_PLL(SCLK_CNTL2, sclk_cntl2); + WREG32_PLL(SCLK_MORE_CNTL, sclk_more_cntl); + + /* set pcie lanes */ + if ((rdev->flags & RADEON_IS_PCIE) && + !(rdev->flags & RADEON_IS_IGP) && + rdev->asic->set_pcie_lanes && + (ps->pcie_lanes != + rdev->pm.power_state[rdev->pm.current_power_state_index].pcie_lanes)) { + radeon_set_pcie_lanes(rdev, + ps->pcie_lanes); + DRM_DEBUG("Setting: p: %d\n", ps->pcie_lanes); + } +} + +void r100_pm_prepare(struct radeon_device *rdev) +{ + struct drm_device *ddev = rdev->ddev; + struct drm_crtc *crtc; + struct radeon_crtc *radeon_crtc; + u32 tmp; + + /* disable any active CRTCs */ + list_for_each_entry(crtc, &ddev->mode_config.crtc_list, head) { + radeon_crtc = to_radeon_crtc(crtc); + if (radeon_crtc->enabled) { + if (radeon_crtc->crtc_id) { + tmp = RREG32(RADEON_CRTC2_GEN_CNTL); + tmp |= RADEON_CRTC2_DISP_REQ_EN_B; + WREG32(RADEON_CRTC2_GEN_CNTL, tmp); + } else { + tmp = RREG32(RADEON_CRTC_GEN_CNTL); + tmp |= RADEON_CRTC_DISP_REQ_EN_B; + WREG32(RADEON_CRTC_GEN_CNTL, tmp); + } + } + } +} + +void r100_pm_finish(struct radeon_device *rdev) +{ + struct drm_device *ddev = rdev->ddev; + struct drm_crtc *crtc; + struct radeon_crtc *radeon_crtc; + u32 tmp; + + /* enable any active CRTCs */ + list_for_each_entry(crtc, &ddev->mode_config.crtc_list, head) { + radeon_crtc = to_radeon_crtc(crtc); + if (radeon_crtc->enabled) { + if (radeon_crtc->crtc_id) { + tmp = RREG32(RADEON_CRTC2_GEN_CNTL); + tmp &= ~RADEON_CRTC2_DISP_REQ_EN_B; + WREG32(RADEON_CRTC2_GEN_CNTL, tmp); + } else { + tmp = RREG32(RADEON_CRTC_GEN_CNTL); + tmp &= ~RADEON_CRTC_DISP_REQ_EN_B; + WREG32(RADEON_CRTC_GEN_CNTL, tmp); + } + } + } +} + +bool r100_gui_idle(struct radeon_device *rdev) +{ + if (RREG32(RADEON_RBBM_STATUS) & RADEON_RBBM_ACTIVE) + return false; + else + return true; +} + /* hpd for digital panel detect/disconnect */ bool r100_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd) { @@ -254,6 +523,9 @@ int r100_irq_set(struct radeon_device *rdev) if (rdev->irq.sw_int) { tmp |= RADEON_SW_INT_ENABLE; } + if (rdev->irq.gui_idle) { + tmp |= RADEON_GUI_IDLE_MASK; + } if (rdev->irq.crtc_vblank_int[0]) { tmp |= RADEON_CRTC_VBLANK_MASK; } @@ -288,6 +560,12 @@ static inline uint32_t r100_irq_ack(struct radeon_device *rdev) RADEON_CRTC_VBLANK_STAT | RADEON_CRTC2_VBLANK_STAT | RADEON_FP_DETECT_STAT | RADEON_FP2_DETECT_STAT; + /* the interrupt works, but the status bit is permanently asserted */ + if (rdev->irq.gui_idle && radeon_gui_idle(rdev)) { + if (!rdev->irq.gui_idle_acked) + irq_mask |= RADEON_GUI_IDLE_STAT; + } + if (irqs) { WREG32(RADEON_GEN_INT_STATUS, irqs); } @@ -299,6 +577,9 @@ int r100_irq_process(struct radeon_device *rdev) uint32_t status, msi_rearm; bool queue_hotplug = false; + /* reset gui idle ack. the status bit is broken */ + rdev->irq.gui_idle_acked = false; + status = r100_irq_ack(rdev); if (!status) { return IRQ_NONE; @@ -311,6 +592,12 @@ int r100_irq_process(struct radeon_device *rdev) if (status & RADEON_SW_INT_TEST) { radeon_fence_process(rdev); } + /* gui idle interrupt */ + if (status & RADEON_GUI_IDLE_STAT) { + rdev->irq.gui_idle_acked = true; + rdev->pm.gui_idle = true; + wake_up(&rdev->irq.idle_queue); + } /* Vertical blank interrupts */ if (status & RADEON_CRTC_VBLANK_STAT) { drm_handle_vblank(rdev->ddev, 0); @@ -332,6 +619,8 @@ int r100_irq_process(struct radeon_device *rdev) } status = r100_irq_ack(rdev); } + /* reset gui idle ack. the status bit is broken */ + rdev->irq.gui_idle_acked = false; if (queue_hotplug) queue_work(rdev->wq, &rdev->hotplug_work); if (rdev->msi_enabled) { @@ -663,26 +952,6 @@ int r100_cp_init(struct radeon_device *rdev, unsigned ring_size) if (r100_debugfs_cp_init(rdev)) { DRM_ERROR("Failed to register debugfs file for CP !\n"); } - /* Reset CP */ - tmp = RREG32(RADEON_CP_CSQ_STAT); - if ((tmp & (1 << 31))) { - DRM_INFO("radeon: cp busy (0x%08X) resetting\n", tmp); - WREG32(RADEON_CP_CSQ_MODE, 0); - WREG32(RADEON_CP_CSQ_CNTL, 0); - WREG32(RADEON_RBBM_SOFT_RESET, RADEON_SOFT_RESET_CP); - tmp = RREG32(RADEON_RBBM_SOFT_RESET); - mdelay(2); - WREG32(RADEON_RBBM_SOFT_RESET, 0); - tmp = RREG32(RADEON_RBBM_SOFT_RESET); - mdelay(2); - tmp = RREG32(RADEON_CP_CSQ_STAT); - if ((tmp & (1 << 31))) { - DRM_INFO("radeon: cp reset failed (0x%08X)\n", tmp); - } - } else { - DRM_INFO("radeon: cp idle (0x%08X)\n", tmp); - } - if (!rdev->me_fw) { r = r100_cp_init_microcode(rdev); if (r) { @@ -787,39 +1056,6 @@ void r100_cp_disable(struct radeon_device *rdev) } } -int r100_cp_reset(struct radeon_device *rdev) -{ - uint32_t tmp; - bool reinit_cp; - int i; - - reinit_cp = rdev->cp.ready; - rdev->cp.ready = false; - WREG32(RADEON_CP_CSQ_MODE, 0); - WREG32(RADEON_CP_CSQ_CNTL, 0); - WREG32(RADEON_RBBM_SOFT_RESET, RADEON_SOFT_RESET_CP); - (void)RREG32(RADEON_RBBM_SOFT_RESET); - udelay(200); - WREG32(RADEON_RBBM_SOFT_RESET, 0); - /* Wait to prevent race in RBBM_STATUS */ - mdelay(1); - for (i = 0; i < rdev->usec_timeout; i++) { - tmp = RREG32(RADEON_RBBM_STATUS); - if (!(tmp & (1 << 16))) { - DRM_INFO("CP reset succeed (RBBM_STATUS=0x%08X)\n", - tmp); - if (reinit_cp) { - return r100_cp_init(rdev, rdev->cp.ring_size); - } - return 0; - } - DRM_UDELAY(1); - } - tmp = RREG32(RADEON_RBBM_STATUS); - DRM_ERROR("Failed to reset CP (RBBM_STATUS=0x%08X)!\n", tmp); - return -1; -} - void r100_cp_commit(struct radeon_device *rdev) { WREG32(RADEON_CP_RB_WPTR, rdev->cp.wptr); @@ -994,7 +1230,6 @@ int r100_cs_packet_parse_vline(struct radeon_cs_parser *p) header = radeon_get_ib_value(p, h_idx); crtc_id = radeon_get_ib_value(p, h_idx + 5); reg = CP_PACKET0_GET_REG(header); - mutex_lock(&p->rdev->ddev->mode_config.mutex); obj = drm_mode_object_find(p->rdev->ddev, crtc_id, DRM_MODE_OBJECT_CRTC); if (!obj) { DRM_ERROR("cannot find crtc %d\n", crtc_id); @@ -1028,7 +1263,6 @@ int r100_cs_packet_parse_vline(struct radeon_cs_parser *p) ib[h_idx + 3] |= RADEON_ENG_DISPLAY_SELECT_CRTC1; } out: - mutex_unlock(&p->rdev->ddev->mode_config.mutex); return r; } @@ -1392,6 +1626,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, case RADEON_TXFORMAT_RGB332: case RADEON_TXFORMAT_Y8: track->textures[i].cpp = 1; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case RADEON_TXFORMAT_AI88: case RADEON_TXFORMAT_ARGB1555: @@ -1403,12 +1638,14 @@ static int r100_packet0_check(struct radeon_cs_parser *p, case RADEON_TXFORMAT_LDUDV655: case RADEON_TXFORMAT_DUDV88: track->textures[i].cpp = 2; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case RADEON_TXFORMAT_ARGB8888: case RADEON_TXFORMAT_RGBA8888: case RADEON_TXFORMAT_SHADOW32: case RADEON_TXFORMAT_LDUDUV8888: track->textures[i].cpp = 4; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case RADEON_TXFORMAT_DXT1: track->textures[i].cpp = 1; @@ -1733,76 +1970,163 @@ int r100_mc_wait_for_idle(struct radeon_device *rdev) return -1; } -void r100_gpu_init(struct radeon_device *rdev) +void r100_gpu_lockup_update(struct r100_gpu_lockup *lockup, struct radeon_cp *cp) { - /* TODO: anythings to do here ? pipes ? */ - r100_hdp_reset(rdev); + lockup->last_cp_rptr = cp->rptr; + lockup->last_jiffies = jiffies; +} + +/** + * r100_gpu_cp_is_lockup() - check if CP is lockup by recording information + * @rdev: radeon device structure + * @lockup: r100_gpu_lockup structure holding CP lockup tracking informations + * @cp: radeon_cp structure holding CP information + * + * We don't need to initialize the lockup tracking information as we will either + * have CP rptr to a different value of jiffies wrap around which will force + * initialization of the lockup tracking informations. + * + * A possible false positivie is if we get call after while and last_cp_rptr == + * the current CP rptr, even if it's unlikely it might happen. To avoid this + * if the elapsed time since last call is bigger than 2 second than we return + * false and update the tracking information. Due to this the caller must call + * r100_gpu_cp_is_lockup several time in less than 2sec for lockup to be reported + * the fencing code should be cautious about that. + * + * Caller should write to the ring to force CP to do something so we don't get + * false positive when CP is just gived nothing to do. + * + **/ +bool r100_gpu_cp_is_lockup(struct radeon_device *rdev, struct r100_gpu_lockup *lockup, struct radeon_cp *cp) +{ + unsigned long cjiffies, elapsed; + + cjiffies = jiffies; + if (!time_after(cjiffies, lockup->last_jiffies)) { + /* likely a wrap around */ + lockup->last_cp_rptr = cp->rptr; + lockup->last_jiffies = jiffies; + return false; + } + if (cp->rptr != lockup->last_cp_rptr) { + /* CP is still working no lockup */ + lockup->last_cp_rptr = cp->rptr; + lockup->last_jiffies = jiffies; + return false; + } + elapsed = jiffies_to_msecs(cjiffies - lockup->last_jiffies); + if (elapsed >= 3000) { + /* very likely the improbable case where current + * rptr is equal to last recorded, a while ago, rptr + * this is more likely a false positive update tracking + * information which should force us to be recall at + * latter point + */ + lockup->last_cp_rptr = cp->rptr; + lockup->last_jiffies = jiffies; + return false; + } + if (elapsed >= 1000) { + dev_err(rdev->dev, "GPU lockup CP stall for more than %lumsec\n", elapsed); + return true; + } + /* give a chance to the GPU ... */ + return false; } -void r100_hdp_reset(struct radeon_device *rdev) +bool r100_gpu_is_lockup(struct radeon_device *rdev) { - uint32_t tmp; + u32 rbbm_status; + int r; - tmp = RREG32(RADEON_HOST_PATH_CNTL) & RADEON_HDP_APER_CNTL; - tmp |= (7 << 28); - WREG32(RADEON_HOST_PATH_CNTL, tmp | RADEON_HDP_SOFT_RESET | RADEON_HDP_READ_BUFFER_INVALIDATE); - (void)RREG32(RADEON_HOST_PATH_CNTL); - udelay(200); - WREG32(RADEON_RBBM_SOFT_RESET, 0); - WREG32(RADEON_HOST_PATH_CNTL, tmp); - (void)RREG32(RADEON_HOST_PATH_CNTL); + rbbm_status = RREG32(R_000E40_RBBM_STATUS); + if (!G_000E40_GUI_ACTIVE(rbbm_status)) { + r100_gpu_lockup_update(&rdev->config.r100.lockup, &rdev->cp); + return false; + } + /* force CP activities */ + r = radeon_ring_lock(rdev, 2); + if (!r) { + /* PACKET2 NOP */ + radeon_ring_write(rdev, 0x80000000); + radeon_ring_write(rdev, 0x80000000); + radeon_ring_unlock_commit(rdev); + } + rdev->cp.rptr = RREG32(RADEON_CP_RB_RPTR); + return r100_gpu_cp_is_lockup(rdev, &rdev->config.r100.lockup, &rdev->cp); } -int r100_rb2d_reset(struct radeon_device *rdev) +void r100_bm_disable(struct radeon_device *rdev) { - uint32_t tmp; - int i; + u32 tmp; - WREG32(RADEON_RBBM_SOFT_RESET, RADEON_SOFT_RESET_E2); - (void)RREG32(RADEON_RBBM_SOFT_RESET); - udelay(200); - WREG32(RADEON_RBBM_SOFT_RESET, 0); - /* Wait to prevent race in RBBM_STATUS */ + /* disable bus mastering */ + tmp = RREG32(R_000030_BUS_CNTL); + WREG32(R_000030_BUS_CNTL, (tmp & 0xFFFFFFFF) | 0x00000044); + mdelay(1); + WREG32(R_000030_BUS_CNTL, (tmp & 0xFFFFFFFF) | 0x00000042); + mdelay(1); + WREG32(R_000030_BUS_CNTL, (tmp & 0xFFFFFFFF) | 0x00000040); + tmp = RREG32(RADEON_BUS_CNTL); + mdelay(1); + pci_read_config_word(rdev->pdev, 0x4, (u16*)&tmp); + pci_write_config_word(rdev->pdev, 0x4, tmp & 0xFFFB); mdelay(1); - for (i = 0; i < rdev->usec_timeout; i++) { - tmp = RREG32(RADEON_RBBM_STATUS); - if (!(tmp & (1 << 26))) { - DRM_INFO("RB2D reset succeed (RBBM_STATUS=0x%08X)\n", - tmp); - return 0; - } - DRM_UDELAY(1); - } - tmp = RREG32(RADEON_RBBM_STATUS); - DRM_ERROR("Failed to reset RB2D (RBBM_STATUS=0x%08X)!\n", tmp); - return -1; } -int r100_gpu_reset(struct radeon_device *rdev) +int r100_asic_reset(struct radeon_device *rdev) { - uint32_t status; + struct r100_mc_save save; + u32 status, tmp; - /* reset order likely matter */ - status = RREG32(RADEON_RBBM_STATUS); - /* reset HDP */ - r100_hdp_reset(rdev); - /* reset rb2d */ - if (status & ((1 << 17) | (1 << 18) | (1 << 27))) { - r100_rb2d_reset(rdev); + r100_mc_stop(rdev, &save); + status = RREG32(R_000E40_RBBM_STATUS); + if (!G_000E40_GUI_ACTIVE(status)) { + return 0; } - /* TODO: reset 3D engine */ + status = RREG32(R_000E40_RBBM_STATUS); + dev_info(rdev->dev, "(%s:%d) RBBM_STATUS=0x%08X\n", __func__, __LINE__, status); + /* stop CP */ + WREG32(RADEON_CP_CSQ_CNTL, 0); + tmp = RREG32(RADEON_CP_RB_CNTL); + WREG32(RADEON_CP_RB_CNTL, tmp | RADEON_RB_RPTR_WR_ENA); + WREG32(RADEON_CP_RB_RPTR_WR, 0); + WREG32(RADEON_CP_RB_WPTR, 0); + WREG32(RADEON_CP_RB_CNTL, tmp); + /* save PCI state */ + pci_save_state(rdev->pdev); + /* disable bus mastering */ + r100_bm_disable(rdev); + WREG32(R_0000F0_RBBM_SOFT_RESET, S_0000F0_SOFT_RESET_SE(1) | + S_0000F0_SOFT_RESET_RE(1) | + S_0000F0_SOFT_RESET_PP(1) | + S_0000F0_SOFT_RESET_RB(1)); + RREG32(R_0000F0_RBBM_SOFT_RESET); + mdelay(500); + WREG32(R_0000F0_RBBM_SOFT_RESET, 0); + mdelay(1); + status = RREG32(R_000E40_RBBM_STATUS); + dev_info(rdev->dev, "(%s:%d) RBBM_STATUS=0x%08X\n", __func__, __LINE__, status); /* reset CP */ - status = RREG32(RADEON_RBBM_STATUS); - if (status & (1 << 16)) { - r100_cp_reset(rdev); - } + WREG32(R_0000F0_RBBM_SOFT_RESET, S_0000F0_SOFT_RESET_CP(1)); + RREG32(R_0000F0_RBBM_SOFT_RESET); + mdelay(500); + WREG32(R_0000F0_RBBM_SOFT_RESET, 0); + mdelay(1); + status = RREG32(R_000E40_RBBM_STATUS); + dev_info(rdev->dev, "(%s:%d) RBBM_STATUS=0x%08X\n", __func__, __LINE__, status); + /* restore PCI & busmastering */ + pci_restore_state(rdev->pdev); + r100_enable_bm(rdev); /* Check if GPU is idle */ - status = RREG32(RADEON_RBBM_STATUS); - if (status & RADEON_RBBM_ACTIVE) { - DRM_ERROR("Failed to reset GPU (RBBM_STATUS=0x%08X)\n", status); + if (G_000E40_SE_BUSY(status) || G_000E40_RE_BUSY(status) || + G_000E40_TAM_BUSY(status) || G_000E40_PB_BUSY(status)) { + dev_err(rdev->dev, "failed to reset GPU\n"); + rdev->gpu_lockup = true; return -1; } - DRM_INFO("GPU reset succeed (RBBM_STATUS=0x%08X)\n", status); + r100_mc_resume(rdev, &save); + dev_info(rdev->dev, "GPU reset succeed\n"); return 0; } @@ -2002,11 +2326,6 @@ void r100_vram_init_sizes(struct radeon_device *rdev) else rdev->mc.mc_vram_size = rdev->mc.real_vram_size; } - /* FIXME remove this once we support unmappable VRAM */ - if (rdev->mc.mc_vram_size > rdev->mc.aper_size) { - rdev->mc.mc_vram_size = rdev->mc.aper_size; - rdev->mc.real_vram_size = rdev->mc.aper_size; - } } void r100_vga_set_state(struct radeon_device *rdev, bool state) @@ -2033,6 +2352,7 @@ void r100_mc_init(struct radeon_device *rdev) if (rdev->flags & RADEON_IS_IGP) base = (RREG32(RADEON_NB_TOM) & 0xffff) << 16; radeon_vram_location(rdev, &rdev->mc, base); + rdev->mc.gtt_base_align = 0; if (!(rdev->flags & RADEON_IS_AGP)) radeon_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); @@ -2286,12 +2606,6 @@ int r100_set_surface_reg(struct radeon_device *rdev, int reg, int surf_index = reg * 16; int flags = 0; - /* r100/r200 divide by 16 */ - if (rdev->family < CHIP_R300) - flags = pitch / 16; - else - flags = pitch / 8; - if (rdev->family <= CHIP_RS200) { if ((tiling_flags & (RADEON_TILING_MACRO|RADEON_TILING_MICRO)) == (RADEON_TILING_MACRO|RADEON_TILING_MICRO)) @@ -2315,6 +2629,20 @@ int r100_set_surface_reg(struct radeon_device *rdev, int reg, if (tiling_flags & RADEON_TILING_SWAP_32BIT) flags |= RADEON_SURF_AP0_SWP_32BPP | RADEON_SURF_AP1_SWP_32BPP; + /* when we aren't tiling the pitch seems to needs to be furtherdivided down. - tested on power5 + rn50 server */ + if (tiling_flags & (RADEON_TILING_SWAP_16BIT | RADEON_TILING_SWAP_32BIT)) { + if (!(tiling_flags & (RADEON_TILING_MACRO | RADEON_TILING_MICRO))) + if (ASIC_IS_RN50(rdev)) + pitch /= 16; + } + + /* r100/r200 divide by 16 */ + if (rdev->family < CHIP_R300) + flags |= pitch / 16; + else + flags |= pitch / 8; + + DRM_DEBUG("writing surface %d %d %x %x\n", reg, flags, offset, offset+obj_size-1); WREG32(RADEON_SURFACE0_INFO + surf_index, flags); WREG32(RADEON_SURFACE0_LOWER_BOUND + surf_index, offset); @@ -2335,53 +2663,53 @@ void r100_bandwidth_update(struct radeon_device *rdev) fixed20_12 peak_disp_bw, mem_bw, pix_clk, pix_clk2, temp_ff, crit_point_ff; uint32_t temp, data, mem_trcd, mem_trp, mem_tras; fixed20_12 memtcas_ff[8] = { - fixed_init(1), - fixed_init(2), - fixed_init(3), - fixed_init(0), - fixed_init_half(1), - fixed_init_half(2), - fixed_init(0), + dfixed_init(1), + dfixed_init(2), + dfixed_init(3), + dfixed_init(0), + dfixed_init_half(1), + dfixed_init_half(2), + dfixed_init(0), }; fixed20_12 memtcas_rs480_ff[8] = { - fixed_init(0), - fixed_init(1), - fixed_init(2), - fixed_init(3), - fixed_init(0), - fixed_init_half(1), - fixed_init_half(2), - fixed_init_half(3), + dfixed_init(0), + dfixed_init(1), + dfixed_init(2), + dfixed_init(3), + dfixed_init(0), + dfixed_init_half(1), + dfixed_init_half(2), + dfixed_init_half(3), }; fixed20_12 memtcas2_ff[8] = { - fixed_init(0), - fixed_init(1), - fixed_init(2), - fixed_init(3), - fixed_init(4), - fixed_init(5), - fixed_init(6), - fixed_init(7), + dfixed_init(0), + dfixed_init(1), + dfixed_init(2), + dfixed_init(3), + dfixed_init(4), + dfixed_init(5), + dfixed_init(6), + dfixed_init(7), }; fixed20_12 memtrbs[8] = { - fixed_init(1), - fixed_init_half(1), - fixed_init(2), - fixed_init_half(2), - fixed_init(3), - fixed_init_half(3), - fixed_init(4), - fixed_init_half(4) + dfixed_init(1), + dfixed_init_half(1), + dfixed_init(2), + dfixed_init_half(2), + dfixed_init(3), + dfixed_init_half(3), + dfixed_init(4), + dfixed_init_half(4) }; fixed20_12 memtrbs_r4xx[8] = { - fixed_init(4), - fixed_init(5), - fixed_init(6), - fixed_init(7), - fixed_init(8), - fixed_init(9), - fixed_init(10), - fixed_init(11) + dfixed_init(4), + dfixed_init(5), + dfixed_init(6), + dfixed_init(7), + dfixed_init(8), + dfixed_init(9), + dfixed_init(10), + dfixed_init(11) }; fixed20_12 min_mem_eff; fixed20_12 mc_latency_sclk, mc_latency_mclk, k1; @@ -2412,7 +2740,7 @@ void r100_bandwidth_update(struct radeon_device *rdev) } } - min_mem_eff.full = rfixed_const_8(0); + min_mem_eff.full = dfixed_const_8(0); /* get modes */ if ((rdev->disp_priority == 2) && ASIC_IS_R300(rdev)) { uint32_t mc_init_misc_lat_timer = RREG32(R300_MC_INIT_MISC_LAT_TIMER); @@ -2433,28 +2761,28 @@ void r100_bandwidth_update(struct radeon_device *rdev) mclk_ff = rdev->pm.mclk; temp = (rdev->mc.vram_width / 8) * (rdev->mc.vram_is_ddr ? 2 : 1); - temp_ff.full = rfixed_const(temp); - mem_bw.full = rfixed_mul(mclk_ff, temp_ff); + temp_ff.full = dfixed_const(temp); + mem_bw.full = dfixed_mul(mclk_ff, temp_ff); pix_clk.full = 0; pix_clk2.full = 0; peak_disp_bw.full = 0; if (mode1) { - temp_ff.full = rfixed_const(1000); - pix_clk.full = rfixed_const(mode1->clock); /* convert to fixed point */ - pix_clk.full = rfixed_div(pix_clk, temp_ff); - temp_ff.full = rfixed_const(pixel_bytes1); - peak_disp_bw.full += rfixed_mul(pix_clk, temp_ff); + temp_ff.full = dfixed_const(1000); + pix_clk.full = dfixed_const(mode1->clock); /* convert to fixed point */ + pix_clk.full = dfixed_div(pix_clk, temp_ff); + temp_ff.full = dfixed_const(pixel_bytes1); + peak_disp_bw.full += dfixed_mul(pix_clk, temp_ff); } if (mode2) { - temp_ff.full = rfixed_const(1000); - pix_clk2.full = rfixed_const(mode2->clock); /* convert to fixed point */ - pix_clk2.full = rfixed_div(pix_clk2, temp_ff); - temp_ff.full = rfixed_const(pixel_bytes2); - peak_disp_bw.full += rfixed_mul(pix_clk2, temp_ff); + temp_ff.full = dfixed_const(1000); + pix_clk2.full = dfixed_const(mode2->clock); /* convert to fixed point */ + pix_clk2.full = dfixed_div(pix_clk2, temp_ff); + temp_ff.full = dfixed_const(pixel_bytes2); + peak_disp_bw.full += dfixed_mul(pix_clk2, temp_ff); } - mem_bw.full = rfixed_mul(mem_bw, min_mem_eff); + mem_bw.full = dfixed_mul(mem_bw, min_mem_eff); if (peak_disp_bw.full >= mem_bw.full) { DRM_ERROR("You may not have enough display bandwidth for current mode\n" "If you have flickering problem, try to lower resolution, refresh rate, or color depth\n"); @@ -2496,9 +2824,9 @@ void r100_bandwidth_update(struct radeon_device *rdev) mem_tras = ((temp >> 12) & 0xf) + 4; } /* convert to FF */ - trcd_ff.full = rfixed_const(mem_trcd); - trp_ff.full = rfixed_const(mem_trp); - tras_ff.full = rfixed_const(mem_tras); + trcd_ff.full = dfixed_const(mem_trcd); + trp_ff.full = dfixed_const(mem_trp); + tras_ff.full = dfixed_const(mem_tras); /* Get values from the MEM_SDRAM_MODE_REG register...converting its */ temp = RREG32(RADEON_MEM_SDRAM_MODE_REG); @@ -2516,7 +2844,7 @@ void r100_bandwidth_update(struct radeon_device *rdev) /* extra cas latency stored in bits 23-25 0-4 clocks */ data = (temp >> 23) & 0x7; if (data < 5) - tcas_ff.full += rfixed_const(data); + tcas_ff.full += dfixed_const(data); } if (ASIC_IS_R300(rdev) && !(rdev->flags & RADEON_IS_IGP)) { @@ -2553,72 +2881,72 @@ void r100_bandwidth_update(struct radeon_device *rdev) if (rdev->flags & RADEON_IS_AGP) { fixed20_12 agpmode_ff; - agpmode_ff.full = rfixed_const(radeon_agpmode); - temp_ff.full = rfixed_const_666(16); - sclk_eff_ff.full -= rfixed_mul(agpmode_ff, temp_ff); + agpmode_ff.full = dfixed_const(radeon_agpmode); + temp_ff.full = dfixed_const_666(16); + sclk_eff_ff.full -= dfixed_mul(agpmode_ff, temp_ff); } /* TODO PCIE lanes may affect this - agpmode == 16?? */ if (ASIC_IS_R300(rdev)) { - sclk_delay_ff.full = rfixed_const(250); + sclk_delay_ff.full = dfixed_const(250); } else { if ((rdev->family == CHIP_RV100) || rdev->flags & RADEON_IS_IGP) { if (rdev->mc.vram_is_ddr) - sclk_delay_ff.full = rfixed_const(41); + sclk_delay_ff.full = dfixed_const(41); else - sclk_delay_ff.full = rfixed_const(33); + sclk_delay_ff.full = dfixed_const(33); } else { if (rdev->mc.vram_width == 128) - sclk_delay_ff.full = rfixed_const(57); + sclk_delay_ff.full = dfixed_const(57); else - sclk_delay_ff.full = rfixed_const(41); + sclk_delay_ff.full = dfixed_const(41); } } - mc_latency_sclk.full = rfixed_div(sclk_delay_ff, sclk_eff_ff); + mc_latency_sclk.full = dfixed_div(sclk_delay_ff, sclk_eff_ff); if (rdev->mc.vram_is_ddr) { if (rdev->mc.vram_width == 32) { - k1.full = rfixed_const(40); + k1.full = dfixed_const(40); c = 3; } else { - k1.full = rfixed_const(20); + k1.full = dfixed_const(20); c = 1; } } else { - k1.full = rfixed_const(40); + k1.full = dfixed_const(40); c = 3; } - temp_ff.full = rfixed_const(2); - mc_latency_mclk.full = rfixed_mul(trcd_ff, temp_ff); - temp_ff.full = rfixed_const(c); - mc_latency_mclk.full += rfixed_mul(tcas_ff, temp_ff); - temp_ff.full = rfixed_const(4); - mc_latency_mclk.full += rfixed_mul(tras_ff, temp_ff); - mc_latency_mclk.full += rfixed_mul(trp_ff, temp_ff); + temp_ff.full = dfixed_const(2); + mc_latency_mclk.full = dfixed_mul(trcd_ff, temp_ff); + temp_ff.full = dfixed_const(c); + mc_latency_mclk.full += dfixed_mul(tcas_ff, temp_ff); + temp_ff.full = dfixed_const(4); + mc_latency_mclk.full += dfixed_mul(tras_ff, temp_ff); + mc_latency_mclk.full += dfixed_mul(trp_ff, temp_ff); mc_latency_mclk.full += k1.full; - mc_latency_mclk.full = rfixed_div(mc_latency_mclk, mclk_ff); - mc_latency_mclk.full += rfixed_div(temp_ff, sclk_eff_ff); + mc_latency_mclk.full = dfixed_div(mc_latency_mclk, mclk_ff); + mc_latency_mclk.full += dfixed_div(temp_ff, sclk_eff_ff); /* HW cursor time assuming worst case of full size colour cursor. */ - temp_ff.full = rfixed_const((2 * (cur_size - (rdev->mc.vram_is_ddr + 1)))); + temp_ff.full = dfixed_const((2 * (cur_size - (rdev->mc.vram_is_ddr + 1)))); temp_ff.full += trcd_ff.full; if (temp_ff.full < tras_ff.full) temp_ff.full = tras_ff.full; - cur_latency_mclk.full = rfixed_div(temp_ff, mclk_ff); + cur_latency_mclk.full = dfixed_div(temp_ff, mclk_ff); - temp_ff.full = rfixed_const(cur_size); - cur_latency_sclk.full = rfixed_div(temp_ff, sclk_eff_ff); + temp_ff.full = dfixed_const(cur_size); + cur_latency_sclk.full = dfixed_div(temp_ff, sclk_eff_ff); /* Find the total latency for the display data. */ - disp_latency_overhead.full = rfixed_const(8); - disp_latency_overhead.full = rfixed_div(disp_latency_overhead, sclk_ff); + disp_latency_overhead.full = dfixed_const(8); + disp_latency_overhead.full = dfixed_div(disp_latency_overhead, sclk_ff); mc_latency_mclk.full += disp_latency_overhead.full + cur_latency_mclk.full; mc_latency_sclk.full += disp_latency_overhead.full + cur_latency_sclk.full; @@ -2646,16 +2974,16 @@ void r100_bandwidth_update(struct radeon_device *rdev) /* Find the drain rate of the display buffer. */ - temp_ff.full = rfixed_const((16/pixel_bytes1)); - disp_drain_rate.full = rfixed_div(pix_clk, temp_ff); + temp_ff.full = dfixed_const((16/pixel_bytes1)); + disp_drain_rate.full = dfixed_div(pix_clk, temp_ff); /* Find the critical point of the display buffer. */ - crit_point_ff.full = rfixed_mul(disp_drain_rate, disp_latency); - crit_point_ff.full += rfixed_const_half(0); + crit_point_ff.full = dfixed_mul(disp_drain_rate, disp_latency); + crit_point_ff.full += dfixed_const_half(0); - critical_point = rfixed_trunc(crit_point_ff); + critical_point = dfixed_trunc(crit_point_ff); if (rdev->disp_priority == 2) { critical_point = 0; @@ -2726,8 +3054,8 @@ void r100_bandwidth_update(struct radeon_device *rdev) /* Find the drain rate of the display buffer. */ - temp_ff.full = rfixed_const((16/pixel_bytes2)); - disp_drain_rate2.full = rfixed_div(pix_clk2, temp_ff); + temp_ff.full = dfixed_const((16/pixel_bytes2)); + disp_drain_rate2.full = dfixed_div(pix_clk2, temp_ff); grph2_cntl = RREG32(RADEON_GRPH2_BUFFER_CNTL); grph2_cntl &= ~(RADEON_GRPH_STOP_REQ_MASK); @@ -2748,8 +3076,8 @@ void r100_bandwidth_update(struct radeon_device *rdev) critical_point2 = 0; else { temp = (rdev->mc.vram_width * rdev->mc.vram_is_ddr + 1)/128; - temp_ff.full = rfixed_const(temp); - temp_ff.full = rfixed_mul(mclk_ff, temp_ff); + temp_ff.full = dfixed_const(temp); + temp_ff.full = dfixed_mul(mclk_ff, temp_ff); if (sclk_ff.full < temp_ff.full) temp_ff.full = sclk_ff.full; @@ -2757,15 +3085,15 @@ void r100_bandwidth_update(struct radeon_device *rdev) if (mode1) { temp_ff.full = read_return_rate.full - disp_drain_rate.full; - time_disp1_drop_priority.full = rfixed_div(crit_point_ff, temp_ff); + time_disp1_drop_priority.full = dfixed_div(crit_point_ff, temp_ff); } else { time_disp1_drop_priority.full = 0; } crit_point_ff.full = disp_latency.full + time_disp1_drop_priority.full + disp_latency.full; - crit_point_ff.full = rfixed_mul(crit_point_ff, disp_drain_rate2); - crit_point_ff.full += rfixed_const_half(0); + crit_point_ff.full = dfixed_mul(crit_point_ff, disp_drain_rate2); + crit_point_ff.full += dfixed_const_half(0); - critical_point2 = rfixed_trunc(crit_point_ff); + critical_point2 = dfixed_trunc(crit_point_ff); if (rdev->disp_priority == 2) { critical_point2 = 0; @@ -2829,33 +3157,6 @@ static inline void r100_cs_track_texture_print(struct r100_cs_track_texture *t) DRM_ERROR("compress format %d\n", t->compress_format); } -static int r100_cs_track_cube(struct radeon_device *rdev, - struct r100_cs_track *track, unsigned idx) -{ - unsigned face, w, h; - struct radeon_bo *cube_robj; - unsigned long size; - - for (face = 0; face < 5; face++) { - cube_robj = track->textures[idx].cube_info[face].robj; - w = track->textures[idx].cube_info[face].width; - h = track->textures[idx].cube_info[face].height; - - size = w * h; - size *= track->textures[idx].cpp; - - size += track->textures[idx].cube_info[face].offset; - - if (size > radeon_bo_size(cube_robj)) { - DRM_ERROR("Cube texture offset greater than object size %lu %lu\n", - size, radeon_bo_size(cube_robj)); - r100_cs_track_texture_print(&track->textures[idx]); - return -1; - } - } - return 0; -} - static int r100_track_compress_size(int compress_format, int w, int h) { int block_width, block_height, block_bytes; @@ -2886,6 +3187,37 @@ static int r100_track_compress_size(int compress_format, int w, int h) return sz; } +static int r100_cs_track_cube(struct radeon_device *rdev, + struct r100_cs_track *track, unsigned idx) +{ + unsigned face, w, h; + struct radeon_bo *cube_robj; + unsigned long size; + unsigned compress_format = track->textures[idx].compress_format; + + for (face = 0; face < 5; face++) { + cube_robj = track->textures[idx].cube_info[face].robj; + w = track->textures[idx].cube_info[face].width; + h = track->textures[idx].cube_info[face].height; + + if (compress_format) { + size = r100_track_compress_size(compress_format, w, h); + } else + size = w * h; + size *= track->textures[idx].cpp; + + size += track->textures[idx].cube_info[face].offset; + + if (size > radeon_bo_size(cube_robj)) { + DRM_ERROR("Cube texture offset greater than object size %lu %lu\n", + size, radeon_bo_size(cube_robj)); + r100_cs_track_texture_print(&track->textures[idx]); + return -1; + } + } + return 0; +} + static int r100_cs_track_texture_check(struct radeon_device *rdev, struct r100_cs_track *track) { @@ -3399,7 +3731,7 @@ static int r100_startup(struct radeon_device *rdev) /* Resume clock */ r100_clock_startup(rdev); /* Initialize GPU configuration (# pipes, ...) */ - r100_gpu_init(rdev); +// r100_gpu_init(rdev); /* Initialize GART (initialize after TTM so we can allocate * memory through TTM but finalize after TTM) */ r100_enable_bm(rdev); @@ -3436,7 +3768,7 @@ int r100_resume(struct radeon_device *rdev) /* Resume clock before doing reset */ r100_clock_startup(rdev); /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), RREG32(R_0007C0_CP_STAT)); @@ -3462,7 +3794,6 @@ int r100_suspend(struct radeon_device *rdev) void r100_fini(struct radeon_device *rdev) { - radeon_pm_fini(rdev); r100_cp_fini(rdev); r100_wb_fini(rdev); r100_ib_fini(rdev); @@ -3505,7 +3836,7 @@ int r100_init(struct radeon_device *rdev) return r; } /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), @@ -3518,8 +3849,6 @@ int r100_init(struct radeon_device *rdev) r100_errata(rdev); /* Initialize clocks */ radeon_get_clock_info(rdev->ddev); - /* Initialize power management */ - radeon_pm_init(rdev); /* initialize AGP */ if (rdev->flags & RADEON_IS_AGP) { r = radeon_agp_init(rdev); diff --git a/drivers/gpu/drm/radeon/r100d.h b/drivers/gpu/drm/radeon/r100d.h index df29a630c466..d016b16fa116 100644 --- a/drivers/gpu/drm/radeon/r100d.h +++ b/drivers/gpu/drm/radeon/r100d.h @@ -74,6 +74,134 @@ #define CP_PACKET3_GET_OPCODE(h) (((h) >> 8) & 0xFF) /* Registers */ +#define R_0000F0_RBBM_SOFT_RESET 0x0000F0 +#define S_0000F0_SOFT_RESET_CP(x) (((x) & 0x1) << 0) +#define G_0000F0_SOFT_RESET_CP(x) (((x) >> 0) & 0x1) +#define C_0000F0_SOFT_RESET_CP 0xFFFFFFFE +#define S_0000F0_SOFT_RESET_HI(x) (((x) & 0x1) << 1) +#define G_0000F0_SOFT_RESET_HI(x) (((x) >> 1) & 0x1) +#define C_0000F0_SOFT_RESET_HI 0xFFFFFFFD +#define S_0000F0_SOFT_RESET_SE(x) (((x) & 0x1) << 2) +#define G_0000F0_SOFT_RESET_SE(x) (((x) >> 2) & 0x1) +#define C_0000F0_SOFT_RESET_SE 0xFFFFFFFB +#define S_0000F0_SOFT_RESET_RE(x) (((x) & 0x1) << 3) +#define G_0000F0_SOFT_RESET_RE(x) (((x) >> 3) & 0x1) +#define C_0000F0_SOFT_RESET_RE 0xFFFFFFF7 +#define S_0000F0_SOFT_RESET_PP(x) (((x) & 0x1) << 4) +#define G_0000F0_SOFT_RESET_PP(x) (((x) >> 4) & 0x1) +#define C_0000F0_SOFT_RESET_PP 0xFFFFFFEF +#define S_0000F0_SOFT_RESET_E2(x) (((x) & 0x1) << 5) +#define G_0000F0_SOFT_RESET_E2(x) (((x) >> 5) & 0x1) +#define C_0000F0_SOFT_RESET_E2 0xFFFFFFDF +#define S_0000F0_SOFT_RESET_RB(x) (((x) & 0x1) << 6) +#define G_0000F0_SOFT_RESET_RB(x) (((x) >> 6) & 0x1) +#define C_0000F0_SOFT_RESET_RB 0xFFFFFFBF +#define S_0000F0_SOFT_RESET_HDP(x) (((x) & 0x1) << 7) +#define G_0000F0_SOFT_RESET_HDP(x) (((x) >> 7) & 0x1) +#define C_0000F0_SOFT_RESET_HDP 0xFFFFFF7F +#define S_0000F0_SOFT_RESET_MC(x) (((x) & 0x1) << 8) +#define G_0000F0_SOFT_RESET_MC(x) (((x) >> 8) & 0x1) +#define C_0000F0_SOFT_RESET_MC 0xFFFFFEFF +#define S_0000F0_SOFT_RESET_AIC(x) (((x) & 0x1) << 9) +#define G_0000F0_SOFT_RESET_AIC(x) (((x) >> 9) & 0x1) +#define C_0000F0_SOFT_RESET_AIC 0xFFFFFDFF +#define S_0000F0_SOFT_RESET_VIP(x) (((x) & 0x1) << 10) +#define G_0000F0_SOFT_RESET_VIP(x) (((x) >> 10) & 0x1) +#define C_0000F0_SOFT_RESET_VIP 0xFFFFFBFF +#define S_0000F0_SOFT_RESET_DISP(x) (((x) & 0x1) << 11) +#define G_0000F0_SOFT_RESET_DISP(x) (((x) >> 11) & 0x1) +#define C_0000F0_SOFT_RESET_DISP 0xFFFFF7FF +#define S_0000F0_SOFT_RESET_CG(x) (((x) & 0x1) << 12) +#define G_0000F0_SOFT_RESET_CG(x) (((x) >> 12) & 0x1) +#define C_0000F0_SOFT_RESET_CG 0xFFFFEFFF +#define R_000030_BUS_CNTL 0x000030 +#define S_000030_BUS_DBL_RESYNC(x) (((x) & 0x1) << 0) +#define G_000030_BUS_DBL_RESYNC(x) (((x) >> 0) & 0x1) +#define C_000030_BUS_DBL_RESYNC 0xFFFFFFFE +#define S_000030_BUS_MSTR_RESET(x) (((x) & 0x1) << 1) +#define G_000030_BUS_MSTR_RESET(x) (((x) >> 1) & 0x1) +#define C_000030_BUS_MSTR_RESET 0xFFFFFFFD +#define S_000030_BUS_FLUSH_BUF(x) (((x) & 0x1) << 2) +#define G_000030_BUS_FLUSH_BUF(x) (((x) >> 2) & 0x1) +#define C_000030_BUS_FLUSH_BUF 0xFFFFFFFB +#define S_000030_BUS_STOP_REQ_DIS(x) (((x) & 0x1) << 3) +#define G_000030_BUS_STOP_REQ_DIS(x) (((x) >> 3) & 0x1) +#define C_000030_BUS_STOP_REQ_DIS 0xFFFFFFF7 +#define S_000030_BUS_PM4_READ_COMBINE_EN(x) (((x) & 0x1) << 4) +#define G_000030_BUS_PM4_READ_COMBINE_EN(x) (((x) >> 4) & 0x1) +#define C_000030_BUS_PM4_READ_COMBINE_EN 0xFFFFFFEF +#define S_000030_BUS_WRT_COMBINE_EN(x) (((x) & 0x1) << 5) +#define G_000030_BUS_WRT_COMBINE_EN(x) (((x) >> 5) & 0x1) +#define C_000030_BUS_WRT_COMBINE_EN 0xFFFFFFDF +#define S_000030_BUS_MASTER_DIS(x) (((x) & 0x1) << 6) +#define G_000030_BUS_MASTER_DIS(x) (((x) >> 6) & 0x1) +#define C_000030_BUS_MASTER_DIS 0xFFFFFFBF +#define S_000030_BIOS_ROM_WRT_EN(x) (((x) & 0x1) << 7) +#define G_000030_BIOS_ROM_WRT_EN(x) (((x) >> 7) & 0x1) +#define C_000030_BIOS_ROM_WRT_EN 0xFFFFFF7F +#define S_000030_BM_DAC_CRIPPLE(x) (((x) & 0x1) << 8) +#define G_000030_BM_DAC_CRIPPLE(x) (((x) >> 8) & 0x1) +#define C_000030_BM_DAC_CRIPPLE 0xFFFFFEFF +#define S_000030_BUS_NON_PM4_READ_COMBINE_EN(x) (((x) & 0x1) << 9) +#define G_000030_BUS_NON_PM4_READ_COMBINE_EN(x) (((x) >> 9) & 0x1) +#define C_000030_BUS_NON_PM4_READ_COMBINE_EN 0xFFFFFDFF +#define S_000030_BUS_XFERD_DISCARD_EN(x) (((x) & 0x1) << 10) +#define G_000030_BUS_XFERD_DISCARD_EN(x) (((x) >> 10) & 0x1) +#define C_000030_BUS_XFERD_DISCARD_EN 0xFFFFFBFF +#define S_000030_BUS_SGL_READ_DISABLE(x) (((x) & 0x1) << 11) +#define G_000030_BUS_SGL_READ_DISABLE(x) (((x) >> 11) & 0x1) +#define C_000030_BUS_SGL_READ_DISABLE 0xFFFFF7FF +#define S_000030_BIOS_DIS_ROM(x) (((x) & 0x1) << 12) +#define G_000030_BIOS_DIS_ROM(x) (((x) >> 12) & 0x1) +#define C_000030_BIOS_DIS_ROM 0xFFFFEFFF +#define S_000030_BUS_PCI_READ_RETRY_EN(x) (((x) & 0x1) << 13) +#define G_000030_BUS_PCI_READ_RETRY_EN(x) (((x) >> 13) & 0x1) +#define C_000030_BUS_PCI_READ_RETRY_EN 0xFFFFDFFF +#define S_000030_BUS_AGP_AD_STEPPING_EN(x) (((x) & 0x1) << 14) +#define G_000030_BUS_AGP_AD_STEPPING_EN(x) (((x) >> 14) & 0x1) +#define C_000030_BUS_AGP_AD_STEPPING_EN 0xFFFFBFFF +#define S_000030_BUS_PCI_WRT_RETRY_EN(x) (((x) & 0x1) << 15) +#define G_000030_BUS_PCI_WRT_RETRY_EN(x) (((x) >> 15) & 0x1) +#define C_000030_BUS_PCI_WRT_RETRY_EN 0xFFFF7FFF +#define S_000030_BUS_RETRY_WS(x) (((x) & 0xF) << 16) +#define G_000030_BUS_RETRY_WS(x) (((x) >> 16) & 0xF) +#define C_000030_BUS_RETRY_WS 0xFFF0FFFF +#define S_000030_BUS_MSTR_RD_MULT(x) (((x) & 0x1) << 20) +#define G_000030_BUS_MSTR_RD_MULT(x) (((x) >> 20) & 0x1) +#define C_000030_BUS_MSTR_RD_MULT 0xFFEFFFFF +#define S_000030_BUS_MSTR_RD_LINE(x) (((x) & 0x1) << 21) +#define G_000030_BUS_MSTR_RD_LINE(x) (((x) >> 21) & 0x1) +#define C_000030_BUS_MSTR_RD_LINE 0xFFDFFFFF +#define S_000030_BUS_SUSPEND(x) (((x) & 0x1) << 22) +#define G_000030_BUS_SUSPEND(x) (((x) >> 22) & 0x1) +#define C_000030_BUS_SUSPEND 0xFFBFFFFF +#define S_000030_LAT_16X(x) (((x) & 0x1) << 23) +#define G_000030_LAT_16X(x) (((x) >> 23) & 0x1) +#define C_000030_LAT_16X 0xFF7FFFFF +#define S_000030_BUS_RD_DISCARD_EN(x) (((x) & 0x1) << 24) +#define G_000030_BUS_RD_DISCARD_EN(x) (((x) >> 24) & 0x1) +#define C_000030_BUS_RD_DISCARD_EN 0xFEFFFFFF +#define S_000030_ENFRCWRDY(x) (((x) & 0x1) << 25) +#define G_000030_ENFRCWRDY(x) (((x) >> 25) & 0x1) +#define C_000030_ENFRCWRDY 0xFDFFFFFF +#define S_000030_BUS_MSTR_WS(x) (((x) & 0x1) << 26) +#define G_000030_BUS_MSTR_WS(x) (((x) >> 26) & 0x1) +#define C_000030_BUS_MSTR_WS 0xFBFFFFFF +#define S_000030_BUS_PARKING_DIS(x) (((x) & 0x1) << 27) +#define G_000030_BUS_PARKING_DIS(x) (((x) >> 27) & 0x1) +#define C_000030_BUS_PARKING_DIS 0xF7FFFFFF +#define S_000030_BUS_MSTR_DISCONNECT_EN(x) (((x) & 0x1) << 28) +#define G_000030_BUS_MSTR_DISCONNECT_EN(x) (((x) >> 28) & 0x1) +#define C_000030_BUS_MSTR_DISCONNECT_EN 0xEFFFFFFF +#define S_000030_SERR_EN(x) (((x) & 0x1) << 29) +#define G_000030_SERR_EN(x) (((x) >> 29) & 0x1) +#define C_000030_SERR_EN 0xDFFFFFFF +#define S_000030_BUS_READ_BURST(x) (((x) & 0x1) << 30) +#define G_000030_BUS_READ_BURST(x) (((x) >> 30) & 0x1) +#define C_000030_BUS_READ_BURST 0xBFFFFFFF +#define S_000030_BUS_RDY_READ_DLY(x) (((x) & 0x1) << 31) +#define G_000030_BUS_RDY_READ_DLY(x) (((x) >> 31) & 0x1) +#define C_000030_BUS_RDY_READ_DLY 0x7FFFFFFF #define R_000040_GEN_INT_CNTL 0x000040 #define S_000040_CRTC_VBLANK(x) (((x) & 0x1) << 0) #define G_000040_CRTC_VBLANK(x) (((x) >> 0) & 0x1) @@ -710,5 +838,41 @@ #define G_00000D_FORCE_RB(x) (((x) >> 28) & 0x1) #define C_00000D_FORCE_RB 0xEFFFFFFF +/* PLL regs */ +#define SCLK_CNTL 0xd +#define FORCE_HDP (1 << 17) +#define CLK_PWRMGT_CNTL 0x14 +#define GLOBAL_PMAN_EN (1 << 10) +#define DISP_PM (1 << 20) +#define PLL_PWRMGT_CNTL 0x15 +#define MPLL_TURNOFF (1 << 0) +#define SPLL_TURNOFF (1 << 1) +#define PPLL_TURNOFF (1 << 2) +#define P2PLL_TURNOFF (1 << 3) +#define TVPLL_TURNOFF (1 << 4) +#define MOBILE_SU (1 << 16) +#define SU_SCLK_USE_BCLK (1 << 17) +#define SCLK_CNTL2 0x1e +#define REDUCED_SPEED_SCLK_MODE (1 << 16) +#define REDUCED_SPEED_SCLK_SEL(x) ((x) << 17) +#define MCLK_MISC 0x1f +#define EN_MCLK_TRISTATE_IN_SUSPEND (1 << 18) +#define SCLK_MORE_CNTL 0x35 +#define REDUCED_SPEED_SCLK_EN (1 << 16) +#define IO_CG_VOLTAGE_DROP (1 << 17) +#define VOLTAGE_DELAY_SEL(x) ((x) << 20) +#define VOLTAGE_DROP_SYNC (1 << 19) + +/* mmreg */ +#define DISP_PWR_MAN 0xd08 +#define DISP_D3_GRPH_RST (1 << 18) +#define DISP_D3_SUBPIC_RST (1 << 19) +#define DISP_D3_OV0_RST (1 << 20) +#define DISP_D1D2_GRPH_RST (1 << 21) +#define DISP_D1D2_SUBPIC_RST (1 << 22) +#define DISP_D1D2_OV0_RST (1 << 23) +#define DISP_DVO_ENABLE_RST (1 << 24) +#define TV_ENABLE_RST (1 << 25) +#define AUTO_PWRUP_EN (1 << 26) #endif diff --git a/drivers/gpu/drm/radeon/r200.c b/drivers/gpu/drm/radeon/r200.c index 85617c311212..0266d72e0a4c 100644 --- a/drivers/gpu/drm/radeon/r200.c +++ b/drivers/gpu/drm/radeon/r200.c @@ -415,6 +415,8 @@ int r200_packet0_check(struct radeon_cs_parser *p, /* 2D, 3D, CUBE */ switch (tmp) { case 0: + case 3: + case 4: case 5: case 6: case 7: @@ -450,6 +452,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, case R200_TXFORMAT_RGB332: case R200_TXFORMAT_Y8: track->textures[i].cpp = 1; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case R200_TXFORMAT_AI88: case R200_TXFORMAT_ARGB1555: @@ -461,6 +464,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, case R200_TXFORMAT_DVDU88: case R200_TXFORMAT_AVYU4444: track->textures[i].cpp = 2; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case R200_TXFORMAT_ARGB8888: case R200_TXFORMAT_RGBA8888: @@ -468,6 +472,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, case R200_TXFORMAT_BGR111110: case R200_TXFORMAT_LDVDU8888: track->textures[i].cpp = 4; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case R200_TXFORMAT_DXT1: track->textures[i].cpp = 1; diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c index a5ff8076b423..19a7ef7ee344 100644 --- a/drivers/gpu/drm/radeon/r300.c +++ b/drivers/gpu/drm/radeon/r300.c @@ -27,8 +27,9 @@ */ #include <linux/seq_file.h> #include <linux/slab.h> -#include "drmP.h" -#include "drm.h" +#include <drm/drmP.h> +#include <drm/drm.h> +#include <drm/drm_crtc_helper.h> #include "radeon_reg.h" #include "radeon.h" #include "radeon_asic.h" @@ -151,6 +152,10 @@ void rv370_pcie_gart_disable(struct radeon_device *rdev) u32 tmp; int r; + WREG32_PCIE(RADEON_PCIE_TX_GART_START_LO, 0); + WREG32_PCIE(RADEON_PCIE_TX_GART_END_LO, 0); + WREG32_PCIE(RADEON_PCIE_TX_GART_START_HI, 0); + WREG32_PCIE(RADEON_PCIE_TX_GART_END_HI, 0); tmp = RREG32_PCIE(RADEON_PCIE_TX_GART_CNTL); tmp |= RADEON_PCIE_TX_GART_UNMAPPED_ACCESS_DISCARD; WREG32_PCIE(RADEON_PCIE_TX_GART_CNTL, tmp & ~RADEON_PCIE_TX_GART_EN); @@ -323,7 +328,6 @@ void r300_gpu_init(struct radeon_device *rdev) { uint32_t gb_tile_config, tmp; - r100_hdp_reset(rdev); if ((rdev->family == CHIP_R300 && rdev->pdev->device != 0x4144) || (rdev->family == CHIP_R350 && rdev->pdev->device != 0x4148)) { /* r300,r350 */ @@ -375,89 +379,85 @@ void r300_gpu_init(struct radeon_device *rdev) rdev->num_gb_pipes, rdev->num_z_pipes); } -int r300_ga_reset(struct radeon_device *rdev) +bool r300_gpu_is_lockup(struct radeon_device *rdev) { - uint32_t tmp; - bool reinit_cp; - int i; + u32 rbbm_status; + int r; - reinit_cp = rdev->cp.ready; - rdev->cp.ready = false; - for (i = 0; i < rdev->usec_timeout; i++) { - WREG32(RADEON_CP_CSQ_MODE, 0); - WREG32(RADEON_CP_CSQ_CNTL, 0); - WREG32(RADEON_RBBM_SOFT_RESET, 0x32005); - (void)RREG32(RADEON_RBBM_SOFT_RESET); - udelay(200); - WREG32(RADEON_RBBM_SOFT_RESET, 0); - /* Wait to prevent race in RBBM_STATUS */ - mdelay(1); - tmp = RREG32(RADEON_RBBM_STATUS); - if (tmp & ((1 << 20) | (1 << 26))) { - DRM_ERROR("VAP & CP still busy (RBBM_STATUS=0x%08X)", tmp); - /* GA still busy soft reset it */ - WREG32(0x429C, 0x200); - WREG32(R300_VAP_PVS_STATE_FLUSH_REG, 0); - WREG32(R300_RE_SCISSORS_TL, 0); - WREG32(R300_RE_SCISSORS_BR, 0); - WREG32(0x24AC, 0); - } - /* Wait to prevent race in RBBM_STATUS */ - mdelay(1); - tmp = RREG32(RADEON_RBBM_STATUS); - if (!(tmp & ((1 << 20) | (1 << 26)))) { - break; - } + rbbm_status = RREG32(R_000E40_RBBM_STATUS); + if (!G_000E40_GUI_ACTIVE(rbbm_status)) { + r100_gpu_lockup_update(&rdev->config.r300.lockup, &rdev->cp); + return false; } - for (i = 0; i < rdev->usec_timeout; i++) { - tmp = RREG32(RADEON_RBBM_STATUS); - if (!(tmp & ((1 << 20) | (1 << 26)))) { - DRM_INFO("GA reset succeed (RBBM_STATUS=0x%08X)\n", - tmp); - if (reinit_cp) { - return r100_cp_init(rdev, rdev->cp.ring_size); - } - return 0; - } - DRM_UDELAY(1); + /* force CP activities */ + r = radeon_ring_lock(rdev, 2); + if (!r) { + /* PACKET2 NOP */ + radeon_ring_write(rdev, 0x80000000); + radeon_ring_write(rdev, 0x80000000); + radeon_ring_unlock_commit(rdev); } - tmp = RREG32(RADEON_RBBM_STATUS); - DRM_ERROR("Failed to reset GA ! (RBBM_STATUS=0x%08X)\n", tmp); - return -1; + rdev->cp.rptr = RREG32(RADEON_CP_RB_RPTR); + return r100_gpu_cp_is_lockup(rdev, &rdev->config.r300.lockup, &rdev->cp); } -int r300_gpu_reset(struct radeon_device *rdev) +int r300_asic_reset(struct radeon_device *rdev) { - uint32_t status; - - /* reset order likely matter */ - status = RREG32(RADEON_RBBM_STATUS); - /* reset HDP */ - r100_hdp_reset(rdev); - /* reset rb2d */ - if (status & ((1 << 17) | (1 << 18) | (1 << 27))) { - r100_rb2d_reset(rdev); - } - /* reset GA */ - if (status & ((1 << 20) | (1 << 26))) { - r300_ga_reset(rdev); - } - /* reset CP */ - status = RREG32(RADEON_RBBM_STATUS); - if (status & (1 << 16)) { - r100_cp_reset(rdev); + struct r100_mc_save save; + u32 status, tmp; + + r100_mc_stop(rdev, &save); + status = RREG32(R_000E40_RBBM_STATUS); + if (!G_000E40_GUI_ACTIVE(status)) { + return 0; } + status = RREG32(R_000E40_RBBM_STATUS); + dev_info(rdev->dev, "(%s:%d) RBBM_STATUS=0x%08X\n", __func__, __LINE__, status); + /* stop CP */ + WREG32(RADEON_CP_CSQ_CNTL, 0); + tmp = RREG32(RADEON_CP_RB_CNTL); + WREG32(RADEON_CP_RB_CNTL, tmp | RADEON_RB_RPTR_WR_ENA); + WREG32(RADEON_CP_RB_RPTR_WR, 0); + WREG32(RADEON_CP_RB_WPTR, 0); + WREG32(RADEON_CP_RB_CNTL, tmp); + /* save PCI state */ + pci_save_state(rdev->pdev); + /* disable bus mastering */ + r100_bm_disable(rdev); + WREG32(R_0000F0_RBBM_SOFT_RESET, S_0000F0_SOFT_RESET_VAP(1) | + S_0000F0_SOFT_RESET_GA(1)); + RREG32(R_0000F0_RBBM_SOFT_RESET); + mdelay(500); + WREG32(R_0000F0_RBBM_SOFT_RESET, 0); + mdelay(1); + status = RREG32(R_000E40_RBBM_STATUS); + dev_info(rdev->dev, "(%s:%d) RBBM_STATUS=0x%08X\n", __func__, __LINE__, status); + /* resetting the CP seems to be problematic sometimes it end up + * hard locking the computer, but it's necessary for successfull + * reset more test & playing is needed on R3XX/R4XX to find a + * reliable (if any solution) + */ + WREG32(R_0000F0_RBBM_SOFT_RESET, S_0000F0_SOFT_RESET_CP(1)); + RREG32(R_0000F0_RBBM_SOFT_RESET); + mdelay(500); + WREG32(R_0000F0_RBBM_SOFT_RESET, 0); + mdelay(1); + status = RREG32(R_000E40_RBBM_STATUS); + dev_info(rdev->dev, "(%s:%d) RBBM_STATUS=0x%08X\n", __func__, __LINE__, status); + /* restore PCI & busmastering */ + pci_restore_state(rdev->pdev); + r100_enable_bm(rdev); /* Check if GPU is idle */ - status = RREG32(RADEON_RBBM_STATUS); - if (status & RADEON_RBBM_ACTIVE) { - DRM_ERROR("Failed to reset GPU (RBBM_STATUS=0x%08X)\n", status); + if (G_000E40_GA_BUSY(status) || G_000E40_VAP_BUSY(status)) { + dev_err(rdev->dev, "failed to reset GPU\n"); + rdev->gpu_lockup = true; return -1; } - DRM_INFO("GPU reset succeed (RBBM_STATUS=0x%08X)\n", status); + r100_mc_resume(rdev, &save); + dev_info(rdev->dev, "GPU reset succeed\n"); return 0; } - /* * r300,r350,rv350,rv380 VRAM info */ @@ -481,6 +481,7 @@ void r300_mc_init(struct radeon_device *rdev) if (rdev->flags & RADEON_IS_IGP) base = (RREG32(RADEON_NB_TOM) & 0xffff) << 16; radeon_vram_location(rdev, &rdev->mc, base); + rdev->mc.gtt_base_align = 0; if (!(rdev->flags & RADEON_IS_AGP)) radeon_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); @@ -881,6 +882,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, case R300_TX_FORMAT_Y4X4: case R300_TX_FORMAT_Z3Y3X2: track->textures[i].cpp = 1; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case R300_TX_FORMAT_X16: case R300_TX_FORMAT_Y8X8: @@ -892,6 +894,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, case R300_TX_FORMAT_B8G8_B8G8: case R300_TX_FORMAT_G8R8_G8B8: track->textures[i].cpp = 2; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case R300_TX_FORMAT_Y16X16: case R300_TX_FORMAT_Z11Y11X10: @@ -902,14 +905,17 @@ static int r300_packet0_check(struct radeon_cs_parser *p, case R300_TX_FORMAT_FL_I32: case 0x1e: track->textures[i].cpp = 4; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case R300_TX_FORMAT_W16Z16Y16X16: case R300_TX_FORMAT_FL_R16G16B16A16: case R300_TX_FORMAT_FL_I32A32: track->textures[i].cpp = 8; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case R300_TX_FORMAT_FL_R32G32B32A32: track->textures[i].cpp = 16; + track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case R300_TX_FORMAT_DXT1: track->textures[i].cpp = 1; @@ -1171,6 +1177,8 @@ int r300_cs_parse(struct radeon_cs_parser *p) int r; track = kzalloc(sizeof(*track), GFP_KERNEL); + if (track == NULL) + return -ENOMEM; r100_cs_track_clear(p->rdev, track); p->track = track; do { @@ -1316,7 +1324,7 @@ int r300_resume(struct radeon_device *rdev) /* Resume clock before doing reset */ r300_clock_startup(rdev); /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), RREG32(R_0007C0_CP_STAT)); @@ -1344,7 +1352,6 @@ int r300_suspend(struct radeon_device *rdev) void r300_fini(struct radeon_device *rdev) { - radeon_pm_fini(rdev); r100_cp_fini(rdev); r100_wb_fini(rdev); r100_ib_fini(rdev); @@ -1387,7 +1394,7 @@ int r300_init(struct radeon_device *rdev) return r; } /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), @@ -1400,8 +1407,6 @@ int r300_init(struct radeon_device *rdev) r300_errata(rdev); /* Initialize clocks */ radeon_get_clock_info(rdev->ddev); - /* Initialize power management */ - radeon_pm_init(rdev); /* initialize AGP */ if (rdev->flags & RADEON_IS_AGP) { r = radeon_agp_init(rdev); diff --git a/drivers/gpu/drm/radeon/r300d.h b/drivers/gpu/drm/radeon/r300d.h index 4c73114f0de9..968a33317fbf 100644 --- a/drivers/gpu/drm/radeon/r300d.h +++ b/drivers/gpu/drm/radeon/r300d.h @@ -209,7 +209,52 @@ #define S_000E40_GUI_ACTIVE(x) (((x) & 0x1) << 31) #define G_000E40_GUI_ACTIVE(x) (((x) >> 31) & 0x1) #define C_000E40_GUI_ACTIVE 0x7FFFFFFF - +#define R_0000F0_RBBM_SOFT_RESET 0x0000F0 +#define S_0000F0_SOFT_RESET_CP(x) (((x) & 0x1) << 0) +#define G_0000F0_SOFT_RESET_CP(x) (((x) >> 0) & 0x1) +#define C_0000F0_SOFT_RESET_CP 0xFFFFFFFE +#define S_0000F0_SOFT_RESET_HI(x) (((x) & 0x1) << 1) +#define G_0000F0_SOFT_RESET_HI(x) (((x) >> 1) & 0x1) +#define C_0000F0_SOFT_RESET_HI 0xFFFFFFFD +#define S_0000F0_SOFT_RESET_VAP(x) (((x) & 0x1) << 2) +#define G_0000F0_SOFT_RESET_VAP(x) (((x) >> 2) & 0x1) +#define C_0000F0_SOFT_RESET_VAP 0xFFFFFFFB +#define S_0000F0_SOFT_RESET_RE(x) (((x) & 0x1) << 3) +#define G_0000F0_SOFT_RESET_RE(x) (((x) >> 3) & 0x1) +#define C_0000F0_SOFT_RESET_RE 0xFFFFFFF7 +#define S_0000F0_SOFT_RESET_PP(x) (((x) & 0x1) << 4) +#define G_0000F0_SOFT_RESET_PP(x) (((x) >> 4) & 0x1) +#define C_0000F0_SOFT_RESET_PP 0xFFFFFFEF +#define S_0000F0_SOFT_RESET_E2(x) (((x) & 0x1) << 5) +#define G_0000F0_SOFT_RESET_E2(x) (((x) >> 5) & 0x1) +#define C_0000F0_SOFT_RESET_E2 0xFFFFFFDF +#define S_0000F0_SOFT_RESET_RB(x) (((x) & 0x1) << 6) +#define G_0000F0_SOFT_RESET_RB(x) (((x) >> 6) & 0x1) +#define C_0000F0_SOFT_RESET_RB 0xFFFFFFBF +#define S_0000F0_SOFT_RESET_HDP(x) (((x) & 0x1) << 7) +#define G_0000F0_SOFT_RESET_HDP(x) (((x) >> 7) & 0x1) +#define C_0000F0_SOFT_RESET_HDP 0xFFFFFF7F +#define S_0000F0_SOFT_RESET_MC(x) (((x) & 0x1) << 8) +#define G_0000F0_SOFT_RESET_MC(x) (((x) >> 8) & 0x1) +#define C_0000F0_SOFT_RESET_MC 0xFFFFFEFF +#define S_0000F0_SOFT_RESET_AIC(x) (((x) & 0x1) << 9) +#define G_0000F0_SOFT_RESET_AIC(x) (((x) >> 9) & 0x1) +#define C_0000F0_SOFT_RESET_AIC 0xFFFFFDFF +#define S_0000F0_SOFT_RESET_VIP(x) (((x) & 0x1) << 10) +#define G_0000F0_SOFT_RESET_VIP(x) (((x) >> 10) & 0x1) +#define C_0000F0_SOFT_RESET_VIP 0xFFFFFBFF +#define S_0000F0_SOFT_RESET_DISP(x) (((x) & 0x1) << 11) +#define G_0000F0_SOFT_RESET_DISP(x) (((x) >> 11) & 0x1) +#define C_0000F0_SOFT_RESET_DISP 0xFFFFF7FF +#define S_0000F0_SOFT_RESET_CG(x) (((x) & 0x1) << 12) +#define G_0000F0_SOFT_RESET_CG(x) (((x) >> 12) & 0x1) +#define C_0000F0_SOFT_RESET_CG 0xFFFFEFFF +#define S_0000F0_SOFT_RESET_GA(x) (((x) & 0x1) << 13) +#define G_0000F0_SOFT_RESET_GA(x) (((x) >> 13) & 0x1) +#define C_0000F0_SOFT_RESET_GA 0xFFFFDFFF +#define S_0000F0_SOFT_RESET_IDCT(x) (((x) & 0x1) << 14) +#define G_0000F0_SOFT_RESET_IDCT(x) (((x) >> 14) & 0x1) +#define C_0000F0_SOFT_RESET_IDCT 0xFFFFBFFF #define R_00000D_SCLK_CNTL 0x00000D #define S_00000D_SCLK_SRC_SEL(x) (((x) & 0x7) << 0) diff --git a/drivers/gpu/drm/radeon/r420.c b/drivers/gpu/drm/radeon/r420.c index c2bda4ad62e7..e6c89142bb4d 100644 --- a/drivers/gpu/drm/radeon/r420.c +++ b/drivers/gpu/drm/radeon/r420.c @@ -36,6 +36,45 @@ #include "r420d.h" #include "r420_reg_safe.h" +void r420_pm_init_profile(struct radeon_device *rdev) +{ + /* default */ + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_cm_idx = 0; + /* low sh */ + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_cm_idx = 0; + /* mid sh */ + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_cm_idx = 0; + /* high sh */ + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_cm_idx = 0; + /* low mh */ + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_cm_idx = 0; + /* mid mh */ + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_cm_idx = 0; + /* high mh */ + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_cm_idx = 0; +} + static void r420_set_reg_safe(struct radeon_device *rdev) { rdev->config.r300.reg_safe_bm = r420_reg_safe_bm; @@ -241,7 +280,7 @@ int r420_resume(struct radeon_device *rdev) /* Resume clock before doing reset */ r420_clock_resume(rdev); /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), RREG32(R_0007C0_CP_STAT)); @@ -274,7 +313,6 @@ int r420_suspend(struct radeon_device *rdev) void r420_fini(struct radeon_device *rdev) { - radeon_pm_fini(rdev); r100_cp_fini(rdev); r100_wb_fini(rdev); r100_ib_fini(rdev); @@ -322,7 +360,7 @@ int r420_init(struct radeon_device *rdev) } } /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), @@ -334,8 +372,6 @@ int r420_init(struct radeon_device *rdev) /* Initialize clocks */ radeon_get_clock_info(rdev->ddev); - /* Initialize power management */ - radeon_pm_init(rdev); /* initialize AGP */ if (rdev->flags & RADEON_IS_AGP) { r = radeon_agp_init(rdev); diff --git a/drivers/gpu/drm/radeon/r500_reg.h b/drivers/gpu/drm/radeon/r500_reg.h index 0cf2ad2a5585..93c9a2bbccf8 100644 --- a/drivers/gpu/drm/radeon/r500_reg.h +++ b/drivers/gpu/drm/radeon/r500_reg.h @@ -347,9 +347,11 @@ #define AVIVO_D1CRTC_CONTROL 0x6080 # define AVIVO_CRTC_EN (1 << 0) +# define AVIVO_CRTC_DISP_READ_REQUEST_DISABLE (1 << 24) #define AVIVO_D1CRTC_BLANK_CONTROL 0x6084 #define AVIVO_D1CRTC_INTERLACE_CONTROL 0x6088 #define AVIVO_D1CRTC_INTERLACE_STATUS 0x608c +#define AVIVO_D1CRTC_STATUS_POSITION 0x60a0 #define AVIVO_D1CRTC_FRAME_COUNT 0x60a4 #define AVIVO_D1CRTC_STEREO_CONTROL 0x60c4 @@ -488,6 +490,7 @@ #define AVIVO_D2CRTC_BLANK_CONTROL 0x6884 #define AVIVO_D2CRTC_INTERLACE_CONTROL 0x6888 #define AVIVO_D2CRTC_INTERLACE_STATUS 0x688c +#define AVIVO_D2CRTC_STATUS_POSITION 0x68a0 #define AVIVO_D2CRTC_FRAME_COUNT 0x68a4 #define AVIVO_D2CRTC_STEREO_CONTROL 0x68c4 diff --git a/drivers/gpu/drm/radeon/r520.c b/drivers/gpu/drm/radeon/r520.c index 3c44b8d39318..694af7cc23ac 100644 --- a/drivers/gpu/drm/radeon/r520.c +++ b/drivers/gpu/drm/radeon/r520.c @@ -53,7 +53,6 @@ static void r520_gpu_init(struct radeon_device *rdev) { unsigned pipe_select_current, gb_pipe_select, tmp; - r100_hdp_reset(rdev); rv515_vga_render_disable(rdev); /* * DST_PIPE_CONFIG 0x170C @@ -126,6 +125,7 @@ void r520_mc_init(struct radeon_device *rdev) r520_vram_get_type(rdev); r100_vram_init_sizes(rdev); radeon_vram_location(rdev, &rdev->mc, 0); + rdev->mc.gtt_base_align = 0; if (!(rdev->flags & RADEON_IS_AGP)) radeon_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); @@ -209,7 +209,7 @@ int r520_resume(struct radeon_device *rdev) /* Resume clock before doing reset */ rv515_clock_startup(rdev); /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), RREG32(R_0007C0_CP_STAT)); @@ -246,7 +246,7 @@ int r520_init(struct radeon_device *rdev) return -EINVAL; } /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), @@ -262,8 +262,6 @@ int r520_init(struct radeon_device *rdev) } /* Initialize clocks */ radeon_get_clock_info(rdev->ddev); - /* Initialize power management */ - radeon_pm_init(rdev); /* initialize AGP */ if (rdev->flags & RADEON_IS_AGP) { r = radeon_agp_init(rdev); diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index 8f3454e2056a..e100f69faeec 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -44,6 +44,9 @@ #define R700_PFP_UCODE_SIZE 848 #define R700_PM4_UCODE_SIZE 1360 #define R700_RLC_UCODE_SIZE 1024 +#define EVERGREEN_PFP_UCODE_SIZE 1120 +#define EVERGREEN_PM4_UCODE_SIZE 1376 +#define EVERGREEN_RLC_UCODE_SIZE 768 /* Firmware Names */ MODULE_FIRMWARE("radeon/R600_pfp.bin"); @@ -68,6 +71,18 @@ MODULE_FIRMWARE("radeon/RV710_pfp.bin"); MODULE_FIRMWARE("radeon/RV710_me.bin"); MODULE_FIRMWARE("radeon/R600_rlc.bin"); MODULE_FIRMWARE("radeon/R700_rlc.bin"); +MODULE_FIRMWARE("radeon/CEDAR_pfp.bin"); +MODULE_FIRMWARE("radeon/CEDAR_me.bin"); +MODULE_FIRMWARE("radeon/CEDAR_rlc.bin"); +MODULE_FIRMWARE("radeon/REDWOOD_pfp.bin"); +MODULE_FIRMWARE("radeon/REDWOOD_me.bin"); +MODULE_FIRMWARE("radeon/REDWOOD_rlc.bin"); +MODULE_FIRMWARE("radeon/JUNIPER_pfp.bin"); +MODULE_FIRMWARE("radeon/JUNIPER_me.bin"); +MODULE_FIRMWARE("radeon/JUNIPER_rlc.bin"); +MODULE_FIRMWARE("radeon/CYPRESS_pfp.bin"); +MODULE_FIRMWARE("radeon/CYPRESS_me.bin"); +MODULE_FIRMWARE("radeon/CYPRESS_rlc.bin"); int r600_debugfs_mc_info_init(struct radeon_device *rdev); @@ -75,6 +90,499 @@ int r600_debugfs_mc_info_init(struct radeon_device *rdev); int r600_mc_wait_for_idle(struct radeon_device *rdev); void r600_gpu_init(struct radeon_device *rdev); void r600_fini(struct radeon_device *rdev); +void r600_irq_disable(struct radeon_device *rdev); + +void r600_pm_get_dynpm_state(struct radeon_device *rdev) +{ + int i; + + rdev->pm.dynpm_can_upclock = true; + rdev->pm.dynpm_can_downclock = true; + + /* power state array is low to high, default is first */ + if ((rdev->flags & RADEON_IS_IGP) || (rdev->family == CHIP_R600)) { + int min_power_state_index = 0; + + if (rdev->pm.num_power_states > 2) + min_power_state_index = 1; + + switch (rdev->pm.dynpm_planned_action) { + case DYNPM_ACTION_MINIMUM: + rdev->pm.requested_power_state_index = min_power_state_index; + rdev->pm.requested_clock_mode_index = 0; + rdev->pm.dynpm_can_downclock = false; + break; + case DYNPM_ACTION_DOWNCLOCK: + if (rdev->pm.current_power_state_index == min_power_state_index) { + rdev->pm.requested_power_state_index = rdev->pm.current_power_state_index; + rdev->pm.dynpm_can_downclock = false; + } else { + if (rdev->pm.active_crtc_count > 1) { + for (i = 0; i < rdev->pm.num_power_states; i++) { + if (rdev->pm.power_state[i].flags & RADEON_PM_STATE_SINGLE_DISPLAY_ONLY) + continue; + else if (i >= rdev->pm.current_power_state_index) { + rdev->pm.requested_power_state_index = + rdev->pm.current_power_state_index; + break; + } else { + rdev->pm.requested_power_state_index = i; + break; + } + } + } else { + if (rdev->pm.current_power_state_index == 0) + rdev->pm.requested_power_state_index = + rdev->pm.num_power_states - 1; + else + rdev->pm.requested_power_state_index = + rdev->pm.current_power_state_index - 1; + } + } + rdev->pm.requested_clock_mode_index = 0; + /* don't use the power state if crtcs are active and no display flag is set */ + if ((rdev->pm.active_crtc_count > 0) && + (rdev->pm.power_state[rdev->pm.requested_power_state_index]. + clock_info[rdev->pm.requested_clock_mode_index].flags & + RADEON_PM_MODE_NO_DISPLAY)) { + rdev->pm.requested_power_state_index++; + } + break; + case DYNPM_ACTION_UPCLOCK: + if (rdev->pm.current_power_state_index == (rdev->pm.num_power_states - 1)) { + rdev->pm.requested_power_state_index = rdev->pm.current_power_state_index; + rdev->pm.dynpm_can_upclock = false; + } else { + if (rdev->pm.active_crtc_count > 1) { + for (i = (rdev->pm.num_power_states - 1); i >= 0; i--) { + if (rdev->pm.power_state[i].flags & RADEON_PM_STATE_SINGLE_DISPLAY_ONLY) + continue; + else if (i <= rdev->pm.current_power_state_index) { + rdev->pm.requested_power_state_index = + rdev->pm.current_power_state_index; + break; + } else { + rdev->pm.requested_power_state_index = i; + break; + } + } + } else + rdev->pm.requested_power_state_index = + rdev->pm.current_power_state_index + 1; + } + rdev->pm.requested_clock_mode_index = 0; + break; + case DYNPM_ACTION_DEFAULT: + rdev->pm.requested_power_state_index = rdev->pm.default_power_state_index; + rdev->pm.requested_clock_mode_index = 0; + rdev->pm.dynpm_can_upclock = false; + break; + case DYNPM_ACTION_NONE: + default: + DRM_ERROR("Requested mode for not defined action\n"); + return; + } + } else { + /* XXX select a power state based on AC/DC, single/dualhead, etc. */ + /* for now just select the first power state and switch between clock modes */ + /* power state array is low to high, default is first (0) */ + if (rdev->pm.active_crtc_count > 1) { + rdev->pm.requested_power_state_index = -1; + /* start at 1 as we don't want the default mode */ + for (i = 1; i < rdev->pm.num_power_states; i++) { + if (rdev->pm.power_state[i].flags & RADEON_PM_STATE_SINGLE_DISPLAY_ONLY) + continue; + else if ((rdev->pm.power_state[i].type == POWER_STATE_TYPE_PERFORMANCE) || + (rdev->pm.power_state[i].type == POWER_STATE_TYPE_BATTERY)) { + rdev->pm.requested_power_state_index = i; + break; + } + } + /* if nothing selected, grab the default state. */ + if (rdev->pm.requested_power_state_index == -1) + rdev->pm.requested_power_state_index = 0; + } else + rdev->pm.requested_power_state_index = 1; + + switch (rdev->pm.dynpm_planned_action) { + case DYNPM_ACTION_MINIMUM: + rdev->pm.requested_clock_mode_index = 0; + rdev->pm.dynpm_can_downclock = false; + break; + case DYNPM_ACTION_DOWNCLOCK: + if (rdev->pm.requested_power_state_index == rdev->pm.current_power_state_index) { + if (rdev->pm.current_clock_mode_index == 0) { + rdev->pm.requested_clock_mode_index = 0; + rdev->pm.dynpm_can_downclock = false; + } else + rdev->pm.requested_clock_mode_index = + rdev->pm.current_clock_mode_index - 1; + } else { + rdev->pm.requested_clock_mode_index = 0; + rdev->pm.dynpm_can_downclock = false; + } + /* don't use the power state if crtcs are active and no display flag is set */ + if ((rdev->pm.active_crtc_count > 0) && + (rdev->pm.power_state[rdev->pm.requested_power_state_index]. + clock_info[rdev->pm.requested_clock_mode_index].flags & + RADEON_PM_MODE_NO_DISPLAY)) { + rdev->pm.requested_clock_mode_index++; + } + break; + case DYNPM_ACTION_UPCLOCK: + if (rdev->pm.requested_power_state_index == rdev->pm.current_power_state_index) { + if (rdev->pm.current_clock_mode_index == + (rdev->pm.power_state[rdev->pm.requested_power_state_index].num_clock_modes - 1)) { + rdev->pm.requested_clock_mode_index = rdev->pm.current_clock_mode_index; + rdev->pm.dynpm_can_upclock = false; + } else + rdev->pm.requested_clock_mode_index = + rdev->pm.current_clock_mode_index + 1; + } else { + rdev->pm.requested_clock_mode_index = + rdev->pm.power_state[rdev->pm.requested_power_state_index].num_clock_modes - 1; + rdev->pm.dynpm_can_upclock = false; + } + break; + case DYNPM_ACTION_DEFAULT: + rdev->pm.requested_power_state_index = rdev->pm.default_power_state_index; + rdev->pm.requested_clock_mode_index = 0; + rdev->pm.dynpm_can_upclock = false; + break; + case DYNPM_ACTION_NONE: + default: + DRM_ERROR("Requested mode for not defined action\n"); + return; + } + } + + DRM_DEBUG("Requested: e: %d m: %d p: %d\n", + rdev->pm.power_state[rdev->pm.requested_power_state_index]. + clock_info[rdev->pm.requested_clock_mode_index].sclk, + rdev->pm.power_state[rdev->pm.requested_power_state_index]. + clock_info[rdev->pm.requested_clock_mode_index].mclk, + rdev->pm.power_state[rdev->pm.requested_power_state_index]. + pcie_lanes); +} + +static int r600_pm_get_type_index(struct radeon_device *rdev, + enum radeon_pm_state_type ps_type, + int instance) +{ + int i; + int found_instance = -1; + + for (i = 0; i < rdev->pm.num_power_states; i++) { + if (rdev->pm.power_state[i].type == ps_type) { + found_instance++; + if (found_instance == instance) + return i; + } + } + /* return default if no match */ + return rdev->pm.default_power_state_index; +} + +void rs780_pm_init_profile(struct radeon_device *rdev) +{ + if (rdev->pm.num_power_states == 2) { + /* default */ + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_cm_idx = 0; + /* low sh */ + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_cm_idx = 0; + /* mid sh */ + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_cm_idx = 0; + /* high sh */ + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_cm_idx = 0; + /* low mh */ + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_cm_idx = 0; + /* mid mh */ + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_cm_idx = 0; + /* high mh */ + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_cm_idx = 0; + } else if (rdev->pm.num_power_states == 3) { + /* default */ + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_cm_idx = 0; + /* low sh */ + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_cm_idx = 0; + /* mid sh */ + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_cm_idx = 0; + /* high sh */ + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_cm_idx = 0; + /* low mh */ + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_cm_idx = 0; + /* mid mh */ + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_cm_idx = 0; + /* high mh */ + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_cm_idx = 0; + } else { + /* default */ + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_cm_idx = 0; + /* low sh */ + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_cm_idx = 0; + /* mid sh */ + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_cm_idx = 0; + /* high sh */ + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_ps_idx = 3; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_cm_idx = 0; + /* low mh */ + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_cm_idx = 0; + /* mid mh */ + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_ps_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_cm_idx = 0; + /* high mh */ + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_ps_idx = 3; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_cm_idx = 0; + } +} + +void r600_pm_init_profile(struct radeon_device *rdev) +{ + if (rdev->family == CHIP_R600) { + /* XXX */ + /* default */ + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_cm_idx = 0; + /* low sh */ + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_cm_idx = 0; + /* mid sh */ + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_cm_idx = 0; + /* high sh */ + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_cm_idx = 0; + /* low mh */ + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_cm_idx = 0; + /* mid mh */ + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_cm_idx = 0; + /* high mh */ + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_cm_idx = 0; + } else { + if (rdev->pm.num_power_states < 4) { + /* default */ + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_cm_idx = 2; + /* low sh */ + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_cm_idx = 0; + /* mid sh */ + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_cm_idx = 1; + /* high sh */ + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_ps_idx = 1; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_cm_idx = 2; + /* low mh */ + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_cm_idx = 0; + /* low mh */ + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_cm_idx = 1; + /* high mh */ + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_ps_idx = 2; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_cm_idx = 2; + } else { + /* default */ + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_ps_idx = rdev->pm.default_power_state_index; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_DEFAULT_IDX].dpms_on_cm_idx = 2; + /* low sh */ + if (rdev->flags & RADEON_IS_MOBILITY) { + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_BATTERY, 0); + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_BATTERY, 0); + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_cm_idx = 0; + } else { + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 0); + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 0); + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_SH_IDX].dpms_on_cm_idx = 0; + } + /* mid sh */ + if (rdev->flags & RADEON_IS_MOBILITY) { + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_BATTERY, 0); + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_BATTERY, 0); + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_cm_idx = 1; + } else { + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 0); + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 0); + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_SH_IDX].dpms_on_cm_idx = 1; + } + /* high sh */ + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 0); + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 0); + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_SH_IDX].dpms_on_cm_idx = 2; + /* low mh */ + if (rdev->flags & RADEON_IS_MOBILITY) { + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_BATTERY, 1); + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_BATTERY, 1); + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_cm_idx = 0; + } else { + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 1); + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 1); + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_LOW_MH_IDX].dpms_on_cm_idx = 0; + } + /* mid mh */ + if (rdev->flags & RADEON_IS_MOBILITY) { + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_BATTERY, 1); + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_BATTERY, 1); + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_cm_idx = 1; + } else { + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 1); + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 1); + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_MID_MH_IDX].dpms_on_cm_idx = 1; + } + /* high mh */ + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 1); + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_ps_idx = + r600_pm_get_type_index(rdev, POWER_STATE_TYPE_PERFORMANCE, 1); + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_off_cm_idx = 0; + rdev->pm.profiles[PM_PROFILE_HIGH_MH_IDX].dpms_on_cm_idx = 2; + } + } +} + +void r600_pm_misc(struct radeon_device *rdev) +{ + int req_ps_idx = rdev->pm.requested_power_state_index; + int req_cm_idx = rdev->pm.requested_clock_mode_index; + struct radeon_power_state *ps = &rdev->pm.power_state[req_ps_idx]; + struct radeon_voltage *voltage = &ps->clock_info[req_cm_idx].voltage; + + if ((voltage->type == VOLTAGE_SW) && voltage->voltage) { + if (voltage->voltage != rdev->pm.current_vddc) { + radeon_atom_set_voltage(rdev, voltage->voltage); + rdev->pm.current_vddc = voltage->voltage; + DRM_DEBUG("Setting: v: %d\n", voltage->voltage); + } + } +} + +bool r600_gui_idle(struct radeon_device *rdev) +{ + if (RREG32(GRBM_STATUS) & GUI_ACTIVE) + return false; + else + return true; +} /* hpd for digital panel detect/disconnect */ bool r600_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd) @@ -594,7 +1102,7 @@ static void r600_mc_program(struct radeon_device *rdev) WREG32(MC_VM_FB_LOCATION, tmp); WREG32(HDP_NONSURFACE_BASE, (rdev->mc.vram_start >> 8)); WREG32(HDP_NONSURFACE_INFO, (2 << 7)); - WREG32(HDP_NONSURFACE_SIZE, rdev->mc.mc_vram_size | 0x3FF); + WREG32(HDP_NONSURFACE_SIZE, 0x3FFFFFFF); if (rdev->flags & RADEON_IS_AGP) { WREG32(MC_VM_AGP_TOP, rdev->mc.gtt_end >> 22); WREG32(MC_VM_AGP_BOT, rdev->mc.gtt_start >> 22); @@ -671,6 +1179,7 @@ void r600_vram_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc) if (rdev->flags & RADEON_IS_IGP) base = (RREG32(MC_VM_FB_LOCATION) & 0xFFFF) << 24; radeon_vram_location(rdev, &rdev->mc, base); + rdev->mc.gtt_base_align = 0; radeon_gtt_location(rdev, mc); } } @@ -714,15 +1223,12 @@ int r600_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.visible_vram_size = rdev->mc.aper_size; - /* FIXME remove this once we support unmappable VRAM */ - if (rdev->mc.mc_vram_size > rdev->mc.aper_size) { - rdev->mc.mc_vram_size = rdev->mc.aper_size; - rdev->mc.real_vram_size = rdev->mc.aper_size; - } r600_vram_gtt_location(rdev, &rdev->mc); - if (rdev->flags & RADEON_IS_IGP) + if (rdev->flags & RADEON_IS_IGP) { + rs690_pm_info(rdev); rdev->mc.igp_sideport_enabled = radeon_atombios_sideport_present(rdev); + } radeon_update_bandwidth_info(rdev); return 0; } @@ -750,7 +1256,6 @@ int r600_gpu_soft_reset(struct radeon_device *rdev) S_008014_DB2_BUSY(1) | S_008014_DB3_BUSY(1) | S_008014_CB0_BUSY(1) | S_008014_CB1_BUSY(1) | S_008014_CB2_BUSY(1) | S_008014_CB3_BUSY(1); - u32 srbm_reset = 0; u32 tmp; dev_info(rdev->dev, "GPU softreset \n"); @@ -765,7 +1270,7 @@ int r600_gpu_soft_reset(struct radeon_device *rdev) dev_warn(rdev->dev, "Wait for MC idle timedout !\n"); } /* Disable CP parsing/prefetching */ - WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(0xff)); + WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(1)); /* Check if any of the rendering block is busy and reset it */ if ((RREG32(R_008010_GRBM_STATUS) & grbm_busy_mask) || (RREG32(R_008014_GRBM_STATUS2) & grbm2_busy_mask)) { @@ -784,72 +1289,56 @@ int r600_gpu_soft_reset(struct radeon_device *rdev) S_008020_SOFT_RESET_VGT(1); dev_info(rdev->dev, " R_008020_GRBM_SOFT_RESET=0x%08X\n", tmp); WREG32(R_008020_GRBM_SOFT_RESET, tmp); - (void)RREG32(R_008020_GRBM_SOFT_RESET); - udelay(50); + RREG32(R_008020_GRBM_SOFT_RESET); + mdelay(15); WREG32(R_008020_GRBM_SOFT_RESET, 0); - (void)RREG32(R_008020_GRBM_SOFT_RESET); } /* Reset CP (we always reset CP) */ tmp = S_008020_SOFT_RESET_CP(1); dev_info(rdev->dev, "R_008020_GRBM_SOFT_RESET=0x%08X\n", tmp); WREG32(R_008020_GRBM_SOFT_RESET, tmp); - (void)RREG32(R_008020_GRBM_SOFT_RESET); - udelay(50); + RREG32(R_008020_GRBM_SOFT_RESET); + mdelay(15); WREG32(R_008020_GRBM_SOFT_RESET, 0); - (void)RREG32(R_008020_GRBM_SOFT_RESET); - /* Reset others GPU block if necessary */ - if (G_000E50_RLC_BUSY(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_RLC(1); - if (G_000E50_GRBM_RQ_PENDING(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_GRBM(1); - if (G_000E50_HI_RQ_PENDING(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_IH(1); - if (G_000E50_VMC_BUSY(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_VMC(1); - if (G_000E50_MCB_BUSY(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_MC(1); - if (G_000E50_MCDZ_BUSY(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_MC(1); - if (G_000E50_MCDY_BUSY(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_MC(1); - if (G_000E50_MCDX_BUSY(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_MC(1); - if (G_000E50_MCDW_BUSY(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_MC(1); - if (G_000E50_RLC_BUSY(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_RLC(1); - if (G_000E50_SEM_BUSY(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_SEM(1); - if (G_000E50_BIF_BUSY(RREG32(R_000E50_SRBM_STATUS))) - srbm_reset |= S_000E60_SOFT_RESET_BIF(1); - dev_info(rdev->dev, " R_000E60_SRBM_SOFT_RESET=0x%08X\n", srbm_reset); - WREG32(R_000E60_SRBM_SOFT_RESET, srbm_reset); - (void)RREG32(R_000E60_SRBM_SOFT_RESET); - udelay(50); - WREG32(R_000E60_SRBM_SOFT_RESET, 0); - (void)RREG32(R_000E60_SRBM_SOFT_RESET); - WREG32(R_000E60_SRBM_SOFT_RESET, srbm_reset); - (void)RREG32(R_000E60_SRBM_SOFT_RESET); - udelay(50); - WREG32(R_000E60_SRBM_SOFT_RESET, 0); - (void)RREG32(R_000E60_SRBM_SOFT_RESET); /* Wait a little for things to settle down */ - udelay(50); + mdelay(1); dev_info(rdev->dev, " R_008010_GRBM_STATUS=0x%08X\n", RREG32(R_008010_GRBM_STATUS)); dev_info(rdev->dev, " R_008014_GRBM_STATUS2=0x%08X\n", RREG32(R_008014_GRBM_STATUS2)); dev_info(rdev->dev, " R_000E50_SRBM_STATUS=0x%08X\n", RREG32(R_000E50_SRBM_STATUS)); - /* After reset we need to reinit the asic as GPU often endup in an - * incoherent state. - */ - atom_asic_init(rdev->mode_info.atom_context); rv515_mc_resume(rdev, &save); return 0; } -int r600_gpu_reset(struct radeon_device *rdev) +bool r600_gpu_is_lockup(struct radeon_device *rdev) +{ + u32 srbm_status; + u32 grbm_status; + u32 grbm_status2; + int r; + + srbm_status = RREG32(R_000E50_SRBM_STATUS); + grbm_status = RREG32(R_008010_GRBM_STATUS); + grbm_status2 = RREG32(R_008014_GRBM_STATUS2); + if (!G_008010_GUI_ACTIVE(grbm_status)) { + r100_gpu_lockup_update(&rdev->config.r300.lockup, &rdev->cp); + return false; + } + /* force CP activities */ + r = radeon_ring_lock(rdev, 2); + if (!r) { + /* PACKET2 NOP */ + radeon_ring_write(rdev, 0x80000000); + radeon_ring_write(rdev, 0x80000000); + radeon_ring_unlock_commit(rdev); + } + rdev->cp.rptr = RREG32(R600_CP_RB_RPTR); + return r100_gpu_cp_is_lockup(rdev, &rdev->config.r300.lockup, &rdev->cp); +} + +int r600_asic_reset(struct radeon_device *rdev) { return r600_gpu_soft_reset(rdev); } @@ -1467,10 +1956,31 @@ int r600_init_microcode(struct radeon_device *rdev) chip_name = "RV710"; rlc_chip_name = "R700"; break; + case CHIP_CEDAR: + chip_name = "CEDAR"; + rlc_chip_name = "CEDAR"; + break; + case CHIP_REDWOOD: + chip_name = "REDWOOD"; + rlc_chip_name = "REDWOOD"; + break; + case CHIP_JUNIPER: + chip_name = "JUNIPER"; + rlc_chip_name = "JUNIPER"; + break; + case CHIP_CYPRESS: + case CHIP_HEMLOCK: + chip_name = "CYPRESS"; + rlc_chip_name = "CYPRESS"; + break; default: BUG(); } - if (rdev->family >= CHIP_RV770) { + if (rdev->family >= CHIP_CEDAR) { + pfp_req_size = EVERGREEN_PFP_UCODE_SIZE * 4; + me_req_size = EVERGREEN_PM4_UCODE_SIZE * 4; + rlc_req_size = EVERGREEN_RLC_UCODE_SIZE * 4; + } else if (rdev->family >= CHIP_RV770) { pfp_req_size = R700_PFP_UCODE_SIZE * 4; me_req_size = R700_PM4_UCODE_SIZE * 4; rlc_req_size = R700_RLC_UCODE_SIZE * 4; @@ -1584,12 +2094,15 @@ int r600_cp_start(struct radeon_device *rdev) } radeon_ring_write(rdev, PACKET3(PACKET3_ME_INITIALIZE, 5)); radeon_ring_write(rdev, 0x1); - if (rdev->family < CHIP_RV770) { - radeon_ring_write(rdev, 0x3); - radeon_ring_write(rdev, rdev->config.r600.max_hw_contexts - 1); - } else { + if (rdev->family >= CHIP_CEDAR) { + radeon_ring_write(rdev, 0x0); + radeon_ring_write(rdev, rdev->config.evergreen.max_hw_contexts - 1); + } else if (rdev->family >= CHIP_RV770) { radeon_ring_write(rdev, 0x0); radeon_ring_write(rdev, rdev->config.rv770.max_hw_contexts - 1); + } else { + radeon_ring_write(rdev, 0x3); + radeon_ring_write(rdev, rdev->config.r600.max_hw_contexts - 1); } radeon_ring_write(rdev, PACKET3_ME_INITIALIZE_DEVICE_ID(1)); radeon_ring_write(rdev, 0); @@ -2051,8 +2564,6 @@ int r600_init(struct radeon_device *rdev) r = radeon_clocks_init(rdev); if (r) return r; - /* Initialize power management */ - radeon_pm_init(rdev); /* Fence driver */ r = radeon_fence_driver_init(rdev); if (r) @@ -2117,7 +2628,6 @@ int r600_init(struct radeon_device *rdev) void r600_fini(struct radeon_device *rdev) { - radeon_pm_fini(rdev); r600_audio_fini(rdev); r600_blit_fini(rdev); r600_cp_fini(rdev); @@ -2290,10 +2800,11 @@ static void r600_ih_ring_fini(struct radeon_device *rdev) } } -static void r600_rlc_stop(struct radeon_device *rdev) +void r600_rlc_stop(struct radeon_device *rdev) { - if (rdev->family >= CHIP_RV770) { + if ((rdev->family >= CHIP_RV770) && + (rdev->family <= CHIP_RV740)) { /* r7xx asics need to soft reset RLC before halting */ WREG32(SRBM_SOFT_RESET, SOFT_RESET_RLC); RREG32(SRBM_SOFT_RESET); @@ -2330,7 +2841,12 @@ static int r600_rlc_init(struct radeon_device *rdev) WREG32(RLC_UCODE_CNTL, 0); fw_data = (const __be32 *)rdev->rlc_fw->data; - if (rdev->family >= CHIP_RV770) { + if (rdev->family >= CHIP_CEDAR) { + for (i = 0; i < EVERGREEN_RLC_UCODE_SIZE; i++) { + WREG32(RLC_UCODE_ADDR, i); + WREG32(RLC_UCODE_DATA, be32_to_cpup(fw_data++)); + } + } else if (rdev->family >= CHIP_RV770) { for (i = 0; i < R700_RLC_UCODE_SIZE; i++) { WREG32(RLC_UCODE_ADDR, i); WREG32(RLC_UCODE_DATA, be32_to_cpup(fw_data++)); @@ -2360,7 +2876,7 @@ static void r600_enable_interrupts(struct radeon_device *rdev) rdev->ih.enabled = true; } -static void r600_disable_interrupts(struct radeon_device *rdev) +void r600_disable_interrupts(struct radeon_device *rdev) { u32 ih_rb_cntl = RREG32(IH_RB_CNTL); u32 ih_cntl = RREG32(IH_CNTL); @@ -2475,7 +2991,10 @@ int r600_irq_init(struct radeon_device *rdev) WREG32(IH_CNTL, ih_cntl); /* force the active interrupt state to all disabled */ - r600_disable_interrupt_state(rdev); + if (rdev->family >= CHIP_CEDAR) + evergreen_disable_interrupt_state(rdev); + else + r600_disable_interrupt_state(rdev); /* enable irqs */ r600_enable_interrupts(rdev); @@ -2485,7 +3004,7 @@ int r600_irq_init(struct radeon_device *rdev) void r600_irq_suspend(struct radeon_device *rdev) { - r600_disable_interrupts(rdev); + r600_irq_disable(rdev); r600_rlc_stop(rdev); } @@ -2500,6 +3019,8 @@ int r600_irq_set(struct radeon_device *rdev) u32 cp_int_cntl = CNTX_BUSY_INT_ENABLE | CNTX_EMPTY_INT_ENABLE; u32 mode_int = 0; u32 hpd1, hpd2, hpd3, hpd4 = 0, hpd5 = 0, hpd6 = 0; + u32 grbm_int_cntl = 0; + u32 hdmi1, hdmi2; if (!rdev->irq.installed) { WARN(1, "Can't enable IRQ/MSI because no handler is installed.\n"); @@ -2513,7 +3034,9 @@ int r600_irq_set(struct radeon_device *rdev) return 0; } + hdmi1 = RREG32(R600_HDMI_BLOCK1 + R600_HDMI_CNTL) & ~R600_HDMI_INT_EN; if (ASIC_IS_DCE3(rdev)) { + hdmi2 = RREG32(R600_HDMI_BLOCK3 + R600_HDMI_CNTL) & ~R600_HDMI_INT_EN; hpd1 = RREG32(DC_HPD1_INT_CONTROL) & ~DC_HPDx_INT_EN; hpd2 = RREG32(DC_HPD2_INT_CONTROL) & ~DC_HPDx_INT_EN; hpd3 = RREG32(DC_HPD3_INT_CONTROL) & ~DC_HPDx_INT_EN; @@ -2523,6 +3046,7 @@ int r600_irq_set(struct radeon_device *rdev) hpd6 = RREG32(DC_HPD6_INT_CONTROL) & ~DC_HPDx_INT_EN; } } else { + hdmi2 = RREG32(R600_HDMI_BLOCK2 + R600_HDMI_CNTL) & ~R600_HDMI_INT_EN; hpd1 = RREG32(DC_HOT_PLUG_DETECT1_INT_CONTROL) & ~DC_HPDx_INT_EN; hpd2 = RREG32(DC_HOT_PLUG_DETECT2_INT_CONTROL) & ~DC_HPDx_INT_EN; hpd3 = RREG32(DC_HOT_PLUG_DETECT3_INT_CONTROL) & ~DC_HPDx_INT_EN; @@ -2564,10 +3088,25 @@ int r600_irq_set(struct radeon_device *rdev) DRM_DEBUG("r600_irq_set: hpd 6\n"); hpd6 |= DC_HPDx_INT_EN; } + if (rdev->irq.hdmi[0]) { + DRM_DEBUG("r600_irq_set: hdmi 1\n"); + hdmi1 |= R600_HDMI_INT_EN; + } + if (rdev->irq.hdmi[1]) { + DRM_DEBUG("r600_irq_set: hdmi 2\n"); + hdmi2 |= R600_HDMI_INT_EN; + } + if (rdev->irq.gui_idle) { + DRM_DEBUG("gui idle\n"); + grbm_int_cntl |= GUI_IDLE_INT_ENABLE; + } WREG32(CP_INT_CNTL, cp_int_cntl); WREG32(DxMODE_INT_MASK, mode_int); + WREG32(GRBM_INT_CNTL, grbm_int_cntl); + WREG32(R600_HDMI_BLOCK1 + R600_HDMI_CNTL, hdmi1); if (ASIC_IS_DCE3(rdev)) { + WREG32(R600_HDMI_BLOCK3 + R600_HDMI_CNTL, hdmi2); WREG32(DC_HPD1_INT_CONTROL, hpd1); WREG32(DC_HPD2_INT_CONTROL, hpd2); WREG32(DC_HPD3_INT_CONTROL, hpd3); @@ -2577,6 +3116,7 @@ int r600_irq_set(struct radeon_device *rdev) WREG32(DC_HPD6_INT_CONTROL, hpd6); } } else { + WREG32(R600_HDMI_BLOCK2 + R600_HDMI_CNTL, hdmi2); WREG32(DC_HOT_PLUG_DETECT1_INT_CONTROL, hpd1); WREG32(DC_HOT_PLUG_DETECT2_INT_CONTROL, hpd2); WREG32(DC_HOT_PLUG_DETECT3_INT_CONTROL, hpd3); @@ -2660,6 +3200,18 @@ static inline void r600_irq_ack(struct radeon_device *rdev, WREG32(DC_HPD6_INT_CONTROL, tmp); } } + if (RREG32(R600_HDMI_BLOCK1 + R600_HDMI_STATUS) & R600_HDMI_INT_PENDING) { + WREG32_P(R600_HDMI_BLOCK1 + R600_HDMI_CNTL, R600_HDMI_INT_ACK, ~R600_HDMI_INT_ACK); + } + if (ASIC_IS_DCE3(rdev)) { + if (RREG32(R600_HDMI_BLOCK3 + R600_HDMI_STATUS) & R600_HDMI_INT_PENDING) { + WREG32_P(R600_HDMI_BLOCK3 + R600_HDMI_CNTL, R600_HDMI_INT_ACK, ~R600_HDMI_INT_ACK); + } + } else { + if (RREG32(R600_HDMI_BLOCK2 + R600_HDMI_STATUS) & R600_HDMI_INT_PENDING) { + WREG32_P(R600_HDMI_BLOCK2 + R600_HDMI_CNTL, R600_HDMI_INT_ACK, ~R600_HDMI_INT_ACK); + } + } } void r600_irq_disable(struct radeon_device *rdev) @@ -2713,6 +3265,8 @@ static inline u32 r600_get_ih_wptr(struct radeon_device *rdev) * 19 1 FP Hot plug detection B * 19 2 DAC A auto-detection * 19 3 DAC B auto-detection + * 21 4 HDMI block A + * 21 5 HDMI block B * 176 - CP_INT RB * 177 - CP_INT IB1 * 178 - CP_INT IB2 @@ -2852,6 +3406,10 @@ restart_ih: break; } break; + case 21: /* HDMI */ + DRM_DEBUG("IH: HDMI: 0x%x\n", src_data); + r600_audio_schedule_polling(rdev); + break; case 176: /* CP_INT in ring buffer */ case 177: /* CP_INT in IB1 */ case 178: /* CP_INT in IB2 */ @@ -2861,6 +3419,11 @@ restart_ih: case 181: /* CP EOP event */ DRM_DEBUG("IH: CP EOP\n"); break; + case 233: /* GUI IDLE */ + DRM_DEBUG("IH: CP EOP\n"); + rdev->pm.gui_idle = true; + wake_up(&rdev->irq.idle_queue); + break; default: DRM_DEBUG("Unhandled interrupt: %d %d\n", src_id, src_data); break; diff --git a/drivers/gpu/drm/radeon/r600_audio.c b/drivers/gpu/drm/radeon/r600_audio.c index 1d898051c631..2b26553c352c 100644 --- a/drivers/gpu/drm/radeon/r600_audio.c +++ b/drivers/gpu/drm/radeon/r600_audio.c @@ -44,7 +44,7 @@ static int r600_audio_chipset_supported(struct radeon_device *rdev) /* * current number of channels */ -static int r600_audio_channels(struct radeon_device *rdev) +int r600_audio_channels(struct radeon_device *rdev) { return (RREG32(R600_AUDIO_RATE_BPS_CHANNEL) & 0x7) + 1; } @@ -52,7 +52,7 @@ static int r600_audio_channels(struct radeon_device *rdev) /* * current bits per sample */ -static int r600_audio_bits_per_sample(struct radeon_device *rdev) +int r600_audio_bits_per_sample(struct radeon_device *rdev) { uint32_t value = (RREG32(R600_AUDIO_RATE_BPS_CHANNEL) & 0xF0) >> 4; switch (value) { @@ -71,7 +71,7 @@ static int r600_audio_bits_per_sample(struct radeon_device *rdev) /* * current sampling rate in HZ */ -static int r600_audio_rate(struct radeon_device *rdev) +int r600_audio_rate(struct radeon_device *rdev) { uint32_t value = RREG32(R600_AUDIO_RATE_BPS_CHANNEL); uint32_t result; @@ -90,7 +90,7 @@ static int r600_audio_rate(struct radeon_device *rdev) /* * iec 60958 status bits */ -static uint8_t r600_audio_status_bits(struct radeon_device *rdev) +uint8_t r600_audio_status_bits(struct radeon_device *rdev) { return RREG32(R600_AUDIO_STATUS_BITS) & 0xff; } @@ -98,12 +98,21 @@ static uint8_t r600_audio_status_bits(struct radeon_device *rdev) /* * iec 60958 category code */ -static uint8_t r600_audio_category_code(struct radeon_device *rdev) +uint8_t r600_audio_category_code(struct radeon_device *rdev) { return (RREG32(R600_AUDIO_STATUS_BITS) >> 8) & 0xff; } /* + * schedule next audio update event + */ +void r600_audio_schedule_polling(struct radeon_device *rdev) +{ + mod_timer(&rdev->audio_timer, + jiffies + msecs_to_jiffies(AUDIO_TIMER_INTERVALL)); +} + +/* * update all hdmi interfaces with current audio parameters */ static void r600_audio_update_hdmi(unsigned long param) @@ -118,7 +127,7 @@ static void r600_audio_update_hdmi(unsigned long param) uint8_t category_code = r600_audio_category_code(rdev); struct drm_encoder *encoder; - int changes = 0; + int changes = 0, still_going = 0; changes |= channels != rdev->audio_channels; changes |= rate != rdev->audio_rate; @@ -135,15 +144,13 @@ static void r600_audio_update_hdmi(unsigned long param) } list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); + still_going |= radeon_encoder->audio_polling_active; if (changes || r600_hdmi_buffer_status_changed(encoder)) - r600_hdmi_update_audio_settings( - encoder, channels, - rate, bps, status_bits, - category_code); + r600_hdmi_update_audio_settings(encoder); } - mod_timer(&rdev->audio_timer, - jiffies + msecs_to_jiffies(AUDIO_TIMER_INTERVALL)); + if(still_going) r600_audio_schedule_polling(rdev); } /* @@ -176,9 +183,34 @@ int r600_audio_init(struct radeon_device *rdev) r600_audio_update_hdmi, (unsigned long)rdev); + return 0; +} + +/* + * enable the polling timer, to check for status changes + */ +void r600_audio_enable_polling(struct drm_encoder *encoder) +{ + struct drm_device *dev = encoder->dev; + struct radeon_device *rdev = dev->dev_private; + struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); + + DRM_DEBUG("r600_audio_enable_polling: %d", radeon_encoder->audio_polling_active); + if (radeon_encoder->audio_polling_active) + return; + + radeon_encoder->audio_polling_active = 1; mod_timer(&rdev->audio_timer, jiffies + 1); +} - return 0; +/* + * disable the polling timer, so we get no more status updates + */ +void r600_audio_disable_polling(struct drm_encoder *encoder) +{ + struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); + DRM_DEBUG("r600_audio_disable_polling: %d", radeon_encoder->audio_polling_active); + radeon_encoder->audio_polling_active = 0; } /* diff --git a/drivers/gpu/drm/radeon/r600_blit.c b/drivers/gpu/drm/radeon/r600_blit.c index f4fb88ece2bb..ca5c29f70779 100644 --- a/drivers/gpu/drm/radeon/r600_blit.c +++ b/drivers/gpu/drm/radeon/r600_blit.c @@ -538,9 +538,12 @@ int r600_prepare_blit_copy(struct drm_device *dev, struct drm_file *file_priv) { drm_radeon_private_t *dev_priv = dev->dev_private; + int ret; DRM_DEBUG("\n"); - r600_nomm_get_vb(dev); + ret = r600_nomm_get_vb(dev); + if (ret) + return ret; dev_priv->blit_vb->file_priv = file_priv; diff --git a/drivers/gpu/drm/radeon/r600_blit_kms.c b/drivers/gpu/drm/radeon/r600_blit_kms.c index f6c6c77db7e0..d13622ae74e9 100644 --- a/drivers/gpu/drm/radeon/r600_blit_kms.c +++ b/drivers/gpu/drm/radeon/r600_blit_kms.c @@ -447,6 +447,9 @@ int r600_blit_init(struct radeon_device *rdev) u32 packet2s[16]; int num_packet2s = 0; + /* don't reinitialize blit */ + if (rdev->r600_blit.shader_obj) + return 0; mutex_init(&rdev->r600_blit.mutex); rdev->r600_blit.state_offset = 0; diff --git a/drivers/gpu/drm/radeon/r600_cs.c b/drivers/gpu/drm/radeon/r600_cs.c index c39c1bc13016..144c32d37136 100644 --- a/drivers/gpu/drm/radeon/r600_cs.c +++ b/drivers/gpu/drm/radeon/r600_cs.c @@ -585,7 +585,7 @@ static int r600_cs_packet_parse_vline(struct radeon_cs_parser *p) header = radeon_get_ib_value(p, h_idx); crtc_id = radeon_get_ib_value(p, h_idx + 2 + 7 + 1); reg = CP_PACKET0_GET_REG(header); - mutex_lock(&p->rdev->ddev->mode_config.mutex); + obj = drm_mode_object_find(p->rdev->ddev, crtc_id, DRM_MODE_OBJECT_CRTC); if (!obj) { DRM_ERROR("cannot find crtc %d\n", crtc_id); @@ -620,7 +620,6 @@ static int r600_cs_packet_parse_vline(struct radeon_cs_parser *p) ib[h_idx + 4] = AVIVO_D2MODE_VLINE_STATUS >> 2; } out: - mutex_unlock(&p->rdev->ddev->mode_config.mutex); return r; } diff --git a/drivers/gpu/drm/radeon/r600_hdmi.c b/drivers/gpu/drm/radeon/r600_hdmi.c index 2616b822ba68..26b4bc9d89a5 100644 --- a/drivers/gpu/drm/radeon/r600_hdmi.c +++ b/drivers/gpu/drm/radeon/r600_hdmi.c @@ -290,17 +290,15 @@ void r600_hdmi_audio_workaround(struct drm_encoder *encoder) if (!offset) return; - if (r600_hdmi_is_audio_buffer_filled(encoder)) { - /* disable audio workaround and start delivering of audio frames */ - WREG32_P(offset+R600_HDMI_CNTL, 0x00000001, ~0x00001001); + if (!radeon_encoder->hdmi_audio_workaround || + r600_hdmi_is_audio_buffer_filled(encoder)) { - } else if (radeon_encoder->hdmi_audio_workaround) { - /* enable audio workaround and start delivering of audio frames */ - WREG32_P(offset+R600_HDMI_CNTL, 0x00001001, ~0x00001001); + /* disable audio workaround */ + WREG32_P(offset+R600_HDMI_CNTL, 0x00000001, ~0x00001001); } else { - /* disable audio workaround and stop delivering of audio frames */ - WREG32_P(offset+R600_HDMI_CNTL, 0x00000000, ~0x00001001); + /* enable audio workaround */ + WREG32_P(offset+R600_HDMI_CNTL, 0x00001001, ~0x00001001); } } @@ -345,25 +343,23 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod /* audio packets per line, does anyone know how to calc this ? */ WREG32_P(offset+R600_HDMI_CNTL, 0x00040000, ~0x001F0000); - - /* update? reset? don't realy know */ - WREG32_P(offset+R600_HDMI_CNTL, 0x14000000, ~0x14000000); } /* * update settings with current parameters from audio engine */ -void r600_hdmi_update_audio_settings(struct drm_encoder *encoder, - int channels, - int rate, - int bps, - uint8_t status_bits, - uint8_t category_code) +void r600_hdmi_update_audio_settings(struct drm_encoder *encoder) { struct drm_device *dev = encoder->dev; struct radeon_device *rdev = dev->dev_private; uint32_t offset = to_radeon_encoder(encoder)->hdmi_offset; + int channels = r600_audio_channels(rdev); + int rate = r600_audio_rate(rdev); + int bps = r600_audio_bits_per_sample(rdev); + uint8_t status_bits = r600_audio_status_bits(rdev); + uint8_t category_code = r600_audio_category_code(rdev); + uint32_t iec; if (!offset) @@ -415,9 +411,6 @@ void r600_hdmi_update_audio_settings(struct drm_encoder *encoder, r600_hdmi_audioinfoframe(encoder, channels-1, 0, 0, 0, 0, 0, 0, 0); r600_hdmi_audio_workaround(encoder); - - /* update? reset? don't realy know */ - WREG32_P(offset+R600_HDMI_CNTL, 0x04000000, ~0x04000000); } static int r600_hdmi_find_free_block(struct drm_device *dev) @@ -486,6 +479,7 @@ void r600_hdmi_enable(struct drm_encoder *encoder) struct drm_device *dev = encoder->dev; struct radeon_device *rdev = dev->dev_private; struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); + uint32_t offset; if (ASIC_IS_DCE4(rdev)) return; @@ -499,10 +493,10 @@ void r600_hdmi_enable(struct drm_encoder *encoder) } } + offset = radeon_encoder->hdmi_offset; if (ASIC_IS_DCE32(rdev) && !ASIC_IS_DCE4(rdev)) { WREG32_P(radeon_encoder->hdmi_config_offset + 0x4, 0x1, ~0x1); } else if (rdev->family >= CHIP_R600 && !ASIC_IS_DCE3(rdev)) { - int offset = radeon_encoder->hdmi_offset; switch (radeon_encoder->encoder_id) { case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1: WREG32_P(AVIVO_TMDSA_CNTL, 0x4, ~0x4); @@ -518,6 +512,21 @@ void r600_hdmi_enable(struct drm_encoder *encoder) } } + if (rdev->irq.installed + && rdev->family != CHIP_RS600 + && rdev->family != CHIP_RS690 + && rdev->family != CHIP_RS740) { + + /* if irq is available use it */ + rdev->irq.hdmi[offset == R600_HDMI_BLOCK1 ? 0 : 1] = true; + radeon_irq_set(rdev); + + r600_audio_disable_polling(encoder); + } else { + /* if not fallback to polling */ + r600_audio_enable_polling(encoder); + } + DRM_DEBUG("Enabling HDMI interface @ 0x%04X for encoder 0x%x\n", radeon_encoder->hdmi_offset, radeon_encoder->encoder_id); } @@ -530,22 +539,30 @@ void r600_hdmi_disable(struct drm_encoder *encoder) struct drm_device *dev = encoder->dev; struct radeon_device *rdev = dev->dev_private; struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); + uint32_t offset; if (ASIC_IS_DCE4(rdev)) return; - if (!radeon_encoder->hdmi_offset) { + offset = radeon_encoder->hdmi_offset; + if (!offset) { dev_err(rdev->dev, "Disabling not enabled HDMI\n"); return; } DRM_DEBUG("Disabling HDMI interface @ 0x%04X for encoder 0x%x\n", - radeon_encoder->hdmi_offset, radeon_encoder->encoder_id); + offset, radeon_encoder->encoder_id); + + /* disable irq */ + rdev->irq.hdmi[offset == R600_HDMI_BLOCK1 ? 0 : 1] = false; + radeon_irq_set(rdev); + + /* disable polling */ + r600_audio_disable_polling(encoder); if (ASIC_IS_DCE32(rdev) && !ASIC_IS_DCE4(rdev)) { WREG32_P(radeon_encoder->hdmi_config_offset + 0x4, 0, ~0x1); } else if (rdev->family >= CHIP_R600 && !ASIC_IS_DCE3(rdev)) { - int offset = radeon_encoder->hdmi_offset; switch (radeon_encoder->encoder_id) { case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1: WREG32_P(AVIVO_TMDSA_CNTL, 0, ~0x4); diff --git a/drivers/gpu/drm/radeon/r600_reg.h b/drivers/gpu/drm/radeon/r600_reg.h index 7b1d22370f6e..d84612ae47e0 100644 --- a/drivers/gpu/drm/radeon/r600_reg.h +++ b/drivers/gpu/drm/radeon/r600_reg.h @@ -157,33 +157,36 @@ #define R600_HDMI_BLOCK3 0x7800 /* HDMI registers */ -#define R600_HDMI_ENABLE 0x00 -#define R600_HDMI_STATUS 0x04 -#define R600_HDMI_CNTL 0x08 -#define R600_HDMI_UNKNOWN_0 0x0C -#define R600_HDMI_AUDIOCNTL 0x10 -#define R600_HDMI_VIDEOCNTL 0x14 -#define R600_HDMI_VERSION 0x18 -#define R600_HDMI_UNKNOWN_1 0x28 -#define R600_HDMI_VIDEOINFOFRAME_0 0x54 -#define R600_HDMI_VIDEOINFOFRAME_1 0x58 -#define R600_HDMI_VIDEOINFOFRAME_2 0x5c -#define R600_HDMI_VIDEOINFOFRAME_3 0x60 -#define R600_HDMI_32kHz_CTS 0xac -#define R600_HDMI_32kHz_N 0xb0 -#define R600_HDMI_44_1kHz_CTS 0xb4 -#define R600_HDMI_44_1kHz_N 0xb8 -#define R600_HDMI_48kHz_CTS 0xbc -#define R600_HDMI_48kHz_N 0xc0 -#define R600_HDMI_AUDIOINFOFRAME_0 0xcc -#define R600_HDMI_AUDIOINFOFRAME_1 0xd0 -#define R600_HDMI_IEC60958_1 0xd4 -#define R600_HDMI_IEC60958_2 0xd8 -#define R600_HDMI_UNKNOWN_2 0xdc -#define R600_HDMI_AUDIO_DEBUG_0 0xe0 -#define R600_HDMI_AUDIO_DEBUG_1 0xe4 -#define R600_HDMI_AUDIO_DEBUG_2 0xe8 -#define R600_HDMI_AUDIO_DEBUG_3 0xec +#define R600_HDMI_ENABLE 0x00 +#define R600_HDMI_STATUS 0x04 +# define R600_HDMI_INT_PENDING (1 << 29) +#define R600_HDMI_CNTL 0x08 +# define R600_HDMI_INT_EN (1 << 28) +# define R600_HDMI_INT_ACK (1 << 29) +#define R600_HDMI_UNKNOWN_0 0x0C +#define R600_HDMI_AUDIOCNTL 0x10 +#define R600_HDMI_VIDEOCNTL 0x14 +#define R600_HDMI_VERSION 0x18 +#define R600_HDMI_UNKNOWN_1 0x28 +#define R600_HDMI_VIDEOINFOFRAME_0 0x54 +#define R600_HDMI_VIDEOINFOFRAME_1 0x58 +#define R600_HDMI_VIDEOINFOFRAME_2 0x5c +#define R600_HDMI_VIDEOINFOFRAME_3 0x60 +#define R600_HDMI_32kHz_CTS 0xac +#define R600_HDMI_32kHz_N 0xb0 +#define R600_HDMI_44_1kHz_CTS 0xb4 +#define R600_HDMI_44_1kHz_N 0xb8 +#define R600_HDMI_48kHz_CTS 0xbc +#define R600_HDMI_48kHz_N 0xc0 +#define R600_HDMI_AUDIOINFOFRAME_0 0xcc +#define R600_HDMI_AUDIOINFOFRAME_1 0xd0 +#define R600_HDMI_IEC60958_1 0xd4 +#define R600_HDMI_IEC60958_2 0xd8 +#define R600_HDMI_UNKNOWN_2 0xdc +#define R600_HDMI_AUDIO_DEBUG_0 0xe0 +#define R600_HDMI_AUDIO_DEBUG_1 0xe4 +#define R600_HDMI_AUDIO_DEBUG_2 0xe8 +#define R600_HDMI_AUDIO_DEBUG_3 0xec /* HDMI additional config base register addresses */ #define R600_HDMI_CONFIG1 0x7600 diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 034218c3dbbb..2f94dc66c183 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -89,7 +89,6 @@ extern int radeon_testing; extern int radeon_connector_table; extern int radeon_tv; extern int radeon_new_pll; -extern int radeon_dynpm; extern int radeon_audio; extern int radeon_disp_priority; extern int radeon_hw_i2c; @@ -99,6 +98,7 @@ extern int radeon_hw_i2c; * symbol; */ #define RADEON_MAX_USEC_TIMEOUT 100000 /* 100 ms */ +#define RADEON_FENCE_JIFFIES_TIMEOUT (HZ / 2) /* RADEON_IB_POOL_SIZE must be a power of 2 */ #define RADEON_IB_POOL_SIZE 16 #define RADEON_DEBUGFS_MAX_NUM_FILES 32 @@ -172,8 +172,12 @@ struct radeon_clock { int radeon_pm_init(struct radeon_device *rdev); void radeon_pm_fini(struct radeon_device *rdev); void radeon_pm_compute_clocks(struct radeon_device *rdev); +void radeon_pm_suspend(struct radeon_device *rdev); +void radeon_pm_resume(struct radeon_device *rdev); void radeon_combios_get_power_modes(struct radeon_device *rdev); void radeon_atombios_get_power_modes(struct radeon_device *rdev); +void radeon_atom_set_voltage(struct radeon_device *rdev, u16 level); +void rs690_pm_info(struct radeon_device *rdev); /* * Fences. @@ -182,7 +186,8 @@ struct radeon_fence_driver { uint32_t scratch_reg; atomic_t seq; uint32_t last_seq; - unsigned long count_timeout; + unsigned long last_jiffies; + unsigned long last_timeout; wait_queue_head_t queue; rwlock_t lock; struct list_head created; @@ -197,7 +202,6 @@ struct radeon_fence { struct list_head list; /* protected by radeon_fence.lock */ uint32_t seq; - unsigned long timeout; bool emited; bool signaled; }; @@ -259,6 +263,7 @@ struct radeon_bo_list { unsigned rdomain; unsigned wdomain; u32 tiling_flags; + bool reserved; }; /* @@ -346,6 +351,7 @@ struct radeon_mc { int vram_mtrr; bool vram_is_ddr; bool igp_sideport_enabled; + u64 gtt_base_align; }; bool radeon_combios_sideport_present(struct radeon_device *rdev); @@ -371,10 +377,15 @@ struct radeon_irq { bool installed; bool sw_int; /* FIXME: use a define max crtc rather than hardcode it */ - bool crtc_vblank_int[2]; + bool crtc_vblank_int[6]; wait_queue_head_t vblank_queue; /* FIXME: use defines for max hpd/dacs */ bool hpd[6]; + bool gui_idle; + bool gui_idle_acked; + wait_queue_head_t idle_queue; + /* FIXME: use defines for max HDMI blocks */ + bool hdmi[2]; spinlock_t sw_lock; int sw_refcount; }; @@ -462,7 +473,9 @@ int radeon_ib_test(struct radeon_device *rdev); extern void radeon_ib_bogus_add(struct radeon_device *rdev, struct radeon_ib *ib); /* Ring access between begin & end cannot sleep */ void radeon_ring_free_size(struct radeon_device *rdev); +int radeon_ring_alloc(struct radeon_device *rdev, unsigned ndw); int radeon_ring_lock(struct radeon_device *rdev, unsigned ndw); +void radeon_ring_commit(struct radeon_device *rdev); void radeon_ring_unlock_commit(struct radeon_device *rdev); void radeon_ring_unlock_undo(struct radeon_device *rdev); int radeon_ring_test(struct radeon_device *rdev); @@ -566,6 +579,7 @@ typedef int (*radeon_packet3_check_t)(struct radeon_cs_parser *p, */ int radeon_agp_init(struct radeon_device *rdev); void radeon_agp_resume(struct radeon_device *rdev); +void radeon_agp_suspend(struct radeon_device *rdev); void radeon_agp_fini(struct radeon_device *rdev); @@ -597,17 +611,25 @@ struct radeon_wb { * Equation between gpu/memory clock and available bandwidth is hw dependent * (type of memory, bus size, efficiency, ...) */ -enum radeon_pm_state { - PM_STATE_DISABLED, - PM_STATE_MINIMUM, - PM_STATE_PAUSED, - PM_STATE_ACTIVE + +enum radeon_pm_method { + PM_METHOD_PROFILE, + PM_METHOD_DYNPM, +}; + +enum radeon_dynpm_state { + DYNPM_STATE_DISABLED, + DYNPM_STATE_MINIMUM, + DYNPM_STATE_PAUSED, + DYNPM_STATE_ACTIVE, + DYNPM_STATE_SUSPENDED, }; -enum radeon_pm_action { - PM_ACTION_NONE, - PM_ACTION_MINIMUM, - PM_ACTION_DOWNCLOCK, - PM_ACTION_UPCLOCK +enum radeon_dynpm_action { + DYNPM_ACTION_NONE, + DYNPM_ACTION_MINIMUM, + DYNPM_ACTION_DOWNCLOCK, + DYNPM_ACTION_UPCLOCK, + DYNPM_ACTION_DEFAULT }; enum radeon_voltage_type { @@ -625,11 +647,28 @@ enum radeon_pm_state_type { POWER_STATE_TYPE_PERFORMANCE, }; -enum radeon_pm_clock_mode_type { - POWER_MODE_TYPE_DEFAULT, - POWER_MODE_TYPE_LOW, - POWER_MODE_TYPE_MID, - POWER_MODE_TYPE_HIGH, +enum radeon_pm_profile_type { + PM_PROFILE_DEFAULT, + PM_PROFILE_AUTO, + PM_PROFILE_LOW, + PM_PROFILE_MID, + PM_PROFILE_HIGH, +}; + +#define PM_PROFILE_DEFAULT_IDX 0 +#define PM_PROFILE_LOW_SH_IDX 1 +#define PM_PROFILE_MID_SH_IDX 2 +#define PM_PROFILE_HIGH_SH_IDX 3 +#define PM_PROFILE_LOW_MH_IDX 4 +#define PM_PROFILE_MID_MH_IDX 5 +#define PM_PROFILE_HIGH_MH_IDX 6 +#define PM_PROFILE_MAX 7 + +struct radeon_pm_profile { + int dpms_off_ps_idx; + int dpms_on_ps_idx; + int dpms_off_cm_idx; + int dpms_on_cm_idx; }; struct radeon_voltage { @@ -646,12 +685,8 @@ struct radeon_voltage { u32 voltage; }; -struct radeon_pm_non_clock_info { - /* pcie lanes */ - int pcie_lanes; - /* standardized non-clock flags */ - u32 flags; -}; +/* clock mode flags */ +#define RADEON_PM_MODE_NO_DISPLAY (1 << 0) struct radeon_pm_clock_info { /* memory clock */ @@ -660,10 +695,13 @@ struct radeon_pm_clock_info { u32 sclk; /* voltage info */ struct radeon_voltage voltage; - /* standardized clock flags - not sure we'll need these */ + /* standardized clock flags */ u32 flags; }; +/* state flags */ +#define RADEON_PM_STATE_SINGLE_DISPLAY_ONLY (1 << 0) + struct radeon_power_state { enum radeon_pm_state_type type; /* XXX: use a define for num clock modes */ @@ -671,9 +709,11 @@ struct radeon_power_state { /* number of valid clock modes in this power state */ int num_clock_modes; struct radeon_pm_clock_info *default_clock_mode; - /* non clock info about this state */ - struct radeon_pm_non_clock_info non_clock_info; - bool voltage_drop_active; + /* standardized state flags */ + u32 flags; + u32 misc; /* vbios specific flags */ + u32 misc2; /* vbios specific flags */ + int pcie_lanes; /* pcie lanes */ }; /* @@ -683,14 +723,11 @@ struct radeon_power_state { struct radeon_pm { struct mutex mutex; - struct delayed_work idle_work; - enum radeon_pm_state state; - enum radeon_pm_action planned_action; - unsigned long action_timeout; - bool downclocked; - int active_crtcs; + u32 active_crtcs; + int active_crtc_count; int req_vblank; bool vblank_sync; + bool gui_idle; fixed20_12 max_bandwidth; fixed20_12 igp_sideport_mclk; fixed20_12 igp_system_mclk; @@ -707,12 +744,28 @@ struct radeon_pm { struct radeon_power_state power_state[8]; /* number of valid power states */ int num_power_states; - struct radeon_power_state *current_power_state; - struct radeon_pm_clock_info *current_clock_mode; - struct radeon_power_state *requested_power_state; - struct radeon_pm_clock_info *requested_clock_mode; - struct radeon_power_state *default_power_state; + int current_power_state_index; + int current_clock_mode_index; + int requested_power_state_index; + int requested_clock_mode_index; + int default_power_state_index; + u32 current_sclk; + u32 current_mclk; + u32 current_vddc; struct radeon_i2c_chan *i2c_bus; + /* selected pm method */ + enum radeon_pm_method pm_method; + /* dynpm power management */ + struct delayed_work dynpm_idle_work; + enum radeon_dynpm_state dynpm_state; + enum radeon_dynpm_action dynpm_planned_action; + unsigned long dynpm_action_timeout; + bool dynpm_can_upclock; + bool dynpm_can_downclock; + /* profile-based power management */ + enum radeon_pm_profile_type profile; + int profile_index; + struct radeon_pm_profile profiles[PM_PROFILE_MAX]; }; @@ -746,7 +799,8 @@ struct radeon_asic { int (*resume)(struct radeon_device *rdev); int (*suspend)(struct radeon_device *rdev); void (*vga_set_state)(struct radeon_device *rdev, bool state); - int (*gpu_reset)(struct radeon_device *rdev); + bool (*gpu_is_lockup)(struct radeon_device *rdev); + int (*asic_reset)(struct radeon_device *rdev); void (*gart_tlb_flush)(struct radeon_device *rdev); int (*gart_set_page)(struct radeon_device *rdev, int i, uint64_t addr); int (*cp_init)(struct radeon_device *rdev, unsigned ring_size); @@ -799,44 +853,84 @@ struct radeon_asic { * through ring. */ void (*ioctl_wait_idle)(struct radeon_device *rdev, struct radeon_bo *bo); + bool (*gui_idle)(struct radeon_device *rdev); + /* power management */ + void (*pm_misc)(struct radeon_device *rdev); + void (*pm_prepare)(struct radeon_device *rdev); + void (*pm_finish)(struct radeon_device *rdev); + void (*pm_init_profile)(struct radeon_device *rdev); + void (*pm_get_dynpm_state)(struct radeon_device *rdev); }; /* * Asic structures */ +struct r100_gpu_lockup { + unsigned long last_jiffies; + u32 last_cp_rptr; +}; + struct r100_asic { - const unsigned *reg_safe_bm; - unsigned reg_safe_bm_size; - u32 hdp_cntl; + const unsigned *reg_safe_bm; + unsigned reg_safe_bm_size; + u32 hdp_cntl; + struct r100_gpu_lockup lockup; }; struct r300_asic { - const unsigned *reg_safe_bm; - unsigned reg_safe_bm_size; - u32 resync_scratch; - u32 hdp_cntl; + const unsigned *reg_safe_bm; + unsigned reg_safe_bm_size; + u32 resync_scratch; + u32 hdp_cntl; + struct r100_gpu_lockup lockup; }; struct r600_asic { - unsigned max_pipes; - unsigned max_tile_pipes; - unsigned max_simds; - unsigned max_backends; - unsigned max_gprs; - unsigned max_threads; - unsigned max_stack_entries; - unsigned max_hw_contexts; - unsigned max_gs_threads; - unsigned sx_max_export_size; - unsigned sx_max_export_pos_size; - unsigned sx_max_export_smx_size; - unsigned sq_num_cf_insts; - unsigned tiling_nbanks; - unsigned tiling_npipes; - unsigned tiling_group_size; + unsigned max_pipes; + unsigned max_tile_pipes; + unsigned max_simds; + unsigned max_backends; + unsigned max_gprs; + unsigned max_threads; + unsigned max_stack_entries; + unsigned max_hw_contexts; + unsigned max_gs_threads; + unsigned sx_max_export_size; + unsigned sx_max_export_pos_size; + unsigned sx_max_export_smx_size; + unsigned sq_num_cf_insts; + unsigned tiling_nbanks; + unsigned tiling_npipes; + unsigned tiling_group_size; + struct r100_gpu_lockup lockup; }; struct rv770_asic { + unsigned max_pipes; + unsigned max_tile_pipes; + unsigned max_simds; + unsigned max_backends; + unsigned max_gprs; + unsigned max_threads; + unsigned max_stack_entries; + unsigned max_hw_contexts; + unsigned max_gs_threads; + unsigned sx_max_export_size; + unsigned sx_max_export_pos_size; + unsigned sx_max_export_smx_size; + unsigned sq_num_cf_insts; + unsigned sx_num_of_sets; + unsigned sc_prim_fifo_size; + unsigned sc_hiz_tile_fifo_size; + unsigned sc_earlyz_tile_fifo_fize; + unsigned tiling_nbanks; + unsigned tiling_npipes; + unsigned tiling_group_size; + struct r100_gpu_lockup lockup; +}; + +struct evergreen_asic { + unsigned num_ses; unsigned max_pipes; unsigned max_tile_pipes; unsigned max_simds; @@ -853,7 +947,7 @@ struct rv770_asic { unsigned sx_num_of_sets; unsigned sc_prim_fifo_size; unsigned sc_hiz_tile_fifo_size; - unsigned sc_earlyz_tile_fifo_fize; + unsigned sc_earlyz_tile_fifo_size; unsigned tiling_nbanks; unsigned tiling_npipes; unsigned tiling_group_size; @@ -864,6 +958,7 @@ union radeon_asic_config { struct r100_asic r100; struct r600_asic r600; struct rv770_asic rv770; + struct evergreen_asic evergreen; }; /* @@ -927,9 +1022,6 @@ struct radeon_device { bool is_atom_bios; uint16_t bios_header_start; struct radeon_bo *stollen_vga_memory; - struct fb_info *fbdev_info; - struct radeon_bo *fbdev_rbo; - struct radeon_framebuffer *fbdev_rfb; /* Register mmio */ resource_size_t rmmio_base; resource_size_t rmmio_size; @@ -974,6 +1066,7 @@ struct radeon_device { struct work_struct hotplug_work; int num_crtc; /* number of crtcs */ struct mutex dc_hw_i2c_mutex; /* display controller hw i2c mutex */ + struct mutex vram_mutex; /* audio stuff */ struct timer_list audio_timer; @@ -984,6 +1077,7 @@ struct radeon_device { uint8_t audio_category_code; bool powered_down; + struct notifier_block acpi_nb; }; int radeon_device_init(struct radeon_device *rdev, @@ -1145,7 +1239,8 @@ static inline void radeon_ring_write(struct radeon_device *rdev, uint32_t v) #define radeon_suspend(rdev) (rdev)->asic->suspend((rdev)) #define radeon_cs_parse(p) rdev->asic->cs_parse((p)) #define radeon_vga_set_state(rdev, state) (rdev)->asic->vga_set_state((rdev), (state)) -#define radeon_gpu_reset(rdev) (rdev)->asic->gpu_reset((rdev)) +#define radeon_gpu_is_lockup(rdev) (rdev)->asic->gpu_is_lockup((rdev)) +#define radeon_asic_reset(rdev) (rdev)->asic->asic_reset((rdev)) #define radeon_gart_tlb_flush(rdev) (rdev)->asic->gart_tlb_flush((rdev)) #define radeon_gart_set_page(rdev, i, p) (rdev)->asic->gart_set_page((rdev), (i), (p)) #define radeon_cp_commit(rdev) (rdev)->asic->cp_commit((rdev)) @@ -1173,9 +1268,16 @@ static inline void radeon_ring_write(struct radeon_device *rdev, uint32_t v) #define radeon_hpd_fini(rdev) (rdev)->asic->hpd_fini((rdev)) #define radeon_hpd_sense(rdev, hpd) (rdev)->asic->hpd_sense((rdev), (hpd)) #define radeon_hpd_set_polarity(rdev, hpd) (rdev)->asic->hpd_set_polarity((rdev), (hpd)) +#define radeon_gui_idle(rdev) (rdev)->asic->gui_idle((rdev)) +#define radeon_pm_misc(rdev) (rdev)->asic->pm_misc((rdev)) +#define radeon_pm_prepare(rdev) (rdev)->asic->pm_prepare((rdev)) +#define radeon_pm_finish(rdev) (rdev)->asic->pm_finish((rdev)) +#define radeon_pm_init_profile(rdev) (rdev)->asic->pm_init_profile((rdev)) +#define radeon_pm_get_dynpm_state(rdev) (rdev)->asic->pm_get_dynpm_state((rdev)) /* Common functions */ /* AGP */ +extern int radeon_gpu_reset(struct radeon_device *rdev); extern void radeon_agp_disable(struct radeon_device *rdev); extern int radeon_gart_table_vram_pin(struct radeon_device *rdev); extern void radeon_gart_restore(struct radeon_device *rdev); @@ -1200,6 +1302,8 @@ extern int radeon_resume_kms(struct drm_device *dev); extern int radeon_suspend_kms(struct drm_device *dev, pm_message_t state); /* r100,rv100,rs100,rv200,rs200,r200,rv250,rs300,rv280 */ +extern void r100_gpu_lockup_update(struct r100_gpu_lockup *lockup, struct radeon_cp *cp); +extern bool r100_gpu_cp_is_lockup(struct radeon_device *rdev, struct r100_gpu_lockup *lockup, struct radeon_cp *cp); /* rv200,rv250,rv280 */ extern void r200_set_safe_registers(struct radeon_device *rdev); @@ -1260,6 +1364,7 @@ extern void rs690_line_buffer_adjust(struct radeon_device *rdev, extern void r600_vram_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc); extern bool r600_card_posted(struct radeon_device *rdev); extern void r600_cp_stop(struct radeon_device *rdev); +extern int r600_cp_start(struct radeon_device *rdev); extern void r600_ring_init(struct radeon_device *rdev, unsigned ring_size); extern int r600_cp_resume(struct radeon_device *rdev); extern void r600_cp_fini(struct radeon_device *rdev); @@ -1276,29 +1381,39 @@ extern void r600_scratch_init(struct radeon_device *rdev); extern int r600_blit_init(struct radeon_device *rdev); extern void r600_blit_fini(struct radeon_device *rdev); extern int r600_init_microcode(struct radeon_device *rdev); -extern int r600_gpu_reset(struct radeon_device *rdev); +extern int r600_asic_reset(struct radeon_device *rdev); /* r600 irq */ extern int r600_irq_init(struct radeon_device *rdev); extern void r600_irq_fini(struct radeon_device *rdev); extern void r600_ih_ring_init(struct radeon_device *rdev, unsigned ring_size); extern int r600_irq_set(struct radeon_device *rdev); extern void r600_irq_suspend(struct radeon_device *rdev); +extern void r600_disable_interrupts(struct radeon_device *rdev); +extern void r600_rlc_stop(struct radeon_device *rdev); /* r600 audio */ extern int r600_audio_init(struct radeon_device *rdev); extern int r600_audio_tmds_index(struct drm_encoder *encoder); extern void r600_audio_set_clock(struct drm_encoder *encoder, int clock); +extern int r600_audio_channels(struct radeon_device *rdev); +extern int r600_audio_bits_per_sample(struct radeon_device *rdev); +extern int r600_audio_rate(struct radeon_device *rdev); +extern uint8_t r600_audio_status_bits(struct radeon_device *rdev); +extern uint8_t r600_audio_category_code(struct radeon_device *rdev); +extern void r600_audio_schedule_polling(struct radeon_device *rdev); +extern void r600_audio_enable_polling(struct drm_encoder *encoder); +extern void r600_audio_disable_polling(struct drm_encoder *encoder); extern void r600_audio_fini(struct radeon_device *rdev); extern void r600_hdmi_init(struct drm_encoder *encoder); extern void r600_hdmi_enable(struct drm_encoder *encoder); extern void r600_hdmi_disable(struct drm_encoder *encoder); extern void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mode); extern int r600_hdmi_buffer_status_changed(struct drm_encoder *encoder); -extern void r600_hdmi_update_audio_settings(struct drm_encoder *encoder, - int channels, - int rate, - int bps, - uint8_t status_bits, - uint8_t category_code); +extern void r600_hdmi_update_audio_settings(struct drm_encoder *encoder); + +extern void r700_cp_stop(struct radeon_device *rdev); +extern void r700_cp_fini(struct radeon_device *rdev); +extern void evergreen_disable_interrupt_state(struct radeon_device *rdev); +extern int evergreen_irq_set(struct radeon_device *rdev); /* evergreen */ struct evergreen_mc_save { diff --git a/drivers/gpu/drm/radeon/radeon_agp.c b/drivers/gpu/drm/radeon/radeon_agp.c index 28e473f1f56f..f40dfb77f9b1 100644 --- a/drivers/gpu/drm/radeon/radeon_agp.c +++ b/drivers/gpu/drm/radeon/radeon_agp.c @@ -270,3 +270,8 @@ void radeon_agp_fini(struct radeon_device *rdev) } #endif } + +void radeon_agp_suspend(struct radeon_device *rdev) +{ + radeon_agp_fini(rdev); +} diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c index a4b4bc9fa322..646f96f97c77 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.c +++ b/drivers/gpu/drm/radeon/radeon_asic.c @@ -134,7 +134,8 @@ static struct radeon_asic r100_asic = { .suspend = &r100_suspend, .resume = &r100_resume, .vga_set_state = &r100_vga_set_state, - .gpu_reset = &r100_gpu_reset, + .gpu_is_lockup = &r100_gpu_is_lockup, + .asic_reset = &r100_asic_reset, .gart_tlb_flush = &r100_pci_gart_tlb_flush, .gart_set_page = &r100_pci_gart_set_page, .cp_commit = &r100_cp_commit, @@ -164,6 +165,12 @@ static struct radeon_asic r100_asic = { .hpd_sense = &r100_hpd_sense, .hpd_set_polarity = &r100_hpd_set_polarity, .ioctl_wait_idle = NULL, + .gui_idle = &r100_gui_idle, + .pm_misc = &r100_pm_misc, + .pm_prepare = &r100_pm_prepare, + .pm_finish = &r100_pm_finish, + .pm_init_profile = &r100_pm_init_profile, + .pm_get_dynpm_state = &r100_pm_get_dynpm_state, }; static struct radeon_asic r200_asic = { @@ -172,7 +179,8 @@ static struct radeon_asic r200_asic = { .suspend = &r100_suspend, .resume = &r100_resume, .vga_set_state = &r100_vga_set_state, - .gpu_reset = &r100_gpu_reset, + .gpu_is_lockup = &r100_gpu_is_lockup, + .asic_reset = &r100_asic_reset, .gart_tlb_flush = &r100_pci_gart_tlb_flush, .gart_set_page = &r100_pci_gart_set_page, .cp_commit = &r100_cp_commit, @@ -201,6 +209,12 @@ static struct radeon_asic r200_asic = { .hpd_sense = &r100_hpd_sense, .hpd_set_polarity = &r100_hpd_set_polarity, .ioctl_wait_idle = NULL, + .gui_idle = &r100_gui_idle, + .pm_misc = &r100_pm_misc, + .pm_prepare = &r100_pm_prepare, + .pm_finish = &r100_pm_finish, + .pm_init_profile = &r100_pm_init_profile, + .pm_get_dynpm_state = &r100_pm_get_dynpm_state, }; static struct radeon_asic r300_asic = { @@ -209,7 +223,8 @@ static struct radeon_asic r300_asic = { .suspend = &r300_suspend, .resume = &r300_resume, .vga_set_state = &r100_vga_set_state, - .gpu_reset = &r300_gpu_reset, + .gpu_is_lockup = &r300_gpu_is_lockup, + .asic_reset = &r300_asic_reset, .gart_tlb_flush = &r100_pci_gart_tlb_flush, .gart_set_page = &r100_pci_gart_set_page, .cp_commit = &r100_cp_commit, @@ -239,6 +254,12 @@ static struct radeon_asic r300_asic = { .hpd_sense = &r100_hpd_sense, .hpd_set_polarity = &r100_hpd_set_polarity, .ioctl_wait_idle = NULL, + .gui_idle = &r100_gui_idle, + .pm_misc = &r100_pm_misc, + .pm_prepare = &r100_pm_prepare, + .pm_finish = &r100_pm_finish, + .pm_init_profile = &r100_pm_init_profile, + .pm_get_dynpm_state = &r100_pm_get_dynpm_state, }; static struct radeon_asic r300_asic_pcie = { @@ -247,7 +268,8 @@ static struct radeon_asic r300_asic_pcie = { .suspend = &r300_suspend, .resume = &r300_resume, .vga_set_state = &r100_vga_set_state, - .gpu_reset = &r300_gpu_reset, + .gpu_is_lockup = &r300_gpu_is_lockup, + .asic_reset = &r300_asic_reset, .gart_tlb_flush = &rv370_pcie_gart_tlb_flush, .gart_set_page = &rv370_pcie_gart_set_page, .cp_commit = &r100_cp_commit, @@ -276,6 +298,12 @@ static struct radeon_asic r300_asic_pcie = { .hpd_sense = &r100_hpd_sense, .hpd_set_polarity = &r100_hpd_set_polarity, .ioctl_wait_idle = NULL, + .gui_idle = &r100_gui_idle, + .pm_misc = &r100_pm_misc, + .pm_prepare = &r100_pm_prepare, + .pm_finish = &r100_pm_finish, + .pm_init_profile = &r100_pm_init_profile, + .pm_get_dynpm_state = &r100_pm_get_dynpm_state, }; static struct radeon_asic r420_asic = { @@ -284,7 +312,8 @@ static struct radeon_asic r420_asic = { .suspend = &r420_suspend, .resume = &r420_resume, .vga_set_state = &r100_vga_set_state, - .gpu_reset = &r300_gpu_reset, + .gpu_is_lockup = &r300_gpu_is_lockup, + .asic_reset = &r300_asic_reset, .gart_tlb_flush = &rv370_pcie_gart_tlb_flush, .gart_set_page = &rv370_pcie_gart_set_page, .cp_commit = &r100_cp_commit, @@ -314,6 +343,12 @@ static struct radeon_asic r420_asic = { .hpd_sense = &r100_hpd_sense, .hpd_set_polarity = &r100_hpd_set_polarity, .ioctl_wait_idle = NULL, + .gui_idle = &r100_gui_idle, + .pm_misc = &r100_pm_misc, + .pm_prepare = &r100_pm_prepare, + .pm_finish = &r100_pm_finish, + .pm_init_profile = &r420_pm_init_profile, + .pm_get_dynpm_state = &r100_pm_get_dynpm_state, }; static struct radeon_asic rs400_asic = { @@ -322,7 +357,8 @@ static struct radeon_asic rs400_asic = { .suspend = &rs400_suspend, .resume = &rs400_resume, .vga_set_state = &r100_vga_set_state, - .gpu_reset = &r300_gpu_reset, + .gpu_is_lockup = &r300_gpu_is_lockup, + .asic_reset = &r300_asic_reset, .gart_tlb_flush = &rs400_gart_tlb_flush, .gart_set_page = &rs400_gart_set_page, .cp_commit = &r100_cp_commit, @@ -352,6 +388,12 @@ static struct radeon_asic rs400_asic = { .hpd_sense = &r100_hpd_sense, .hpd_set_polarity = &r100_hpd_set_polarity, .ioctl_wait_idle = NULL, + .gui_idle = &r100_gui_idle, + .pm_misc = &r100_pm_misc, + .pm_prepare = &r100_pm_prepare, + .pm_finish = &r100_pm_finish, + .pm_init_profile = &r100_pm_init_profile, + .pm_get_dynpm_state = &r100_pm_get_dynpm_state, }; static struct radeon_asic rs600_asic = { @@ -360,7 +402,8 @@ static struct radeon_asic rs600_asic = { .suspend = &rs600_suspend, .resume = &rs600_resume, .vga_set_state = &r100_vga_set_state, - .gpu_reset = &r300_gpu_reset, + .gpu_is_lockup = &r300_gpu_is_lockup, + .asic_reset = &rs600_asic_reset, .gart_tlb_flush = &rs600_gart_tlb_flush, .gart_set_page = &rs600_gart_set_page, .cp_commit = &r100_cp_commit, @@ -390,6 +433,12 @@ static struct radeon_asic rs600_asic = { .hpd_sense = &rs600_hpd_sense, .hpd_set_polarity = &rs600_hpd_set_polarity, .ioctl_wait_idle = NULL, + .gui_idle = &r100_gui_idle, + .pm_misc = &rs600_pm_misc, + .pm_prepare = &rs600_pm_prepare, + .pm_finish = &rs600_pm_finish, + .pm_init_profile = &r420_pm_init_profile, + .pm_get_dynpm_state = &r100_pm_get_dynpm_state, }; static struct radeon_asic rs690_asic = { @@ -398,7 +447,8 @@ static struct radeon_asic rs690_asic = { .suspend = &rs690_suspend, .resume = &rs690_resume, .vga_set_state = &r100_vga_set_state, - .gpu_reset = &r300_gpu_reset, + .gpu_is_lockup = &r300_gpu_is_lockup, + .asic_reset = &rs600_asic_reset, .gart_tlb_flush = &rs400_gart_tlb_flush, .gart_set_page = &rs400_gart_set_page, .cp_commit = &r100_cp_commit, @@ -428,6 +478,12 @@ static struct radeon_asic rs690_asic = { .hpd_sense = &rs600_hpd_sense, .hpd_set_polarity = &rs600_hpd_set_polarity, .ioctl_wait_idle = NULL, + .gui_idle = &r100_gui_idle, + .pm_misc = &rs600_pm_misc, + .pm_prepare = &rs600_pm_prepare, + .pm_finish = &rs600_pm_finish, + .pm_init_profile = &r420_pm_init_profile, + .pm_get_dynpm_state = &r100_pm_get_dynpm_state, }; static struct radeon_asic rv515_asic = { @@ -436,7 +492,8 @@ static struct radeon_asic rv515_asic = { .suspend = &rv515_suspend, .resume = &rv515_resume, .vga_set_state = &r100_vga_set_state, - .gpu_reset = &rv515_gpu_reset, + .gpu_is_lockup = &r300_gpu_is_lockup, + .asic_reset = &rs600_asic_reset, .gart_tlb_flush = &rv370_pcie_gart_tlb_flush, .gart_set_page = &rv370_pcie_gart_set_page, .cp_commit = &r100_cp_commit, @@ -466,6 +523,12 @@ static struct radeon_asic rv515_asic = { .hpd_sense = &rs600_hpd_sense, .hpd_set_polarity = &rs600_hpd_set_polarity, .ioctl_wait_idle = NULL, + .gui_idle = &r100_gui_idle, + .pm_misc = &rs600_pm_misc, + .pm_prepare = &rs600_pm_prepare, + .pm_finish = &rs600_pm_finish, + .pm_init_profile = &r420_pm_init_profile, + .pm_get_dynpm_state = &r100_pm_get_dynpm_state, }; static struct radeon_asic r520_asic = { @@ -474,7 +537,8 @@ static struct radeon_asic r520_asic = { .suspend = &rv515_suspend, .resume = &r520_resume, .vga_set_state = &r100_vga_set_state, - .gpu_reset = &rv515_gpu_reset, + .gpu_is_lockup = &r300_gpu_is_lockup, + .asic_reset = &rs600_asic_reset, .gart_tlb_flush = &rv370_pcie_gart_tlb_flush, .gart_set_page = &rv370_pcie_gart_set_page, .cp_commit = &r100_cp_commit, @@ -504,6 +568,12 @@ static struct radeon_asic r520_asic = { .hpd_sense = &rs600_hpd_sense, .hpd_set_polarity = &rs600_hpd_set_polarity, .ioctl_wait_idle = NULL, + .gui_idle = &r100_gui_idle, + .pm_misc = &rs600_pm_misc, + .pm_prepare = &rs600_pm_prepare, + .pm_finish = &rs600_pm_finish, + .pm_init_profile = &r420_pm_init_profile, + .pm_get_dynpm_state = &r100_pm_get_dynpm_state, }; static struct radeon_asic r600_asic = { @@ -513,7 +583,8 @@ static struct radeon_asic r600_asic = { .resume = &r600_resume, .cp_commit = &r600_cp_commit, .vga_set_state = &r600_vga_set_state, - .gpu_reset = &r600_gpu_reset, + .gpu_is_lockup = &r600_gpu_is_lockup, + .asic_reset = &r600_asic_reset, .gart_tlb_flush = &r600_pcie_gart_tlb_flush, .gart_set_page = &rs600_gart_set_page, .ring_test = &r600_ring_test, @@ -541,6 +612,12 @@ static struct radeon_asic r600_asic = { .hpd_sense = &r600_hpd_sense, .hpd_set_polarity = &r600_hpd_set_polarity, .ioctl_wait_idle = r600_ioctl_wait_idle, + .gui_idle = &r600_gui_idle, + .pm_misc = &r600_pm_misc, + .pm_prepare = &rs600_pm_prepare, + .pm_finish = &rs600_pm_finish, + .pm_init_profile = &r600_pm_init_profile, + .pm_get_dynpm_state = &r600_pm_get_dynpm_state, }; static struct radeon_asic rs780_asic = { @@ -549,8 +626,9 @@ static struct radeon_asic rs780_asic = { .suspend = &r600_suspend, .resume = &r600_resume, .cp_commit = &r600_cp_commit, + .gpu_is_lockup = &r600_gpu_is_lockup, .vga_set_state = &r600_vga_set_state, - .gpu_reset = &r600_gpu_reset, + .asic_reset = &r600_asic_reset, .gart_tlb_flush = &r600_pcie_gart_tlb_flush, .gart_set_page = &rs600_gart_set_page, .ring_test = &r600_ring_test, @@ -578,6 +656,12 @@ static struct radeon_asic rs780_asic = { .hpd_sense = &r600_hpd_sense, .hpd_set_polarity = &r600_hpd_set_polarity, .ioctl_wait_idle = r600_ioctl_wait_idle, + .gui_idle = &r600_gui_idle, + .pm_misc = &r600_pm_misc, + .pm_prepare = &rs600_pm_prepare, + .pm_finish = &rs600_pm_finish, + .pm_init_profile = &rs780_pm_init_profile, + .pm_get_dynpm_state = &r600_pm_get_dynpm_state, }; static struct radeon_asic rv770_asic = { @@ -586,7 +670,8 @@ static struct radeon_asic rv770_asic = { .suspend = &rv770_suspend, .resume = &rv770_resume, .cp_commit = &r600_cp_commit, - .gpu_reset = &rv770_gpu_reset, + .asic_reset = &r600_asic_reset, + .gpu_is_lockup = &r600_gpu_is_lockup, .vga_set_state = &r600_vga_set_state, .gart_tlb_flush = &r600_pcie_gart_tlb_flush, .gart_set_page = &rs600_gart_set_page, @@ -615,6 +700,12 @@ static struct radeon_asic rv770_asic = { .hpd_sense = &r600_hpd_sense, .hpd_set_polarity = &r600_hpd_set_polarity, .ioctl_wait_idle = r600_ioctl_wait_idle, + .gui_idle = &r600_gui_idle, + .pm_misc = &rv770_pm_misc, + .pm_prepare = &rs600_pm_prepare, + .pm_finish = &rs600_pm_finish, + .pm_init_profile = &r600_pm_init_profile, + .pm_get_dynpm_state = &r600_pm_get_dynpm_state, }; static struct radeon_asic evergreen_asic = { @@ -622,18 +713,19 @@ static struct radeon_asic evergreen_asic = { .fini = &evergreen_fini, .suspend = &evergreen_suspend, .resume = &evergreen_resume, - .cp_commit = NULL, - .gpu_reset = &evergreen_gpu_reset, + .cp_commit = &r600_cp_commit, + .gpu_is_lockup = &evergreen_gpu_is_lockup, + .asic_reset = &evergreen_asic_reset, .vga_set_state = &r600_vga_set_state, - .gart_tlb_flush = &r600_pcie_gart_tlb_flush, + .gart_tlb_flush = &evergreen_pcie_gart_tlb_flush, .gart_set_page = &rs600_gart_set_page, - .ring_test = NULL, - .ring_ib_execute = NULL, - .irq_set = NULL, - .irq_process = NULL, - .get_vblank_counter = NULL, - .fence_ring_emit = NULL, - .cs_parse = NULL, + .ring_test = &r600_ring_test, + .ring_ib_execute = &r600_ring_ib_execute, + .irq_set = &evergreen_irq_set, + .irq_process = &evergreen_irq_process, + .get_vblank_counter = &evergreen_get_vblank_counter, + .fence_ring_emit = &r600_fence_ring_emit, + .cs_parse = &evergreen_cs_parse, .copy_blit = NULL, .copy_dma = NULL, .copy = NULL, @@ -650,6 +742,12 @@ static struct radeon_asic evergreen_asic = { .hpd_fini = &evergreen_hpd_fini, .hpd_sense = &evergreen_hpd_sense, .hpd_set_polarity = &evergreen_hpd_set_polarity, + .gui_idle = &r600_gui_idle, + .pm_misc = &evergreen_pm_misc, + .pm_prepare = &evergreen_pm_prepare, + .pm_finish = &evergreen_pm_finish, + .pm_init_profile = &r600_pm_init_profile, + .pm_get_dynpm_state = &r600_pm_get_dynpm_state, }; int radeon_asic_init(struct radeon_device *rdev) @@ -682,6 +780,13 @@ int radeon_asic_init(struct radeon_device *rdev) case CHIP_R423: case CHIP_RV410: rdev->asic = &r420_asic; + /* handle macs */ + if (rdev->bios == NULL) { + rdev->asic->get_engine_clock = &radeon_legacy_get_engine_clock; + rdev->asic->set_engine_clock = &radeon_legacy_set_engine_clock; + rdev->asic->get_memory_clock = &radeon_legacy_get_memory_clock; + rdev->asic->set_memory_clock = NULL; + } break; case CHIP_RS400: case CHIP_RS480: diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h index a0b8280663d1..c0bbaa64157a 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.h +++ b/drivers/gpu/drm/radeon/radeon_asic.h @@ -60,7 +60,8 @@ int r100_resume(struct radeon_device *rdev); uint32_t r100_mm_rreg(struct radeon_device *rdev, uint32_t reg); void r100_mm_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); void r100_vga_set_state(struct radeon_device *rdev, bool state); -int r100_gpu_reset(struct radeon_device *rdev); +bool r100_gpu_is_lockup(struct radeon_device *rdev); +int r100_asic_reset(struct radeon_device *rdev); u32 r100_get_vblank_counter(struct radeon_device *rdev, int crtc); void r100_pci_gart_tlb_flush(struct radeon_device *rdev); int r100_pci_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr); @@ -110,8 +111,6 @@ void r100_vram_init_sizes(struct radeon_device *rdev); void r100_wb_disable(struct radeon_device *rdev); void r100_wb_fini(struct radeon_device *rdev); int r100_wb_init(struct radeon_device *rdev); -void r100_hdp_reset(struct radeon_device *rdev); -int r100_rb2d_reset(struct radeon_device *rdev); int r100_cp_reset(struct radeon_device *rdev); void r100_vga_render_disable(struct radeon_device *rdev); int r100_cs_track_check_pkt3_indx_buffer(struct radeon_cs_parser *p, @@ -126,6 +125,13 @@ int r100_cs_packet_parse(struct radeon_cs_parser *p, unsigned idx); void r100_enable_bm(struct radeon_device *rdev); void r100_set_common_regs(struct radeon_device *rdev); +void r100_bm_disable(struct radeon_device *rdev); +extern bool r100_gui_idle(struct radeon_device *rdev); +extern void r100_pm_misc(struct radeon_device *rdev); +extern void r100_pm_prepare(struct radeon_device *rdev); +extern void r100_pm_finish(struct radeon_device *rdev); +extern void r100_pm_init_profile(struct radeon_device *rdev); +extern void r100_pm_get_dynpm_state(struct radeon_device *rdev); /* * r200,rv250,rs300,rv280 @@ -134,7 +140,7 @@ extern int r200_copy_dma(struct radeon_device *rdev, uint64_t src_offset, uint64_t dst_offset, unsigned num_pages, - struct radeon_fence *fence); + struct radeon_fence *fence); /* * r300,r350,rv350,rv380 @@ -143,7 +149,8 @@ extern int r300_init(struct radeon_device *rdev); extern void r300_fini(struct radeon_device *rdev); extern int r300_suspend(struct radeon_device *rdev); extern int r300_resume(struct radeon_device *rdev); -extern int r300_gpu_reset(struct radeon_device *rdev); +extern bool r300_gpu_is_lockup(struct radeon_device *rdev); +extern int r300_asic_reset(struct radeon_device *rdev); extern void r300_ring_start(struct radeon_device *rdev); extern void r300_fence_ring_emit(struct radeon_device *rdev, struct radeon_fence *fence); @@ -162,6 +169,7 @@ extern int r420_init(struct radeon_device *rdev); extern void r420_fini(struct radeon_device *rdev); extern int r420_suspend(struct radeon_device *rdev); extern int r420_resume(struct radeon_device *rdev); +extern void r420_pm_init_profile(struct radeon_device *rdev); /* * rs400,rs480 @@ -178,6 +186,7 @@ void rs400_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); /* * rs600. */ +extern int rs600_asic_reset(struct radeon_device *rdev); extern int rs600_init(struct radeon_device *rdev); extern void rs600_fini(struct radeon_device *rdev); extern int rs600_suspend(struct radeon_device *rdev); @@ -195,6 +204,9 @@ void rs600_hpd_fini(struct radeon_device *rdev); bool rs600_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd); void rs600_hpd_set_polarity(struct radeon_device *rdev, enum radeon_hpd_id hpd); +extern void rs600_pm_misc(struct radeon_device *rdev); +extern void rs600_pm_prepare(struct radeon_device *rdev); +extern void rs600_pm_finish(struct radeon_device *rdev); /* * rs690,rs740 @@ -212,7 +224,6 @@ void rs690_bandwidth_update(struct radeon_device *rdev); */ int rv515_init(struct radeon_device *rdev); void rv515_fini(struct radeon_device *rdev); -int rv515_gpu_reset(struct radeon_device *rdev); uint32_t rv515_mc_rreg(struct radeon_device *rdev, uint32_t reg); void rv515_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); void rv515_ring_start(struct radeon_device *rdev); @@ -252,7 +263,8 @@ int r600_copy_dma(struct radeon_device *rdev, struct radeon_fence *fence); int r600_irq_process(struct radeon_device *rdev); int r600_irq_set(struct radeon_device *rdev); -int r600_gpu_reset(struct radeon_device *rdev); +bool r600_gpu_is_lockup(struct radeon_device *rdev); +int r600_asic_reset(struct radeon_device *rdev); int r600_set_surface_reg(struct radeon_device *rdev, int reg, uint32_t tiling_flags, uint32_t pitch, uint32_t offset, uint32_t obj_size); @@ -268,6 +280,11 @@ bool r600_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd); void r600_hpd_set_polarity(struct radeon_device *rdev, enum radeon_hpd_id hpd); extern void r600_ioctl_wait_idle(struct radeon_device *rdev, struct radeon_bo *bo); +extern bool r600_gui_idle(struct radeon_device *rdev); +extern void r600_pm_misc(struct radeon_device *rdev); +extern void r600_pm_init_profile(struct radeon_device *rdev); +extern void rs780_pm_init_profile(struct radeon_device *rdev); +extern void r600_pm_get_dynpm_state(struct radeon_device *rdev); /* * rv770,rv730,rv710,rv740 @@ -276,20 +293,30 @@ int rv770_init(struct radeon_device *rdev); void rv770_fini(struct radeon_device *rdev); int rv770_suspend(struct radeon_device *rdev); int rv770_resume(struct radeon_device *rdev); -int rv770_gpu_reset(struct radeon_device *rdev); +extern void rv770_pm_misc(struct radeon_device *rdev); /* * evergreen */ +void evergreen_pcie_gart_tlb_flush(struct radeon_device *rdev); int evergreen_init(struct radeon_device *rdev); void evergreen_fini(struct radeon_device *rdev); int evergreen_suspend(struct radeon_device *rdev); int evergreen_resume(struct radeon_device *rdev); -int evergreen_gpu_reset(struct radeon_device *rdev); +bool evergreen_gpu_is_lockup(struct radeon_device *rdev); +int evergreen_asic_reset(struct radeon_device *rdev); void evergreen_bandwidth_update(struct radeon_device *rdev); void evergreen_hpd_init(struct radeon_device *rdev); void evergreen_hpd_fini(struct radeon_device *rdev); bool evergreen_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd); void evergreen_hpd_set_polarity(struct radeon_device *rdev, enum radeon_hpd_id hpd); +u32 evergreen_get_vblank_counter(struct radeon_device *rdev, int crtc); +int evergreen_irq_set(struct radeon_device *rdev); +int evergreen_irq_process(struct radeon_device *rdev); +extern int evergreen_cs_parse(struct radeon_cs_parser *p); +extern void evergreen_pm_misc(struct radeon_device *rdev); +extern void evergreen_pm_prepare(struct radeon_device *rdev); +extern void evergreen_pm_finish(struct radeon_device *rdev); + #endif diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c index 9916d825401c..10673ae59cfa 100644 --- a/drivers/gpu/drm/radeon/radeon_atombios.c +++ b/drivers/gpu/drm/radeon/radeon_atombios.c @@ -280,6 +280,15 @@ static bool radeon_atom_apply_quirks(struct drm_device *dev, } } + /* ASUS HD 3600 board lists the DVI port as HDMI */ + if ((dev->pdev->device == 0x9598) && + (dev->pdev->subsystem_vendor == 0x1043) && + (dev->pdev->subsystem_device == 0x01e4)) { + if (*connector_type == DRM_MODE_CONNECTOR_HDMIA) { + *connector_type = DRM_MODE_CONNECTOR_DVII; + } + } + /* ASUS HD 3450 board lists the DVI port as HDMI */ if ((dev->pdev->device == 0x95C5) && (dev->pdev->subsystem_vendor == 0x1043) && @@ -530,6 +539,8 @@ bool radeon_get_atom_connector_info_from_object_table(struct drm_device *dev) } /* look up gpio for ddc, hpd */ + ddc_bus.valid = false; + hpd.hpd = RADEON_HPD_NONE; if ((le16_to_cpu(path->usDeviceTag) & (ATOM_DEVICE_TV_SUPPORT | ATOM_DEVICE_CV_SUPPORT)) == 0) { for (j = 0; j < con_obj->ucNumberOfObjects; j++) { @@ -547,7 +558,6 @@ bool radeon_get_atom_connector_info_from_object_table(struct drm_device *dev) ATOM_I2C_RECORD *i2c_record; ATOM_HPD_INT_RECORD *hpd_record; ATOM_I2C_ID_CONFIG_ACCESS *i2c_config; - hpd.hpd = RADEON_HPD_NONE; while (record->ucRecordType > 0 && record-> @@ -585,13 +595,10 @@ bool radeon_get_atom_connector_info_from_object_table(struct drm_device *dev) break; } } - } else { - hpd.hpd = RADEON_HPD_NONE; - ddc_bus.valid = false; } /* needed for aux chan transactions */ - ddc_bus.hpd_id = hpd.hpd ? (hpd.hpd - 1) : 0; + ddc_bus.hpd = hpd.hpd; conn_id = le16_to_cpu(path->usConnObjectId); @@ -682,11 +689,19 @@ bool radeon_get_atom_connector_info_from_supported_devices_table(struct uint8_t dac; union atom_supported_devices *supported_devices; int i, j, max_device; - struct bios_connector bios_connectors[ATOM_MAX_SUPPORTED_DEVICE]; + struct bios_connector *bios_connectors; + size_t bc_size = sizeof(*bios_connectors) * ATOM_MAX_SUPPORTED_DEVICE; - if (!atom_parse_data_header(ctx, index, &size, &frev, &crev, &data_offset)) + bios_connectors = kzalloc(bc_size, GFP_KERNEL); + if (!bios_connectors) return false; + if (!atom_parse_data_header(ctx, index, &size, &frev, &crev, + &data_offset)) { + kfree(bios_connectors); + return false; + } + supported_devices = (union atom_supported_devices *)(ctx->bios + data_offset); @@ -853,6 +868,7 @@ bool radeon_get_atom_connector_info_from_supported_devices_table(struct radeon_link_encoder_connector(dev); + kfree(bios_connectors); return true; } @@ -1022,8 +1038,15 @@ bool radeon_atombios_sideport_present(struct radeon_device *rdev) data_offset); switch (crev) { case 1: - if (igp_info->info.ucMemoryType & 0xf0) - return true; + /* AMD IGPS */ + if ((rdev->family == CHIP_RS690) || + (rdev->family == CHIP_RS740)) { + if (igp_info->info.ulBootUpMemoryClock) + return true; + } else { + if (igp_info->info.ucMemoryType & 0xf0) + return true; + } break; case 2: if (igp_info->info_2.ucMemoryType & 0x0f) @@ -1174,7 +1197,7 @@ struct radeon_encoder_atom_dig *radeon_atombios_get_lvds_info(struct lvds->native_mode.vtotal = lvds->native_mode.vdisplay + le16_to_cpu(lvds_info->info.sLCDTiming.usVBlanking_Time); lvds->native_mode.vsync_start = lvds->native_mode.vdisplay + - le16_to_cpu(lvds_info->info.sLCDTiming.usVSyncWidth); + le16_to_cpu(lvds_info->info.sLCDTiming.usVSyncOffset); lvds->native_mode.vsync_end = lvds->native_mode.vsync_start + le16_to_cpu(lvds_info->info.sLCDTiming.usVSyncWidth); lvds->panel_pwr_delay = @@ -1442,26 +1465,30 @@ radeon_atombios_get_tv_dac_info(struct radeon_encoder *encoder) static const char *thermal_controller_names[] = { "NONE", - "LM63", - "ADM1032", - "ADM1030", - "MUA6649", - "LM64", - "F75375", - "ASC7512", + "lm63", + "adm1032", + "adm1030", + "max6649", + "lm64", + "f75375", + "asc7xxx", }; static const char *pp_lib_thermal_controller_names[] = { "NONE", - "LM63", - "ADM1032", - "ADM1030", - "MUA6649", - "LM64", - "F75375", + "lm63", + "adm1032", + "adm1030", + "max6649", + "lm64", + "f75375", "RV6xx", "RV770", - "ADT7473", + "adt7473", + "External GPIO", + "Evergreen", + "adt7473 with internal", + }; union power_info { @@ -1485,7 +1512,7 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) int state_index = 0, mode_index = 0; struct radeon_i2c_bus_rec i2c_bus; - rdev->pm.default_power_state = NULL; + rdev->pm.default_power_state_index = -1; if (atom_parse_data_header(mode_info->atom_context, index, NULL, &frev, &crev, &data_offset)) { @@ -1498,10 +1525,19 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) power_info->info.ucOverdriveControllerAddress >> 1); i2c_bus = radeon_lookup_i2c_gpio(rdev, power_info->info.ucOverdriveI2cLine); rdev->pm.i2c_bus = radeon_i2c_create(rdev->ddev, &i2c_bus, "Thermal"); + if (rdev->pm.i2c_bus) { + struct i2c_board_info info = { }; + const char *name = thermal_controller_names[power_info->info. + ucOverdriveThermalController]; + info.addr = power_info->info.ucOverdriveControllerAddress >> 1; + strlcpy(info.type, name, sizeof(info.type)); + i2c_new_device(&rdev->pm.i2c_bus->adapter, &info); + } } num_modes = power_info->info.ucNumOfPowerModeEntries; if (num_modes > ATOM_MAX_NUMBEROF_POWER_BLOCK) num_modes = ATOM_MAX_NUMBEROF_POWER_BLOCK; + /* last mode is usually default, array is low to high */ for (i = 0; i < num_modes; i++) { rdev->pm.power_state[state_index].clock_info[0].voltage.type = VOLTAGE_NONE; switch (frev) { @@ -1515,16 +1551,11 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) if ((rdev->pm.power_state[state_index].clock_info[0].mclk == 0) || (rdev->pm.power_state[state_index].clock_info[0].sclk == 0)) continue; - /* skip overclock modes for now */ - if ((rdev->pm.power_state[state_index].clock_info[0].mclk > - rdev->clock.default_mclk + RADEON_MODE_OVERCLOCK_MARGIN) || - (rdev->pm.power_state[state_index].clock_info[0].sclk > - rdev->clock.default_sclk + RADEON_MODE_OVERCLOCK_MARGIN)) - continue; - rdev->pm.power_state[state_index].non_clock_info.pcie_lanes = + rdev->pm.power_state[state_index].pcie_lanes = power_info->info.asPowerPlayInfo[i].ucNumPciELanes; misc = le32_to_cpu(power_info->info.asPowerPlayInfo[i].ulMiscInfo); - if (misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_SUPPORT) { + if ((misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_SUPPORT) || + (misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_ACTIVE_HIGH)) { rdev->pm.power_state[state_index].clock_info[0].voltage.type = VOLTAGE_GPIO; rdev->pm.power_state[state_index].clock_info[0].voltage.gpio = @@ -1542,6 +1573,8 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) rdev->pm.power_state[state_index].clock_info[0].voltage.vddc_id = power_info->info.asPowerPlayInfo[i].ucVoltageDropIndex; } + rdev->pm.power_state[state_index].flags = RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; + rdev->pm.power_state[state_index].misc = misc; /* order matters! */ if (misc & ATOM_PM_MISCINFO_POWER_SAVING_MODE) rdev->pm.power_state[state_index].type = @@ -1555,15 +1588,23 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) if (misc & ATOM_PM_MISCINFO_LOAD_BALANCE_EN) rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_BALANCED; - if (misc & ATOM_PM_MISCINFO_3D_ACCELERATION_EN) + if (misc & ATOM_PM_MISCINFO_3D_ACCELERATION_EN) { rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_PERFORMANCE; + rdev->pm.power_state[state_index].flags &= + ~RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; + } if (misc & ATOM_PM_MISCINFO_DRIVER_DEFAULT_MODE) { rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_DEFAULT; - rdev->pm.default_power_state = &rdev->pm.power_state[state_index]; + rdev->pm.default_power_state_index = state_index; rdev->pm.power_state[state_index].default_clock_mode = &rdev->pm.power_state[state_index].clock_info[0]; + rdev->pm.power_state[state_index].flags &= + ~RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; + } else if (state_index == 0) { + rdev->pm.power_state[state_index].clock_info[0].flags |= + RADEON_PM_MODE_NO_DISPLAY; } state_index++; break; @@ -1577,17 +1618,12 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) if ((rdev->pm.power_state[state_index].clock_info[0].mclk == 0) || (rdev->pm.power_state[state_index].clock_info[0].sclk == 0)) continue; - /* skip overclock modes for now */ - if ((rdev->pm.power_state[state_index].clock_info[0].mclk > - rdev->clock.default_mclk + RADEON_MODE_OVERCLOCK_MARGIN) || - (rdev->pm.power_state[state_index].clock_info[0].sclk > - rdev->clock.default_sclk + RADEON_MODE_OVERCLOCK_MARGIN)) - continue; - rdev->pm.power_state[state_index].non_clock_info.pcie_lanes = + rdev->pm.power_state[state_index].pcie_lanes = power_info->info_2.asPowerPlayInfo[i].ucNumPciELanes; misc = le32_to_cpu(power_info->info_2.asPowerPlayInfo[i].ulMiscInfo); misc2 = le32_to_cpu(power_info->info_2.asPowerPlayInfo[i].ulMiscInfo2); - if (misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_SUPPORT) { + if ((misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_SUPPORT) || + (misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_ACTIVE_HIGH)) { rdev->pm.power_state[state_index].clock_info[0].voltage.type = VOLTAGE_GPIO; rdev->pm.power_state[state_index].clock_info[0].voltage.gpio = @@ -1605,6 +1641,9 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) rdev->pm.power_state[state_index].clock_info[0].voltage.vddc_id = power_info->info_2.asPowerPlayInfo[i].ucVoltageDropIndex; } + rdev->pm.power_state[state_index].flags = RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; + rdev->pm.power_state[state_index].misc = misc; + rdev->pm.power_state[state_index].misc2 = misc2; /* order matters! */ if (misc & ATOM_PM_MISCINFO_POWER_SAVING_MODE) rdev->pm.power_state[state_index].type = @@ -1618,18 +1657,29 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) if (misc & ATOM_PM_MISCINFO_LOAD_BALANCE_EN) rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_BALANCED; - if (misc & ATOM_PM_MISCINFO_3D_ACCELERATION_EN) + if (misc & ATOM_PM_MISCINFO_3D_ACCELERATION_EN) { rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_PERFORMANCE; + rdev->pm.power_state[state_index].flags &= + ~RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; + } if (misc2 & ATOM_PM_MISCINFO2_SYSTEM_AC_LITE_MODE) rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_BALANCED; + if (misc2 & ATOM_PM_MISCINFO2_MULTI_DISPLAY_SUPPORT) + rdev->pm.power_state[state_index].flags &= + ~RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; if (misc & ATOM_PM_MISCINFO_DRIVER_DEFAULT_MODE) { rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_DEFAULT; - rdev->pm.default_power_state = &rdev->pm.power_state[state_index]; + rdev->pm.default_power_state_index = state_index; rdev->pm.power_state[state_index].default_clock_mode = &rdev->pm.power_state[state_index].clock_info[0]; + rdev->pm.power_state[state_index].flags &= + ~RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; + } else if (state_index == 0) { + rdev->pm.power_state[state_index].clock_info[0].flags |= + RADEON_PM_MODE_NO_DISPLAY; } state_index++; break; @@ -1643,17 +1693,12 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) if ((rdev->pm.power_state[state_index].clock_info[0].mclk == 0) || (rdev->pm.power_state[state_index].clock_info[0].sclk == 0)) continue; - /* skip overclock modes for now */ - if ((rdev->pm.power_state[state_index].clock_info[0].mclk > - rdev->clock.default_mclk + RADEON_MODE_OVERCLOCK_MARGIN) || - (rdev->pm.power_state[state_index].clock_info[0].sclk > - rdev->clock.default_sclk + RADEON_MODE_OVERCLOCK_MARGIN)) - continue; - rdev->pm.power_state[state_index].non_clock_info.pcie_lanes = + rdev->pm.power_state[state_index].pcie_lanes = power_info->info_3.asPowerPlayInfo[i].ucNumPciELanes; misc = le32_to_cpu(power_info->info_3.asPowerPlayInfo[i].ulMiscInfo); misc2 = le32_to_cpu(power_info->info_3.asPowerPlayInfo[i].ulMiscInfo2); - if (misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_SUPPORT) { + if ((misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_SUPPORT) || + (misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_ACTIVE_HIGH)) { rdev->pm.power_state[state_index].clock_info[0].voltage.type = VOLTAGE_GPIO; rdev->pm.power_state[state_index].clock_info[0].voltage.gpio = @@ -1677,6 +1722,9 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) power_info->info_3.asPowerPlayInfo[i].ucVDDCI_VoltageDropIndex; } } + rdev->pm.power_state[state_index].flags = RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; + rdev->pm.power_state[state_index].misc = misc; + rdev->pm.power_state[state_index].misc2 = misc2; /* order matters! */ if (misc & ATOM_PM_MISCINFO_POWER_SAVING_MODE) rdev->pm.power_state[state_index].type = @@ -1690,42 +1738,89 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) if (misc & ATOM_PM_MISCINFO_LOAD_BALANCE_EN) rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_BALANCED; - if (misc & ATOM_PM_MISCINFO_3D_ACCELERATION_EN) + if (misc & ATOM_PM_MISCINFO_3D_ACCELERATION_EN) { rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_PERFORMANCE; + rdev->pm.power_state[state_index].flags &= + ~RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; + } if (misc2 & ATOM_PM_MISCINFO2_SYSTEM_AC_LITE_MODE) rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_BALANCED; if (misc & ATOM_PM_MISCINFO_DRIVER_DEFAULT_MODE) { rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_DEFAULT; - rdev->pm.default_power_state = &rdev->pm.power_state[state_index]; + rdev->pm.default_power_state_index = state_index; rdev->pm.power_state[state_index].default_clock_mode = &rdev->pm.power_state[state_index].clock_info[0]; + } else if (state_index == 0) { + rdev->pm.power_state[state_index].clock_info[0].flags |= + RADEON_PM_MODE_NO_DISPLAY; } state_index++; break; } } - } else if (frev == 4) { + /* last mode is usually default */ + if (rdev->pm.default_power_state_index == -1) { + rdev->pm.power_state[state_index - 1].type = + POWER_STATE_TYPE_DEFAULT; + rdev->pm.default_power_state_index = state_index - 1; + rdev->pm.power_state[state_index - 1].default_clock_mode = + &rdev->pm.power_state[state_index - 1].clock_info[0]; + rdev->pm.power_state[state_index].flags &= + ~RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; + rdev->pm.power_state[state_index].misc = 0; + rdev->pm.power_state[state_index].misc2 = 0; + } + } else { + int fw_index = GetIndexIntoMasterTable(DATA, FirmwareInfo); + uint8_t fw_frev, fw_crev; + uint16_t fw_data_offset, vddc = 0; + union firmware_info *firmware_info; + ATOM_PPLIB_THERMALCONTROLLER *controller = &power_info->info_4.sThermalController; + + if (atom_parse_data_header(mode_info->atom_context, fw_index, NULL, + &fw_frev, &fw_crev, &fw_data_offset)) { + firmware_info = + (union firmware_info *)(mode_info->atom_context->bios + + fw_data_offset); + vddc = firmware_info->info_14.usBootUpVDDCVoltage; + } + /* add the i2c bus for thermal/fan chip */ /* no support for internal controller yet */ - if (power_info->info_4.sThermalController.ucType > 0) { - if ((power_info->info_4.sThermalController.ucType == ATOM_PP_THERMALCONTROLLER_RV6xx) || - (power_info->info_4.sThermalController.ucType == ATOM_PP_THERMALCONTROLLER_RV770)) { + if (controller->ucType > 0) { + if ((controller->ucType == ATOM_PP_THERMALCONTROLLER_RV6xx) || + (controller->ucType == ATOM_PP_THERMALCONTROLLER_RV770) || + (controller->ucType == ATOM_PP_THERMALCONTROLLER_EVERGREEN)) { DRM_INFO("Internal thermal controller %s fan control\n", - (power_info->info_4.sThermalController.ucFanParameters & + (controller->ucFanParameters & ATOM_PP_FANPARAMETERS_NOFAN) ? "without" : "with"); + } else if ((controller->ucType == + ATOM_PP_THERMALCONTROLLER_EXTERNAL_GPIO) || + (controller->ucType == + ATOM_PP_THERMALCONTROLLER_ADT7473_WITH_INTERNAL)) { + DRM_INFO("Special thermal controller config\n"); } else { DRM_INFO("Possible %s thermal controller at 0x%02x %s fan control\n", - pp_lib_thermal_controller_names[power_info->info_4.sThermalController.ucType], - power_info->info_4.sThermalController.ucI2cAddress >> 1, - (power_info->info_4.sThermalController.ucFanParameters & + pp_lib_thermal_controller_names[controller->ucType], + controller->ucI2cAddress >> 1, + (controller->ucFanParameters & ATOM_PP_FANPARAMETERS_NOFAN) ? "without" : "with"); - i2c_bus = radeon_lookup_i2c_gpio(rdev, power_info->info_4.sThermalController.ucI2cLine); + i2c_bus = radeon_lookup_i2c_gpio(rdev, controller->ucI2cLine); rdev->pm.i2c_bus = radeon_i2c_create(rdev->ddev, &i2c_bus, "Thermal"); + if (rdev->pm.i2c_bus) { + struct i2c_board_info info = { }; + const char *name = pp_lib_thermal_controller_names[controller->ucType]; + info.addr = controller->ucI2cAddress >> 1; + strlcpy(info.type, name, sizeof(info.type)); + i2c_new_device(&rdev->pm.i2c_bus->adapter, &info); + } + } } + /* first mode is usually default, followed by low to high */ for (i = 0; i < power_info->info_4.ucNumStates; i++) { mode_index = 0; power_state = (struct _ATOM_PPLIB_STATE *) @@ -1754,14 +1849,31 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) /* skip invalid modes */ if (rdev->pm.power_state[state_index].clock_info[mode_index].sclk == 0) continue; - /* skip overclock modes for now */ - if (rdev->pm.power_state[state_index].clock_info[mode_index].sclk > - rdev->clock.default_sclk + RADEON_MODE_OVERCLOCK_MARGIN) + /* voltage works differently on IGPs */ + mode_index++; + } else if (ASIC_IS_DCE4(rdev)) { + struct _ATOM_PPLIB_EVERGREEN_CLOCK_INFO *clock_info = + (struct _ATOM_PPLIB_EVERGREEN_CLOCK_INFO *) + (mode_info->atom_context->bios + + data_offset + + le16_to_cpu(power_info->info_4.usClockInfoArrayOffset) + + (power_state->ucClockStateIndices[j] * + power_info->info_4.ucClockInfoSize)); + sclk = le16_to_cpu(clock_info->usEngineClockLow); + sclk |= clock_info->ucEngineClockHigh << 16; + mclk = le16_to_cpu(clock_info->usMemoryClockLow); + mclk |= clock_info->ucMemoryClockHigh << 16; + rdev->pm.power_state[state_index].clock_info[mode_index].mclk = mclk; + rdev->pm.power_state[state_index].clock_info[mode_index].sclk = sclk; + /* skip invalid modes */ + if ((rdev->pm.power_state[state_index].clock_info[mode_index].mclk == 0) || + (rdev->pm.power_state[state_index].clock_info[mode_index].sclk == 0)) continue; rdev->pm.power_state[state_index].clock_info[mode_index].voltage.type = VOLTAGE_SW; rdev->pm.power_state[state_index].clock_info[mode_index].voltage.voltage = clock_info->usVDDC; + /* XXX usVDDCI */ mode_index++; } else { struct _ATOM_PPLIB_R600_CLOCK_INFO *clock_info = @@ -1781,12 +1893,6 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) if ((rdev->pm.power_state[state_index].clock_info[mode_index].mclk == 0) || (rdev->pm.power_state[state_index].clock_info[mode_index].sclk == 0)) continue; - /* skip overclock modes for now */ - if ((rdev->pm.power_state[state_index].clock_info[mode_index].mclk > - rdev->clock.default_mclk + RADEON_MODE_OVERCLOCK_MARGIN) || - (rdev->pm.power_state[state_index].clock_info[mode_index].sclk > - rdev->clock.default_sclk + RADEON_MODE_OVERCLOCK_MARGIN)) - continue; rdev->pm.power_state[state_index].clock_info[mode_index].voltage.type = VOLTAGE_SW; rdev->pm.power_state[state_index].clock_info[mode_index].voltage.voltage = @@ -1798,7 +1904,9 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) if (mode_index) { misc = le32_to_cpu(non_clock_info->ulCapsAndSettings); misc2 = le16_to_cpu(non_clock_info->usClassification); - rdev->pm.power_state[state_index].non_clock_info.pcie_lanes = + rdev->pm.power_state[state_index].misc = misc; + rdev->pm.power_state[state_index].misc2 = misc2; + rdev->pm.power_state[state_index].pcie_lanes = ((misc & ATOM_PPLIB_PCIE_LINK_WIDTH_MASK) >> ATOM_PPLIB_PCIE_LINK_WIDTH_SHIFT) + 1; switch (misc2 & ATOM_PPLIB_CLASSIFICATION_UI_MASK) { @@ -1815,22 +1923,46 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) POWER_STATE_TYPE_PERFORMANCE; break; } + rdev->pm.power_state[state_index].flags = 0; + if (misc & ATOM_PPLIB_SINGLE_DISPLAY_ONLY) + rdev->pm.power_state[state_index].flags |= + RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; if (misc2 & ATOM_PPLIB_CLASSIFICATION_BOOT) { rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_DEFAULT; - rdev->pm.default_power_state = &rdev->pm.power_state[state_index]; + rdev->pm.default_power_state_index = state_index; rdev->pm.power_state[state_index].default_clock_mode = &rdev->pm.power_state[state_index].clock_info[mode_index - 1]; + /* patch the table values with the default slck/mclk from firmware info */ + for (j = 0; j < mode_index; j++) { + rdev->pm.power_state[state_index].clock_info[j].mclk = + rdev->clock.default_mclk; + rdev->pm.power_state[state_index].clock_info[j].sclk = + rdev->clock.default_sclk; + if (vddc) + rdev->pm.power_state[state_index].clock_info[j].voltage.voltage = + vddc; + } } state_index++; } } + /* if multiple clock modes, mark the lowest as no display */ + for (i = 0; i < state_index; i++) { + if (rdev->pm.power_state[i].num_clock_modes > 1) + rdev->pm.power_state[i].clock_info[0].flags |= + RADEON_PM_MODE_NO_DISPLAY; + } + /* first mode is usually default */ + if (rdev->pm.default_power_state_index == -1) { + rdev->pm.power_state[0].type = + POWER_STATE_TYPE_DEFAULT; + rdev->pm.default_power_state_index = 0; + rdev->pm.power_state[0].default_clock_mode = + &rdev->pm.power_state[0].clock_info[0]; + } } } else { - /* XXX figure out some good default low power mode for cards w/out power tables */ - } - - if (rdev->pm.default_power_state == NULL) { /* add the default mode */ rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_DEFAULT; @@ -1840,18 +1972,17 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev) rdev->pm.power_state[state_index].default_clock_mode = &rdev->pm.power_state[state_index].clock_info[0]; rdev->pm.power_state[state_index].clock_info[0].voltage.type = VOLTAGE_NONE; - if (rdev->asic->get_pcie_lanes) - rdev->pm.power_state[state_index].non_clock_info.pcie_lanes = radeon_get_pcie_lanes(rdev); - else - rdev->pm.power_state[state_index].non_clock_info.pcie_lanes = 16; - rdev->pm.default_power_state = &rdev->pm.power_state[state_index]; + rdev->pm.power_state[state_index].pcie_lanes = 16; + rdev->pm.default_power_state_index = state_index; + rdev->pm.power_state[state_index].flags = 0; state_index++; } + rdev->pm.num_power_states = state_index; - rdev->pm.current_power_state = rdev->pm.default_power_state; - rdev->pm.current_clock_mode = - rdev->pm.default_power_state->default_clock_mode; + rdev->pm.current_power_state_index = rdev->pm.default_power_state_index; + rdev->pm.current_clock_mode_index = 0; + rdev->pm.current_vddc = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.voltage; } void radeon_atom_set_clock_gating(struct radeon_device *rdev, int enable) @@ -1907,6 +2038,42 @@ void radeon_atom_set_memory_clock(struct radeon_device *rdev, atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); } +union set_voltage { + struct _SET_VOLTAGE_PS_ALLOCATION alloc; + struct _SET_VOLTAGE_PARAMETERS v1; + struct _SET_VOLTAGE_PARAMETERS_V2 v2; +}; + +void radeon_atom_set_voltage(struct radeon_device *rdev, u16 level) +{ + union set_voltage args; + int index = GetIndexIntoMasterTable(COMMAND, SetVoltage); + u8 frev, crev, volt_index = level; + + if (!atom_parse_cmd_header(rdev->mode_info.atom_context, index, &frev, &crev)) + return; + + switch (crev) { + case 1: + args.v1.ucVoltageType = SET_VOLTAGE_TYPE_ASIC_VDDC; + args.v1.ucVoltageMode = SET_ASIC_VOLTAGE_MODE_ALL_SOURCE; + args.v1.ucVoltageIndex = volt_index; + break; + case 2: + args.v2.ucVoltageType = SET_VOLTAGE_TYPE_ASIC_VDDC; + args.v2.ucVoltageMode = SET_ASIC_VOLTAGE_MODE_SET_VOLTAGE; + args.v2.usVoltageLevel = cpu_to_le16(level); + break; + default: + DRM_ERROR("Unknown table version %d, %d\n", frev, crev); + return; + } + + atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); +} + + + void radeon_atom_initialize_bios_scratch_regs(struct drm_device *dev) { struct radeon_device *rdev = dev->dev_private; diff --git a/drivers/gpu/drm/radeon/radeon_bios.c b/drivers/gpu/drm/radeon/radeon_bios.c index 8ad71f701316..2c9213739999 100644 --- a/drivers/gpu/drm/radeon/radeon_bios.c +++ b/drivers/gpu/drm/radeon/radeon_bios.c @@ -48,6 +48,10 @@ static bool igp_read_bios_from_vram(struct radeon_device *rdev) resource_size_t vram_base; resource_size_t size = 256 * 1024; /* ??? */ + if (!(rdev->flags & RADEON_IS_IGP)) + if (!radeon_card_posted(rdev)) + return false; + rdev->bios = NULL; vram_base = drm_get_resource_start(rdev->ddev, 0); bios = ioremap(vram_base, size); @@ -85,12 +89,11 @@ static bool radeon_read_bios(struct radeon_device *rdev) pci_unmap_rom(rdev->pdev, bios); return false; } - rdev->bios = kmalloc(size, GFP_KERNEL); + rdev->bios = kmemdup(bios, size, GFP_KERNEL); if (rdev->bios == NULL) { pci_unmap_rom(rdev->pdev, bios); return false; } - memcpy(rdev->bios, bios, size); pci_unmap_rom(rdev->pdev, bios); return true; } diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c index 37db8adb2748..2417d7b06fdb 100644 --- a/drivers/gpu/drm/radeon/radeon_combios.c +++ b/drivers/gpu/drm/radeon/radeon_combios.c @@ -450,17 +450,17 @@ bool radeon_combios_check_hardcoded_edid(struct radeon_device *rdev) { int edid_info; struct edid *edid; + unsigned char *raw; edid_info = combios_get_table_offset(rdev->ddev, COMBIOS_HARDCODED_EDID_TABLE); if (!edid_info) return false; - edid = kmalloc(EDID_LENGTH * (DRM_MAX_EDID_EXT_NUM + 1), - GFP_KERNEL); + raw = rdev->bios + edid_info; + edid = kmalloc(EDID_LENGTH * (raw[0x7e] + 1), GFP_KERNEL); if (edid == NULL) return false; - memcpy((unsigned char *)edid, - (unsigned char *)(rdev->bios + edid_info), EDID_LENGTH); + memcpy((unsigned char *)edid, raw, EDID_LENGTH * (raw[0x7e] + 1)); if (!drm_edid_is_valid(edid)) { kfree(edid); @@ -600,7 +600,7 @@ static struct radeon_i2c_bus_rec combios_setup_i2c_bus(struct radeon_device *rde } i2c.mm_i2c = false; i2c.i2c_id = 0; - i2c.hpd_id = 0; + i2c.hpd = RADEON_HPD_NONE; if (ddc_line) i2c.valid = true; @@ -1113,18 +1113,20 @@ struct radeon_encoder_lvds *radeon_combios_get_lvds_info(struct radeon_encoder break; if ((RBIOS16(tmp) == lvds->native_mode.hdisplay) && - (RBIOS16(tmp + 2) == - lvds->native_mode.vdisplay)) { - lvds->native_mode.htotal = RBIOS16(tmp + 17) * 8; - lvds->native_mode.hsync_start = RBIOS16(tmp + 21) * 8; - lvds->native_mode.hsync_end = (RBIOS8(tmp + 23) + - RBIOS16(tmp + 21)) * 8; - - lvds->native_mode.vtotal = RBIOS16(tmp + 24); - lvds->native_mode.vsync_start = RBIOS16(tmp + 28) & 0x7ff; - lvds->native_mode.vsync_end = - ((RBIOS16(tmp + 28) & 0xf800) >> 11) + - (RBIOS16(tmp + 28) & 0x7ff); + (RBIOS16(tmp + 2) == lvds->native_mode.vdisplay)) { + lvds->native_mode.htotal = lvds->native_mode.hdisplay + + (RBIOS16(tmp + 17) - RBIOS16(tmp + 19)) * 8; + lvds->native_mode.hsync_start = lvds->native_mode.hdisplay + + (RBIOS16(tmp + 21) - RBIOS16(tmp + 19) - 1) * 8; + lvds->native_mode.hsync_end = lvds->native_mode.hsync_start + + (RBIOS8(tmp + 23) * 8); + + lvds->native_mode.vtotal = lvds->native_mode.vdisplay + + (RBIOS16(tmp + 24) - RBIOS16(tmp + 26)); + lvds->native_mode.vsync_start = lvds->native_mode.vdisplay + + ((RBIOS16(tmp + 28) & 0x7ff) - RBIOS16(tmp + 26)); + lvds->native_mode.vsync_end = lvds->native_mode.vsync_start + + ((RBIOS16(tmp + 28) & 0xf800) >> 11); lvds->native_mode.clock = RBIOS16(tmp + 9) * 10; lvds->native_mode.flags = 0; @@ -1409,6 +1411,11 @@ bool radeon_get_legacy_connector_info_from_table(struct drm_device *dev) rdev->mode_info.connector_table = CT_IMAC_G5_ISIGHT; } else #endif /* CONFIG_PPC_PMAC */ +#ifdef CONFIG_PPC64 + if (ASIC_IS_RN50(rdev)) + rdev->mode_info.connector_table = CT_RN50_POWER; + else +#endif rdev->mode_info.connector_table = CT_GENERIC; } @@ -1851,6 +1858,33 @@ bool radeon_get_legacy_connector_info_from_table(struct drm_device *dev) CONNECTOR_OBJECT_ID_SVIDEO, &hpd); break; + case CT_RN50_POWER: + DRM_INFO("Connector Table: %d (rn50-power)\n", + rdev->mode_info.connector_table); + /* VGA - primary dac */ + ddc_i2c = combios_setup_i2c_bus(rdev, RADEON_GPIO_VGA_DDC); + hpd.hpd = RADEON_HPD_NONE; + radeon_add_legacy_encoder(dev, + radeon_get_encoder_id(dev, + ATOM_DEVICE_CRT1_SUPPORT, + 1), + ATOM_DEVICE_CRT1_SUPPORT); + radeon_add_legacy_connector(dev, 0, ATOM_DEVICE_CRT1_SUPPORT, + DRM_MODE_CONNECTOR_VGA, &ddc_i2c, + CONNECTOR_OBJECT_ID_VGA, + &hpd); + ddc_i2c = combios_setup_i2c_bus(rdev, RADEON_GPIO_CRT2_DDC); + hpd.hpd = RADEON_HPD_NONE; + radeon_add_legacy_encoder(dev, + radeon_get_encoder_id(dev, + ATOM_DEVICE_CRT2_SUPPORT, + 2), + ATOM_DEVICE_CRT2_SUPPORT); + radeon_add_legacy_connector(dev, 1, ATOM_DEVICE_CRT2_SUPPORT, + DRM_MODE_CONNECTOR_VGA, &ddc_i2c, + CONNECTOR_OBJECT_ID_VGA, + &hpd); + break; default: DRM_INFO("Connector table: %d (invalid)\n", rdev->mode_info.connector_table); @@ -1904,15 +1938,6 @@ static bool radeon_apply_legacy_quirks(struct drm_device *dev, return false; } - /* Some RV100 cards with 2 VGA ports show up with DVI+VGA */ - if (dev->pdev->device == 0x5159 && - dev->pdev->subsystem_vendor == 0x1002 && - dev->pdev->subsystem_device == 0x013a) { - if (*legacy_connector == CONNECTOR_DVI_I_LEGACY) - *legacy_connector = CONNECTOR_CRT_LEGACY; - - } - /* X300 card with extra non-existent DVI port */ if (dev->pdev->device == 0x5B60 && dev->pdev->subsystem_vendor == 0x17af && @@ -2024,6 +2049,7 @@ bool radeon_get_legacy_connector_info_from_bios(struct drm_device *dev) combios_setup_i2c_bus(rdev, RADEON_GPIO_CRT2_DDC); break; default: + ddc_i2c.valid = false; break; } @@ -2196,7 +2222,7 @@ bool radeon_get_legacy_connector_info_from_bios(struct drm_device *dev) ATOM_DEVICE_DFP1_SUPPORT); ddc_i2c = combios_setup_i2c_bus(rdev, RADEON_GPIO_DVI_DDC); - hpd.hpd = RADEON_HPD_NONE; + hpd.hpd = RADEON_HPD_1; radeon_add_legacy_connector(dev, 0, ATOM_DEVICE_CRT1_SUPPORT | @@ -2337,6 +2363,7 @@ bool radeon_get_legacy_connector_info_from_bios(struct drm_device *dev) if (RBIOS8(tv_info + 6) == 'T') { if (radeon_apply_legacy_tv_quirks(dev)) { hpd.hpd = RADEON_HPD_NONE; + ddc_i2c.valid = false; radeon_add_legacy_encoder(dev, radeon_get_encoder_id (dev, @@ -2366,7 +2393,7 @@ void radeon_combios_get_power_modes(struct radeon_device *rdev) u8 rev, blocks, tmp; int state_index = 0; - rdev->pm.default_power_state = NULL; + rdev->pm.default_power_state_index = -1; if (rdev->flags & RADEON_IS_MOBILITY) { offset = combios_get_table_offset(dev, COMBIOS_POWERPLAY_INFO_TABLE); @@ -2380,17 +2407,13 @@ void radeon_combios_get_power_modes(struct radeon_device *rdev) if ((rdev->pm.power_state[state_index].clock_info[0].mclk == 0) || (rdev->pm.power_state[state_index].clock_info[0].sclk == 0)) goto default_mode; - /* skip overclock modes for now */ - if ((rdev->pm.power_state[state_index].clock_info[0].mclk > - rdev->clock.default_mclk + RADEON_MODE_OVERCLOCK_MARGIN) || - (rdev->pm.power_state[state_index].clock_info[0].sclk > - rdev->clock.default_sclk + RADEON_MODE_OVERCLOCK_MARGIN)) - goto default_mode; rdev->pm.power_state[state_index].type = POWER_STATE_TYPE_BATTERY; misc = RBIOS16(offset + 0x5 + 0x0); if (rev > 4) misc2 = RBIOS16(offset + 0x5 + 0xe); + rdev->pm.power_state[state_index].misc = misc; + rdev->pm.power_state[state_index].misc2 = misc2; if (misc & 0x4) { rdev->pm.power_state[state_index].clock_info[0].voltage.type = VOLTAGE_GPIO; if (misc & 0x8) @@ -2437,8 +2460,9 @@ void radeon_combios_get_power_modes(struct radeon_device *rdev) } else rdev->pm.power_state[state_index].clock_info[0].voltage.type = VOLTAGE_NONE; if (rev > 6) - rdev->pm.power_state[state_index].non_clock_info.pcie_lanes = + rdev->pm.power_state[state_index].pcie_lanes = RBIOS8(offset + 0x5 + 0x10); + rdev->pm.power_state[state_index].flags = RADEON_PM_STATE_SINGLE_DISPLAY_ONLY; state_index++; } else { /* XXX figure out some good default low power mode for mobility cards w/out power tables */ @@ -2455,17 +2479,19 @@ default_mode: rdev->pm.power_state[state_index].clock_info[0].mclk = rdev->clock.default_mclk; rdev->pm.power_state[state_index].clock_info[0].sclk = rdev->clock.default_sclk; rdev->pm.power_state[state_index].default_clock_mode = &rdev->pm.power_state[state_index].clock_info[0]; - rdev->pm.power_state[state_index].clock_info[0].voltage.type = VOLTAGE_NONE; - if (rdev->asic->get_pcie_lanes) - rdev->pm.power_state[state_index].non_clock_info.pcie_lanes = radeon_get_pcie_lanes(rdev); + if ((state_index > 0) && + (rdev->pm.power_state[0].clock_info[0].voltage.type == VOLTAGE_GPIO)) + rdev->pm.power_state[state_index].clock_info[0].voltage = + rdev->pm.power_state[0].clock_info[0].voltage; else - rdev->pm.power_state[state_index].non_clock_info.pcie_lanes = 16; - rdev->pm.default_power_state = &rdev->pm.power_state[state_index]; + rdev->pm.power_state[state_index].clock_info[0].voltage.type = VOLTAGE_NONE; + rdev->pm.power_state[state_index].pcie_lanes = 16; + rdev->pm.power_state[state_index].flags = 0; + rdev->pm.default_power_state_index = state_index; rdev->pm.num_power_states = state_index + 1; - rdev->pm.current_power_state = rdev->pm.default_power_state; - rdev->pm.current_clock_mode = - rdev->pm.default_power_state->default_clock_mode; + rdev->pm.current_power_state_index = rdev->pm.default_power_state_index; + rdev->pm.current_clock_mode_index = 0; } void radeon_external_tmds_setup(struct drm_encoder *encoder) @@ -3016,6 +3042,22 @@ void radeon_combios_asic_init(struct drm_device *dev) combios_write_ram_size(dev); } + /* quirk for rs4xx HP nx6125 laptop to make it resume + * - it hangs on resume inside the dynclk 1 table. + */ + if (rdev->family == CHIP_RS480 && + rdev->pdev->subsystem_vendor == 0x103c && + rdev->pdev->subsystem_device == 0x308b) + return; + + /* quirk for rs4xx HP dv5000 laptop to make it resume + * - it hangs on resume inside the dynclk 1 table. + */ + if (rdev->family == CHIP_RS480 && + rdev->pdev->subsystem_vendor == 0x103c && + rdev->pdev->subsystem_device == 0x30a4) + return; + /* DYN CLK 1 */ table = combios_get_table_offset(dev, COMBIOS_DYN_CLK_1_TABLE); if (table) diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c index 4559a53d5e57..adccbc2c202c 100644 --- a/drivers/gpu/drm/radeon/radeon_connectors.c +++ b/drivers/gpu/drm/radeon/radeon_connectors.c @@ -771,30 +771,27 @@ static enum drm_connector_status radeon_dvi_detect(struct drm_connector *connect } else ret = connector_status_connected; - /* multiple connectors on the same encoder with the same ddc line - * This tends to be HDMI and DVI on the same encoder with the - * same ddc line. If the edid says HDMI, consider the HDMI port - * connected and the DVI port disconnected. If the edid doesn't - * say HDMI, vice versa. + /* This gets complicated. We have boards with VGA + HDMI with a + * shared DDC line and we have boards with DVI-D + HDMI with a shared + * DDC line. The latter is more complex because with DVI<->HDMI adapters + * you don't really know what's connected to which port as both are digital. */ if (radeon_connector->shared_ddc && (ret == connector_status_connected)) { struct drm_device *dev = connector->dev; + struct radeon_device *rdev = dev->dev_private; struct drm_connector *list_connector; struct radeon_connector *list_radeon_connector; list_for_each_entry(list_connector, &dev->mode_config.connector_list, head) { if (connector == list_connector) continue; list_radeon_connector = to_radeon_connector(list_connector); - if (radeon_connector->devices == list_radeon_connector->devices) { - if (drm_detect_hdmi_monitor(radeon_connector->edid)) { - if (connector->connector_type == DRM_MODE_CONNECTOR_DVID) { - kfree(radeon_connector->edid); - radeon_connector->edid = NULL; - ret = connector_status_disconnected; - } - } else { - if ((connector->connector_type == DRM_MODE_CONNECTOR_HDMIA) || - (connector->connector_type == DRM_MODE_CONNECTOR_HDMIB)) { + if (list_radeon_connector->shared_ddc && + (list_radeon_connector->ddc_bus->rec.i2c_id == + radeon_connector->ddc_bus->rec.i2c_id)) { + /* cases where both connectors are digital */ + if (list_connector->connector_type != DRM_MODE_CONNECTOR_VGA) { + /* hpd is our only option in this case */ + if (!radeon_hpd_sense(rdev, radeon_connector->hpd.hpd)) { kfree(radeon_connector->edid); radeon_connector->edid = NULL; ret = connector_status_disconnected; @@ -1041,7 +1038,6 @@ radeon_add_atom_connector(struct drm_device *dev, struct radeon_connector_atom_dig *radeon_dig_connector; uint32_t subpixel_order = SubPixelNone; bool shared_ddc = false; - int ret; /* fixme - tv/cv/din */ if (connector_type == DRM_MODE_CONNECTOR_Unknown) @@ -1076,9 +1072,7 @@ radeon_add_atom_connector(struct drm_device *dev, switch (connector_type) { case DRM_MODE_CONNECTOR_VGA: drm_connector_init(dev, &radeon_connector->base, &radeon_vga_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_vga_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_vga_connector_helper_funcs); if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_create(dev, i2c_bus, "VGA"); if (!radeon_connector->ddc_bus) @@ -1088,12 +1082,11 @@ radeon_add_atom_connector(struct drm_device *dev, drm_connector_attach_property(&radeon_connector->base, rdev->mode_info.load_detect_property, 1); + connector->polled = DRM_CONNECTOR_POLL_CONNECT; break; case DRM_MODE_CONNECTOR_DVIA: drm_connector_init(dev, &radeon_connector->base, &radeon_vga_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_vga_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_vga_connector_helper_funcs); if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_create(dev, i2c_bus, "DVI"); if (!radeon_connector->ddc_bus) @@ -1113,9 +1106,7 @@ radeon_add_atom_connector(struct drm_device *dev, radeon_dig_connector->igp_lane_info = igp_lane_info; radeon_connector->con_priv = radeon_dig_connector; drm_connector_init(dev, &radeon_connector->base, &radeon_dvi_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_dvi_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_dvi_connector_helper_funcs); if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_create(dev, i2c_bus, "DVI"); if (!radeon_connector->ddc_bus) @@ -1141,9 +1132,7 @@ radeon_add_atom_connector(struct drm_device *dev, radeon_dig_connector->igp_lane_info = igp_lane_info; radeon_connector->con_priv = radeon_dig_connector; drm_connector_init(dev, &radeon_connector->base, &radeon_dvi_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_dvi_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_dvi_connector_helper_funcs); if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_create(dev, i2c_bus, "HDMI"); if (!radeon_connector->ddc_bus) @@ -1163,9 +1152,7 @@ radeon_add_atom_connector(struct drm_device *dev, radeon_dig_connector->igp_lane_info = igp_lane_info; radeon_connector->con_priv = radeon_dig_connector; drm_connector_init(dev, &radeon_connector->base, &radeon_dp_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_dp_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_dp_connector_helper_funcs); if (i2c_bus->valid) { /* add DP i2c bus */ if (connector_type == DRM_MODE_CONNECTOR_eDP) @@ -1191,9 +1178,7 @@ radeon_add_atom_connector(struct drm_device *dev, case DRM_MODE_CONNECTOR_9PinDIN: if (radeon_tv == 1) { drm_connector_init(dev, &radeon_connector->base, &radeon_tv_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_tv_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_tv_connector_helper_funcs); radeon_connector->dac_load_detect = true; drm_connector_attach_property(&radeon_connector->base, rdev->mode_info.load_detect_property, @@ -1211,9 +1196,7 @@ radeon_add_atom_connector(struct drm_device *dev, radeon_dig_connector->igp_lane_info = igp_lane_info; radeon_connector->con_priv = radeon_dig_connector; drm_connector_init(dev, &radeon_connector->base, &radeon_lvds_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_lvds_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_lvds_connector_helper_funcs); if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_create(dev, i2c_bus, "LVDS"); if (!radeon_connector->ddc_bus) @@ -1226,6 +1209,12 @@ radeon_add_atom_connector(struct drm_device *dev, break; } + if (hpd->hpd == RADEON_HPD_NONE) { + if (i2c_bus->valid) + connector->polled = DRM_CONNECTOR_POLL_CONNECT; + } else + connector->polled = DRM_CONNECTOR_POLL_HPD; + connector->display_info.subpixel_order = subpixel_order; drm_sysfs_connector_add(connector); return; @@ -1250,7 +1239,6 @@ radeon_add_legacy_connector(struct drm_device *dev, struct drm_connector *connector; struct radeon_connector *radeon_connector; uint32_t subpixel_order = SubPixelNone; - int ret; /* fixme - tv/cv/din */ if (connector_type == DRM_MODE_CONNECTOR_Unknown) @@ -1278,9 +1266,7 @@ radeon_add_legacy_connector(struct drm_device *dev, switch (connector_type) { case DRM_MODE_CONNECTOR_VGA: drm_connector_init(dev, &radeon_connector->base, &radeon_vga_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_vga_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_vga_connector_helper_funcs); if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_create(dev, i2c_bus, "VGA"); if (!radeon_connector->ddc_bus) @@ -1290,12 +1276,11 @@ radeon_add_legacy_connector(struct drm_device *dev, drm_connector_attach_property(&radeon_connector->base, rdev->mode_info.load_detect_property, 1); + connector->polled = DRM_CONNECTOR_POLL_CONNECT; break; case DRM_MODE_CONNECTOR_DVIA: drm_connector_init(dev, &radeon_connector->base, &radeon_vga_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_vga_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_vga_connector_helper_funcs); if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_create(dev, i2c_bus, "DVI"); if (!radeon_connector->ddc_bus) @@ -1309,9 +1294,7 @@ radeon_add_legacy_connector(struct drm_device *dev, case DRM_MODE_CONNECTOR_DVII: case DRM_MODE_CONNECTOR_DVID: drm_connector_init(dev, &radeon_connector->base, &radeon_dvi_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_dvi_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_dvi_connector_helper_funcs); if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_create(dev, i2c_bus, "DVI"); if (!radeon_connector->ddc_bus) @@ -1330,9 +1313,7 @@ radeon_add_legacy_connector(struct drm_device *dev, case DRM_MODE_CONNECTOR_9PinDIN: if (radeon_tv == 1) { drm_connector_init(dev, &radeon_connector->base, &radeon_tv_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_tv_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_tv_connector_helper_funcs); radeon_connector->dac_load_detect = true; /* RS400,RC410,RS480 chipset seems to report a lot * of false positive on load detect, we haven't yet @@ -1351,9 +1332,7 @@ radeon_add_legacy_connector(struct drm_device *dev, break; case DRM_MODE_CONNECTOR_LVDS: drm_connector_init(dev, &radeon_connector->base, &radeon_lvds_connector_funcs, connector_type); - ret = drm_connector_helper_add(&radeon_connector->base, &radeon_lvds_connector_helper_funcs); - if (ret) - goto failed; + drm_connector_helper_add(&radeon_connector->base, &radeon_lvds_connector_helper_funcs); if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_create(dev, i2c_bus, "LVDS"); if (!radeon_connector->ddc_bus) @@ -1366,6 +1345,11 @@ radeon_add_legacy_connector(struct drm_device *dev, break; } + if (hpd->hpd == RADEON_HPD_NONE) { + if (i2c_bus->valid) + connector->polled = DRM_CONNECTOR_POLL_CONNECT; + } else + connector->polled = DRM_CONNECTOR_POLL_HPD; connector->display_info.subpixel_order = subpixel_order; drm_sysfs_connector_add(connector); return; diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c index f9b0fe002c0a..ae0fb7356e62 100644 --- a/drivers/gpu/drm/radeon/radeon_cs.c +++ b/drivers/gpu/drm/radeon/radeon_cs.c @@ -220,10 +220,6 @@ int radeon_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) int r; mutex_lock(&rdev->cs_mutex); - if (rdev->gpu_lockup) { - mutex_unlock(&rdev->cs_mutex); - return -EINVAL; - } /* initialize parser */ memset(&parser, 0, sizeof(struct radeon_cs_parser)); parser.filp = filp; diff --git a/drivers/gpu/drm/radeon/radeon_cursor.c b/drivers/gpu/drm/radeon/radeon_cursor.c index b7023fff89eb..4eb67c0e0996 100644 --- a/drivers/gpu/drm/radeon/radeon_cursor.c +++ b/drivers/gpu/drm/radeon/radeon_cursor.c @@ -194,7 +194,7 @@ unpin: fail: drm_gem_object_unreference_unlocked(obj); - return 0; + return ret; } int radeon_crtc_cursor_move(struct drm_crtc *crtc, diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 7b629e305560..dd279da90546 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -226,20 +226,20 @@ void radeon_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc) { u64 size_af, size_bf; - size_af = 0xFFFFFFFF - mc->vram_end; - size_bf = mc->vram_start; + size_af = ((0xFFFFFFFF - mc->vram_end) + mc->gtt_base_align) & ~mc->gtt_base_align; + size_bf = mc->vram_start & ~mc->gtt_base_align; if (size_bf > size_af) { if (mc->gtt_size > size_bf) { dev_warn(rdev->dev, "limiting GTT\n"); mc->gtt_size = size_bf; } - mc->gtt_start = mc->vram_start - mc->gtt_size; + mc->gtt_start = (mc->vram_start & ~mc->gtt_base_align) - mc->gtt_size; } else { if (mc->gtt_size > size_af) { dev_warn(rdev->dev, "limiting GTT\n"); mc->gtt_size = size_af; } - mc->gtt_start = mc->vram_end + 1; + mc->gtt_start = (mc->vram_end + 1 + mc->gtt_base_align) & ~mc->gtt_base_align; } mc->gtt_end = mc->gtt_start + mc->gtt_size - 1; dev_info(rdev->dev, "GTT: %lluM 0x%08llX - 0x%08llX\n", @@ -299,24 +299,24 @@ void radeon_update_bandwidth_info(struct radeon_device *rdev) sclk = radeon_get_engine_clock(rdev); mclk = rdev->clock.default_mclk; - a.full = rfixed_const(100); - rdev->pm.sclk.full = rfixed_const(sclk); - rdev->pm.sclk.full = rfixed_div(rdev->pm.sclk, a); - rdev->pm.mclk.full = rfixed_const(mclk); - rdev->pm.mclk.full = rfixed_div(rdev->pm.mclk, a); + a.full = dfixed_const(100); + rdev->pm.sclk.full = dfixed_const(sclk); + rdev->pm.sclk.full = dfixed_div(rdev->pm.sclk, a); + rdev->pm.mclk.full = dfixed_const(mclk); + rdev->pm.mclk.full = dfixed_div(rdev->pm.mclk, a); - a.full = rfixed_const(16); + a.full = dfixed_const(16); /* core_bandwidth = sclk(Mhz) * 16 */ - rdev->pm.core_bandwidth.full = rfixed_div(rdev->pm.sclk, a); + rdev->pm.core_bandwidth.full = dfixed_div(rdev->pm.sclk, a); } else { sclk = radeon_get_engine_clock(rdev); mclk = radeon_get_memory_clock(rdev); - a.full = rfixed_const(100); - rdev->pm.sclk.full = rfixed_const(sclk); - rdev->pm.sclk.full = rfixed_div(rdev->pm.sclk, a); - rdev->pm.mclk.full = rfixed_const(mclk); - rdev->pm.mclk.full = rfixed_div(rdev->pm.mclk, a); + a.full = dfixed_const(100); + rdev->pm.sclk.full = dfixed_const(sclk); + rdev->pm.sclk.full = dfixed_div(rdev->pm.sclk, a); + rdev->pm.mclk.full = dfixed_const(mclk); + rdev->pm.mclk.full = dfixed_div(rdev->pm.mclk, a); } } @@ -546,8 +546,10 @@ static void radeon_switcheroo_set_state(struct pci_dev *pdev, enum vga_switchero /* don't suspend or resume card normally */ rdev->powered_down = false; radeon_resume_kms(dev); + drm_kms_helper_poll_enable(dev); } else { printk(KERN_INFO "radeon: switched off\n"); + drm_kms_helper_poll_disable(dev); radeon_suspend_kms(dev, pmm); /* don't suspend or resume card normally */ rdev->powered_down = true; @@ -599,9 +601,11 @@ int radeon_device_init(struct radeon_device *rdev, spin_lock_init(&rdev->ih.lock); mutex_init(&rdev->gem.mutex); mutex_init(&rdev->pm.mutex); + mutex_init(&rdev->vram_mutex); rwlock_init(&rdev->fence_drv.lock); INIT_LIST_HEAD(&rdev->gem.objects); init_waitqueue_head(&rdev->irq.vblank_queue); + init_waitqueue_head(&rdev->irq.idle_queue); /* setup workqueue */ rdev->wq = create_workqueue("radeon"); @@ -671,7 +675,7 @@ int radeon_device_init(struct radeon_device *rdev, /* Acceleration not working on AGP card try again * with fallback to PCI or PCIE GART */ - radeon_gpu_reset(rdev); + radeon_asic_reset(rdev); radeon_fini(rdev); radeon_agp_disable(rdev); r = radeon_init(rdev); @@ -691,6 +695,8 @@ void radeon_device_fini(struct radeon_device *rdev) { DRM_INFO("radeon: finishing device.\n"); rdev->shutdown = true; + /* evict vram memory */ + radeon_bo_evict_vram(rdev); radeon_fini(rdev); destroy_workqueue(rdev->wq); vga_switcheroo_unregister_client(rdev->pdev); @@ -707,6 +713,7 @@ int radeon_suspend_kms(struct drm_device *dev, pm_message_t state) { struct radeon_device *rdev; struct drm_crtc *crtc; + struct drm_connector *connector; int r; if (dev == NULL || dev->dev_private == NULL) { @@ -719,6 +726,12 @@ int radeon_suspend_kms(struct drm_device *dev, pm_message_t state) if (rdev->powered_down) return 0; + + /* turn off display hw */ + list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF); + } + /* unpin the front buffers */ list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { struct radeon_framebuffer *rfb = to_radeon_framebuffer(crtc->fb); @@ -728,9 +741,10 @@ int radeon_suspend_kms(struct drm_device *dev, pm_message_t state) continue; } robj = rfb->obj->driver_private; - if (robj != rdev->fbdev_rbo) { + /* don't unpin kernel fb objects */ + if (!radeon_fbdev_robj_is_fb(rdev, robj)) { r = radeon_bo_reserve(robj, false); - if (unlikely(r == 0)) { + if (r == 0) { radeon_bo_unpin(robj); radeon_bo_unreserve(robj); } @@ -743,11 +757,14 @@ int radeon_suspend_kms(struct drm_device *dev, pm_message_t state) radeon_save_bios_scratch_regs(rdev); + radeon_pm_suspend(rdev); radeon_suspend(rdev); radeon_hpd_fini(rdev); /* evict remaining vram memory */ radeon_bo_evict_vram(rdev); + radeon_agp_suspend(rdev); + pci_save_state(dev->pdev); if (state.event == PM_EVENT_SUSPEND) { /* Shut down the device */ @@ -755,13 +772,14 @@ int radeon_suspend_kms(struct drm_device *dev, pm_message_t state) pci_set_power_state(dev->pdev, PCI_D3hot); } acquire_console_sem(); - fb_set_suspend(rdev->fbdev_info, 1); + radeon_fbdev_set_suspend(rdev, 1); release_console_sem(); return 0; } int radeon_resume_kms(struct drm_device *dev) { + struct drm_connector *connector; struct radeon_device *rdev = dev->dev_private; if (rdev->powered_down) @@ -778,8 +796,15 @@ int radeon_resume_kms(struct drm_device *dev) /* resume AGP if in use */ radeon_agp_resume(rdev); radeon_resume(rdev); + radeon_pm_resume(rdev); radeon_restore_bios_scratch_regs(rdev); - fb_set_suspend(rdev->fbdev_info, 0); + + /* turn on display hw */ + list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON); + } + + radeon_fbdev_set_suspend(rdev, 0); release_console_sem(); /* reset hpd state */ @@ -789,6 +814,26 @@ int radeon_resume_kms(struct drm_device *dev) return 0; } +int radeon_gpu_reset(struct radeon_device *rdev) +{ + int r; + + radeon_save_bios_scratch_regs(rdev); + radeon_suspend(rdev); + + r = radeon_asic_reset(rdev); + if (!r) { + dev_info(rdev->dev, "GPU reset succeed\n"); + radeon_resume(rdev); + radeon_restore_bios_scratch_regs(rdev); + drm_helper_resume_force_mode(rdev->ddev); + return 0; + } + /* bad news, how to tell it to userspace ? */ + dev_info(rdev->dev, "GPU reset failed\n"); + return r; +} + /* * Debugfs diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index bb1c122cad21..8154cdf796e4 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -284,8 +284,7 @@ static const char *connector_names[15] = { "eDP", }; -static const char *hpd_names[7] = { - "NONE", +static const char *hpd_names[6] = { "HPD1", "HPD2", "HPD3", @@ -633,37 +632,37 @@ calc_fb_div(struct radeon_pll *pll, vco_freq = freq * post_div; /* feedback_divider = vco_freq * ref_div / pll->reference_freq; */ - a.full = rfixed_const(pll->reference_freq); - feedback_divider.full = rfixed_const(vco_freq); - feedback_divider.full = rfixed_div(feedback_divider, a); - a.full = rfixed_const(ref_div); - feedback_divider.full = rfixed_mul(feedback_divider, a); + a.full = dfixed_const(pll->reference_freq); + feedback_divider.full = dfixed_const(vco_freq); + feedback_divider.full = dfixed_div(feedback_divider, a); + a.full = dfixed_const(ref_div); + feedback_divider.full = dfixed_mul(feedback_divider, a); if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV) { /* feedback_divider = floor((feedback_divider * 10.0) + 0.5) * 0.1; */ - a.full = rfixed_const(10); - feedback_divider.full = rfixed_mul(feedback_divider, a); - feedback_divider.full += rfixed_const_half(0); - feedback_divider.full = rfixed_floor(feedback_divider); - feedback_divider.full = rfixed_div(feedback_divider, a); + a.full = dfixed_const(10); + feedback_divider.full = dfixed_mul(feedback_divider, a); + feedback_divider.full += dfixed_const_half(0); + feedback_divider.full = dfixed_floor(feedback_divider); + feedback_divider.full = dfixed_div(feedback_divider, a); /* *fb_div = floor(feedback_divider); */ - a.full = rfixed_floor(feedback_divider); - *fb_div = rfixed_trunc(a); + a.full = dfixed_floor(feedback_divider); + *fb_div = dfixed_trunc(a); /* *fb_div_frac = fmod(feedback_divider, 1.0) * 10.0; */ - a.full = rfixed_const(10); - b.full = rfixed_mul(feedback_divider, a); + a.full = dfixed_const(10); + b.full = dfixed_mul(feedback_divider, a); - feedback_divider.full = rfixed_floor(feedback_divider); - feedback_divider.full = rfixed_mul(feedback_divider, a); + feedback_divider.full = dfixed_floor(feedback_divider); + feedback_divider.full = dfixed_mul(feedback_divider, a); feedback_divider.full = b.full - feedback_divider.full; - *fb_div_frac = rfixed_trunc(feedback_divider); + *fb_div_frac = dfixed_trunc(feedback_divider); } else { /* *fb_div = floor(feedback_divider + 0.5); */ - feedback_divider.full += rfixed_const_half(0); - feedback_divider.full = rfixed_floor(feedback_divider); + feedback_divider.full += dfixed_const_half(0); + feedback_divider.full = dfixed_floor(feedback_divider); - *fb_div = rfixed_trunc(feedback_divider); + *fb_div = dfixed_trunc(feedback_divider); *fb_div_frac = 0; } @@ -693,10 +692,10 @@ calc_fb_ref_div(struct radeon_pll *pll, pll_out_max = pll->pll_out_max; } - ffreq.full = rfixed_const(freq); + ffreq.full = dfixed_const(freq); /* max_error = ffreq * 0.0025; */ - a.full = rfixed_const(400); - max_error.full = rfixed_div(ffreq, a); + a.full = dfixed_const(400); + max_error.full = dfixed_div(ffreq, a); for ((*ref_div) = pll->min_ref_div; (*ref_div) < pll->max_ref_div; ++(*ref_div)) { if (calc_fb_div(pll, freq, post_div, (*ref_div), fb_div, fb_div_frac)) { @@ -707,9 +706,9 @@ calc_fb_ref_div(struct radeon_pll *pll, continue; /* pll_out = vco / post_div; */ - a.full = rfixed_const(post_div); - pll_out.full = rfixed_const(vco); - pll_out.full = rfixed_div(pll_out, a); + a.full = dfixed_const(post_div); + pll_out.full = dfixed_const(vco); + pll_out.full = dfixed_div(pll_out, a); if (pll_out.full >= ffreq.full) { error.full = pll_out.full - ffreq.full; @@ -831,10 +830,6 @@ void radeon_compute_pll(struct radeon_pll *pll, static void radeon_user_framebuffer_destroy(struct drm_framebuffer *fb) { struct radeon_framebuffer *radeon_fb = to_radeon_framebuffer(fb); - struct drm_device *dev = fb->dev; - - if (fb->fbdev) - radeonfb_remove(dev, fb); if (radeon_fb->obj) drm_gem_object_unreference_unlocked(radeon_fb->obj); @@ -856,21 +851,15 @@ static const struct drm_framebuffer_funcs radeon_fb_funcs = { .create_handle = radeon_user_framebuffer_create_handle, }; -struct drm_framebuffer * -radeon_framebuffer_create(struct drm_device *dev, - struct drm_mode_fb_cmd *mode_cmd, - struct drm_gem_object *obj) +void +radeon_framebuffer_init(struct drm_device *dev, + struct radeon_framebuffer *rfb, + struct drm_mode_fb_cmd *mode_cmd, + struct drm_gem_object *obj) { - struct radeon_framebuffer *radeon_fb; - - radeon_fb = kzalloc(sizeof(*radeon_fb), GFP_KERNEL); - if (radeon_fb == NULL) { - return NULL; - } - drm_framebuffer_init(dev, &radeon_fb->base, &radeon_fb_funcs); - drm_helper_mode_fill_fb_struct(&radeon_fb->base, mode_cmd); - radeon_fb->obj = obj; - return &radeon_fb->base; + rfb->obj = obj; + drm_framebuffer_init(dev, &rfb->base, &radeon_fb_funcs); + drm_helper_mode_fill_fb_struct(&rfb->base, mode_cmd); } static struct drm_framebuffer * @@ -879,6 +868,7 @@ radeon_user_framebuffer_create(struct drm_device *dev, struct drm_mode_fb_cmd *mode_cmd) { struct drm_gem_object *obj; + struct radeon_framebuffer *radeon_fb; obj = drm_gem_object_lookup(dev, file_priv, mode_cmd->handle); if (obj == NULL) { @@ -886,12 +876,26 @@ radeon_user_framebuffer_create(struct drm_device *dev, "can't create framebuffer\n", mode_cmd->handle); return NULL; } - return radeon_framebuffer_create(dev, mode_cmd, obj); + + radeon_fb = kzalloc(sizeof(*radeon_fb), GFP_KERNEL); + if (radeon_fb == NULL) { + return NULL; + } + + radeon_framebuffer_init(dev, radeon_fb, mode_cmd, obj); + + return &radeon_fb->base; +} + +static void radeon_output_poll_changed(struct drm_device *dev) +{ + struct radeon_device *rdev = dev->dev_private; + radeon_fb_output_poll_changed(rdev); } static const struct drm_mode_config_funcs radeon_mode_funcs = { .fb_create = radeon_user_framebuffer_create, - .fb_changed = radeonfb_probe, + .output_poll_changed = radeon_output_poll_changed }; struct drm_prop_enum_list { @@ -978,8 +982,11 @@ void radeon_update_display_priority(struct radeon_device *rdev) /* set display priority to high for r3xx, rv515 chips * this avoids flickering due to underflow to the * display controllers during heavy acceleration. + * Don't force high on rs4xx igp chips as it seems to + * affect the sound card. See kernel bug 15982. */ - if (ASIC_IS_R300(rdev) || (rdev->family == CHIP_RV515)) + if ((ASIC_IS_R300(rdev) || (rdev->family == CHIP_RV515)) && + !(rdev->flags & RADEON_IS_IGP)) rdev->disp_priority = 2; else rdev->disp_priority = 0; @@ -1031,15 +1038,24 @@ int radeon_modeset_init(struct radeon_device *rdev) } /* initialize hpd */ radeon_hpd_init(rdev); - drm_helper_initial_config(rdev->ddev); + + /* Initialize power management */ + radeon_pm_init(rdev); + + radeon_fbdev_init(rdev); + drm_kms_helper_poll_init(rdev->ddev); + return 0; } void radeon_modeset_fini(struct radeon_device *rdev) { + radeon_fbdev_fini(rdev); kfree(rdev->mode_info.bios_hardcoded_edid); + radeon_pm_fini(rdev); if (rdev->mode_info.mode_config_initialized) { + drm_kms_helper_poll_fini(rdev->ddev); radeon_hpd_fini(rdev); drm_mode_config_cleanup(rdev->ddev); rdev->mode_info.mode_config_initialized = false; @@ -1089,15 +1105,15 @@ bool radeon_crtc_scaling_mode_fixup(struct drm_crtc *crtc, } if (radeon_crtc->rmx_type != RMX_OFF) { fixed20_12 a, b; - a.full = rfixed_const(crtc->mode.vdisplay); - b.full = rfixed_const(radeon_crtc->native_mode.hdisplay); - radeon_crtc->vsc.full = rfixed_div(a, b); - a.full = rfixed_const(crtc->mode.hdisplay); - b.full = rfixed_const(radeon_crtc->native_mode.vdisplay); - radeon_crtc->hsc.full = rfixed_div(a, b); + a.full = dfixed_const(crtc->mode.vdisplay); + b.full = dfixed_const(radeon_crtc->native_mode.hdisplay); + radeon_crtc->vsc.full = dfixed_div(a, b); + a.full = dfixed_const(crtc->mode.hdisplay); + b.full = dfixed_const(radeon_crtc->native_mode.vdisplay); + radeon_crtc->hsc.full = dfixed_div(a, b); } else { - radeon_crtc->vsc.full = rfixed_const(1); - radeon_crtc->hsc.full = rfixed_const(1); + radeon_crtc->vsc.full = dfixed_const(1); + radeon_crtc->hsc.full = dfixed_const(1); } return true; } diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index b3749d47be7b..e166fe4d7c30 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -44,9 +44,11 @@ * - 2.1.0 - add square tiling interface * - 2.2.0 - add r6xx/r7xx const buffer support * - 2.3.0 - add MSPOS + 3D texture + r500 VAP regs + * - 2.4.0 - add crtc id query + * - 2.5.0 - add get accel 2 to work around ddx breakage for evergreen */ #define KMS_DRIVER_MAJOR 2 -#define KMS_DRIVER_MINOR 3 +#define KMS_DRIVER_MINOR 5 #define KMS_DRIVER_PATCHLEVEL 0 int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags); int radeon_driver_unload_kms(struct drm_device *dev); @@ -91,7 +93,6 @@ int radeon_testing = 0; int radeon_connector_table = 0; int radeon_tv = 1; int radeon_new_pll = -1; -int radeon_dynpm = -1; int radeon_audio = 1; int radeon_disp_priority = 0; int radeon_hw_i2c = 0; @@ -132,9 +133,6 @@ module_param_named(tv, radeon_tv, int, 0444); MODULE_PARM_DESC(new_pll, "Select new PLL code"); module_param_named(new_pll, radeon_new_pll, int, 0444); -MODULE_PARM_DESC(dynpm, "Disable/Enable dynamic power management (1 = enable)"); -module_param_named(dynpm, radeon_dynpm, int, 0444); - MODULE_PARM_DESC(audio, "Audio enable (0 = disable)"); module_param_named(audio, radeon_audio, int, 0444); diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c index c5ddaf58563a..e0b30b264c28 100644 --- a/drivers/gpu/drm/radeon/radeon_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_encoders.c @@ -309,9 +309,6 @@ static bool radeon_atom_mode_fixup(struct drm_encoder *encoder, struct drm_device *dev = encoder->dev; struct radeon_device *rdev = dev->dev_private; - /* adjust pm to upcoming mode change */ - radeon_pm_compute_clocks(rdev); - /* set the active encoder to connector routing */ radeon_encoder_set_active_device(encoder); drm_mode_set_crtcinfo(adjusted_mode, 0); @@ -1075,6 +1072,8 @@ radeon_atom_encoder_dpms(struct drm_encoder *encoder, int mode) if (is_dig) { switch (mode) { case DRM_MODE_DPMS_ON: + if (!ASIC_IS_DCE4(rdev)) + atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_ENABLE_OUTPUT, 0, 0); if (atombios_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_DP) { struct drm_connector *connector = radeon_get_connector_for_encoder(encoder); @@ -1082,8 +1081,6 @@ radeon_atom_encoder_dpms(struct drm_encoder *encoder, int mode) if (ASIC_IS_DCE4(rdev)) atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_DP_VIDEO_ON); } - if (!ASIC_IS_DCE4(rdev)) - atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_ENABLE_OUTPUT, 0, 0); break; case DRM_MODE_DPMS_STANDBY: case DRM_MODE_DPMS_SUSPEND: @@ -1111,8 +1108,6 @@ radeon_atom_encoder_dpms(struct drm_encoder *encoder, int mode) } radeon_atombios_encoder_dpms_scratch_regs(encoder, (mode == DRM_MODE_DPMS_ON) ? true : false); - /* adjust pm to dpms change */ - radeon_pm_compute_clocks(rdev); } union crtc_source_param { @@ -1546,10 +1541,49 @@ static void radeon_atom_encoder_commit(struct drm_encoder *encoder) static void radeon_atom_encoder_disable(struct drm_encoder *encoder) { + struct drm_device *dev = encoder->dev; + struct radeon_device *rdev = dev->dev_private; struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); struct radeon_encoder_atom_dig *dig; radeon_atom_encoder_dpms(encoder, DRM_MODE_DPMS_OFF); + switch (radeon_encoder->encoder_id) { + case ENCODER_OBJECT_ID_INTERNAL_TMDS1: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1: + case ENCODER_OBJECT_ID_INTERNAL_LVDS: + case ENCODER_OBJECT_ID_INTERNAL_LVTM1: + atombios_digital_setup(encoder, PANEL_ENCODER_ACTION_DISABLE); + break; + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA: + if (ASIC_IS_DCE4(rdev)) + /* disable the transmitter */ + atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE, 0, 0); + else { + /* disable the encoder and transmitter */ + atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE, 0, 0); + atombios_dig_encoder_setup(encoder, ATOM_DISABLE); + } + break; + case ENCODER_OBJECT_ID_INTERNAL_DDI: + atombios_ddia_setup(encoder, ATOM_DISABLE); + break; + case ENCODER_OBJECT_ID_INTERNAL_DVO1: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1: + atombios_external_tmds_setup(encoder, ATOM_DISABLE); + break; + case ENCODER_OBJECT_ID_INTERNAL_DAC1: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC1: + case ENCODER_OBJECT_ID_INTERNAL_DAC2: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC2: + atombios_dac_setup(encoder, ATOM_DISABLE); + if (radeon_encoder->devices & (ATOM_DEVICE_TV_SUPPORT | ATOM_DEVICE_CV_SUPPORT)) + atombios_tv_setup(encoder, ATOM_DISABLE); + break; + } + if (radeon_encoder_is_digital(encoder)) { if (atombios_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_HDMI) r600_hdmi_disable(encoder); diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c index 9ac57a09784b..dc1634bb0c11 100644 --- a/drivers/gpu/drm/radeon/radeon_fb.c +++ b/drivers/gpu/drm/radeon/radeon_fb.c @@ -23,10 +23,6 @@ * Authors: * David Airlie */ - /* - * Modularization - */ - #include <linux/module.h> #include <linux/slab.h> #include <linux/fb.h> @@ -42,17 +38,21 @@ #include <linux/vga_switcheroo.h> -struct radeon_fb_device { +/* object hierarchy - + this contains a helper + a radeon fb + the helper contains a pointer to radeon framebuffer baseclass. +*/ +struct radeon_fbdev { struct drm_fb_helper helper; - struct radeon_framebuffer *rfb; - struct radeon_device *rdev; + struct radeon_framebuffer rfb; + struct list_head fbdev_list; + struct radeon_device *rdev; }; static struct fb_ops radeonfb_ops = { .owner = THIS_MODULE, .fb_check_var = drm_fb_helper_check_var, .fb_set_par = drm_fb_helper_set_par, - .fb_setcolreg = drm_fb_helper_setcolreg, .fb_fillrect = cfb_fillrect, .fb_copyarea = cfb_copyarea, .fb_imageblit = cfb_imageblit, @@ -61,45 +61,6 @@ static struct fb_ops radeonfb_ops = { .fb_setcmap = drm_fb_helper_setcmap, }; -/** - * Currently it is assumed that the old framebuffer is reused. - * - * LOCKING - * caller should hold the mode config lock. - * - */ -int radeonfb_resize(struct drm_device *dev, struct drm_crtc *crtc) -{ - struct fb_info *info; - struct drm_framebuffer *fb; - struct drm_display_mode *mode = crtc->desired_mode; - - fb = crtc->fb; - if (fb == NULL) { - return 1; - } - info = fb->fbdev; - if (info == NULL) { - return 1; - } - if (mode == NULL) { - return 1; - } - info->var.xres = mode->hdisplay; - info->var.right_margin = mode->hsync_start - mode->hdisplay; - info->var.hsync_len = mode->hsync_end - mode->hsync_start; - info->var.left_margin = mode->htotal - mode->hsync_end; - info->var.yres = mode->vdisplay; - info->var.lower_margin = mode->vsync_start - mode->vdisplay; - info->var.vsync_len = mode->vsync_end - mode->vsync_start; - info->var.upper_margin = mode->vtotal - mode->vsync_end; - info->var.pixclock = 10000000 / mode->htotal * 1000 / mode->vtotal * 100; - /* avoid overflow */ - info->var.pixclock = info->var.pixclock * 1000 / mode->vrefresh; - - return 0; -} -EXPORT_SYMBOL(radeonfb_resize); static int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bool tiled) { @@ -125,57 +86,44 @@ static int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bo return aligned; } -static struct drm_fb_helper_funcs radeon_fb_helper_funcs = { - .gamma_set = radeon_crtc_fb_gamma_set, - .gamma_get = radeon_crtc_fb_gamma_get, -}; +static void radeonfb_destroy_pinned_object(struct drm_gem_object *gobj) +{ + struct radeon_bo *rbo = gobj->driver_private; + int ret; + + ret = radeon_bo_reserve(rbo, false); + if (likely(ret == 0)) { + radeon_bo_kunmap(rbo); + radeon_bo_unreserve(rbo); + } + drm_gem_object_unreference_unlocked(gobj); +} -int radeonfb_create(struct drm_device *dev, - uint32_t fb_width, uint32_t fb_height, - uint32_t surface_width, uint32_t surface_height, - uint32_t surface_depth, uint32_t surface_bpp, - struct drm_framebuffer **fb_p) +static int radeonfb_create_pinned_object(struct radeon_fbdev *rfbdev, + struct drm_mode_fb_cmd *mode_cmd, + struct drm_gem_object **gobj_p) { - struct radeon_device *rdev = dev->dev_private; - struct fb_info *info; - struct radeon_fb_device *rfbdev; - struct drm_framebuffer *fb = NULL; - struct radeon_framebuffer *rfb; - struct drm_mode_fb_cmd mode_cmd; + struct radeon_device *rdev = rfbdev->rdev; struct drm_gem_object *gobj = NULL; struct radeon_bo *rbo = NULL; - struct device *device = &rdev->pdev->dev; - int size, aligned_size, ret; - u64 fb_gpuaddr; - void *fbptr = NULL; - unsigned long tmp; bool fb_tiled = false; /* useful for testing */ u32 tiling_flags = 0; + int ret; + int aligned_size, size; - mode_cmd.width = surface_width; - mode_cmd.height = surface_height; - - /* avivo can't scanout real 24bpp */ - if ((surface_bpp == 24) && ASIC_IS_AVIVO(rdev)) - surface_bpp = 32; - - mode_cmd.bpp = surface_bpp; /* need to align pitch with crtc limits */ - mode_cmd.pitch = radeon_align_pitch(rdev, mode_cmd.width, mode_cmd.bpp, fb_tiled) * ((mode_cmd.bpp + 1) / 8); - mode_cmd.depth = surface_depth; + mode_cmd->pitch = radeon_align_pitch(rdev, mode_cmd->width, mode_cmd->bpp, fb_tiled) * ((mode_cmd->bpp + 1) / 8); - size = mode_cmd.pitch * mode_cmd.height; + size = mode_cmd->pitch * mode_cmd->height; aligned_size = ALIGN(size, PAGE_SIZE); - ret = radeon_gem_object_create(rdev, aligned_size, 0, - RADEON_GEM_DOMAIN_VRAM, - false, ttm_bo_type_kernel, - &gobj); + RADEON_GEM_DOMAIN_VRAM, + false, ttm_bo_type_kernel, + &gobj); if (ret) { - printk(KERN_ERR "failed to allocate framebuffer (%d %d)\n", - surface_width, surface_height); - ret = -ENOMEM; - goto out; + printk(KERN_ERR "failed to allocate framebuffer (%d)\n", + aligned_size); + return -ENOMEM; } rbo = gobj->driver_private; @@ -183,7 +131,7 @@ int radeonfb_create(struct drm_device *dev, tiling_flags = RADEON_TILING_MACRO; #ifdef __BIG_ENDIAN - switch (mode_cmd.bpp) { + switch (mode_cmd->bpp) { case 32: tiling_flags |= RADEON_TILING_SWAP_32BIT; break; @@ -196,57 +144,81 @@ int radeonfb_create(struct drm_device *dev, if (tiling_flags) { ret = radeon_bo_set_tiling_flags(rbo, - tiling_flags | RADEON_TILING_SURFACE, - mode_cmd.pitch); + tiling_flags | RADEON_TILING_SURFACE, + mode_cmd->pitch); if (ret) dev_err(rdev->dev, "FB failed to set tiling flags\n"); } - mutex_lock(&rdev->ddev->struct_mutex); - fb = radeon_framebuffer_create(rdev->ddev, &mode_cmd, gobj); - if (fb == NULL) { - DRM_ERROR("failed to allocate fb.\n"); - ret = -ENOMEM; - goto out_unref; - } + + ret = radeon_bo_reserve(rbo, false); if (unlikely(ret != 0)) goto out_unref; - ret = radeon_bo_pin(rbo, RADEON_GEM_DOMAIN_VRAM, &fb_gpuaddr); + ret = radeon_bo_pin(rbo, RADEON_GEM_DOMAIN_VRAM, NULL); if (ret) { radeon_bo_unreserve(rbo); goto out_unref; } if (fb_tiled) radeon_bo_check_tiling(rbo, 0, 0); - ret = radeon_bo_kmap(rbo, &fbptr); + ret = radeon_bo_kmap(rbo, NULL); radeon_bo_unreserve(rbo); if (ret) { goto out_unref; } - list_add(&fb->filp_head, &rdev->ddev->mode_config.fb_kernel_list); + *gobj_p = gobj; + return 0; +out_unref: + radeonfb_destroy_pinned_object(gobj); + *gobj_p = NULL; + return ret; +} - *fb_p = fb; - rfb = to_radeon_framebuffer(fb); - rdev->fbdev_rfb = rfb; - rdev->fbdev_rbo = rbo; +static int radeonfb_create(struct radeon_fbdev *rfbdev, + struct drm_fb_helper_surface_size *sizes) +{ + struct radeon_device *rdev = rfbdev->rdev; + struct fb_info *info; + struct drm_framebuffer *fb = NULL; + struct drm_mode_fb_cmd mode_cmd; + struct drm_gem_object *gobj = NULL; + struct radeon_bo *rbo = NULL; + struct device *device = &rdev->pdev->dev; + int ret; + unsigned long tmp; + + mode_cmd.width = sizes->surface_width; + mode_cmd.height = sizes->surface_height; + + /* avivo can't scanout real 24bpp */ + if ((sizes->surface_bpp == 24) && ASIC_IS_AVIVO(rdev)) + sizes->surface_bpp = 32; - info = framebuffer_alloc(sizeof(struct radeon_fb_device), device); + mode_cmd.bpp = sizes->surface_bpp; + mode_cmd.depth = sizes->surface_depth; + + ret = radeonfb_create_pinned_object(rfbdev, &mode_cmd, &gobj); + rbo = gobj->driver_private; + + /* okay we have an object now allocate the framebuffer */ + info = framebuffer_alloc(0, device); if (info == NULL) { ret = -ENOMEM; goto out_unref; } - rdev->fbdev_info = info; - rfbdev = info->par; - rfbdev->helper.funcs = &radeon_fb_helper_funcs; - rfbdev->helper.dev = dev; - ret = drm_fb_helper_init_crtc_count(&rfbdev->helper, rdev->num_crtc, - RADEONFB_CONN_LIMIT); - if (ret) - goto out_unref; + info->par = rfbdev; + + radeon_framebuffer_init(rdev->ddev, &rfbdev->rfb, &mode_cmd, gobj); + + fb = &rfbdev->rfb.base; + + /* setup helper */ + rfbdev->helper.fb = fb; + rfbdev->helper.fbdev = info; - memset_io(fbptr, 0x0, aligned_size); + memset_io(rbo->kptr, 0x0, radeon_bo_size(rbo)); strcpy(info->fix.id, "radeondrmfb"); @@ -255,17 +227,22 @@ int radeonfb_create(struct drm_device *dev, info->flags = FBINFO_DEFAULT; info->fbops = &radeonfb_ops; - tmp = fb_gpuaddr - rdev->mc.vram_start; + tmp = radeon_bo_gpu_offset(rbo) - rdev->mc.vram_start; info->fix.smem_start = rdev->mc.aper_base + tmp; - info->fix.smem_len = size; - info->screen_base = fbptr; - info->screen_size = size; + info->fix.smem_len = radeon_bo_size(rbo); + info->screen_base = rbo->kptr; + info->screen_size = radeon_bo_size(rbo); - drm_fb_helper_fill_var(info, fb, fb_width, fb_height); + drm_fb_helper_fill_var(info, &rfbdev->helper, sizes->fb_width, sizes->fb_height); /* setup aperture base/size for vesafb takeover */ - info->aperture_base = rdev->ddev->mode_config.fb_base; - info->aperture_size = rdev->mc.real_vram_size; + info->apertures = alloc_apertures(1); + if (!info->apertures) { + ret = -ENOMEM; + goto out_unref; + } + info->apertures->ranges[0].base = rdev->ddev->mode_config.fb_base; + info->apertures->ranges[0].size = rdev->mc.real_vram_size; info->fix.mmio_start = 0; info->fix.mmio_len = 0; @@ -274,44 +251,55 @@ int radeonfb_create(struct drm_device *dev, info->pixmap.access_align = 32; info->pixmap.flags = FB_PIXMAP_SYSTEM; info->pixmap.scan_align = 1; + if (info->screen_base == NULL) { ret = -ENOSPC; goto out_unref; } + + ret = fb_alloc_cmap(&info->cmap, 256, 0); + if (ret) { + ret = -ENOMEM; + goto out_unref; + } + DRM_INFO("fb mappable at 0x%lX\n", info->fix.smem_start); DRM_INFO("vram apper at 0x%lX\n", (unsigned long)rdev->mc.aper_base); - DRM_INFO("size %lu\n", (unsigned long)size); + DRM_INFO("size %lu\n", (unsigned long)radeon_bo_size(rbo)); DRM_INFO("fb depth is %d\n", fb->depth); DRM_INFO(" pitch is %d\n", fb->pitch); - fb->fbdev = info; - rfbdev->rfb = rfb; - rfbdev->rdev = rdev; - - mutex_unlock(&rdev->ddev->struct_mutex); vga_switcheroo_client_fb_set(rdev->ddev->pdev, info); return 0; out_unref: if (rbo) { - ret = radeon_bo_reserve(rbo, false); - if (likely(ret == 0)) { - radeon_bo_kunmap(rbo); - radeon_bo_unreserve(rbo); - } + } if (fb && ret) { - list_del(&fb->filp_head); drm_gem_object_unreference(gobj); drm_framebuffer_cleanup(fb); kfree(fb); } - drm_gem_object_unreference(gobj); - mutex_unlock(&rdev->ddev->struct_mutex); -out: return ret; } +static int radeon_fb_find_or_create_single(struct drm_fb_helper *helper, + struct drm_fb_helper_surface_size *sizes) +{ + struct radeon_fbdev *rfbdev = (struct radeon_fbdev *)helper; + int new_fb = 0; + int ret; + + if (!helper->fb) { + ret = radeonfb_create(rfbdev, sizes); + if (ret) + return ret; + new_fb = 1; + } + return new_fb; +} + static char *mode_option; int radeon_parse_options(char *options) { @@ -328,46 +316,108 @@ int radeon_parse_options(char *options) return 0; } -int radeonfb_probe(struct drm_device *dev) +void radeon_fb_output_poll_changed(struct radeon_device *rdev) { - struct radeon_device *rdev = dev->dev_private; - int bpp_sel = 32; - - /* select 8 bpp console on RN50 or 16MB cards */ - if (ASIC_IS_RN50(rdev) || rdev->mc.real_vram_size <= (32*1024*1024)) - bpp_sel = 8; - - return drm_fb_helper_single_fb_probe(dev, bpp_sel, &radeonfb_create); + drm_fb_helper_hotplug_event(&rdev->mode_info.rfbdev->helper); } -int radeonfb_remove(struct drm_device *dev, struct drm_framebuffer *fb) +static int radeon_fbdev_destroy(struct drm_device *dev, struct radeon_fbdev *rfbdev) { struct fb_info *info; - struct radeon_framebuffer *rfb = to_radeon_framebuffer(fb); + struct radeon_framebuffer *rfb = &rfbdev->rfb; struct radeon_bo *rbo; int r; - if (!fb) { - return -EINVAL; + if (rfbdev->helper.fbdev) { + info = rfbdev->helper.fbdev; + + unregister_framebuffer(info); + if (info->cmap.len) + fb_dealloc_cmap(&info->cmap); + framebuffer_release(info); } - info = fb->fbdev; - if (info) { - struct radeon_fb_device *rfbdev = info->par; + + if (rfb->obj) { rbo = rfb->obj->driver_private; - unregister_framebuffer(info); r = radeon_bo_reserve(rbo, false); if (likely(r == 0)) { radeon_bo_kunmap(rbo); radeon_bo_unpin(rbo); radeon_bo_unreserve(rbo); } - drm_fb_helper_free(&rfbdev->helper); - framebuffer_release(info); + drm_gem_object_unreference_unlocked(rfb->obj); } + drm_fb_helper_fini(&rfbdev->helper); + drm_framebuffer_cleanup(&rfb->base); - printk(KERN_INFO "unregistered panic notifier\n"); + return 0; +} +static struct drm_fb_helper_funcs radeon_fb_helper_funcs = { + .gamma_set = radeon_crtc_fb_gamma_set, + .gamma_get = radeon_crtc_fb_gamma_get, + .fb_probe = radeon_fb_find_or_create_single, +}; + +int radeon_fbdev_init(struct radeon_device *rdev) +{ + struct radeon_fbdev *rfbdev; + int bpp_sel = 32; + int ret; + + /* select 8 bpp console on RN50 or 16MB cards */ + if (ASIC_IS_RN50(rdev) || rdev->mc.real_vram_size <= (32*1024*1024)) + bpp_sel = 8; + + rfbdev = kzalloc(sizeof(struct radeon_fbdev), GFP_KERNEL); + if (!rfbdev) + return -ENOMEM; + + rfbdev->rdev = rdev; + rdev->mode_info.rfbdev = rfbdev; + rfbdev->helper.funcs = &radeon_fb_helper_funcs; + + ret = drm_fb_helper_init(rdev->ddev, &rfbdev->helper, + rdev->num_crtc, + RADEONFB_CONN_LIMIT); + if (ret) { + kfree(rfbdev); + return ret; + } + + drm_fb_helper_single_add_all_connectors(&rfbdev->helper); + drm_fb_helper_initial_config(&rfbdev->helper, bpp_sel); return 0; } -EXPORT_SYMBOL(radeonfb_remove); -MODULE_LICENSE("GPL"); + +void radeon_fbdev_fini(struct radeon_device *rdev) +{ + if (!rdev->mode_info.rfbdev) + return; + + radeon_fbdev_destroy(rdev->ddev, rdev->mode_info.rfbdev); + kfree(rdev->mode_info.rfbdev); + rdev->mode_info.rfbdev = NULL; +} + +void radeon_fbdev_set_suspend(struct radeon_device *rdev, int state) +{ + fb_set_suspend(rdev->mode_info.rfbdev->helper.fbdev, state); +} + +int radeon_fbdev_total_size(struct radeon_device *rdev) +{ + struct radeon_bo *robj; + int size = 0; + + robj = rdev->mode_info.rfbdev->rfb.obj->driver_private; + size += radeon_bo_size(robj); + return size; +} + +bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj) +{ + if (robj == rdev->mode_info.rfbdev->rfb.obj->driver_private) + return true; + return false; +} diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index d90f95b405c5..b1f9a81b5d1d 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -58,7 +58,6 @@ int radeon_fence_emit(struct radeon_device *rdev, struct radeon_fence *fence) radeon_fence_ring_emit(rdev, fence); fence->emited = true; - fence->timeout = jiffies + ((2000 * HZ) / 1000); list_del(&fence->list); list_add_tail(&fence->list, &rdev->fence_drv.emited); write_unlock_irqrestore(&rdev->fence_drv.lock, irq_flags); @@ -71,15 +70,34 @@ static bool radeon_fence_poll_locked(struct radeon_device *rdev) struct list_head *i, *n; uint32_t seq; bool wake = false; + unsigned long cjiffies; - if (rdev == NULL) { - return true; - } - if (rdev->shutdown) { - return true; - } seq = RREG32(rdev->fence_drv.scratch_reg); - rdev->fence_drv.last_seq = seq; + if (seq != rdev->fence_drv.last_seq) { + rdev->fence_drv.last_seq = seq; + rdev->fence_drv.last_jiffies = jiffies; + rdev->fence_drv.last_timeout = RADEON_FENCE_JIFFIES_TIMEOUT; + } else { + cjiffies = jiffies; + if (time_after(cjiffies, rdev->fence_drv.last_jiffies)) { + cjiffies -= rdev->fence_drv.last_jiffies; + if (time_after(rdev->fence_drv.last_timeout, cjiffies)) { + /* update the timeout */ + rdev->fence_drv.last_timeout -= cjiffies; + } else { + /* the 500ms timeout is elapsed we should test + * for GPU lockup + */ + rdev->fence_drv.last_timeout = 1; + } + } else { + /* wrap around update last jiffies, we will just wait + * a little longer + */ + rdev->fence_drv.last_jiffies = cjiffies; + } + return false; + } n = NULL; list_for_each(i, &rdev->fence_drv.emited) { fence = list_entry(i, struct radeon_fence, list); @@ -171,9 +189,8 @@ bool radeon_fence_signaled(struct radeon_fence *fence) int radeon_fence_wait(struct radeon_fence *fence, bool intr) { struct radeon_device *rdev; - unsigned long cur_jiffies; - unsigned long timeout; - bool expired = false; + unsigned long irq_flags, timeout; + u32 seq; int r; if (fence == NULL) { @@ -184,21 +201,18 @@ int radeon_fence_wait(struct radeon_fence *fence, bool intr) if (radeon_fence_signaled(fence)) { return 0; } - + timeout = rdev->fence_drv.last_timeout; retry: - cur_jiffies = jiffies; - timeout = HZ / 100; - if (time_after(fence->timeout, cur_jiffies)) { - timeout = fence->timeout - cur_jiffies; - } - + /* save current sequence used to check for GPU lockup */ + seq = rdev->fence_drv.last_seq; if (intr) { radeon_irq_kms_sw_irq_get(rdev); r = wait_event_interruptible_timeout(rdev->fence_drv.queue, radeon_fence_signaled(fence), timeout); radeon_irq_kms_sw_irq_put(rdev); - if (unlikely(r < 0)) + if (unlikely(r < 0)) { return r; + } } else { radeon_irq_kms_sw_irq_get(rdev); r = wait_event_timeout(rdev->fence_drv.queue, @@ -206,38 +220,36 @@ retry: radeon_irq_kms_sw_irq_put(rdev); } if (unlikely(!radeon_fence_signaled(fence))) { - if (unlikely(r == 0)) { - expired = true; + /* we were interrupted for some reason and fence isn't + * isn't signaled yet, resume wait + */ + if (r) { + timeout = r; + goto retry; } - if (unlikely(expired)) { - timeout = 1; - if (time_after(cur_jiffies, fence->timeout)) { - timeout = cur_jiffies - fence->timeout; - } - timeout = jiffies_to_msecs(timeout); - if (timeout > 500) { - DRM_ERROR("fence(%p:0x%08X) %lums timeout " - "going to reset GPU\n", - fence, fence->seq, timeout); - radeon_gpu_reset(rdev); - WREG32(rdev->fence_drv.scratch_reg, fence->seq); - } + /* don't protect read access to rdev->fence_drv.last_seq + * if we experiencing a lockup the value doesn't change + */ + if (seq == rdev->fence_drv.last_seq && radeon_gpu_is_lockup(rdev)) { + /* good news we believe it's a lockup */ + WARN(1, "GPU lockup (waiting for 0x%08X last fence id 0x%08X)\n", fence->seq, seq); + /* FIXME: what should we do ? marking everyone + * as signaled for now + */ + rdev->gpu_lockup = true; + r = radeon_gpu_reset(rdev); + if (r) + return r; + WREG32(rdev->fence_drv.scratch_reg, fence->seq); + rdev->gpu_lockup = false; } + timeout = RADEON_FENCE_JIFFIES_TIMEOUT; + write_lock_irqsave(&rdev->fence_drv.lock, irq_flags); + rdev->fence_drv.last_timeout = RADEON_FENCE_JIFFIES_TIMEOUT; + rdev->fence_drv.last_jiffies = jiffies; + write_unlock_irqrestore(&rdev->fence_drv.lock, irq_flags); goto retry; } - if (unlikely(expired)) { - rdev->fence_drv.count_timeout++; - cur_jiffies = jiffies; - timeout = 1; - if (time_after(cur_jiffies, fence->timeout)) { - timeout = cur_jiffies - fence->timeout; - } - timeout = jiffies_to_msecs(timeout); - DRM_ERROR("fence(%p:0x%08X) %lums timeout\n", - fence, fence->seq, timeout); - DRM_ERROR("last signaled fence(0x%08X)\n", - rdev->fence_drv.last_seq); - } return 0; } @@ -333,7 +345,6 @@ int radeon_fence_driver_init(struct radeon_device *rdev) INIT_LIST_HEAD(&rdev->fence_drv.created); INIT_LIST_HEAD(&rdev->fence_drv.emited); INIT_LIST_HEAD(&rdev->fence_drv.signaled); - rdev->fence_drv.count_timeout = 0; init_waitqueue_head(&rdev->fence_drv.queue); rdev->fence_drv.initialized = true; write_unlock_irqrestore(&rdev->fence_drv.lock, irq_flags); diff --git a/drivers/gpu/drm/radeon/radeon_fixed.h b/drivers/gpu/drm/radeon/radeon_fixed.h deleted file mode 100644 index 3d4d84e078ac..000000000000 --- a/drivers/gpu/drm/radeon/radeon_fixed.h +++ /dev/null @@ -1,67 +0,0 @@ -/* - * Copyright 2009 Red Hat Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: Dave Airlie - */ -#ifndef RADEON_FIXED_H -#define RADEON_FIXED_H - -typedef union rfixed { - u32 full; -} fixed20_12; - - -#define rfixed_const(A) (u32)(((A) << 12))/* + ((B + 0.000122)*4096)) */ -#define rfixed_const_half(A) (u32)(((A) << 12) + 2048) -#define rfixed_const_666(A) (u32)(((A) << 12) + 2731) -#define rfixed_const_8(A) (u32)(((A) << 12) + 3277) -#define rfixed_mul(A, B) ((u64)((u64)(A).full * (B).full + 2048) >> 12) -#define fixed_init(A) { .full = rfixed_const((A)) } -#define fixed_init_half(A) { .full = rfixed_const_half((A)) } -#define rfixed_trunc(A) ((A).full >> 12) - -static inline u32 rfixed_floor(fixed20_12 A) -{ - u32 non_frac = rfixed_trunc(A); - - return rfixed_const(non_frac); -} - -static inline u32 rfixed_ceil(fixed20_12 A) -{ - u32 non_frac = rfixed_trunc(A); - - if (A.full > rfixed_const(non_frac)) - return rfixed_const(non_frac + 1); - else - return rfixed_const(non_frac); -} - -static inline u32 rfixed_div(fixed20_12 A, fixed20_12 B) -{ - u64 tmp = ((u64)A.full << 13); - - do_div(tmp, B.full); - tmp += 1; - tmp /= 2; - return lower_32_bits(tmp); -} -#endif diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c index 1770d3c07fd0..e65b90317fab 100644 --- a/drivers/gpu/drm/radeon/radeon_gart.c +++ b/drivers/gpu/drm/radeon/radeon_gart.c @@ -173,7 +173,7 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset, int i, j; if (!rdev->gart.ready) { - DRM_ERROR("trying to bind memory to unitialized GART !\n"); + WARN(1, "trying to bind memory to unitialized GART !\n"); return -EINVAL; } t = offset / RADEON_GPU_PAGE_SIZE; diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c index ef92d147d8f0..a72a3ee5d69b 100644 --- a/drivers/gpu/drm/radeon/radeon_gem.c +++ b/drivers/gpu/drm/radeon/radeon_gem.c @@ -44,6 +44,9 @@ void radeon_gem_object_free(struct drm_gem_object *gobj) if (robj) { radeon_bo_unref(&robj); } + + drm_gem_object_release(gobj); + kfree(gobj); } int radeon_gem_object_create(struct radeon_device *rdev, int size, @@ -158,8 +161,7 @@ int radeon_gem_info_ioctl(struct drm_device *dev, void *data, args->vram_visible = rdev->mc.real_vram_size; if (rdev->stollen_vga_memory) args->vram_visible -= radeon_bo_size(rdev->stollen_vga_memory); - if (rdev->fbdev_rbo) - args->vram_visible -= radeon_bo_size(rdev->fbdev_rbo); + args->vram_visible -= radeon_fbdev_total_size(rdev); args->gart_size = rdev->mc.gtt_size - rdev->cp.ring_size - 4096 - RADEON_IB_POOL_SIZE*64*1024; return 0; diff --git a/drivers/gpu/drm/radeon/radeon_irq_kms.c b/drivers/gpu/drm/radeon/radeon_irq_kms.c index a212041e8b0b..059bfa4098d7 100644 --- a/drivers/gpu/drm/radeon/radeon_irq_kms.c +++ b/drivers/gpu/drm/radeon/radeon_irq_kms.c @@ -26,6 +26,7 @@ * Jerome Glisse */ #include "drmP.h" +#include "drm_crtc_helper.h" #include "radeon_drm.h" #include "radeon_reg.h" #include "radeon.h" @@ -55,7 +56,7 @@ static void radeon_hotplug_work_func(struct work_struct *work) radeon_connector_hotplug(connector); } /* Just fire off a uevent and let userspace tell us what to do */ - drm_sysfs_hotplug_event(dev); + drm_helper_hpd_irq_event(dev); } void radeon_driver_irq_preinstall_kms(struct drm_device *dev) @@ -67,6 +68,7 @@ void radeon_driver_irq_preinstall_kms(struct drm_device *dev) /* Disable *all* interrupts */ rdev->irq.sw_int = false; + rdev->irq.gui_idle = false; for (i = 0; i < rdev->num_crtc; i++) rdev->irq.crtc_vblank_int[i] = false; for (i = 0; i < 6; i++) @@ -96,6 +98,7 @@ void radeon_driver_irq_uninstall_kms(struct drm_device *dev) } /* Disable *all* interrupts */ rdev->irq.sw_int = false; + rdev->irq.gui_idle = false; for (i = 0; i < rdev->num_crtc; i++) rdev->irq.crtc_vblank_int[i] = false; for (i = 0; i < 6; i++) diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index c633319f98ed..ab389f89fa8d 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -98,11 +98,15 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) { struct radeon_device *rdev = dev->dev_private; struct drm_radeon_info *info; + struct radeon_mode_info *minfo = &rdev->mode_info; uint32_t *value_ptr; uint32_t value; + struct drm_crtc *crtc; + int i, found; info = data; value_ptr = (uint32_t *)((unsigned long)info->value); + value = *value_ptr; switch (info->request) { case RADEON_INFO_DEVICE_ID: value = dev->pci_device; @@ -114,6 +118,28 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) value = rdev->num_z_pipes; break; case RADEON_INFO_ACCEL_WORKING: + /* xf86-video-ati 6.13.0 relies on this being false for evergreen */ + if ((rdev->family >= CHIP_CEDAR) && (rdev->family <= CHIP_HEMLOCK)) + value = false; + else + value = rdev->accel_working; + break; + case RADEON_INFO_CRTC_FROM_ID: + for (i = 0, found = 0; i < rdev->num_crtc; i++) { + crtc = (struct drm_crtc *)minfo->crtcs[i]; + if (crtc && crtc->base.id == value) { + struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); + value = radeon_crtc->crtc_id; + found = 1; + break; + } + } + if (!found) { + DRM_DEBUG("unknown crtc id %d\n", value); + return -EINVAL; + } + break; + case RADEON_INFO_ACCEL_WORKING2: value = rdev->accel_working; break; default: diff --git a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c index 88865e38fe30..e1e5255396ac 100644 --- a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c +++ b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c @@ -26,7 +26,7 @@ #include <drm/drmP.h> #include <drm/drm_crtc_helper.h> #include <drm/radeon_drm.h> -#include "radeon_fixed.h" +#include <drm/drm_fixed.h> #include "radeon.h" #include "atom.h" @@ -314,6 +314,9 @@ void radeon_crtc_dpms(struct drm_crtc *crtc, int mode) switch (mode) { case DRM_MODE_DPMS_ON: + radeon_crtc->enabled = true; + /* adjust pm to dpms changes BEFORE enabling crtcs */ + radeon_pm_compute_clocks(rdev); if (radeon_crtc->crtc_id) WREG32_P(RADEON_CRTC2_GEN_CNTL, RADEON_CRTC2_EN, ~(RADEON_CRTC2_EN | mask)); else { @@ -335,6 +338,9 @@ void radeon_crtc_dpms(struct drm_crtc *crtc, int mode) RADEON_CRTC_DISP_REQ_EN_B)); WREG32_P(RADEON_CRTC_EXT_CNTL, mask, ~mask); } + radeon_crtc->enabled = false; + /* adjust pm to dpms changes AFTER disabling crtcs */ + radeon_pm_compute_clocks(rdev); break; } } @@ -966,6 +972,12 @@ static bool radeon_crtc_mode_fixup(struct drm_crtc *crtc, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode) { + struct drm_device *dev = crtc->dev; + struct radeon_device *rdev = dev->dev_private; + + /* adjust pm to upcoming mode change */ + radeon_pm_compute_clocks(rdev); + if (!radeon_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) return false; return true; diff --git a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c index 0274abe17ad9..5688a0cf6bbe 100644 --- a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c @@ -108,6 +108,7 @@ static void radeon_legacy_lvds_dpms(struct drm_encoder *encoder, int mode) udelay(panel_pwr_delay * 1000); WREG32(RADEON_LVDS_GEN_CNTL, lvds_gen_cntl); WREG32_PLL(RADEON_PIXCLKS_CNTL, pixclks_cntl); + udelay(panel_pwr_delay * 1000); break; } @@ -116,8 +117,6 @@ static void radeon_legacy_lvds_dpms(struct drm_encoder *encoder, int mode) else radeon_combios_encoder_dpms_scratch_regs(encoder, (mode == DRM_MODE_DPMS_ON) ? true : false); - /* adjust pm to dpms change */ - radeon_pm_compute_clocks(rdev); } static void radeon_legacy_lvds_prepare(struct drm_encoder *encoder) @@ -217,11 +216,6 @@ static bool radeon_legacy_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *adjusted_mode) { struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); - struct drm_device *dev = encoder->dev; - struct radeon_device *rdev = dev->dev_private; - - /* adjust pm to upcoming mode change */ - radeon_pm_compute_clocks(rdev); /* set the active encoder to connector routing */ radeon_encoder_set_active_device(encoder); @@ -286,8 +280,6 @@ static void radeon_legacy_primary_dac_dpms(struct drm_encoder *encoder, int mode else radeon_combios_encoder_dpms_scratch_regs(encoder, (mode == DRM_MODE_DPMS_ON) ? true : false); - /* adjust pm to dpms change */ - radeon_pm_compute_clocks(rdev); } static void radeon_legacy_primary_dac_prepare(struct drm_encoder *encoder) @@ -474,8 +466,6 @@ static void radeon_legacy_tmds_int_dpms(struct drm_encoder *encoder, int mode) else radeon_combios_encoder_dpms_scratch_regs(encoder, (mode == DRM_MODE_DPMS_ON) ? true : false); - /* adjust pm to dpms change */ - radeon_pm_compute_clocks(rdev); } static void radeon_legacy_tmds_int_prepare(struct drm_encoder *encoder) @@ -642,8 +632,6 @@ static void radeon_legacy_tmds_ext_dpms(struct drm_encoder *encoder, int mode) else radeon_combios_encoder_dpms_scratch_regs(encoder, (mode == DRM_MODE_DPMS_ON) ? true : false); - /* adjust pm to dpms change */ - radeon_pm_compute_clocks(rdev); } static void radeon_legacy_tmds_ext_prepare(struct drm_encoder *encoder) @@ -852,8 +840,6 @@ static void radeon_legacy_tv_dac_dpms(struct drm_encoder *encoder, int mode) else radeon_combios_encoder_dpms_scratch_regs(encoder, (mode == DRM_MODE_DPMS_ON) ? true : false); - /* adjust pm to dpms change */ - radeon_pm_compute_clocks(rdev); } static void radeon_legacy_tv_dac_prepare(struct drm_encoder *encoder) @@ -943,16 +929,14 @@ static void radeon_legacy_tv_dac_mode_set(struct drm_encoder *encoder, if (ASIC_IS_R300(rdev)) { gpiopad_a = RREG32(RADEON_GPIOPAD_A) | 1; disp_output_cntl = RREG32(RADEON_DISP_OUTPUT_CNTL); - } - - if (rdev->family == CHIP_R200 || ASIC_IS_R300(rdev)) - disp_tv_out_cntl = RREG32(RADEON_DISP_TV_OUT_CNTL); - else + } else if (rdev->family != CHIP_R200) disp_hw_debug = RREG32(RADEON_DISP_HW_DEBUG); - - if (rdev->family == CHIP_R200) + else if (rdev->family == CHIP_R200) fp2_gen_cntl = RREG32(RADEON_FP2_GEN_CNTL); + if (rdev->family >= CHIP_R200) + disp_tv_out_cntl = RREG32(RADEON_DISP_TV_OUT_CNTL); + if (is_tv) { uint32_t dac_cntl; @@ -1017,15 +1001,13 @@ static void radeon_legacy_tv_dac_mode_set(struct drm_encoder *encoder, if (ASIC_IS_R300(rdev)) { WREG32_P(RADEON_GPIOPAD_A, gpiopad_a, ~1); WREG32(RADEON_DISP_OUTPUT_CNTL, disp_output_cntl); - } + } else if (rdev->family != CHIP_R200) + WREG32(RADEON_DISP_HW_DEBUG, disp_hw_debug); + else if (rdev->family == CHIP_R200) + WREG32(RADEON_FP2_GEN_CNTL, fp2_gen_cntl); if (rdev->family >= CHIP_R200) WREG32(RADEON_DISP_TV_OUT_CNTL, disp_tv_out_cntl); - else - WREG32(RADEON_DISP_HW_DEBUG, disp_hw_debug); - - if (rdev->family == CHIP_R200) - WREG32(RADEON_FP2_GEN_CNTL, fp2_gen_cntl); if (is_tv) radeon_legacy_tv_mode_set(encoder, mode, adjusted_mode); @@ -1183,6 +1165,17 @@ static enum drm_connector_status radeon_legacy_tv_dac_detect(struct drm_encoder struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); struct radeon_encoder_tv_dac *tv_dac = radeon_encoder->enc_priv; bool color = true; + struct drm_crtc *crtc; + + /* find out if crtc2 is in use or if this encoder is using it */ + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); + if ((radeon_crtc->crtc_id == 1) && crtc->enabled) { + if (encoder->crtc != crtc) { + return connector_status_disconnected; + } + } + } if (connector->connector_type == DRM_MODE_CONNECTOR_SVIDEO || connector->connector_type == DRM_MODE_CONNECTOR_Composite || diff --git a/drivers/gpu/drm/radeon/radeon_legacy_tv.c b/drivers/gpu/drm/radeon/radeon_legacy_tv.c index f2ed27c8055b..032040397743 100644 --- a/drivers/gpu/drm/radeon/radeon_legacy_tv.c +++ b/drivers/gpu/drm/radeon/radeon_legacy_tv.c @@ -642,8 +642,8 @@ void radeon_legacy_tv_mode_set(struct drm_encoder *encoder, } flicker_removal = (tmp + 500) / 1000; - if (flicker_removal < 2) - flicker_removal = 2; + if (flicker_removal < 3) + flicker_removal = 3; for (i = 0; i < ARRAY_SIZE(SLOPE_limit); ++i) { if (flicker_removal == SLOPE_limit[i]) break; diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index 5413fcd63086..95696aa57ac8 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -34,11 +34,12 @@ #include <drm_mode.h> #include <drm_edid.h> #include <drm_dp_helper.h> +#include <drm_fixed.h> #include <linux/i2c.h> #include <linux/i2c-id.h> #include <linux/i2c-algo-bit.h> -#include "radeon_fixed.h" +struct radeon_bo; struct radeon_device; #define to_radeon_crtc(x) container_of(x, struct radeon_crtc, base) @@ -65,6 +66,16 @@ enum radeon_tv_std { TV_STD_PAL_N, }; +enum radeon_hpd_id { + RADEON_HPD_1 = 0, + RADEON_HPD_2, + RADEON_HPD_3, + RADEON_HPD_4, + RADEON_HPD_5, + RADEON_HPD_6, + RADEON_HPD_NONE = 0xff, +}; + /* radeon gpio-based i2c * 1. "mask" reg and bits * grabs the gpio pins for software use @@ -84,7 +95,7 @@ struct radeon_i2c_bus_rec { /* id used by atom */ uint8_t i2c_id; /* id used by atom */ - uint8_t hpd_id; + enum radeon_hpd_id hpd; /* can be used with hw i2c engine */ bool hw_capable; /* uses multi-media i2c engine */ @@ -195,6 +206,7 @@ enum radeon_connector_table { CT_MINI_INTERNAL, CT_IMAC_G5_ISIGHT, CT_EMAC, + CT_RN50_POWER, }; enum radeon_dvo_chip { @@ -202,6 +214,8 @@ enum radeon_dvo_chip { DVO_SIL1178, }; +struct radeon_fbdev; + struct radeon_mode_info { struct atom_context *atom_context; struct card_info *atom_card_info; @@ -218,6 +232,9 @@ struct radeon_mode_info { struct drm_property *tmds_pll_property; /* hardcoded DFP edid from BIOS */ struct edid *bios_hardcoded_edid; + + /* pointer to fbdev info structure */ + struct radeon_fbdev *rfbdev; }; #define MAX_H_CODE_TIMING_LEN 32 @@ -339,6 +356,7 @@ struct radeon_encoder { enum radeon_rmx_type rmx_type; struct drm_display_mode native_mode; void *enc_priv; + int audio_polling_active; int hdmi_offset; int hdmi_config_offset; int hdmi_audio_workaround; @@ -363,16 +381,6 @@ struct radeon_gpio_rec { u32 mask; }; -enum radeon_hpd_id { - RADEON_HPD_NONE = 0, - RADEON_HPD_1, - RADEON_HPD_2, - RADEON_HPD_3, - RADEON_HPD_4, - RADEON_HPD_5, - RADEON_HPD_6, -}; - struct radeon_hpd { enum radeon_hpd_id hpd; u8 plugged_state; @@ -532,11 +540,10 @@ extern void radeon_crtc_fb_gamma_set(struct drm_crtc *crtc, u16 red, u16 green, u16 blue, int regno); extern void radeon_crtc_fb_gamma_get(struct drm_crtc *crtc, u16 *red, u16 *green, u16 *blue, int regno); -struct drm_framebuffer *radeon_framebuffer_create(struct drm_device *dev, - struct drm_mode_fb_cmd *mode_cmd, - struct drm_gem_object *obj); - -int radeonfb_probe(struct drm_device *dev); +void radeon_framebuffer_init(struct drm_device *dev, + struct radeon_framebuffer *rfb, + struct drm_mode_fb_cmd *mode_cmd, + struct drm_gem_object *obj); int radeonfb_remove(struct drm_device *dev, struct drm_framebuffer *fb); bool radeon_get_legacy_connector_info_from_bios(struct drm_device *dev); @@ -575,4 +582,13 @@ void radeon_legacy_tv_adjust_pll2(struct drm_encoder *encoder, void radeon_legacy_tv_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode); + +/* fbdev layer */ +int radeon_fbdev_init(struct radeon_device *rdev); +void radeon_fbdev_fini(struct radeon_device *rdev); +void radeon_fbdev_set_suspend(struct radeon_device *rdev, int state); +int radeon_fbdev_total_size(struct radeon_device *rdev); +bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj); + +void radeon_fb_output_poll_changed(struct radeon_device *rdev); #endif diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index 122774742bd5..d5b9373ce06c 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -112,9 +112,11 @@ int radeon_bo_create(struct radeon_device *rdev, struct drm_gem_object *gobj, radeon_ttm_placement_from_domain(bo, domain); /* Kernel allocation are uninterruptible */ + mutex_lock(&rdev->vram_mutex); r = ttm_bo_init(&rdev->mman.bdev, &bo->tbo, size, type, &bo->placement, 0, 0, !kernel, NULL, size, &radeon_ttm_bo_destroy); + mutex_unlock(&rdev->vram_mutex); if (unlikely(r != 0)) { if (r != -ERESTARTSYS) dev_err(rdev->dev, @@ -166,11 +168,15 @@ void radeon_bo_kunmap(struct radeon_bo *bo) void radeon_bo_unref(struct radeon_bo **bo) { struct ttm_buffer_object *tbo; + struct radeon_device *rdev; if ((*bo) == NULL) return; + rdev = (*bo)->rdev; tbo = &((*bo)->tbo); + mutex_lock(&rdev->vram_mutex); ttm_bo_unref(&tbo); + mutex_unlock(&rdev->vram_mutex); if (tbo == NULL) *bo = NULL; } @@ -192,7 +198,7 @@ int radeon_bo_pin(struct radeon_bo *bo, u32 domain, u64 *gpu_addr) } for (i = 0; i < bo->placement.num_placement; i++) bo->placements[i] |= TTM_PL_FLAG_NO_EVICT; - r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false); + r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false, false); if (likely(r == 0)) { bo->pin_count = 1; if (gpu_addr != NULL) @@ -216,7 +222,7 @@ int radeon_bo_unpin(struct radeon_bo *bo) return 0; for (i = 0; i < bo->placement.num_placement; i++) bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT; - r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false); + r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false, false); if (unlikely(r != 0)) dev_err(bo->rdev->dev, "%p validate failed for unpin\n", bo); return r; @@ -295,6 +301,7 @@ int radeon_bo_list_reserve(struct list_head *head) r = radeon_bo_reserve(lobj->bo, false); if (unlikely(r != 0)) return r; + lobj->reserved = true; } return 0; } @@ -305,7 +312,7 @@ void radeon_bo_list_unreserve(struct list_head *head) list_for_each_entry(lobj, head, list) { /* only unreserve object we successfully reserved */ - if (radeon_bo_is_reserved(lobj->bo)) + if (lobj->reserved && radeon_bo_is_reserved(lobj->bo)) radeon_bo_unreserve(lobj->bo); } } @@ -316,6 +323,9 @@ int radeon_bo_list_validate(struct list_head *head) struct radeon_bo *bo; int r; + list_for_each_entry(lobj, head, list) { + lobj->reserved = false; + } r = radeon_bo_list_reserve(head); if (unlikely(r != 0)) { return r; @@ -331,7 +341,7 @@ int radeon_bo_list_validate(struct list_head *head) lobj->rdomain); } r = ttm_bo_validate(&bo->tbo, &bo->placement, - true, false); + true, false, false); if (unlikely(r)) return r; } @@ -499,11 +509,33 @@ void radeon_bo_move_notify(struct ttm_buffer_object *bo, radeon_bo_check_tiling(rbo, 0, 1); } -void radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo) +int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo) { + struct radeon_device *rdev; struct radeon_bo *rbo; + unsigned long offset, size; + int r; + if (!radeon_ttm_bo_is_radeon_bo(bo)) - return; + return 0; rbo = container_of(bo, struct radeon_bo, tbo); radeon_bo_check_tiling(rbo, 0, 0); + rdev = rbo->rdev; + if (bo->mem.mem_type == TTM_PL_VRAM) { + size = bo->mem.num_pages << PAGE_SHIFT; + offset = bo->mem.mm_node->start << PAGE_SHIFT; + if ((offset + size) > rdev->mc.visible_vram_size) { + /* hurrah the memory is not visible ! */ + radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM); + rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT; + r = ttm_bo_validate(bo, &rbo->placement, false, true, false); + if (unlikely(r != 0)) + return r; + offset = bo->mem.mm_node->start << PAGE_SHIFT; + /* this should not happen */ + if ((offset + size) > rdev->mc.visible_vram_size) + return -EINVAL; + } + } + return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_object.h b/drivers/gpu/drm/radeon/radeon_object.h index 7ab43de1e244..353998dc2c03 100644 --- a/drivers/gpu/drm/radeon/radeon_object.h +++ b/drivers/gpu/drm/radeon/radeon_object.h @@ -168,6 +168,6 @@ extern int radeon_bo_check_tiling(struct radeon_bo *bo, bool has_moved, bool force_drop); extern void radeon_bo_move_notify(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem); -extern void radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo); +extern int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo); extern int radeon_bo_get_surface_reg(struct radeon_bo *bo); #endif diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c index a4b57493aa78..3fa6984d9896 100644 --- a/drivers/gpu/drm/radeon/radeon_pm.c +++ b/drivers/gpu/drm/radeon/radeon_pm.c @@ -23,25 +23,17 @@ #include "drmP.h" #include "radeon.h" #include "avivod.h" +#ifdef CONFIG_ACPI +#include <linux/acpi.h> +#endif +#include <linux/power_supply.h> #define RADEON_IDLE_LOOP_MS 100 #define RADEON_RECLOCK_DELAY_MS 200 #define RADEON_WAIT_VBLANK_TIMEOUT 200 +#define RADEON_WAIT_IDLE_TIMEOUT 200 -static bool radeon_pm_debug_check_in_vbl(struct radeon_device *rdev, bool finish); -static void radeon_pm_set_clocks_locked(struct radeon_device *rdev); -static void radeon_pm_set_clocks(struct radeon_device *rdev); -static void radeon_pm_idle_work_handler(struct work_struct *work); -static int radeon_debugfs_pm_init(struct radeon_device *rdev); - -static const char *pm_state_names[4] = { - "PM_STATE_DISABLED", - "PM_STATE_MINIMUM", - "PM_STATE_PAUSED", - "PM_STATE_ACTIVE" -}; - -static const char *pm_state_types[5] = { +static const char *radeon_pm_state_type_name[5] = { "Default", "Powersave", "Battery", @@ -49,138 +41,109 @@ static const char *pm_state_types[5] = { "Performance", }; -static void radeon_print_power_mode_info(struct radeon_device *rdev) +static void radeon_dynpm_idle_work_handler(struct work_struct *work); +static int radeon_debugfs_pm_init(struct radeon_device *rdev); +static bool radeon_pm_in_vbl(struct radeon_device *rdev); +static bool radeon_pm_debug_check_in_vbl(struct radeon_device *rdev, bool finish); +static void radeon_pm_update_profile(struct radeon_device *rdev); +static void radeon_pm_set_clocks(struct radeon_device *rdev); + +#define ACPI_AC_CLASS "ac_adapter" + +#ifdef CONFIG_ACPI +static int radeon_acpi_event(struct notifier_block *nb, + unsigned long val, + void *data) { - int i, j; - bool is_default; + struct radeon_device *rdev = container_of(nb, struct radeon_device, acpi_nb); + struct acpi_bus_event *entry = (struct acpi_bus_event *)data; - DRM_INFO("%d Power State(s)\n", rdev->pm.num_power_states); - for (i = 0; i < rdev->pm.num_power_states; i++) { - if (rdev->pm.default_power_state == &rdev->pm.power_state[i]) - is_default = true; + if (strcmp(entry->device_class, ACPI_AC_CLASS) == 0) { + if (power_supply_is_system_supplied() > 0) + DRM_DEBUG("pm: AC\n"); else - is_default = false; - DRM_INFO("State %d %s %s\n", i, - pm_state_types[rdev->pm.power_state[i].type], - is_default ? "(default)" : ""); - if ((rdev->flags & RADEON_IS_PCIE) && !(rdev->flags & RADEON_IS_IGP)) - DRM_INFO("\t%d PCIE Lanes\n", rdev->pm.power_state[i].non_clock_info.pcie_lanes); - DRM_INFO("\t%d Clock Mode(s)\n", rdev->pm.power_state[i].num_clock_modes); - for (j = 0; j < rdev->pm.power_state[i].num_clock_modes; j++) { - if (rdev->flags & RADEON_IS_IGP) - DRM_INFO("\t\t%d engine: %d\n", - j, - rdev->pm.power_state[i].clock_info[j].sclk * 10); - else - DRM_INFO("\t\t%d engine/memory: %d/%d\n", - j, - rdev->pm.power_state[i].clock_info[j].sclk * 10, - rdev->pm.power_state[i].clock_info[j].mclk * 10); + DRM_DEBUG("pm: DC\n"); + + if (rdev->pm.pm_method == PM_METHOD_PROFILE) { + if (rdev->pm.profile == PM_PROFILE_AUTO) { + mutex_lock(&rdev->pm.mutex); + radeon_pm_update_profile(rdev); + radeon_pm_set_clocks(rdev); + mutex_unlock(&rdev->pm.mutex); + } } } + + return NOTIFY_OK; } +#endif -static struct radeon_power_state * radeon_pick_power_state(struct radeon_device *rdev, - enum radeon_pm_state_type type) +static void radeon_pm_update_profile(struct radeon_device *rdev) { - int i, j; - enum radeon_pm_state_type wanted_types[2]; - int wanted_count; - - switch (type) { - case POWER_STATE_TYPE_DEFAULT: - default: - return rdev->pm.default_power_state; - case POWER_STATE_TYPE_POWERSAVE: - if (rdev->flags & RADEON_IS_MOBILITY) { - wanted_types[0] = POWER_STATE_TYPE_POWERSAVE; - wanted_types[1] = POWER_STATE_TYPE_BATTERY; - wanted_count = 2; - } else { - wanted_types[0] = POWER_STATE_TYPE_PERFORMANCE; - wanted_count = 1; - } + switch (rdev->pm.profile) { + case PM_PROFILE_DEFAULT: + rdev->pm.profile_index = PM_PROFILE_DEFAULT_IDX; break; - case POWER_STATE_TYPE_BATTERY: - if (rdev->flags & RADEON_IS_MOBILITY) { - wanted_types[0] = POWER_STATE_TYPE_BATTERY; - wanted_types[1] = POWER_STATE_TYPE_POWERSAVE; - wanted_count = 2; + case PM_PROFILE_AUTO: + if (power_supply_is_system_supplied() > 0) { + if (rdev->pm.active_crtc_count > 1) + rdev->pm.profile_index = PM_PROFILE_HIGH_MH_IDX; + else + rdev->pm.profile_index = PM_PROFILE_HIGH_SH_IDX; } else { - wanted_types[0] = POWER_STATE_TYPE_PERFORMANCE; - wanted_count = 1; + if (rdev->pm.active_crtc_count > 1) + rdev->pm.profile_index = PM_PROFILE_MID_MH_IDX; + else + rdev->pm.profile_index = PM_PROFILE_MID_SH_IDX; } break; - case POWER_STATE_TYPE_BALANCED: - case POWER_STATE_TYPE_PERFORMANCE: - wanted_types[0] = type; - wanted_count = 1; + case PM_PROFILE_LOW: + if (rdev->pm.active_crtc_count > 1) + rdev->pm.profile_index = PM_PROFILE_LOW_MH_IDX; + else + rdev->pm.profile_index = PM_PROFILE_LOW_SH_IDX; break; - } - - for (i = 0; i < wanted_count; i++) { - for (j = 0; j < rdev->pm.num_power_states; j++) { - if (rdev->pm.power_state[j].type == wanted_types[i]) - return &rdev->pm.power_state[j]; - } - } - - return rdev->pm.default_power_state; -} - -static struct radeon_pm_clock_info * radeon_pick_clock_mode(struct radeon_device *rdev, - struct radeon_power_state *power_state, - enum radeon_pm_clock_mode_type type) -{ - switch (type) { - case POWER_MODE_TYPE_DEFAULT: - default: - return power_state->default_clock_mode; - case POWER_MODE_TYPE_LOW: - return &power_state->clock_info[0]; - case POWER_MODE_TYPE_MID: - if (power_state->num_clock_modes > 2) - return &power_state->clock_info[1]; + case PM_PROFILE_MID: + if (rdev->pm.active_crtc_count > 1) + rdev->pm.profile_index = PM_PROFILE_MID_MH_IDX; else - return &power_state->clock_info[0]; + rdev->pm.profile_index = PM_PROFILE_MID_SH_IDX; + break; + case PM_PROFILE_HIGH: + if (rdev->pm.active_crtc_count > 1) + rdev->pm.profile_index = PM_PROFILE_HIGH_MH_IDX; + else + rdev->pm.profile_index = PM_PROFILE_HIGH_SH_IDX; break; - case POWER_MODE_TYPE_HIGH: - return &power_state->clock_info[power_state->num_clock_modes - 1]; } + if (rdev->pm.active_crtc_count == 0) { + rdev->pm.requested_power_state_index = + rdev->pm.profiles[rdev->pm.profile_index].dpms_off_ps_idx; + rdev->pm.requested_clock_mode_index = + rdev->pm.profiles[rdev->pm.profile_index].dpms_off_cm_idx; + } else { + rdev->pm.requested_power_state_index = + rdev->pm.profiles[rdev->pm.profile_index].dpms_on_ps_idx; + rdev->pm.requested_clock_mode_index = + rdev->pm.profiles[rdev->pm.profile_index].dpms_on_cm_idx; + } } -static void radeon_get_power_state(struct radeon_device *rdev, - enum radeon_pm_action action) +static void radeon_unmap_vram_bos(struct radeon_device *rdev) { - switch (action) { - case PM_ACTION_MINIMUM: - rdev->pm.requested_power_state = radeon_pick_power_state(rdev, POWER_STATE_TYPE_BATTERY); - rdev->pm.requested_clock_mode = - radeon_pick_clock_mode(rdev, rdev->pm.requested_power_state, POWER_MODE_TYPE_LOW); - break; - case PM_ACTION_DOWNCLOCK: - rdev->pm.requested_power_state = radeon_pick_power_state(rdev, POWER_STATE_TYPE_POWERSAVE); - rdev->pm.requested_clock_mode = - radeon_pick_clock_mode(rdev, rdev->pm.requested_power_state, POWER_MODE_TYPE_MID); - break; - case PM_ACTION_UPCLOCK: - rdev->pm.requested_power_state = radeon_pick_power_state(rdev, POWER_STATE_TYPE_DEFAULT); - rdev->pm.requested_clock_mode = - radeon_pick_clock_mode(rdev, rdev->pm.requested_power_state, POWER_MODE_TYPE_HIGH); - break; - case PM_ACTION_NONE: - default: - DRM_ERROR("Requested mode for not defined action\n"); + struct radeon_bo *bo, *n; + + if (list_empty(&rdev->gem.objects)) return; + + list_for_each_entry_safe(bo, n, &rdev->gem.objects, list) { + if (bo->tbo.mem.mem_type == TTM_PL_VRAM) + ttm_bo_unmap_virtual(&bo->tbo); } - DRM_INFO("Requested: e: %d m: %d p: %d\n", - rdev->pm.requested_clock_mode->sclk, - rdev->pm.requested_clock_mode->mclk, - rdev->pm.requested_power_state->non_clock_info.pcie_lanes); } -static inline void radeon_sync_with_vblank(struct radeon_device *rdev) +static void radeon_sync_with_vblank(struct radeon_device *rdev) { if (rdev->pm.active_crtcs) { rdev->pm.vblank_sync = false; @@ -192,73 +155,387 @@ static inline void radeon_sync_with_vblank(struct radeon_device *rdev) static void radeon_set_power_state(struct radeon_device *rdev) { - /* if *_clock_mode are the same, *_power_state are as well */ - if (rdev->pm.requested_clock_mode == rdev->pm.current_clock_mode) - return; + u32 sclk, mclk; + bool misc_after = false; - DRM_INFO("Setting: e: %d m: %d p: %d\n", - rdev->pm.requested_clock_mode->sclk, - rdev->pm.requested_clock_mode->mclk, - rdev->pm.requested_power_state->non_clock_info.pcie_lanes); + if ((rdev->pm.requested_clock_mode_index == rdev->pm.current_clock_mode_index) && + (rdev->pm.requested_power_state_index == rdev->pm.current_power_state_index)) + return; - /* set pcie lanes */ - /* TODO */ + if (radeon_gui_idle(rdev)) { + sclk = rdev->pm.power_state[rdev->pm.requested_power_state_index]. + clock_info[rdev->pm.requested_clock_mode_index].sclk; + if (sclk > rdev->clock.default_sclk) + sclk = rdev->clock.default_sclk; - /* set voltage */ - /* TODO */ + mclk = rdev->pm.power_state[rdev->pm.requested_power_state_index]. + clock_info[rdev->pm.requested_clock_mode_index].mclk; + if (mclk > rdev->clock.default_mclk) + mclk = rdev->clock.default_mclk; - /* set engine clock */ - radeon_sync_with_vblank(rdev); - radeon_pm_debug_check_in_vbl(rdev, false); - radeon_set_engine_clock(rdev, rdev->pm.requested_clock_mode->sclk); - radeon_pm_debug_check_in_vbl(rdev, true); + /* upvolt before raising clocks, downvolt after lowering clocks */ + if (sclk < rdev->pm.current_sclk) + misc_after = true; -#if 0 - /* set memory clock */ - if (rdev->asic->set_memory_clock) { radeon_sync_with_vblank(rdev); - radeon_pm_debug_check_in_vbl(rdev, false); - radeon_set_memory_clock(rdev, rdev->pm.requested_clock_mode->mclk); - radeon_pm_debug_check_in_vbl(rdev, true); + + if (rdev->pm.pm_method == PM_METHOD_DYNPM) { + if (!radeon_pm_in_vbl(rdev)) + return; + } + + radeon_pm_prepare(rdev); + + if (!misc_after) + /* voltage, pcie lanes, etc.*/ + radeon_pm_misc(rdev); + + /* set engine clock */ + if (sclk != rdev->pm.current_sclk) { + radeon_pm_debug_check_in_vbl(rdev, false); + radeon_set_engine_clock(rdev, sclk); + radeon_pm_debug_check_in_vbl(rdev, true); + rdev->pm.current_sclk = sclk; + DRM_DEBUG("Setting: e: %d\n", sclk); + } + + /* set memory clock */ + if (rdev->asic->set_memory_clock && (mclk != rdev->pm.current_mclk)) { + radeon_pm_debug_check_in_vbl(rdev, false); + radeon_set_memory_clock(rdev, mclk); + radeon_pm_debug_check_in_vbl(rdev, true); + rdev->pm.current_mclk = mclk; + DRM_DEBUG("Setting: m: %d\n", mclk); + } + + if (misc_after) + /* voltage, pcie lanes, etc.*/ + radeon_pm_misc(rdev); + + radeon_pm_finish(rdev); + + rdev->pm.current_power_state_index = rdev->pm.requested_power_state_index; + rdev->pm.current_clock_mode_index = rdev->pm.requested_clock_mode_index; + } else + DRM_DEBUG("pm: GUI not idle!!!\n"); +} + +static void radeon_pm_set_clocks(struct radeon_device *rdev) +{ + int i; + + mutex_lock(&rdev->ddev->struct_mutex); + mutex_lock(&rdev->vram_mutex); + mutex_lock(&rdev->cp.mutex); + + /* gui idle int has issues on older chips it seems */ + if (rdev->family >= CHIP_R600) { + if (rdev->irq.installed) { + /* wait for GPU idle */ + rdev->pm.gui_idle = false; + rdev->irq.gui_idle = true; + radeon_irq_set(rdev); + wait_event_interruptible_timeout( + rdev->irq.idle_queue, rdev->pm.gui_idle, + msecs_to_jiffies(RADEON_WAIT_IDLE_TIMEOUT)); + rdev->irq.gui_idle = false; + radeon_irq_set(rdev); + } + } else { + if (rdev->cp.ready) { + struct radeon_fence *fence; + radeon_ring_alloc(rdev, 64); + radeon_fence_create(rdev, &fence); + radeon_fence_emit(rdev, fence); + radeon_ring_commit(rdev); + radeon_fence_wait(fence, false); + radeon_fence_unref(&fence); + } } -#endif + radeon_unmap_vram_bos(rdev); + + if (rdev->irq.installed) { + for (i = 0; i < rdev->num_crtc; i++) { + if (rdev->pm.active_crtcs & (1 << i)) { + rdev->pm.req_vblank |= (1 << i); + drm_vblank_get(rdev->ddev, i); + } + } + } + + radeon_set_power_state(rdev); + + if (rdev->irq.installed) { + for (i = 0; i < rdev->num_crtc; i++) { + if (rdev->pm.req_vblank & (1 << i)) { + rdev->pm.req_vblank &= ~(1 << i); + drm_vblank_put(rdev->ddev, i); + } + } + } + + /* update display watermarks based on new power state */ + radeon_update_bandwidth_info(rdev); + if (rdev->pm.active_crtc_count) + radeon_bandwidth_update(rdev); + + rdev->pm.dynpm_planned_action = DYNPM_ACTION_NONE; + + mutex_unlock(&rdev->cp.mutex); + mutex_unlock(&rdev->vram_mutex); + mutex_unlock(&rdev->ddev->struct_mutex); +} + +static void radeon_pm_print_states(struct radeon_device *rdev) +{ + int i, j; + struct radeon_power_state *power_state; + struct radeon_pm_clock_info *clock_info; + + DRM_DEBUG("%d Power State(s)\n", rdev->pm.num_power_states); + for (i = 0; i < rdev->pm.num_power_states; i++) { + power_state = &rdev->pm.power_state[i]; + DRM_DEBUG("State %d: %s\n", i, + radeon_pm_state_type_name[power_state->type]); + if (i == rdev->pm.default_power_state_index) + DRM_DEBUG("\tDefault"); + if ((rdev->flags & RADEON_IS_PCIE) && !(rdev->flags & RADEON_IS_IGP)) + DRM_DEBUG("\t%d PCIE Lanes\n", power_state->pcie_lanes); + if (power_state->flags & RADEON_PM_STATE_SINGLE_DISPLAY_ONLY) + DRM_DEBUG("\tSingle display only\n"); + DRM_DEBUG("\t%d Clock Mode(s)\n", power_state->num_clock_modes); + for (j = 0; j < power_state->num_clock_modes; j++) { + clock_info = &(power_state->clock_info[j]); + if (rdev->flags & RADEON_IS_IGP) + DRM_DEBUG("\t\t%d e: %d%s\n", + j, + clock_info->sclk * 10, + clock_info->flags & RADEON_PM_MODE_NO_DISPLAY ? "\tNo display only" : ""); + else + DRM_DEBUG("\t\t%d e: %d\tm: %d\tv: %d%s\n", + j, + clock_info->sclk * 10, + clock_info->mclk * 10, + clock_info->voltage.voltage, + clock_info->flags & RADEON_PM_MODE_NO_DISPLAY ? "\tNo display only" : ""); + } + } +} - rdev->pm.current_power_state = rdev->pm.requested_power_state; - rdev->pm.current_clock_mode = rdev->pm.requested_clock_mode; +static ssize_t radeon_get_pm_profile(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = pci_get_drvdata(to_pci_dev(dev)); + struct radeon_device *rdev = ddev->dev_private; + int cp = rdev->pm.profile; + + return snprintf(buf, PAGE_SIZE, "%s\n", + (cp == PM_PROFILE_AUTO) ? "auto" : + (cp == PM_PROFILE_LOW) ? "low" : + (cp == PM_PROFILE_MID) ? "mid" : + (cp == PM_PROFILE_HIGH) ? "high" : "default"); +} + +static ssize_t radeon_set_pm_profile(struct device *dev, + struct device_attribute *attr, + const char *buf, + size_t count) +{ + struct drm_device *ddev = pci_get_drvdata(to_pci_dev(dev)); + struct radeon_device *rdev = ddev->dev_private; + + mutex_lock(&rdev->pm.mutex); + if (rdev->pm.pm_method == PM_METHOD_PROFILE) { + if (strncmp("default", buf, strlen("default")) == 0) + rdev->pm.profile = PM_PROFILE_DEFAULT; + else if (strncmp("auto", buf, strlen("auto")) == 0) + rdev->pm.profile = PM_PROFILE_AUTO; + else if (strncmp("low", buf, strlen("low")) == 0) + rdev->pm.profile = PM_PROFILE_LOW; + else if (strncmp("mid", buf, strlen("mid")) == 0) + rdev->pm.profile = PM_PROFILE_MID; + else if (strncmp("high", buf, strlen("high")) == 0) + rdev->pm.profile = PM_PROFILE_HIGH; + else { + DRM_ERROR("invalid power profile!\n"); + goto fail; + } + radeon_pm_update_profile(rdev); + radeon_pm_set_clocks(rdev); + } +fail: + mutex_unlock(&rdev->pm.mutex); + + return count; +} + +static ssize_t radeon_get_pm_method(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = pci_get_drvdata(to_pci_dev(dev)); + struct radeon_device *rdev = ddev->dev_private; + int pm = rdev->pm.pm_method; + + return snprintf(buf, PAGE_SIZE, "%s\n", + (pm == PM_METHOD_DYNPM) ? "dynpm" : "profile"); +} + +static ssize_t radeon_set_pm_method(struct device *dev, + struct device_attribute *attr, + const char *buf, + size_t count) +{ + struct drm_device *ddev = pci_get_drvdata(to_pci_dev(dev)); + struct radeon_device *rdev = ddev->dev_private; + + + if (strncmp("dynpm", buf, strlen("dynpm")) == 0) { + mutex_lock(&rdev->pm.mutex); + rdev->pm.pm_method = PM_METHOD_DYNPM; + rdev->pm.dynpm_state = DYNPM_STATE_PAUSED; + rdev->pm.dynpm_planned_action = DYNPM_ACTION_DEFAULT; + mutex_unlock(&rdev->pm.mutex); + } else if (strncmp("profile", buf, strlen("profile")) == 0) { + bool flush_wq = false; + + mutex_lock(&rdev->pm.mutex); + if (rdev->pm.pm_method == PM_METHOD_DYNPM) { + cancel_delayed_work(&rdev->pm.dynpm_idle_work); + flush_wq = true; + } + /* disable dynpm */ + rdev->pm.dynpm_state = DYNPM_STATE_DISABLED; + rdev->pm.dynpm_planned_action = DYNPM_ACTION_NONE; + rdev->pm.pm_method = PM_METHOD_PROFILE; + mutex_unlock(&rdev->pm.mutex); + if (flush_wq) + flush_workqueue(rdev->wq); + } else { + DRM_ERROR("invalid power method!\n"); + goto fail; + } + radeon_pm_compute_clocks(rdev); +fail: + return count; +} + +static DEVICE_ATTR(power_profile, S_IRUGO | S_IWUSR, radeon_get_pm_profile, radeon_set_pm_profile); +static DEVICE_ATTR(power_method, S_IRUGO | S_IWUSR, radeon_get_pm_method, radeon_set_pm_method); + +void radeon_pm_suspend(struct radeon_device *rdev) +{ + bool flush_wq = false; + + mutex_lock(&rdev->pm.mutex); + if (rdev->pm.pm_method == PM_METHOD_DYNPM) { + cancel_delayed_work(&rdev->pm.dynpm_idle_work); + if (rdev->pm.dynpm_state == DYNPM_STATE_ACTIVE) + rdev->pm.dynpm_state = DYNPM_STATE_SUSPENDED; + flush_wq = true; + } + mutex_unlock(&rdev->pm.mutex); + if (flush_wq) + flush_workqueue(rdev->wq); +} + +void radeon_pm_resume(struct radeon_device *rdev) +{ + /* asic init will reset the default power state */ + mutex_lock(&rdev->pm.mutex); + rdev->pm.current_power_state_index = rdev->pm.default_power_state_index; + rdev->pm.current_clock_mode_index = 0; + rdev->pm.current_sclk = rdev->clock.default_sclk; + rdev->pm.current_mclk = rdev->clock.default_mclk; + rdev->pm.current_vddc = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.voltage; + if (rdev->pm.pm_method == PM_METHOD_DYNPM + && rdev->pm.dynpm_state == DYNPM_STATE_SUSPENDED) { + rdev->pm.dynpm_state = DYNPM_STATE_ACTIVE; + queue_delayed_work(rdev->wq, &rdev->pm.dynpm_idle_work, + msecs_to_jiffies(RADEON_IDLE_LOOP_MS)); + } + mutex_unlock(&rdev->pm.mutex); + radeon_pm_compute_clocks(rdev); } int radeon_pm_init(struct radeon_device *rdev) { - rdev->pm.state = PM_STATE_DISABLED; - rdev->pm.planned_action = PM_ACTION_NONE; - rdev->pm.downclocked = false; + int ret; + /* default to profile method */ + rdev->pm.pm_method = PM_METHOD_PROFILE; + rdev->pm.profile = PM_PROFILE_DEFAULT; + rdev->pm.dynpm_state = DYNPM_STATE_DISABLED; + rdev->pm.dynpm_planned_action = DYNPM_ACTION_NONE; + rdev->pm.dynpm_can_upclock = true; + rdev->pm.dynpm_can_downclock = true; + rdev->pm.current_sclk = rdev->clock.default_sclk; + rdev->pm.current_mclk = rdev->clock.default_mclk; if (rdev->bios) { if (rdev->is_atom_bios) radeon_atombios_get_power_modes(rdev); else radeon_combios_get_power_modes(rdev); - radeon_print_power_mode_info(rdev); + radeon_pm_print_states(rdev); + radeon_pm_init_profile(rdev); } - if (radeon_debugfs_pm_init(rdev)) { - DRM_ERROR("Failed to register debugfs file for PM!\n"); - } + if (rdev->pm.num_power_states > 1) { + /* where's the best place to put these? */ + ret = device_create_file(rdev->dev, &dev_attr_power_profile); + if (ret) + DRM_ERROR("failed to create device file for power profile\n"); + ret = device_create_file(rdev->dev, &dev_attr_power_method); + if (ret) + DRM_ERROR("failed to create device file for power method\n"); + +#ifdef CONFIG_ACPI + rdev->acpi_nb.notifier_call = radeon_acpi_event; + register_acpi_notifier(&rdev->acpi_nb); +#endif + INIT_DELAYED_WORK(&rdev->pm.dynpm_idle_work, radeon_dynpm_idle_work_handler); - INIT_DELAYED_WORK(&rdev->pm.idle_work, radeon_pm_idle_work_handler); + if (radeon_debugfs_pm_init(rdev)) { + DRM_ERROR("Failed to register debugfs file for PM!\n"); + } - if (radeon_dynpm != -1 && radeon_dynpm) { - rdev->pm.state = PM_STATE_PAUSED; - DRM_INFO("radeon: dynamic power management enabled\n"); + DRM_INFO("radeon: power management initialized\n"); } - DRM_INFO("radeon: power management initialized\n"); - return 0; } void radeon_pm_fini(struct radeon_device *rdev) { + if (rdev->pm.num_power_states > 1) { + bool flush_wq = false; + + mutex_lock(&rdev->pm.mutex); + if (rdev->pm.pm_method == PM_METHOD_PROFILE) { + rdev->pm.profile = PM_PROFILE_DEFAULT; + radeon_pm_update_profile(rdev); + radeon_pm_set_clocks(rdev); + } else if (rdev->pm.pm_method == PM_METHOD_DYNPM) { + /* cancel work */ + cancel_delayed_work(&rdev->pm.dynpm_idle_work); + flush_wq = true; + /* reset default clocks */ + rdev->pm.dynpm_state = DYNPM_STATE_DISABLED; + rdev->pm.dynpm_planned_action = DYNPM_ACTION_DEFAULT; + radeon_pm_set_clocks(rdev); + } + mutex_unlock(&rdev->pm.mutex); + if (flush_wq) + flush_workqueue(rdev->wq); + + device_remove_file(rdev->dev, &dev_attr_power_profile); + device_remove_file(rdev->dev, &dev_attr_power_method); +#ifdef CONFIG_ACPI + unregister_acpi_notifier(&rdev->acpi_nb); +#endif + } + if (rdev->pm.i2c_bus) radeon_i2c_destroy(rdev->pm.i2c_bus); } @@ -266,146 +543,167 @@ void radeon_pm_fini(struct radeon_device *rdev) void radeon_pm_compute_clocks(struct radeon_device *rdev) { struct drm_device *ddev = rdev->ddev; - struct drm_connector *connector; + struct drm_crtc *crtc; struct radeon_crtc *radeon_crtc; - int count = 0; - if (rdev->pm.state == PM_STATE_DISABLED) + if (rdev->pm.num_power_states < 2) return; mutex_lock(&rdev->pm.mutex); rdev->pm.active_crtcs = 0; - list_for_each_entry(connector, - &ddev->mode_config.connector_list, head) { - if (connector->encoder && - connector->encoder->crtc && - connector->dpms != DRM_MODE_DPMS_OFF) { - radeon_crtc = to_radeon_crtc(connector->encoder->crtc); + rdev->pm.active_crtc_count = 0; + list_for_each_entry(crtc, + &ddev->mode_config.crtc_list, head) { + radeon_crtc = to_radeon_crtc(crtc); + if (radeon_crtc->enabled) { rdev->pm.active_crtcs |= (1 << radeon_crtc->crtc_id); - ++count; + rdev->pm.active_crtc_count++; } } - if (count > 1) { - if (rdev->pm.state == PM_STATE_ACTIVE) { - cancel_delayed_work(&rdev->pm.idle_work); - - rdev->pm.state = PM_STATE_PAUSED; - rdev->pm.planned_action = PM_ACTION_UPCLOCK; - if (rdev->pm.downclocked) - radeon_pm_set_clocks(rdev); - - DRM_DEBUG("radeon: dynamic power management deactivated\n"); - } - } else if (count == 1) { - /* TODO: Increase clocks if needed for current mode */ - - if (rdev->pm.state == PM_STATE_MINIMUM) { - rdev->pm.state = PM_STATE_ACTIVE; - rdev->pm.planned_action = PM_ACTION_UPCLOCK; - radeon_pm_set_clocks(rdev); - - queue_delayed_work(rdev->wq, &rdev->pm.idle_work, - msecs_to_jiffies(RADEON_IDLE_LOOP_MS)); - } - else if (rdev->pm.state == PM_STATE_PAUSED) { - rdev->pm.state = PM_STATE_ACTIVE; - queue_delayed_work(rdev->wq, &rdev->pm.idle_work, - msecs_to_jiffies(RADEON_IDLE_LOOP_MS)); - DRM_DEBUG("radeon: dynamic power management activated\n"); - } - } - else { /* count == 0 */ - if (rdev->pm.state != PM_STATE_MINIMUM) { - cancel_delayed_work(&rdev->pm.idle_work); - - rdev->pm.state = PM_STATE_MINIMUM; - rdev->pm.planned_action = PM_ACTION_MINIMUM; - radeon_pm_set_clocks(rdev); + if (rdev->pm.pm_method == PM_METHOD_PROFILE) { + radeon_pm_update_profile(rdev); + radeon_pm_set_clocks(rdev); + } else if (rdev->pm.pm_method == PM_METHOD_DYNPM) { + if (rdev->pm.dynpm_state != DYNPM_STATE_DISABLED) { + if (rdev->pm.active_crtc_count > 1) { + if (rdev->pm.dynpm_state == DYNPM_STATE_ACTIVE) { + cancel_delayed_work(&rdev->pm.dynpm_idle_work); + + rdev->pm.dynpm_state = DYNPM_STATE_PAUSED; + rdev->pm.dynpm_planned_action = DYNPM_ACTION_DEFAULT; + radeon_pm_get_dynpm_state(rdev); + radeon_pm_set_clocks(rdev); + + DRM_DEBUG("radeon: dynamic power management deactivated\n"); + } + } else if (rdev->pm.active_crtc_count == 1) { + /* TODO: Increase clocks if needed for current mode */ + + if (rdev->pm.dynpm_state == DYNPM_STATE_MINIMUM) { + rdev->pm.dynpm_state = DYNPM_STATE_ACTIVE; + rdev->pm.dynpm_planned_action = DYNPM_ACTION_UPCLOCK; + radeon_pm_get_dynpm_state(rdev); + radeon_pm_set_clocks(rdev); + + queue_delayed_work(rdev->wq, &rdev->pm.dynpm_idle_work, + msecs_to_jiffies(RADEON_IDLE_LOOP_MS)); + } else if (rdev->pm.dynpm_state == DYNPM_STATE_PAUSED) { + rdev->pm.dynpm_state = DYNPM_STATE_ACTIVE; + queue_delayed_work(rdev->wq, &rdev->pm.dynpm_idle_work, + msecs_to_jiffies(RADEON_IDLE_LOOP_MS)); + DRM_DEBUG("radeon: dynamic power management activated\n"); + } + } else { /* count == 0 */ + if (rdev->pm.dynpm_state != DYNPM_STATE_MINIMUM) { + cancel_delayed_work(&rdev->pm.dynpm_idle_work); + + rdev->pm.dynpm_state = DYNPM_STATE_MINIMUM; + rdev->pm.dynpm_planned_action = DYNPM_ACTION_MINIMUM; + radeon_pm_get_dynpm_state(rdev); + radeon_pm_set_clocks(rdev); + } + } } } mutex_unlock(&rdev->pm.mutex); } -static bool radeon_pm_debug_check_in_vbl(struct radeon_device *rdev, bool finish) +static bool radeon_pm_in_vbl(struct radeon_device *rdev) { - u32 stat_crtc1 = 0, stat_crtc2 = 0; + u32 stat_crtc = 0, vbl = 0, position = 0; bool in_vbl = true; - if (ASIC_IS_AVIVO(rdev)) { + if (ASIC_IS_DCE4(rdev)) { + if (rdev->pm.active_crtcs & (1 << 0)) { + vbl = RREG32(EVERGREEN_CRTC_V_BLANK_START_END + + EVERGREEN_CRTC0_REGISTER_OFFSET) & 0xfff; + position = RREG32(EVERGREEN_CRTC_STATUS_POSITION + + EVERGREEN_CRTC0_REGISTER_OFFSET) & 0xfff; + } + if (rdev->pm.active_crtcs & (1 << 1)) { + vbl = RREG32(EVERGREEN_CRTC_V_BLANK_START_END + + EVERGREEN_CRTC1_REGISTER_OFFSET) & 0xfff; + position = RREG32(EVERGREEN_CRTC_STATUS_POSITION + + EVERGREEN_CRTC1_REGISTER_OFFSET) & 0xfff; + } + if (rdev->pm.active_crtcs & (1 << 2)) { + vbl = RREG32(EVERGREEN_CRTC_V_BLANK_START_END + + EVERGREEN_CRTC2_REGISTER_OFFSET) & 0xfff; + position = RREG32(EVERGREEN_CRTC_STATUS_POSITION + + EVERGREEN_CRTC2_REGISTER_OFFSET) & 0xfff; + } + if (rdev->pm.active_crtcs & (1 << 3)) { + vbl = RREG32(EVERGREEN_CRTC_V_BLANK_START_END + + EVERGREEN_CRTC3_REGISTER_OFFSET) & 0xfff; + position = RREG32(EVERGREEN_CRTC_STATUS_POSITION + + EVERGREEN_CRTC3_REGISTER_OFFSET) & 0xfff; + } + if (rdev->pm.active_crtcs & (1 << 4)) { + vbl = RREG32(EVERGREEN_CRTC_V_BLANK_START_END + + EVERGREEN_CRTC4_REGISTER_OFFSET) & 0xfff; + position = RREG32(EVERGREEN_CRTC_STATUS_POSITION + + EVERGREEN_CRTC4_REGISTER_OFFSET) & 0xfff; + } + if (rdev->pm.active_crtcs & (1 << 5)) { + vbl = RREG32(EVERGREEN_CRTC_V_BLANK_START_END + + EVERGREEN_CRTC5_REGISTER_OFFSET) & 0xfff; + position = RREG32(EVERGREEN_CRTC_STATUS_POSITION + + EVERGREEN_CRTC5_REGISTER_OFFSET) & 0xfff; + } + } else if (ASIC_IS_AVIVO(rdev)) { + if (rdev->pm.active_crtcs & (1 << 0)) { + vbl = RREG32(AVIVO_D1CRTC_V_BLANK_START_END) & 0xfff; + position = RREG32(AVIVO_D1CRTC_STATUS_POSITION) & 0xfff; + } + if (rdev->pm.active_crtcs & (1 << 1)) { + vbl = RREG32(AVIVO_D2CRTC_V_BLANK_START_END) & 0xfff; + position = RREG32(AVIVO_D2CRTC_STATUS_POSITION) & 0xfff; + } + if (position < vbl && position > 1) + in_vbl = false; + } else { if (rdev->pm.active_crtcs & (1 << 0)) { - stat_crtc1 = RREG32(D1CRTC_STATUS); - if (!(stat_crtc1 & 1)) + stat_crtc = RREG32(RADEON_CRTC_STATUS); + if (!(stat_crtc & 1)) in_vbl = false; } if (rdev->pm.active_crtcs & (1 << 1)) { - stat_crtc2 = RREG32(D2CRTC_STATUS); - if (!(stat_crtc2 & 1)) + stat_crtc = RREG32(RADEON_CRTC2_STATUS); + if (!(stat_crtc & 1)) in_vbl = false; } } - if (in_vbl == false) - DRM_INFO("not in vbl for pm change %08x %08x at %s\n", stat_crtc1, - stat_crtc2, finish ? "exit" : "entry"); - return in_vbl; -} -static void radeon_pm_set_clocks_locked(struct radeon_device *rdev) -{ - /*radeon_fence_wait_last(rdev);*/ - switch (rdev->pm.planned_action) { - case PM_ACTION_UPCLOCK: - rdev->pm.downclocked = false; - break; - case PM_ACTION_DOWNCLOCK: - rdev->pm.downclocked = true; - break; - case PM_ACTION_MINIMUM: - break; - case PM_ACTION_NONE: - DRM_ERROR("%s: PM_ACTION_NONE\n", __func__); - break; - } - radeon_set_power_state(rdev); - rdev->pm.planned_action = PM_ACTION_NONE; + if (position < vbl && position > 1) + in_vbl = false; + + return in_vbl; } -static void radeon_pm_set_clocks(struct radeon_device *rdev) +static bool radeon_pm_debug_check_in_vbl(struct radeon_device *rdev, bool finish) { - radeon_get_power_state(rdev, rdev->pm.planned_action); - mutex_lock(&rdev->cp.mutex); + u32 stat_crtc = 0; + bool in_vbl = radeon_pm_in_vbl(rdev); - if (rdev->pm.active_crtcs & (1 << 0)) { - rdev->pm.req_vblank |= (1 << 0); - drm_vblank_get(rdev->ddev, 0); - } - if (rdev->pm.active_crtcs & (1 << 1)) { - rdev->pm.req_vblank |= (1 << 1); - drm_vblank_get(rdev->ddev, 1); - } - radeon_pm_set_clocks_locked(rdev); - if (rdev->pm.req_vblank & (1 << 0)) { - rdev->pm.req_vblank &= ~(1 << 0); - drm_vblank_put(rdev->ddev, 0); - } - if (rdev->pm.req_vblank & (1 << 1)) { - rdev->pm.req_vblank &= ~(1 << 1); - drm_vblank_put(rdev->ddev, 1); - } - - mutex_unlock(&rdev->cp.mutex); + if (in_vbl == false) + DRM_DEBUG("not in vbl for pm change %08x at %s\n", stat_crtc, + finish ? "exit" : "entry"); + return in_vbl; } -static void radeon_pm_idle_work_handler(struct work_struct *work) +static void radeon_dynpm_idle_work_handler(struct work_struct *work) { struct radeon_device *rdev; + int resched; rdev = container_of(work, struct radeon_device, - pm.idle_work.work); + pm.dynpm_idle_work.work); + resched = ttm_bo_lock_delayed_workqueue(&rdev->mman.bdev); mutex_lock(&rdev->pm.mutex); - if (rdev->pm.state == PM_STATE_ACTIVE) { + if (rdev->pm.dynpm_state == DYNPM_STATE_ACTIVE) { unsigned long irq_flags; int not_processed = 0; @@ -421,36 +719,41 @@ static void radeon_pm_idle_work_handler(struct work_struct *work) read_unlock_irqrestore(&rdev->fence_drv.lock, irq_flags); if (not_processed >= 3) { /* should upclock */ - if (rdev->pm.planned_action == PM_ACTION_DOWNCLOCK) { - rdev->pm.planned_action = PM_ACTION_NONE; - } else if (rdev->pm.planned_action == PM_ACTION_NONE && - rdev->pm.downclocked) { - rdev->pm.planned_action = - PM_ACTION_UPCLOCK; - rdev->pm.action_timeout = jiffies + + if (rdev->pm.dynpm_planned_action == DYNPM_ACTION_DOWNCLOCK) { + rdev->pm.dynpm_planned_action = DYNPM_ACTION_NONE; + } else if (rdev->pm.dynpm_planned_action == DYNPM_ACTION_NONE && + rdev->pm.dynpm_can_upclock) { + rdev->pm.dynpm_planned_action = + DYNPM_ACTION_UPCLOCK; + rdev->pm.dynpm_action_timeout = jiffies + msecs_to_jiffies(RADEON_RECLOCK_DELAY_MS); } } else if (not_processed == 0) { /* should downclock */ - if (rdev->pm.planned_action == PM_ACTION_UPCLOCK) { - rdev->pm.planned_action = PM_ACTION_NONE; - } else if (rdev->pm.planned_action == PM_ACTION_NONE && - !rdev->pm.downclocked) { - rdev->pm.planned_action = - PM_ACTION_DOWNCLOCK; - rdev->pm.action_timeout = jiffies + + if (rdev->pm.dynpm_planned_action == DYNPM_ACTION_UPCLOCK) { + rdev->pm.dynpm_planned_action = DYNPM_ACTION_NONE; + } else if (rdev->pm.dynpm_planned_action == DYNPM_ACTION_NONE && + rdev->pm.dynpm_can_downclock) { + rdev->pm.dynpm_planned_action = + DYNPM_ACTION_DOWNCLOCK; + rdev->pm.dynpm_action_timeout = jiffies + msecs_to_jiffies(RADEON_RECLOCK_DELAY_MS); } } - if (rdev->pm.planned_action != PM_ACTION_NONE && - jiffies > rdev->pm.action_timeout) { + /* Note, radeon_pm_set_clocks is called with static_switch set + * to false since we want to wait for vbl to avoid flicker. + */ + if (rdev->pm.dynpm_planned_action != DYNPM_ACTION_NONE && + jiffies > rdev->pm.dynpm_action_timeout) { + radeon_pm_get_dynpm_state(rdev); radeon_pm_set_clocks(rdev); } - } - mutex_unlock(&rdev->pm.mutex); - queue_delayed_work(rdev->wq, &rdev->pm.idle_work, + queue_delayed_work(rdev->wq, &rdev->pm.dynpm_idle_work, msecs_to_jiffies(RADEON_IDLE_LOOP_MS)); + } + mutex_unlock(&rdev->pm.mutex); + ttm_bo_unlock_delayed_workqueue(&rdev->mman.bdev, resched); } /* @@ -464,12 +767,13 @@ static int radeon_debugfs_pm_info(struct seq_file *m, void *data) struct drm_device *dev = node->minor->dev; struct radeon_device *rdev = dev->dev_private; - seq_printf(m, "state: %s\n", pm_state_names[rdev->pm.state]); seq_printf(m, "default engine clock: %u0 kHz\n", rdev->clock.default_sclk); seq_printf(m, "current engine clock: %u0 kHz\n", radeon_get_engine_clock(rdev)); seq_printf(m, "default memory clock: %u0 kHz\n", rdev->clock.default_mclk); if (rdev->asic->get_memory_clock) seq_printf(m, "current memory clock: %u0 kHz\n", radeon_get_memory_clock(rdev)); + if (rdev->pm.current_vddc) + seq_printf(m, "voltage: %u mV\n", rdev->pm.current_vddc); if (rdev->asic->get_pcie_lanes) seq_printf(m, "PCIE lanes: %d\n", radeon_get_pcie_lanes(rdev)); diff --git a/drivers/gpu/drm/radeon/radeon_reg.h b/drivers/gpu/drm/radeon/radeon_reg.h index eabbc9cf30a7..c332f46340d5 100644 --- a/drivers/gpu/drm/radeon/radeon_reg.h +++ b/drivers/gpu/drm/radeon/radeon_reg.h @@ -553,7 +553,6 @@ # define RADEON_CRTC_CRNT_VLINE_MASK (0x7ff << 16) #define RADEON_CRTC2_CRNT_FRAME 0x0314 #define RADEON_CRTC2_GUI_TRIG_VLINE 0x0318 -#define RADEON_CRTC2_STATUS 0x03fc #define RADEON_CRTC2_VLINE_CRNT_VLINE 0x0310 #define RADEON_CRTC8_DATA 0x03d5 /* VGA, 0x3b5 */ #define RADEON_CRTC8_IDX 0x03d4 /* VGA, 0x3b4 */ @@ -995,6 +994,7 @@ # define RADEON_FP_DETECT_MASK (1 << 4) # define RADEON_CRTC2_VBLANK_MASK (1 << 9) # define RADEON_FP2_DETECT_MASK (1 << 10) +# define RADEON_GUI_IDLE_MASK (1 << 19) # define RADEON_SW_INT_ENABLE (1 << 25) #define RADEON_GEN_INT_STATUS 0x0044 # define AVIVO_DISPLAY_INT_STATUS (1 << 0) @@ -1006,6 +1006,8 @@ # define RADEON_CRTC2_VBLANK_STAT_ACK (1 << 9) # define RADEON_FP2_DETECT_STAT (1 << 10) # define RADEON_FP2_DETECT_STAT_ACK (1 << 10) +# define RADEON_GUI_IDLE_STAT (1 << 19) +# define RADEON_GUI_IDLE_STAT_ACK (1 << 19) # define RADEON_SW_INT_FIRE (1 << 26) # define RADEON_SW_INT_TEST (1 << 25) # define RADEON_SW_INT_TEST_ACK (1 << 25) diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c index f6e1e8d4d986..261e98a276db 100644 --- a/drivers/gpu/drm/radeon/radeon_ring.c +++ b/drivers/gpu/drm/radeon/radeon_ring.c @@ -219,24 +219,26 @@ int radeon_ib_pool_init(struct radeon_device *rdev) void radeon_ib_pool_fini(struct radeon_device *rdev) { int r; + struct radeon_bo *robj; if (!rdev->ib_pool.ready) { return; } mutex_lock(&rdev->ib_pool.mutex); radeon_ib_bogus_cleanup(rdev); + robj = rdev->ib_pool.robj; + rdev->ib_pool.robj = NULL; + mutex_unlock(&rdev->ib_pool.mutex); - if (rdev->ib_pool.robj) { - r = radeon_bo_reserve(rdev->ib_pool.robj, false); + if (robj) { + r = radeon_bo_reserve(robj, false); if (likely(r == 0)) { - radeon_bo_kunmap(rdev->ib_pool.robj); - radeon_bo_unpin(rdev->ib_pool.robj); - radeon_bo_unreserve(rdev->ib_pool.robj); + radeon_bo_kunmap(robj); + radeon_bo_unpin(robj); + radeon_bo_unreserve(robj); } - radeon_bo_unref(&rdev->ib_pool.robj); - rdev->ib_pool.robj = NULL; + radeon_bo_unref(&robj); } - mutex_unlock(&rdev->ib_pool.mutex); } @@ -258,31 +260,41 @@ void radeon_ring_free_size(struct radeon_device *rdev) } } -int radeon_ring_lock(struct radeon_device *rdev, unsigned ndw) +int radeon_ring_alloc(struct radeon_device *rdev, unsigned ndw) { int r; /* Align requested size with padding so unlock_commit can * pad safely */ ndw = (ndw + rdev->cp.align_mask) & ~rdev->cp.align_mask; - mutex_lock(&rdev->cp.mutex); while (ndw > (rdev->cp.ring_free_dw - 1)) { radeon_ring_free_size(rdev); if (ndw < rdev->cp.ring_free_dw) { break; } r = radeon_fence_wait_next(rdev); - if (r) { - mutex_unlock(&rdev->cp.mutex); + if (r) return r; - } } rdev->cp.count_dw = ndw; rdev->cp.wptr_old = rdev->cp.wptr; return 0; } -void radeon_ring_unlock_commit(struct radeon_device *rdev) +int radeon_ring_lock(struct radeon_device *rdev, unsigned ndw) +{ + int r; + + mutex_lock(&rdev->cp.mutex); + r = radeon_ring_alloc(rdev, ndw); + if (r) { + mutex_unlock(&rdev->cp.mutex); + return r; + } + return 0; +} + +void radeon_ring_commit(struct radeon_device *rdev) { unsigned count_dw_pad; unsigned i; @@ -295,6 +307,11 @@ void radeon_ring_unlock_commit(struct radeon_device *rdev) } DRM_MEMORYBARRIER(); radeon_cp_commit(rdev); +} + +void radeon_ring_unlock_commit(struct radeon_device *rdev) +{ + radeon_ring_commit(rdev); mutex_unlock(&rdev->cp.mutex); } @@ -344,20 +361,23 @@ int radeon_ring_init(struct radeon_device *rdev, unsigned ring_size) void radeon_ring_fini(struct radeon_device *rdev) { int r; + struct radeon_bo *ring_obj; mutex_lock(&rdev->cp.mutex); - if (rdev->cp.ring_obj) { - r = radeon_bo_reserve(rdev->cp.ring_obj, false); + ring_obj = rdev->cp.ring_obj; + rdev->cp.ring = NULL; + rdev->cp.ring_obj = NULL; + mutex_unlock(&rdev->cp.mutex); + + if (ring_obj) { + r = radeon_bo_reserve(ring_obj, false); if (likely(r == 0)) { - radeon_bo_kunmap(rdev->cp.ring_obj); - radeon_bo_unpin(rdev->cp.ring_obj); - radeon_bo_unreserve(rdev->cp.ring_obj); + radeon_bo_kunmap(ring_obj); + radeon_bo_unpin(ring_obj); + radeon_bo_unreserve(ring_obj); } - radeon_bo_unref(&rdev->cp.ring_obj); - rdev->cp.ring = NULL; - rdev->cp.ring_obj = NULL; + radeon_bo_unref(&ring_obj); } - mutex_unlock(&rdev->cp.mutex); } diff --git a/drivers/gpu/drm/radeon/radeon_state.c b/drivers/gpu/drm/radeon/radeon_state.c index cc5316dcf580..b3ba44c0a818 100644 --- a/drivers/gpu/drm/radeon/radeon_state.c +++ b/drivers/gpu/drm/radeon/radeon_state.c @@ -900,9 +900,10 @@ static void radeon_cp_dispatch_clear(struct drm_device * dev, flags |= RADEON_FRONT; } if (flags & (RADEON_DEPTH|RADEON_STENCIL)) { - if (!dev_priv->have_z_offset) + if (!dev_priv->have_z_offset) { printk_once(KERN_ERR "radeon: illegal depth clear request. Buggy mesa detected - please update.\n"); - flags &= ~(RADEON_DEPTH | RADEON_STENCIL); + flags &= ~(RADEON_DEPTH | RADEON_STENCIL); + } } if (flags & (RADEON_FRONT | RADEON_BACK)) { diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index d031b6863082..e9918d88f5b0 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -33,6 +33,7 @@ #include <ttm/ttm_bo_driver.h> #include <ttm/ttm_placement.h> #include <ttm/ttm_module.h> +#include <ttm/ttm_page_alloc.h> #include <drm/drmP.h> #include <drm/radeon_drm.h> #include <linux/seq_file.h> @@ -162,34 +163,21 @@ static int radeon_init_mem_type(struct ttm_bo_device *bdev, uint32_t type, (unsigned)type); return -EINVAL; } - man->io_offset = rdev->mc.agp_base; - man->io_size = rdev->mc.gtt_size; - man->io_addr = NULL; if (!rdev->ddev->agp->cant_use_aperture) - man->flags = TTM_MEMTYPE_FLAG_NEEDS_IOREMAP | - TTM_MEMTYPE_FLAG_MAPPABLE; + man->flags = TTM_MEMTYPE_FLAG_MAPPABLE; man->available_caching = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_WC; man->default_caching = TTM_PL_FLAG_WC; - } else -#endif - { - man->io_offset = 0; - man->io_size = 0; - man->io_addr = NULL; } +#endif break; case TTM_PL_VRAM: /* "On-card" video ram */ man->gpu_offset = rdev->mc.vram_start; man->flags = TTM_MEMTYPE_FLAG_FIXED | - TTM_MEMTYPE_FLAG_NEEDS_IOREMAP | TTM_MEMTYPE_FLAG_MAPPABLE; man->available_caching = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_WC; man->default_caching = TTM_PL_FLAG_WC; - man->io_addr = NULL; - man->io_offset = rdev->mc.aper_base; - man->io_size = rdev->mc.aper_size; break; default: DRM_ERROR("Unsupported memory type %u\n", (unsigned)type); @@ -244,9 +232,9 @@ static void radeon_move_null(struct ttm_buffer_object *bo, } static int radeon_move_blit(struct ttm_buffer_object *bo, - bool evict, int no_wait, - struct ttm_mem_reg *new_mem, - struct ttm_mem_reg *old_mem) + bool evict, int no_wait_reserve, bool no_wait_gpu, + struct ttm_mem_reg *new_mem, + struct ttm_mem_reg *old_mem) { struct radeon_device *rdev; uint64_t old_start, new_start; @@ -290,13 +278,14 @@ static int radeon_move_blit(struct ttm_buffer_object *bo, r = radeon_copy(rdev, old_start, new_start, new_mem->num_pages, fence); /* FIXME: handle copy error */ r = ttm_bo_move_accel_cleanup(bo, (void *)fence, NULL, - evict, no_wait, new_mem); + evict, no_wait_reserve, no_wait_gpu, new_mem); radeon_fence_unref(&fence); return r; } static int radeon_move_vram_ram(struct ttm_buffer_object *bo, - bool evict, bool interruptible, bool no_wait, + bool evict, bool interruptible, + bool no_wait_reserve, bool no_wait_gpu, struct ttm_mem_reg *new_mem) { struct radeon_device *rdev; @@ -317,7 +306,7 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo, placement.busy_placement = &placements; placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; r = ttm_bo_mem_space(bo, &placement, &tmp_mem, - interruptible, no_wait); + interruptible, no_wait_reserve, no_wait_gpu); if (unlikely(r)) { return r; } @@ -331,11 +320,11 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo, if (unlikely(r)) { goto out_cleanup; } - r = radeon_move_blit(bo, true, no_wait, &tmp_mem, old_mem); + r = radeon_move_blit(bo, true, no_wait_reserve, no_wait_gpu, &tmp_mem, old_mem); if (unlikely(r)) { goto out_cleanup; } - r = ttm_bo_move_ttm(bo, true, no_wait, new_mem); + r = ttm_bo_move_ttm(bo, true, no_wait_reserve, no_wait_gpu, new_mem); out_cleanup: if (tmp_mem.mm_node) { struct ttm_bo_global *glob = rdev->mman.bdev.glob; @@ -349,7 +338,8 @@ out_cleanup: } static int radeon_move_ram_vram(struct ttm_buffer_object *bo, - bool evict, bool interruptible, bool no_wait, + bool evict, bool interruptible, + bool no_wait_reserve, bool no_wait_gpu, struct ttm_mem_reg *new_mem) { struct radeon_device *rdev; @@ -369,15 +359,15 @@ static int radeon_move_ram_vram(struct ttm_buffer_object *bo, placement.num_busy_placement = 1; placement.busy_placement = &placements; placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; - r = ttm_bo_mem_space(bo, &placement, &tmp_mem, interruptible, no_wait); + r = ttm_bo_mem_space(bo, &placement, &tmp_mem, interruptible, no_wait_reserve, no_wait_gpu); if (unlikely(r)) { return r; } - r = ttm_bo_move_ttm(bo, true, no_wait, &tmp_mem); + r = ttm_bo_move_ttm(bo, true, no_wait_reserve, no_wait_gpu, &tmp_mem); if (unlikely(r)) { goto out_cleanup; } - r = radeon_move_blit(bo, true, no_wait, new_mem, old_mem); + r = radeon_move_blit(bo, true, no_wait_reserve, no_wait_gpu, new_mem, old_mem); if (unlikely(r)) { goto out_cleanup; } @@ -394,8 +384,9 @@ out_cleanup: } static int radeon_bo_move(struct ttm_buffer_object *bo, - bool evict, bool interruptible, bool no_wait, - struct ttm_mem_reg *new_mem) + bool evict, bool interruptible, + bool no_wait_reserve, bool no_wait_gpu, + struct ttm_mem_reg *new_mem) { struct radeon_device *rdev; struct ttm_mem_reg *old_mem = &bo->mem; @@ -422,23 +413,66 @@ static int radeon_bo_move(struct ttm_buffer_object *bo, if (old_mem->mem_type == TTM_PL_VRAM && new_mem->mem_type == TTM_PL_SYSTEM) { r = radeon_move_vram_ram(bo, evict, interruptible, - no_wait, new_mem); + no_wait_reserve, no_wait_gpu, new_mem); } else if (old_mem->mem_type == TTM_PL_SYSTEM && new_mem->mem_type == TTM_PL_VRAM) { r = radeon_move_ram_vram(bo, evict, interruptible, - no_wait, new_mem); + no_wait_reserve, no_wait_gpu, new_mem); } else { - r = radeon_move_blit(bo, evict, no_wait, new_mem, old_mem); + r = radeon_move_blit(bo, evict, no_wait_reserve, no_wait_gpu, new_mem, old_mem); } if (r) { memcpy: - r = ttm_bo_move_memcpy(bo, evict, no_wait, new_mem); + r = ttm_bo_move_memcpy(bo, evict, no_wait_reserve, no_wait_gpu, new_mem); } - return r; } +static int radeon_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem) +{ + struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type]; + struct radeon_device *rdev = radeon_get_rdev(bdev); + + mem->bus.addr = NULL; + mem->bus.offset = 0; + mem->bus.size = mem->num_pages << PAGE_SHIFT; + mem->bus.base = 0; + mem->bus.is_iomem = false; + if (!(man->flags & TTM_MEMTYPE_FLAG_MAPPABLE)) + return -EINVAL; + switch (mem->mem_type) { + case TTM_PL_SYSTEM: + /* system memory */ + return 0; + case TTM_PL_TT: +#if __OS_HAS_AGP + if (rdev->flags & RADEON_IS_AGP) { + /* RADEON_IS_AGP is set only if AGP is active */ + mem->bus.offset = mem->mm_node->start << PAGE_SHIFT; + mem->bus.base = rdev->mc.agp_base; + mem->bus.is_iomem = !rdev->ddev->agp->cant_use_aperture; + } +#endif + break; + case TTM_PL_VRAM: + mem->bus.offset = mem->mm_node->start << PAGE_SHIFT; + /* check if it's visible */ + if ((mem->bus.offset + mem->bus.size) > rdev->mc.visible_vram_size) + return -EINVAL; + mem->bus.base = rdev->mc.aper_base; + mem->bus.is_iomem = true; + break; + default: + return -EINVAL; + } + return 0; +} + +static void radeon_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem) +{ +} + static int radeon_sync_obj_wait(void *sync_obj, void *sync_arg, bool lazy, bool interruptible) { @@ -479,6 +513,8 @@ static struct ttm_bo_driver radeon_bo_driver = { .sync_obj_ref = &radeon_sync_obj_ref, .move_notify = &radeon_bo_move_notify, .fault_reserve_notify = &radeon_bo_fault_reserve_notify, + .io_mem_reserve = &radeon_ttm_io_mem_reserve, + .io_mem_free = &radeon_ttm_io_mem_free, }; int radeon_ttm_init(struct radeon_device *rdev) @@ -571,13 +607,17 @@ static const struct vm_operations_struct *ttm_vm_ops = NULL; static int radeon_ttm_fault(struct vm_area_struct *vma, struct vm_fault *vmf) { struct ttm_buffer_object *bo; + struct radeon_device *rdev; int r; - bo = (struct ttm_buffer_object *)vma->vm_private_data; + bo = (struct ttm_buffer_object *)vma->vm_private_data; if (bo == NULL) { return VM_FAULT_NOPAGE; } + rdev = radeon_get_rdev(bo->bdev); + mutex_lock(&rdev->vram_mutex); r = ttm_vm_ops->fault(vma, vmf); + mutex_unlock(&rdev->vram_mutex); return r; } @@ -745,8 +785,8 @@ static int radeon_mm_dump_table(struct seq_file *m, void *data) static int radeon_ttm_debugfs_init(struct radeon_device *rdev) { #if defined(CONFIG_DEBUG_FS) - static struct drm_info_list radeon_mem_types_list[RADEON_DEBUGFS_MEM_TYPES]; - static char radeon_mem_types_names[RADEON_DEBUGFS_MEM_TYPES][32]; + static struct drm_info_list radeon_mem_types_list[RADEON_DEBUGFS_MEM_TYPES+1]; + static char radeon_mem_types_names[RADEON_DEBUGFS_MEM_TYPES+1][32]; unsigned i; for (i = 0; i < RADEON_DEBUGFS_MEM_TYPES; i++) { @@ -763,7 +803,13 @@ static int radeon_ttm_debugfs_init(struct radeon_device *rdev) radeon_mem_types_list[i].data = &rdev->mman.bdev.man[TTM_PL_TT].manager; } - return radeon_debugfs_add_files(rdev, radeon_mem_types_list, RADEON_DEBUGFS_MEM_TYPES); + /* Add ttm page pool to debugfs */ + sprintf(radeon_mem_types_names[i], "ttm_page_pool"); + radeon_mem_types_list[i].name = radeon_mem_types_names[i]; + radeon_mem_types_list[i].show = &ttm_page_alloc_debugfs; + radeon_mem_types_list[i].driver_features = 0; + radeon_mem_types_list[i].data = NULL; + return radeon_debugfs_add_files(rdev, radeon_mem_types_list, RADEON_DEBUGFS_MEM_TYPES+1); #endif return 0; diff --git a/drivers/gpu/drm/radeon/reg_srcs/evergreen b/drivers/gpu/drm/radeon/reg_srcs/evergreen new file mode 100644 index 000000000000..f78fd592544d --- /dev/null +++ b/drivers/gpu/drm/radeon/reg_srcs/evergreen @@ -0,0 +1,611 @@ +evergreen 0x9400 +0x00008040 WAIT_UNTIL +0x00008044 WAIT_UNTIL_POLL_CNTL +0x00008048 WAIT_UNTIL_POLL_MASK +0x0000804c WAIT_UNTIL_POLL_REFDATA +0x000088B0 VGT_VTX_VECT_EJECT_REG +0x000088C4 VGT_CACHE_INVALIDATION +0x000088D4 VGT_GS_VERTEX_REUSE +0x00008958 VGT_PRIMITIVE_TYPE +0x0000895C VGT_INDEX_TYPE +0x00008970 VGT_NUM_INDICES +0x00008974 VGT_NUM_INSTANCES +0x00008990 VGT_COMPUTE_DIM_X +0x00008994 VGT_COMPUTE_DIM_Y +0x00008998 VGT_COMPUTE_DIM_Z +0x0000899C VGT_COMPUTE_START_X +0x000089A0 VGT_COMPUTE_START_Y +0x000089A4 VGT_COMPUTE_START_Z +0x000089AC VGT_COMPUTE_THREAD_GOURP_SIZE +0x00008A14 PA_CL_ENHANCE +0x00008A60 PA_SC_LINE_STIPPLE_VALUE +0x00008B10 PA_SC_LINE_STIPPLE_STATE +0x00008BF0 PA_SC_ENHANCE +0x00008D8C SQ_DYN_GPR_CNTL_PS_FLUSH_REQ +0x00008C00 SQ_CONFIG +0x00008C04 SQ_GPR_RESOURCE_MGMT_1 +0x00008C08 SQ_GPR_RESOURCE_MGMT_2 +0x00008C0C SQ_GPR_RESOURCE_MGMT_3 +0x00008C10 SQ_GLOBAL_GPR_RESOURCE_MGMT_1 +0x00008C14 SQ_GLOBAL_GPR_RESOURCE_MGMT_2 +0x00008C18 SQ_THREAD_RESOURCE_MGMT +0x00008C1C SQ_THREAD_RESOURCE_MGMT_2 +0x00008C20 SQ_STACK_RESOURCE_MGMT_1 +0x00008C24 SQ_STACK_RESOURCE_MGMT_2 +0x00008C28 SQ_STACK_RESOURCE_MGMT_3 +0x00008DF8 SQ_CONST_MEM_BASE +0x00008E48 SQ_EX_ALLOC_TABLE_SLOTS +0x00009100 SPI_CONFIG_CNTL +0x0000913C SPI_CONFIG_CNTL_1 +0x00009700 VC_CNTL +0x00009714 VC_ENHANCE +0x00009830 DB_DEBUG +0x00009834 DB_DEBUG2 +0x00009838 DB_DEBUG3 +0x0000983C DB_DEBUG4 +0x00009854 DB_WATERMARKS +0x0000A400 TD_PS_BORDER_COLOR_INDEX +0x0000A404 TD_PS_BORDER_COLOR_RED +0x0000A408 TD_PS_BORDER_COLOR_GREEN +0x0000A40C TD_PS_BORDER_COLOR_BLUE +0x0000A410 TD_PS_BORDER_COLOR_ALPHA +0x0000A414 TD_VS_BORDER_COLOR_INDEX +0x0000A418 TD_VS_BORDER_COLOR_RED +0x0000A41C TD_VS_BORDER_COLOR_GREEN +0x0000A420 TD_VS_BORDER_COLOR_BLUE +0x0000A424 TD_VS_BORDER_COLOR_ALPHA +0x0000A428 TD_GS_BORDER_COLOR_INDEX +0x0000A42C TD_GS_BORDER_COLOR_RED +0x0000A430 TD_GS_BORDER_COLOR_GREEN +0x0000A434 TD_GS_BORDER_COLOR_BLUE +0x0000A438 TD_GS_BORDER_COLOR_ALPHA +0x0000A43C TD_HS_BORDER_COLOR_INDEX +0x0000A440 TD_HS_BORDER_COLOR_RED +0x0000A444 TD_HS_BORDER_COLOR_GREEN +0x0000A448 TD_HS_BORDER_COLOR_BLUE +0x0000A44C TD_HS_BORDER_COLOR_ALPHA +0x0000A450 TD_LS_BORDER_COLOR_INDEX +0x0000A454 TD_LS_BORDER_COLOR_RED +0x0000A458 TD_LS_BORDER_COLOR_GREEN +0x0000A45C TD_LS_BORDER_COLOR_BLUE +0x0000A460 TD_LS_BORDER_COLOR_ALPHA +0x0000A464 TD_CS_BORDER_COLOR_INDEX +0x0000A468 TD_CS_BORDER_COLOR_RED +0x0000A46C TD_CS_BORDER_COLOR_GREEN +0x0000A470 TD_CS_BORDER_COLOR_BLUE +0x0000A474 TD_CS_BORDER_COLOR_ALPHA +0x00028000 DB_RENDER_CONTROL +0x00028004 DB_COUNT_CONTROL +0x0002800C DB_RENDER_OVERRIDE +0x00028010 DB_RENDER_OVERRIDE2 +0x00028028 DB_STENCIL_CLEAR +0x0002802C DB_DEPTH_CLEAR +0x00028030 PA_SC_SCREEN_SCISSOR_TL +0x00028034 PA_SC_SCREEN_SCISSOR_BR +0x0002805C DB_DEPTH_SLICE +0x00028140 SQ_ALU_CONST_BUFFER_SIZE_PS_0 +0x00028144 SQ_ALU_CONST_BUFFER_SIZE_PS_1 +0x00028148 SQ_ALU_CONST_BUFFER_SIZE_PS_2 +0x0002814C SQ_ALU_CONST_BUFFER_SIZE_PS_3 +0x00028150 SQ_ALU_CONST_BUFFER_SIZE_PS_4 +0x00028154 SQ_ALU_CONST_BUFFER_SIZE_PS_5 +0x00028158 SQ_ALU_CONST_BUFFER_SIZE_PS_6 +0x0002815C SQ_ALU_CONST_BUFFER_SIZE_PS_7 +0x00028160 SQ_ALU_CONST_BUFFER_SIZE_PS_8 +0x00028164 SQ_ALU_CONST_BUFFER_SIZE_PS_9 +0x00028168 SQ_ALU_CONST_BUFFER_SIZE_PS_10 +0x0002816C SQ_ALU_CONST_BUFFER_SIZE_PS_11 +0x00028170 SQ_ALU_CONST_BUFFER_SIZE_PS_12 +0x00028174 SQ_ALU_CONST_BUFFER_SIZE_PS_13 +0x00028178 SQ_ALU_CONST_BUFFER_SIZE_PS_14 +0x0002817C SQ_ALU_CONST_BUFFER_SIZE_PS_15 +0x00028180 SQ_ALU_CONST_BUFFER_SIZE_VS_0 +0x00028184 SQ_ALU_CONST_BUFFER_SIZE_VS_1 +0x00028188 SQ_ALU_CONST_BUFFER_SIZE_VS_2 +0x0002818C SQ_ALU_CONST_BUFFER_SIZE_VS_3 +0x00028190 SQ_ALU_CONST_BUFFER_SIZE_VS_4 +0x00028194 SQ_ALU_CONST_BUFFER_SIZE_VS_5 +0x00028198 SQ_ALU_CONST_BUFFER_SIZE_VS_6 +0x0002819C SQ_ALU_CONST_BUFFER_SIZE_VS_7 +0x000281A0 SQ_ALU_CONST_BUFFER_SIZE_VS_8 +0x000281A4 SQ_ALU_CONST_BUFFER_SIZE_VS_9 +0x000281A8 SQ_ALU_CONST_BUFFER_SIZE_VS_10 +0x000281AC SQ_ALU_CONST_BUFFER_SIZE_VS_11 +0x000281B0 SQ_ALU_CONST_BUFFER_SIZE_VS_12 +0x000281B4 SQ_ALU_CONST_BUFFER_SIZE_VS_13 +0x000281B8 SQ_ALU_CONST_BUFFER_SIZE_VS_14 +0x000281BC SQ_ALU_CONST_BUFFER_SIZE_VS_15 +0x000281C0 SQ_ALU_CONST_BUFFER_SIZE_GS_0 +0x000281C4 SQ_ALU_CONST_BUFFER_SIZE_GS_1 +0x000281C8 SQ_ALU_CONST_BUFFER_SIZE_GS_2 +0x000281CC SQ_ALU_CONST_BUFFER_SIZE_GS_3 +0x000281D0 SQ_ALU_CONST_BUFFER_SIZE_GS_4 +0x000281D4 SQ_ALU_CONST_BUFFER_SIZE_GS_5 +0x000281D8 SQ_ALU_CONST_BUFFER_SIZE_GS_6 +0x000281DC SQ_ALU_CONST_BUFFER_SIZE_GS_7 +0x000281E0 SQ_ALU_CONST_BUFFER_SIZE_GS_8 +0x000281E4 SQ_ALU_CONST_BUFFER_SIZE_GS_9 +0x000281E8 SQ_ALU_CONST_BUFFER_SIZE_GS_10 +0x000281EC SQ_ALU_CONST_BUFFER_SIZE_GS_11 +0x000281F0 SQ_ALU_CONST_BUFFER_SIZE_GS_12 +0x000281F4 SQ_ALU_CONST_BUFFER_SIZE_GS_13 +0x000281F8 SQ_ALU_CONST_BUFFER_SIZE_GS_14 +0x000281FC SQ_ALU_CONST_BUFFER_SIZE_GS_15 +0x00028200 PA_SC_WINDOW_OFFSET +0x00028204 PA_SC_WINDOW_SCISSOR_TL +0x00028208 PA_SC_WINDOW_SCISSOR_BR +0x0002820C PA_SC_CLIPRECT_RULE +0x00028210 PA_SC_CLIPRECT_0_TL +0x00028214 PA_SC_CLIPRECT_0_BR +0x00028218 PA_SC_CLIPRECT_1_TL +0x0002821C PA_SC_CLIPRECT_1_BR +0x00028220 PA_SC_CLIPRECT_2_TL +0x00028224 PA_SC_CLIPRECT_2_BR +0x00028228 PA_SC_CLIPRECT_3_TL +0x0002822C PA_SC_CLIPRECT_3_BR +0x00028230 PA_SC_EDGERULE +0x00028234 PA_SU_HARDWARE_SCREEN_OFFSET +0x00028240 PA_SC_GENERIC_SCISSOR_TL +0x00028244 PA_SC_GENERIC_SCISSOR_BR +0x00028250 PA_SC_VPORT_SCISSOR_0_TL +0x00028254 PA_SC_VPORT_SCISSOR_0_BR +0x00028258 PA_SC_VPORT_SCISSOR_1_TL +0x0002825C PA_SC_VPORT_SCISSOR_1_BR +0x00028260 PA_SC_VPORT_SCISSOR_2_TL +0x00028264 PA_SC_VPORT_SCISSOR_2_BR +0x00028268 PA_SC_VPORT_SCISSOR_3_TL +0x0002826C PA_SC_VPORT_SCISSOR_3_BR +0x00028270 PA_SC_VPORT_SCISSOR_4_TL +0x00028274 PA_SC_VPORT_SCISSOR_4_BR +0x00028278 PA_SC_VPORT_SCISSOR_5_TL +0x0002827C PA_SC_VPORT_SCISSOR_5_BR +0x00028280 PA_SC_VPORT_SCISSOR_6_TL +0x00028284 PA_SC_VPORT_SCISSOR_6_BR +0x00028288 PA_SC_VPORT_SCISSOR_7_TL +0x0002828C PA_SC_VPORT_SCISSOR_7_BR +0x00028290 PA_SC_VPORT_SCISSOR_8_TL +0x00028294 PA_SC_VPORT_SCISSOR_8_BR +0x00028298 PA_SC_VPORT_SCISSOR_9_TL +0x0002829C PA_SC_VPORT_SCISSOR_9_BR +0x000282A0 PA_SC_VPORT_SCISSOR_10_TL +0x000282A4 PA_SC_VPORT_SCISSOR_10_BR +0x000282A8 PA_SC_VPORT_SCISSOR_11_TL +0x000282AC PA_SC_VPORT_SCISSOR_11_BR +0x000282B0 PA_SC_VPORT_SCISSOR_12_TL +0x000282B4 PA_SC_VPORT_SCISSOR_12_BR +0x000282B8 PA_SC_VPORT_SCISSOR_13_TL +0x000282BC PA_SC_VPORT_SCISSOR_13_BR +0x000282C0 PA_SC_VPORT_SCISSOR_14_TL +0x000282C4 PA_SC_VPORT_SCISSOR_14_BR +0x000282C8 PA_SC_VPORT_SCISSOR_15_TL +0x000282CC PA_SC_VPORT_SCISSOR_15_BR +0x000282D0 PA_SC_VPORT_ZMIN_0 +0x000282D4 PA_SC_VPORT_ZMAX_0 +0x000282D8 PA_SC_VPORT_ZMIN_1 +0x000282DC PA_SC_VPORT_ZMAX_1 +0x000282E0 PA_SC_VPORT_ZMIN_2 +0x000282E4 PA_SC_VPORT_ZMAX_2 +0x000282E8 PA_SC_VPORT_ZMIN_3 +0x000282EC PA_SC_VPORT_ZMAX_3 +0x000282F0 PA_SC_VPORT_ZMIN_4 +0x000282F4 PA_SC_VPORT_ZMAX_4 +0x000282F8 PA_SC_VPORT_ZMIN_5 +0x000282FC PA_SC_VPORT_ZMAX_5 +0x00028300 PA_SC_VPORT_ZMIN_6 +0x00028304 PA_SC_VPORT_ZMAX_6 +0x00028308 PA_SC_VPORT_ZMIN_7 +0x0002830C PA_SC_VPORT_ZMAX_7 +0x00028310 PA_SC_VPORT_ZMIN_8 +0x00028314 PA_SC_VPORT_ZMAX_8 +0x00028318 PA_SC_VPORT_ZMIN_9 +0x0002831C PA_SC_VPORT_ZMAX_9 +0x00028320 PA_SC_VPORT_ZMIN_10 +0x00028324 PA_SC_VPORT_ZMAX_10 +0x00028328 PA_SC_VPORT_ZMIN_11 +0x0002832C PA_SC_VPORT_ZMAX_11 +0x00028330 PA_SC_VPORT_ZMIN_12 +0x00028334 PA_SC_VPORT_ZMAX_12 +0x00028338 PA_SC_VPORT_ZMIN_13 +0x0002833C PA_SC_VPORT_ZMAX_13 +0x00028340 PA_SC_VPORT_ZMIN_14 +0x00028344 PA_SC_VPORT_ZMAX_14 +0x00028348 PA_SC_VPORT_ZMIN_15 +0x0002834C PA_SC_VPORT_ZMAX_15 +0x00028350 SX_MISC +0x00028380 SQ_VTX_SEMANTIC_0 +0x00028384 SQ_VTX_SEMANTIC_1 +0x00028388 SQ_VTX_SEMANTIC_2 +0x0002838C SQ_VTX_SEMANTIC_3 +0x00028390 SQ_VTX_SEMANTIC_4 +0x00028394 SQ_VTX_SEMANTIC_5 +0x00028398 SQ_VTX_SEMANTIC_6 +0x0002839C SQ_VTX_SEMANTIC_7 +0x000283A0 SQ_VTX_SEMANTIC_8 +0x000283A4 SQ_VTX_SEMANTIC_9 +0x000283A8 SQ_VTX_SEMANTIC_10 +0x000283AC SQ_VTX_SEMANTIC_11 +0x000283B0 SQ_VTX_SEMANTIC_12 +0x000283B4 SQ_VTX_SEMANTIC_13 +0x000283B8 SQ_VTX_SEMANTIC_14 +0x000283BC SQ_VTX_SEMANTIC_15 +0x000283C0 SQ_VTX_SEMANTIC_16 +0x000283C4 SQ_VTX_SEMANTIC_17 +0x000283C8 SQ_VTX_SEMANTIC_18 +0x000283CC SQ_VTX_SEMANTIC_19 +0x000283D0 SQ_VTX_SEMANTIC_20 +0x000283D4 SQ_VTX_SEMANTIC_21 +0x000283D8 SQ_VTX_SEMANTIC_22 +0x000283DC SQ_VTX_SEMANTIC_23 +0x000283E0 SQ_VTX_SEMANTIC_24 +0x000283E4 SQ_VTX_SEMANTIC_25 +0x000283E8 SQ_VTX_SEMANTIC_26 +0x000283EC SQ_VTX_SEMANTIC_27 +0x000283F0 SQ_VTX_SEMANTIC_28 +0x000283F4 SQ_VTX_SEMANTIC_29 +0x000283F8 SQ_VTX_SEMANTIC_30 +0x000283FC SQ_VTX_SEMANTIC_31 +0x00028400 VGT_MAX_VTX_INDX +0x00028404 VGT_MIN_VTX_INDX +0x00028408 VGT_INDX_OFFSET +0x0002840C VGT_MULTI_PRIM_IB_RESET_INDX +0x00028410 SX_ALPHA_TEST_CONTROL +0x00028414 CB_BLEND_RED +0x00028418 CB_BLEND_GREEN +0x0002841C CB_BLEND_BLUE +0x00028420 CB_BLEND_ALPHA +0x00028430 DB_STENCILREFMASK +0x00028434 DB_STENCILREFMASK_BF +0x00028438 SX_ALPHA_REF +0x0002843C PA_CL_VPORT_XSCALE_0 +0x00028440 PA_CL_VPORT_XOFFSET_0 +0x00028444 PA_CL_VPORT_YSCALE_0 +0x00028448 PA_CL_VPORT_YOFFSET_0 +0x0002844C PA_CL_VPORT_ZSCALE_0 +0x00028450 PA_CL_VPORT_ZOFFSET_0 +0x00028454 PA_CL_VPORT_XSCALE_1 +0x00028458 PA_CL_VPORT_XOFFSET_1 +0x0002845C PA_CL_VPORT_YSCALE_1 +0x00028460 PA_CL_VPORT_YOFFSET_1 +0x00028464 PA_CL_VPORT_ZSCALE_1 +0x00028468 PA_CL_VPORT_ZOFFSET_1 +0x0002846C PA_CL_VPORT_XSCALE_2 +0x00028470 PA_CL_VPORT_XOFFSET_2 +0x00028474 PA_CL_VPORT_YSCALE_2 +0x00028478 PA_CL_VPORT_YOFFSET_2 +0x0002847C PA_CL_VPORT_ZSCALE_2 +0x00028480 PA_CL_VPORT_ZOFFSET_2 +0x00028484 PA_CL_VPORT_XSCALE_3 +0x00028488 PA_CL_VPORT_XOFFSET_3 +0x0002848C PA_CL_VPORT_YSCALE_3 +0x00028490 PA_CL_VPORT_YOFFSET_3 +0x00028494 PA_CL_VPORT_ZSCALE_3 +0x00028498 PA_CL_VPORT_ZOFFSET_3 +0x0002849C PA_CL_VPORT_XSCALE_4 +0x000284A0 PA_CL_VPORT_XOFFSET_4 +0x000284A4 PA_CL_VPORT_YSCALE_4 +0x000284A8 PA_CL_VPORT_YOFFSET_4 +0x000284AC PA_CL_VPORT_ZSCALE_4 +0x000284B0 PA_CL_VPORT_ZOFFSET_4 +0x000284B4 PA_CL_VPORT_XSCALE_5 +0x000284B8 PA_CL_VPORT_XOFFSET_5 +0x000284BC PA_CL_VPORT_YSCALE_5 +0x000284C0 PA_CL_VPORT_YOFFSET_5 +0x000284C4 PA_CL_VPORT_ZSCALE_5 +0x000284C8 PA_CL_VPORT_ZOFFSET_5 +0x000284CC PA_CL_VPORT_XSCALE_6 +0x000284D0 PA_CL_VPORT_XOFFSET_6 +0x000284D4 PA_CL_VPORT_YSCALE_6 +0x000284D8 PA_CL_VPORT_YOFFSET_6 +0x000284DC PA_CL_VPORT_ZSCALE_6 +0x000284E0 PA_CL_VPORT_ZOFFSET_6 +0x000284E4 PA_CL_VPORT_XSCALE_7 +0x000284E8 PA_CL_VPORT_XOFFSET_7 +0x000284EC PA_CL_VPORT_YSCALE_7 +0x000284F0 PA_CL_VPORT_YOFFSET_7 +0x000284F4 PA_CL_VPORT_ZSCALE_7 +0x000284F8 PA_CL_VPORT_ZOFFSET_7 +0x000284FC PA_CL_VPORT_XSCALE_8 +0x00028500 PA_CL_VPORT_XOFFSET_8 +0x00028504 PA_CL_VPORT_YSCALE_8 +0x00028508 PA_CL_VPORT_YOFFSET_8 +0x0002850C PA_CL_VPORT_ZSCALE_8 +0x00028510 PA_CL_VPORT_ZOFFSET_8 +0x00028514 PA_CL_VPORT_XSCALE_9 +0x00028518 PA_CL_VPORT_XOFFSET_9 +0x0002851C PA_CL_VPORT_YSCALE_9 +0x00028520 PA_CL_VPORT_YOFFSET_9 +0x00028524 PA_CL_VPORT_ZSCALE_9 +0x00028528 PA_CL_VPORT_ZOFFSET_9 +0x0002852C PA_CL_VPORT_XSCALE_10 +0x00028530 PA_CL_VPORT_XOFFSET_10 +0x00028534 PA_CL_VPORT_YSCALE_10 +0x00028538 PA_CL_VPORT_YOFFSET_10 +0x0002853C PA_CL_VPORT_ZSCALE_10 +0x00028540 PA_CL_VPORT_ZOFFSET_10 +0x00028544 PA_CL_VPORT_XSCALE_11 +0x00028548 PA_CL_VPORT_XOFFSET_11 +0x0002854C PA_CL_VPORT_YSCALE_11 +0x00028550 PA_CL_VPORT_YOFFSET_11 +0x00028554 PA_CL_VPORT_ZSCALE_11 +0x00028558 PA_CL_VPORT_ZOFFSET_11 +0x0002855C PA_CL_VPORT_XSCALE_12 +0x00028560 PA_CL_VPORT_XOFFSET_12 +0x00028564 PA_CL_VPORT_YSCALE_12 +0x00028568 PA_CL_VPORT_YOFFSET_12 +0x0002856C PA_CL_VPORT_ZSCALE_12 +0x00028570 PA_CL_VPORT_ZOFFSET_12 +0x00028574 PA_CL_VPORT_XSCALE_13 +0x00028578 PA_CL_VPORT_XOFFSET_13 +0x0002857C PA_CL_VPORT_YSCALE_13 +0x00028580 PA_CL_VPORT_YOFFSET_13 +0x00028584 PA_CL_VPORT_ZSCALE_13 +0x00028588 PA_CL_VPORT_ZOFFSET_13 +0x0002858C PA_CL_VPORT_XSCALE_14 +0x00028590 PA_CL_VPORT_XOFFSET_14 +0x00028594 PA_CL_VPORT_YSCALE_14 +0x00028598 PA_CL_VPORT_YOFFSET_14 +0x0002859C PA_CL_VPORT_ZSCALE_14 +0x000285A0 PA_CL_VPORT_ZOFFSET_14 +0x000285A4 PA_CL_VPORT_XSCALE_15 +0x000285A8 PA_CL_VPORT_XOFFSET_15 +0x000285AC PA_CL_VPORT_YSCALE_15 +0x000285B0 PA_CL_VPORT_YOFFSET_15 +0x000285B4 PA_CL_VPORT_ZSCALE_15 +0x000285B8 PA_CL_VPORT_ZOFFSET_15 +0x000285BC PA_CL_UCP_0_X +0x000285C0 PA_CL_UCP_0_Y +0x000285C4 PA_CL_UCP_0_Z +0x000285C8 PA_CL_UCP_0_W +0x000285CC PA_CL_UCP_1_X +0x000285D0 PA_CL_UCP_1_Y +0x000285D4 PA_CL_UCP_1_Z +0x000285D8 PA_CL_UCP_1_W +0x000285DC PA_CL_UCP_2_X +0x000285E0 PA_CL_UCP_2_Y +0x000285E4 PA_CL_UCP_2_Z +0x000285E8 PA_CL_UCP_2_W +0x000285EC PA_CL_UCP_3_X +0x000285F0 PA_CL_UCP_3_Y +0x000285F4 PA_CL_UCP_3_Z +0x000285F8 PA_CL_UCP_3_W +0x000285FC PA_CL_UCP_4_X +0x00028600 PA_CL_UCP_4_Y +0x00028604 PA_CL_UCP_4_Z +0x00028608 PA_CL_UCP_4_W +0x0002860C PA_CL_UCP_5_X +0x00028610 PA_CL_UCP_5_Y +0x00028614 PA_CL_UCP_5_Z +0x00028618 PA_CL_UCP_5_W +0x0002861C SPI_VS_OUT_ID_0 +0x00028620 SPI_VS_OUT_ID_1 +0x00028624 SPI_VS_OUT_ID_2 +0x00028628 SPI_VS_OUT_ID_3 +0x0002862C SPI_VS_OUT_ID_4 +0x00028630 SPI_VS_OUT_ID_5 +0x00028634 SPI_VS_OUT_ID_6 +0x00028638 SPI_VS_OUT_ID_7 +0x0002863C SPI_VS_OUT_ID_8 +0x00028640 SPI_VS_OUT_ID_9 +0x00028644 SPI_PS_INPUT_CNTL_0 +0x00028648 SPI_PS_INPUT_CNTL_1 +0x0002864C SPI_PS_INPUT_CNTL_2 +0x00028650 SPI_PS_INPUT_CNTL_3 +0x00028654 SPI_PS_INPUT_CNTL_4 +0x00028658 SPI_PS_INPUT_CNTL_5 +0x0002865C SPI_PS_INPUT_CNTL_6 +0x00028660 SPI_PS_INPUT_CNTL_7 +0x00028664 SPI_PS_INPUT_CNTL_8 +0x00028668 SPI_PS_INPUT_CNTL_9 +0x0002866C SPI_PS_INPUT_CNTL_10 +0x00028670 SPI_PS_INPUT_CNTL_11 +0x00028674 SPI_PS_INPUT_CNTL_12 +0x00028678 SPI_PS_INPUT_CNTL_13 +0x0002867C SPI_PS_INPUT_CNTL_14 +0x00028680 SPI_PS_INPUT_CNTL_15 +0x00028684 SPI_PS_INPUT_CNTL_16 +0x00028688 SPI_PS_INPUT_CNTL_17 +0x0002868C SPI_PS_INPUT_CNTL_18 +0x00028690 SPI_PS_INPUT_CNTL_19 +0x00028694 SPI_PS_INPUT_CNTL_20 +0x00028698 SPI_PS_INPUT_CNTL_21 +0x0002869C SPI_PS_INPUT_CNTL_22 +0x000286A0 SPI_PS_INPUT_CNTL_23 +0x000286A4 SPI_PS_INPUT_CNTL_24 +0x000286A8 SPI_PS_INPUT_CNTL_25 +0x000286AC SPI_PS_INPUT_CNTL_26 +0x000286B0 SPI_PS_INPUT_CNTL_27 +0x000286B4 SPI_PS_INPUT_CNTL_28 +0x000286B8 SPI_PS_INPUT_CNTL_29 +0x000286BC SPI_PS_INPUT_CNTL_30 +0x000286C0 SPI_PS_INPUT_CNTL_31 +0x000286C4 SPI_VS_OUT_CONFIG +0x000286C8 SPI_THREAD_GROUPING +0x000286CC SPI_PS_IN_CONTROL_0 +0x000286D0 SPI_PS_IN_CONTROL_1 +0x000286D4 SPI_INTERP_CONTROL_0 +0x000286D8 SPI_INPUT_Z +0x000286DC SPI_FOG_CNTL +0x000286E0 SPI_BARYC_CNTL +0x000286E4 SPI_PS_IN_CONTROL_2 +0x000286E8 SPI_COMPUTE_INPUT_CNTL +0x000286EC SPI_COMPUTE_NUM_THREAD_X +0x000286F0 SPI_COMPUTE_NUM_THREAD_Y +0x000286F4 SPI_COMPUTE_NUM_THREAD_Z +0x000286F8 GDS_ADDR_SIZE +0x00028780 CB_BLEND0_CONTROL +0x00028784 CB_BLEND1_CONTROL +0x00028788 CB_BLEND2_CONTROL +0x0002878C CB_BLEND3_CONTROL +0x00028790 CB_BLEND4_CONTROL +0x00028794 CB_BLEND5_CONTROL +0x00028798 CB_BLEND6_CONTROL +0x0002879C CB_BLEND7_CONTROL +0x000287CC CS_COPY_STATE +0x000287D0 GFX_COPY_STATE +0x000287D4 PA_CL_POINT_X_RAD +0x000287D8 PA_CL_POINT_Y_RAD +0x000287DC PA_CL_POINT_SIZE +0x000287E0 PA_CL_POINT_CULL_RAD +0x00028808 CB_COLOR_CONTROL +0x0002880C DB_SHADER_CONTROL +0x00028810 PA_CL_CLIP_CNTL +0x00028814 PA_SU_SC_MODE_CNTL +0x00028818 PA_CL_VTE_CNTL +0x0002881C PA_CL_VS_OUT_CNTL +0x00028820 PA_CL_NANINF_CNTL +0x00028824 PA_SU_LINE_STIPPLE_CNTL +0x00028828 PA_SU_LINE_STIPPLE_SCALE +0x0002882C PA_SU_PRIM_FILTER_CNTL +0x00028838 SQ_DYN_GPR_RESOURCE_LIMIT_1 +0x00028844 SQ_PGM_RESOURCES_PS +0x00028848 SQ_PGM_RESOURCES_2_PS +0x0002884C SQ_PGM_EXPORTS_PS +0x00028860 SQ_PGM_RESOURCES_VS +0x00028864 SQ_PGM_RESOURCES_2_VS +0x00028878 SQ_PGM_RESOURCES_GS +0x0002887C SQ_PGM_RESOURCES_2_GS +0x00028890 SQ_PGM_RESOURCES_ES +0x00028894 SQ_PGM_RESOURCES_2_ES +0x000288A8 SQ_PGM_RESOURCES_FS +0x000288BC SQ_PGM_RESOURCES_HS +0x000288C0 SQ_PGM_RESOURCES_2_HS +0x000288D4 SQ_PGM_RESOURCES_LS +0x000288D8 SQ_PGM_RESOURCES_2_LS +0x000288E8 SQ_LDS_ALLOC +0x000288EC SQ_LDS_ALLOC_PS +0x000288F0 SQ_VTX_SEMANTIC_CLEAR +0x00028A00 PA_SU_POINT_SIZE +0x00028A04 PA_SU_POINT_MINMAX +0x00028A08 PA_SU_LINE_CNTL +0x00028A0C PA_SC_LINE_STIPPLE +0x00028A10 VGT_OUTPUT_PATH_CNTL +0x00028A14 VGT_HOS_CNTL +0x00028A18 VGT_HOS_MAX_TESS_LEVEL +0x00028A1C VGT_HOS_MIN_TESS_LEVEL +0x00028A20 VGT_HOS_REUSE_DEPTH +0x00028A24 VGT_GROUP_PRIM_TYPE +0x00028A28 VGT_GROUP_FIRST_DECR +0x00028A2C VGT_GROUP_DECR +0x00028A30 VGT_GROUP_VECT_0_CNTL +0x00028A34 VGT_GROUP_VECT_1_CNTL +0x00028A38 VGT_GROUP_VECT_0_FMT_CNTL +0x00028A3C VGT_GROUP_VECT_1_FMT_CNTL +0x00028A40 VGT_GS_MODE +0x00028A48 PA_SC_MODE_CNTL_0 +0x00028A4C PA_SC_MODE_CNTL_1 +0x00028A50 VGT_ENHANCE +0x00028A54 VGT_GS_PER_ES +0x00028A58 VGT_ES_PER_GS +0x00028A5C VGT_GS_PER_VS +0x00028A6C VGT_GS_OUT_PRIM_TYPE +0x00028A84 VGT_PRIMITIVEID_EN +0x00028A94 VGT_MULTI_PRIM_IB_RESET_EN +0x00028AA0 VGT_INSTANCE_STEP_RATE_0 +0x00028AA4 VGT_INSTANCE_STEP_RATE_1 +0x00028AB4 VGT_REUSE_OFF +0x00028AB8 VGT_VTX_CNT_EN +0x00028ABC DB_HTILE_SURFACE +0x00028AC0 DB_SRESULTS_COMPARE_STATE0 +0x00028AC4 DB_SRESULTS_COMPARE_STATE1 +0x00028AC8 DB_PRELOAD_CONTROL +0x00028B38 VGT_GS_MAX_VERT_OUT +0x00028B54 VGT_SHADER_STAGES_EN +0x00028B58 VGT_LS_HS_CONFIG +0x00028B5C VGT_LS_SIZE +0x00028B60 VGT_HS_SIZE +0x00028B64 VGT_LS_HS_ALLOC +0x00028B68 VGT_HS_PATCH_CONST +0x00028B6C VGT_TF_PARAM +0x00028B70 DB_ALPHA_TO_MASK +0x00028B74 VGT_DISPATCH_INITIATOR +0x00028B78 PA_SU_POLY_OFFSET_DB_FMT_CNTL +0x00028B7C PA_SU_POLY_OFFSET_CLAMP +0x00028B80 PA_SU_POLY_OFFSET_FRONT_SCALE +0x00028B84 PA_SU_POLY_OFFSET_FRONT_OFFSET +0x00028B88 PA_SU_POLY_OFFSET_BACK_SCALE +0x00028B8C PA_SU_POLY_OFFSET_BACK_OFFSET +0x00028B74 VGT_GS_INSTANCE_CNT +0x00028C00 PA_SC_LINE_CNTL +0x00028C08 PA_SU_VTX_CNTL +0x00028C0C PA_CL_GB_VERT_CLIP_ADJ +0x00028C10 PA_CL_GB_VERT_DISC_ADJ +0x00028C14 PA_CL_GB_HORZ_CLIP_ADJ +0x00028C18 PA_CL_GB_HORZ_DISC_ADJ +0x00028C1C PA_SC_AA_SAMPLE_LOCS_0 +0x00028C20 PA_SC_AA_SAMPLE_LOCS_1 +0x00028C24 PA_SC_AA_SAMPLE_LOCS_2 +0x00028C28 PA_SC_AA_SAMPLE_LOCS_3 +0x00028C2C PA_SC_AA_SAMPLE_LOCS_4 +0x00028C30 PA_SC_AA_SAMPLE_LOCS_5 +0x00028C34 PA_SC_AA_SAMPLE_LOCS_6 +0x00028C38 PA_SC_AA_SAMPLE_LOCS_7 +0x00028C3C PA_SC_AA_MASK +0x00028C8C CB_COLOR0_CLEAR_WORD0 +0x00028C90 CB_COLOR0_CLEAR_WORD1 +0x00028C94 CB_COLOR0_CLEAR_WORD2 +0x00028C98 CB_COLOR0_CLEAR_WORD3 +0x00028CC8 CB_COLOR1_CLEAR_WORD0 +0x00028CCC CB_COLOR1_CLEAR_WORD1 +0x00028CD0 CB_COLOR1_CLEAR_WORD2 +0x00028CD4 CB_COLOR1_CLEAR_WORD3 +0x00028D04 CB_COLOR2_CLEAR_WORD0 +0x00028D08 CB_COLOR2_CLEAR_WORD1 +0x00028D0C CB_COLOR2_CLEAR_WORD2 +0x00028D10 CB_COLOR2_CLEAR_WORD3 +0x00028D40 CB_COLOR3_CLEAR_WORD0 +0x00028D44 CB_COLOR3_CLEAR_WORD1 +0x00028D48 CB_COLOR3_CLEAR_WORD2 +0x00028D4C CB_COLOR3_CLEAR_WORD3 +0x00028D7C CB_COLOR4_CLEAR_WORD0 +0x00028D80 CB_COLOR4_CLEAR_WORD1 +0x00028D84 CB_COLOR4_CLEAR_WORD2 +0x00028D88 CB_COLOR4_CLEAR_WORD3 +0x00028DB8 CB_COLOR5_CLEAR_WORD0 +0x00028DBC CB_COLOR5_CLEAR_WORD1 +0x00028DC0 CB_COLOR5_CLEAR_WORD2 +0x00028DC4 CB_COLOR5_CLEAR_WORD3 +0x00028DF4 CB_COLOR6_CLEAR_WORD0 +0x00028DF8 CB_COLOR6_CLEAR_WORD1 +0x00028DFC CB_COLOR6_CLEAR_WORD2 +0x00028E00 CB_COLOR6_CLEAR_WORD3 +0x00028E30 CB_COLOR7_CLEAR_WORD0 +0x00028E34 CB_COLOR7_CLEAR_WORD1 +0x00028E38 CB_COLOR7_CLEAR_WORD2 +0x00028E3C CB_COLOR7_CLEAR_WORD3 +0x00028F80 SQ_ALU_CONST_BUFFER_SIZE_HS_0 +0x00028F84 SQ_ALU_CONST_BUFFER_SIZE_HS_1 +0x00028F88 SQ_ALU_CONST_BUFFER_SIZE_HS_2 +0x00028F8C SQ_ALU_CONST_BUFFER_SIZE_HS_3 +0x00028F90 SQ_ALU_CONST_BUFFER_SIZE_HS_4 +0x00028F94 SQ_ALU_CONST_BUFFER_SIZE_HS_5 +0x00028F98 SQ_ALU_CONST_BUFFER_SIZE_HS_6 +0x00028F9C SQ_ALU_CONST_BUFFER_SIZE_HS_7 +0x00028FA0 SQ_ALU_CONST_BUFFER_SIZE_HS_8 +0x00028FA4 SQ_ALU_CONST_BUFFER_SIZE_HS_9 +0x00028FA8 SQ_ALU_CONST_BUFFER_SIZE_HS_10 +0x00028FAC SQ_ALU_CONST_BUFFER_SIZE_HS_11 +0x00028FB0 SQ_ALU_CONST_BUFFER_SIZE_HS_12 +0x00028FB4 SQ_ALU_CONST_BUFFER_SIZE_HS_13 +0x00028FB8 SQ_ALU_CONST_BUFFER_SIZE_HS_14 +0x00028FBC SQ_ALU_CONST_BUFFER_SIZE_HS_15 +0x00028FC0 SQ_ALU_CONST_BUFFER_SIZE_LS_0 +0x00028FC4 SQ_ALU_CONST_BUFFER_SIZE_LS_1 +0x00028FC8 SQ_ALU_CONST_BUFFER_SIZE_LS_2 +0x00028FCC SQ_ALU_CONST_BUFFER_SIZE_LS_3 +0x00028FD0 SQ_ALU_CONST_BUFFER_SIZE_LS_4 +0x00028FD4 SQ_ALU_CONST_BUFFER_SIZE_LS_5 +0x00028FD8 SQ_ALU_CONST_BUFFER_SIZE_LS_6 +0x00028FDC SQ_ALU_CONST_BUFFER_SIZE_LS_7 +0x00028FE0 SQ_ALU_CONST_BUFFER_SIZE_LS_8 +0x00028FE4 SQ_ALU_CONST_BUFFER_SIZE_LS_9 +0x00028FE8 SQ_ALU_CONST_BUFFER_SIZE_LS_10 +0x00028FEC SQ_ALU_CONST_BUFFER_SIZE_LS_11 +0x00028FF0 SQ_ALU_CONST_BUFFER_SIZE_LS_12 +0x00028FF4 SQ_ALU_CONST_BUFFER_SIZE_LS_13 +0x00028FF8 SQ_ALU_CONST_BUFFER_SIZE_LS_14 +0x00028FFC SQ_ALU_CONST_BUFFER_SIZE_LS_15 +0x0003CFF0 SQ_VTX_BASE_VTX_LOC +0x0003CFF4 SQ_VTX_START_INST_LOC +0x0003FF00 SQ_TEX_SAMPLER_CLEAR +0x0003FF04 SQ_TEX_RESOURCE_CLEAR +0x0003FF08 SQ_LOOP_BOOL_CLEAR diff --git a/drivers/gpu/drm/radeon/rs400.c b/drivers/gpu/drm/radeon/rs400.c index 1a41cb268b72..f454c9a5e7f2 100644 --- a/drivers/gpu/drm/radeon/rs400.c +++ b/drivers/gpu/drm/radeon/rs400.c @@ -57,7 +57,9 @@ void rs400_gart_adjust_size(struct radeon_device *rdev) } if (rdev->family == CHIP_RS400 || rdev->family == CHIP_RS480) { /* FIXME: RS400 & RS480 seems to have issue with GART size - * if 4G of system memory (needs more testing) */ + * if 4G of system memory (needs more testing) + */ + /* XXX is this still an issue with proper alignment? */ rdev->mc.gtt_size = 32 * 1024 * 1024; DRM_ERROR("Forcing to 32M GART size (because of ASIC bug ?)\n"); } @@ -243,8 +245,6 @@ int rs400_mc_wait_for_idle(struct radeon_device *rdev) void rs400_gpu_init(struct radeon_device *rdev) { - /* FIXME: HDP same place on rs400 ? */ - r100_hdp_reset(rdev); /* FIXME: is this correct ? */ r420_pipes_init(rdev); if (rs400_mc_wait_for_idle(rdev)) { @@ -265,6 +265,7 @@ void rs400_mc_init(struct radeon_device *rdev) r100_vram_init_sizes(rdev); base = (RREG32(RADEON_NB_TOM) & 0xffff) << 16; radeon_vram_location(rdev, &rdev->mc, base); + rdev->mc.gtt_base_align = rdev->mc.gtt_size - 1; radeon_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); } @@ -433,7 +434,7 @@ int rs400_resume(struct radeon_device *rdev) /* setup MC before calling post tables */ rs400_mc_program(rdev); /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), RREG32(R_0007C0_CP_STAT)); @@ -458,7 +459,6 @@ int rs400_suspend(struct radeon_device *rdev) void rs400_fini(struct radeon_device *rdev) { - radeon_pm_fini(rdev); r100_cp_fini(rdev); r100_wb_fini(rdev); r100_ib_fini(rdev); @@ -497,7 +497,7 @@ int rs400_init(struct radeon_device *rdev) return r; } /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), @@ -509,8 +509,6 @@ int rs400_init(struct radeon_device *rdev) /* Initialize clocks */ radeon_get_clock_info(rdev->ddev); - /* Initialize power management */ - radeon_pm_init(rdev); /* initialize memory controller */ rs400_mc_init(rdev); /* Fence driver */ diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c index a81bc7a21e14..6dc15ea8ba33 100644 --- a/drivers/gpu/drm/radeon/rs600.c +++ b/drivers/gpu/drm/radeon/rs600.c @@ -46,6 +46,136 @@ void rs600_gpu_init(struct radeon_device *rdev); int rs600_mc_wait_for_idle(struct radeon_device *rdev); +void rs600_pm_misc(struct radeon_device *rdev) +{ + int requested_index = rdev->pm.requested_power_state_index; + struct radeon_power_state *ps = &rdev->pm.power_state[requested_index]; + struct radeon_voltage *voltage = &ps->clock_info[0].voltage; + u32 tmp, dyn_pwrmgt_sclk_length, dyn_sclk_vol_cntl; + u32 hdp_dyn_cntl, /*mc_host_dyn_cntl,*/ dyn_backbias_cntl; + + if ((voltage->type == VOLTAGE_GPIO) && (voltage->gpio.valid)) { + if (ps->misc & ATOM_PM_MISCINFO_VOLTAGE_DROP_SUPPORT) { + tmp = RREG32(voltage->gpio.reg); + if (voltage->active_high) + tmp |= voltage->gpio.mask; + else + tmp &= ~(voltage->gpio.mask); + WREG32(voltage->gpio.reg, tmp); + if (voltage->delay) + udelay(voltage->delay); + } else { + tmp = RREG32(voltage->gpio.reg); + if (voltage->active_high) + tmp &= ~voltage->gpio.mask; + else + tmp |= voltage->gpio.mask; + WREG32(voltage->gpio.reg, tmp); + if (voltage->delay) + udelay(voltage->delay); + } + } else if (voltage->type == VOLTAGE_VDDC) + radeon_atom_set_voltage(rdev, voltage->vddc_id); + + dyn_pwrmgt_sclk_length = RREG32_PLL(DYN_PWRMGT_SCLK_LENGTH); + dyn_pwrmgt_sclk_length &= ~REDUCED_POWER_SCLK_HILEN(0xf); + dyn_pwrmgt_sclk_length &= ~REDUCED_POWER_SCLK_LOLEN(0xf); + if (ps->misc & ATOM_PM_MISCINFO_ASIC_REDUCED_SPEED_SCLK_EN) { + if (ps->misc & ATOM_PM_MISCINFO_DYNAMIC_CLOCK_DIVIDER_BY_2) { + dyn_pwrmgt_sclk_length |= REDUCED_POWER_SCLK_HILEN(2); + dyn_pwrmgt_sclk_length |= REDUCED_POWER_SCLK_LOLEN(2); + } else if (ps->misc & ATOM_PM_MISCINFO_DYNAMIC_CLOCK_DIVIDER_BY_4) { + dyn_pwrmgt_sclk_length |= REDUCED_POWER_SCLK_HILEN(4); + dyn_pwrmgt_sclk_length |= REDUCED_POWER_SCLK_LOLEN(4); + } + } else { + dyn_pwrmgt_sclk_length |= REDUCED_POWER_SCLK_HILEN(1); + dyn_pwrmgt_sclk_length |= REDUCED_POWER_SCLK_LOLEN(1); + } + WREG32_PLL(DYN_PWRMGT_SCLK_LENGTH, dyn_pwrmgt_sclk_length); + + dyn_sclk_vol_cntl = RREG32_PLL(DYN_SCLK_VOL_CNTL); + if (ps->misc & ATOM_PM_MISCINFO_ASIC_DYNAMIC_VOLTAGE_EN) { + dyn_sclk_vol_cntl |= IO_CG_VOLTAGE_DROP; + if (voltage->delay) { + dyn_sclk_vol_cntl |= VOLTAGE_DROP_SYNC; + dyn_sclk_vol_cntl |= VOLTAGE_DELAY_SEL(voltage->delay); + } else + dyn_sclk_vol_cntl &= ~VOLTAGE_DROP_SYNC; + } else + dyn_sclk_vol_cntl &= ~IO_CG_VOLTAGE_DROP; + WREG32_PLL(DYN_SCLK_VOL_CNTL, dyn_sclk_vol_cntl); + + hdp_dyn_cntl = RREG32_PLL(HDP_DYN_CNTL); + if (ps->misc & ATOM_PM_MISCINFO_DYNAMIC_HDP_BLOCK_EN) + hdp_dyn_cntl &= ~HDP_FORCEON; + else + hdp_dyn_cntl |= HDP_FORCEON; + WREG32_PLL(HDP_DYN_CNTL, hdp_dyn_cntl); +#if 0 + /* mc_host_dyn seems to cause hangs from time to time */ + mc_host_dyn_cntl = RREG32_PLL(MC_HOST_DYN_CNTL); + if (ps->misc & ATOM_PM_MISCINFO_DYNAMIC_MC_HOST_BLOCK_EN) + mc_host_dyn_cntl &= ~MC_HOST_FORCEON; + else + mc_host_dyn_cntl |= MC_HOST_FORCEON; + WREG32_PLL(MC_HOST_DYN_CNTL, mc_host_dyn_cntl); +#endif + dyn_backbias_cntl = RREG32_PLL(DYN_BACKBIAS_CNTL); + if (ps->misc & ATOM_PM_MISCINFO2_DYNAMIC_BACK_BIAS_EN) + dyn_backbias_cntl |= IO_CG_BACKBIAS_EN; + else + dyn_backbias_cntl &= ~IO_CG_BACKBIAS_EN; + WREG32_PLL(DYN_BACKBIAS_CNTL, dyn_backbias_cntl); + + /* set pcie lanes */ + if ((rdev->flags & RADEON_IS_PCIE) && + !(rdev->flags & RADEON_IS_IGP) && + rdev->asic->set_pcie_lanes && + (ps->pcie_lanes != + rdev->pm.power_state[rdev->pm.current_power_state_index].pcie_lanes)) { + radeon_set_pcie_lanes(rdev, + ps->pcie_lanes); + DRM_DEBUG("Setting: p: %d\n", ps->pcie_lanes); + } +} + +void rs600_pm_prepare(struct radeon_device *rdev) +{ + struct drm_device *ddev = rdev->ddev; + struct drm_crtc *crtc; + struct radeon_crtc *radeon_crtc; + u32 tmp; + + /* disable any active CRTCs */ + list_for_each_entry(crtc, &ddev->mode_config.crtc_list, head) { + radeon_crtc = to_radeon_crtc(crtc); + if (radeon_crtc->enabled) { + tmp = RREG32(AVIVO_D1CRTC_CONTROL + radeon_crtc->crtc_offset); + tmp |= AVIVO_CRTC_DISP_READ_REQUEST_DISABLE; + WREG32(AVIVO_D1CRTC_CONTROL + radeon_crtc->crtc_offset, tmp); + } + } +} + +void rs600_pm_finish(struct radeon_device *rdev) +{ + struct drm_device *ddev = rdev->ddev; + struct drm_crtc *crtc; + struct radeon_crtc *radeon_crtc; + u32 tmp; + + /* enable any active CRTCs */ + list_for_each_entry(crtc, &ddev->mode_config.crtc_list, head) { + radeon_crtc = to_radeon_crtc(crtc); + if (radeon_crtc->enabled) { + tmp = RREG32(AVIVO_D1CRTC_CONTROL + radeon_crtc->crtc_offset); + tmp &= ~AVIVO_CRTC_DISP_READ_REQUEST_DISABLE; + WREG32(AVIVO_D1CRTC_CONTROL + radeon_crtc->crtc_offset, tmp); + } + } +} + /* hpd for digital panel detect/disconnect */ bool rs600_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd) { @@ -147,6 +277,78 @@ void rs600_hpd_fini(struct radeon_device *rdev) } } +void rs600_bm_disable(struct radeon_device *rdev) +{ + u32 tmp; + + /* disable bus mastering */ + pci_read_config_word(rdev->pdev, 0x4, (u16*)&tmp); + pci_write_config_word(rdev->pdev, 0x4, tmp & 0xFFFB); + mdelay(1); +} + +int rs600_asic_reset(struct radeon_device *rdev) +{ + u32 status, tmp; + + struct rv515_mc_save save; + + /* Stops all mc clients */ + rv515_mc_stop(rdev, &save); + status = RREG32(R_000E40_RBBM_STATUS); + if (!G_000E40_GUI_ACTIVE(status)) { + return 0; + } + status = RREG32(R_000E40_RBBM_STATUS); + dev_info(rdev->dev, "(%s:%d) RBBM_STATUS=0x%08X\n", __func__, __LINE__, status); + /* stop CP */ + WREG32(RADEON_CP_CSQ_CNTL, 0); + tmp = RREG32(RADEON_CP_RB_CNTL); + WREG32(RADEON_CP_RB_CNTL, tmp | RADEON_RB_RPTR_WR_ENA); + WREG32(RADEON_CP_RB_RPTR_WR, 0); + WREG32(RADEON_CP_RB_WPTR, 0); + WREG32(RADEON_CP_RB_CNTL, tmp); + pci_save_state(rdev->pdev); + /* disable bus mastering */ + rs600_bm_disable(rdev); + /* reset GA+VAP */ + WREG32(R_0000F0_RBBM_SOFT_RESET, S_0000F0_SOFT_RESET_VAP(1) | + S_0000F0_SOFT_RESET_GA(1)); + RREG32(R_0000F0_RBBM_SOFT_RESET); + mdelay(500); + WREG32(R_0000F0_RBBM_SOFT_RESET, 0); + mdelay(1); + status = RREG32(R_000E40_RBBM_STATUS); + dev_info(rdev->dev, "(%s:%d) RBBM_STATUS=0x%08X\n", __func__, __LINE__, status); + /* reset CP */ + WREG32(R_0000F0_RBBM_SOFT_RESET, S_0000F0_SOFT_RESET_CP(1)); + RREG32(R_0000F0_RBBM_SOFT_RESET); + mdelay(500); + WREG32(R_0000F0_RBBM_SOFT_RESET, 0); + mdelay(1); + status = RREG32(R_000E40_RBBM_STATUS); + dev_info(rdev->dev, "(%s:%d) RBBM_STATUS=0x%08X\n", __func__, __LINE__, status); + /* reset MC */ + WREG32(R_0000F0_RBBM_SOFT_RESET, S_0000F0_SOFT_RESET_MC(1)); + RREG32(R_0000F0_RBBM_SOFT_RESET); + mdelay(500); + WREG32(R_0000F0_RBBM_SOFT_RESET, 0); + mdelay(1); + status = RREG32(R_000E40_RBBM_STATUS); + dev_info(rdev->dev, "(%s:%d) RBBM_STATUS=0x%08X\n", __func__, __LINE__, status); + /* restore PCI & busmastering */ + pci_restore_state(rdev->pdev); + /* Check if GPU is idle */ + if (G_000E40_GA_BUSY(status) || G_000E40_VAP_BUSY(status)) { + dev_err(rdev->dev, "failed to reset GPU\n"); + rdev->gpu_lockup = true; + return -1; + } + rv515_mc_resume(rdev, &save); + dev_info(rdev->dev, "GPU reset succeed\n"); + return 0; +} + /* * GART. */ @@ -310,6 +512,9 @@ int rs600_irq_set(struct radeon_device *rdev) if (rdev->irq.sw_int) { tmp |= S_000040_SW_INT_EN(1); } + if (rdev->irq.gui_idle) { + tmp |= S_000040_GUI_IDLE(1); + } if (rdev->irq.crtc_vblank_int[0]) { mode_int |= S_006540_D1MODE_VBLANK_INT_MASK(1); } @@ -332,9 +537,15 @@ int rs600_irq_set(struct radeon_device *rdev) static inline uint32_t rs600_irq_ack(struct radeon_device *rdev, u32 *r500_disp_int) { uint32_t irqs = RREG32(R_000044_GEN_INT_STATUS); - uint32_t irq_mask = ~C_000044_SW_INT; + uint32_t irq_mask = S_000044_SW_INT(1); u32 tmp; + /* the interrupt works, but the status bit is permanently asserted */ + if (rdev->irq.gui_idle && radeon_gui_idle(rdev)) { + if (!rdev->irq.gui_idle_acked) + irq_mask |= S_000044_GUI_IDLE_STAT(1); + } + if (G_000044_DISPLAY_INT_STAT(irqs)) { *r500_disp_int = RREG32(R_007EDC_DISP_INTERRUPT_STATUS); if (G_007EDC_LB_D1_VBLANK_INTERRUPT(*r500_disp_int)) { @@ -382,6 +593,9 @@ int rs600_irq_process(struct radeon_device *rdev) uint32_t r500_disp_int; bool queue_hotplug = false; + /* reset gui idle ack. the status bit is broken */ + rdev->irq.gui_idle_acked = false; + status = rs600_irq_ack(rdev, &r500_disp_int); if (!status && !r500_disp_int) { return IRQ_NONE; @@ -390,6 +604,12 @@ int rs600_irq_process(struct radeon_device *rdev) /* SW interrupt */ if (G_000044_SW_INT(status)) radeon_fence_process(rdev); + /* GUI idle */ + if (G_000040_GUI_IDLE(status)) { + rdev->irq.gui_idle_acked = true; + rdev->pm.gui_idle = true; + wake_up(&rdev->irq.idle_queue); + } /* Vertical blank interrupts */ if (G_007EDC_LB_D1_VBLANK_INTERRUPT(r500_disp_int)) { drm_handle_vblank(rdev->ddev, 0); @@ -411,6 +631,8 @@ int rs600_irq_process(struct radeon_device *rdev) } status = rs600_irq_ack(rdev, &r500_disp_int); } + /* reset gui idle ack. the status bit is broken */ + rdev->irq.gui_idle_acked = false; if (queue_hotplug) queue_work(rdev->wq, &rdev->hotplug_work); if (rdev->msi_enabled) { @@ -454,7 +676,6 @@ int rs600_mc_wait_for_idle(struct radeon_device *rdev) void rs600_gpu_init(struct radeon_device *rdev) { - r100_hdp_reset(rdev); r420_pipes_init(rdev); /* Wait for mc idle */ if (rs600_mc_wait_for_idle(rdev)) @@ -477,6 +698,7 @@ void rs600_mc_init(struct radeon_device *rdev) base = G_000004_MC_FB_START(base) << 16; rdev->mc.igp_sideport_enabled = radeon_atombios_sideport_present(rdev); radeon_vram_location(rdev, &rdev->mc, base); + rdev->mc.gtt_base_align = 0; radeon_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); } @@ -601,7 +823,7 @@ int rs600_resume(struct radeon_device *rdev) /* Resume clock before doing reset */ rv515_clock_startup(rdev); /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), RREG32(R_0007C0_CP_STAT)); @@ -626,7 +848,6 @@ int rs600_suspend(struct radeon_device *rdev) void rs600_fini(struct radeon_device *rdev) { - radeon_pm_fini(rdev); r100_cp_fini(rdev); r100_wb_fini(rdev); r100_ib_fini(rdev); @@ -664,7 +885,7 @@ int rs600_init(struct radeon_device *rdev) return -EINVAL; } /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), @@ -676,8 +897,6 @@ int rs600_init(struct radeon_device *rdev) /* Initialize clocks */ radeon_get_clock_info(rdev->ddev); - /* Initialize power management */ - radeon_pm_init(rdev); /* initialize memory controller */ rs600_mc_init(rdev); rs600_debugfs(rdev); diff --git a/drivers/gpu/drm/radeon/rs600d.h b/drivers/gpu/drm/radeon/rs600d.h index e52d2695510b..a27c13ac47c3 100644 --- a/drivers/gpu/drm/radeon/rs600d.h +++ b/drivers/gpu/drm/radeon/rs600d.h @@ -178,6 +178,52 @@ #define S_000074_MC_IND_DATA(x) (((x) & 0xFFFFFFFF) << 0) #define G_000074_MC_IND_DATA(x) (((x) >> 0) & 0xFFFFFFFF) #define C_000074_MC_IND_DATA 0x00000000 +#define R_0000F0_RBBM_SOFT_RESET 0x0000F0 +#define S_0000F0_SOFT_RESET_CP(x) (((x) & 0x1) << 0) +#define G_0000F0_SOFT_RESET_CP(x) (((x) >> 0) & 0x1) +#define C_0000F0_SOFT_RESET_CP 0xFFFFFFFE +#define S_0000F0_SOFT_RESET_HI(x) (((x) & 0x1) << 1) +#define G_0000F0_SOFT_RESET_HI(x) (((x) >> 1) & 0x1) +#define C_0000F0_SOFT_RESET_HI 0xFFFFFFFD +#define S_0000F0_SOFT_RESET_VAP(x) (((x) & 0x1) << 2) +#define G_0000F0_SOFT_RESET_VAP(x) (((x) >> 2) & 0x1) +#define C_0000F0_SOFT_RESET_VAP 0xFFFFFFFB +#define S_0000F0_SOFT_RESET_RE(x) (((x) & 0x1) << 3) +#define G_0000F0_SOFT_RESET_RE(x) (((x) >> 3) & 0x1) +#define C_0000F0_SOFT_RESET_RE 0xFFFFFFF7 +#define S_0000F0_SOFT_RESET_PP(x) (((x) & 0x1) << 4) +#define G_0000F0_SOFT_RESET_PP(x) (((x) >> 4) & 0x1) +#define C_0000F0_SOFT_RESET_PP 0xFFFFFFEF +#define S_0000F0_SOFT_RESET_E2(x) (((x) & 0x1) << 5) +#define G_0000F0_SOFT_RESET_E2(x) (((x) >> 5) & 0x1) +#define C_0000F0_SOFT_RESET_E2 0xFFFFFFDF +#define S_0000F0_SOFT_RESET_RB(x) (((x) & 0x1) << 6) +#define G_0000F0_SOFT_RESET_RB(x) (((x) >> 6) & 0x1) +#define C_0000F0_SOFT_RESET_RB 0xFFFFFFBF +#define S_0000F0_SOFT_RESET_HDP(x) (((x) & 0x1) << 7) +#define G_0000F0_SOFT_RESET_HDP(x) (((x) >> 7) & 0x1) +#define C_0000F0_SOFT_RESET_HDP 0xFFFFFF7F +#define S_0000F0_SOFT_RESET_MC(x) (((x) & 0x1) << 8) +#define G_0000F0_SOFT_RESET_MC(x) (((x) >> 8) & 0x1) +#define C_0000F0_SOFT_RESET_MC 0xFFFFFEFF +#define S_0000F0_SOFT_RESET_AIC(x) (((x) & 0x1) << 9) +#define G_0000F0_SOFT_RESET_AIC(x) (((x) >> 9) & 0x1) +#define C_0000F0_SOFT_RESET_AIC 0xFFFFFDFF +#define S_0000F0_SOFT_RESET_VIP(x) (((x) & 0x1) << 10) +#define G_0000F0_SOFT_RESET_VIP(x) (((x) >> 10) & 0x1) +#define C_0000F0_SOFT_RESET_VIP 0xFFFFFBFF +#define S_0000F0_SOFT_RESET_DISP(x) (((x) & 0x1) << 11) +#define G_0000F0_SOFT_RESET_DISP(x) (((x) >> 11) & 0x1) +#define C_0000F0_SOFT_RESET_DISP 0xFFFFF7FF +#define S_0000F0_SOFT_RESET_CG(x) (((x) & 0x1) << 12) +#define G_0000F0_SOFT_RESET_CG(x) (((x) >> 12) & 0x1) +#define C_0000F0_SOFT_RESET_CG 0xFFFFEFFF +#define S_0000F0_SOFT_RESET_GA(x) (((x) & 0x1) << 13) +#define G_0000F0_SOFT_RESET_GA(x) (((x) >> 13) & 0x1) +#define C_0000F0_SOFT_RESET_GA 0xFFFFDFFF +#define S_0000F0_SOFT_RESET_IDCT(x) (((x) & 0x1) << 14) +#define G_0000F0_SOFT_RESET_IDCT(x) (((x) >> 14) & 0x1) +#define C_0000F0_SOFT_RESET_IDCT 0xFFFFBFFF #define R_000134_HDP_FB_LOCATION 0x000134 #define S_000134_HDP_FB_START(x) (((x) & 0xFFFF) << 0) #define G_000134_HDP_FB_START(x) (((x) >> 0) & 0xFFFF) @@ -588,4 +634,38 @@ #define G_006D4C_D2MODE_PRIORITY_B_FORCE_MASK(x) (((x) >> 24) & 0x1) #define C_006D4C_D2MODE_PRIORITY_B_FORCE_MASK 0xFEFFFFFF +/* PLL regs */ +#define GENERAL_PWRMGT 0x8 +#define GLOBAL_PWRMGT_EN (1 << 0) +#define MOBILE_SU (1 << 2) +#define DYN_PWRMGT_SCLK_LENGTH 0xc +#define NORMAL_POWER_SCLK_HILEN(x) ((x) << 0) +#define NORMAL_POWER_SCLK_LOLEN(x) ((x) << 4) +#define REDUCED_POWER_SCLK_HILEN(x) ((x) << 8) +#define REDUCED_POWER_SCLK_LOLEN(x) ((x) << 12) +#define POWER_D1_SCLK_HILEN(x) ((x) << 16) +#define POWER_D1_SCLK_LOLEN(x) ((x) << 20) +#define STATIC_SCREEN_HILEN(x) ((x) << 24) +#define STATIC_SCREEN_LOLEN(x) ((x) << 28) +#define DYN_SCLK_VOL_CNTL 0xe +#define IO_CG_VOLTAGE_DROP (1 << 0) +#define VOLTAGE_DROP_SYNC (1 << 2) +#define VOLTAGE_DELAY_SEL(x) ((x) << 3) +#define HDP_DYN_CNTL 0x10 +#define HDP_FORCEON (1 << 0) +#define MC_HOST_DYN_CNTL 0x1e +#define MC_HOST_FORCEON (1 << 0) +#define DYN_BACKBIAS_CNTL 0x29 +#define IO_CG_BACKBIAS_EN (1 << 0) + +/* mmreg */ +#define DOUT_POWER_MANAGEMENT_CNTL 0x7ee0 +#define PWRDN_WAIT_BUSY_OFF (1 << 0) +#define PWRDN_WAIT_PWRSEQ_OFF (1 << 4) +#define PWRDN_WAIT_PPLL_OFF (1 << 8) +#define PWRUP_WAIT_PPLL_ON (1 << 12) +#define PWRUP_WAIT_MEM_INIT_DONE (1 << 16) +#define PM_ASSERT_RESET (1 << 20) +#define PM_PWRDN_PPLL (1 << 24) + #endif diff --git a/drivers/gpu/drm/radeon/rs690.c b/drivers/gpu/drm/radeon/rs690.c index bbf3da790fd5..ce4ecbe10816 100644 --- a/drivers/gpu/drm/radeon/rs690.c +++ b/drivers/gpu/drm/radeon/rs690.c @@ -48,8 +48,6 @@ static int rs690_mc_wait_for_idle(struct radeon_device *rdev) static void rs690_gpu_init(struct radeon_device *rdev) { - /* FIXME: HDP same place on rs690 ? */ - r100_hdp_reset(rdev); /* FIXME: is this correct ? */ r420_pipes_init(rdev); if (rs690_mc_wait_for_idle(rdev)) { @@ -78,59 +76,62 @@ void rs690_pm_info(struct radeon_device *rdev) /* Get various system informations from bios */ switch (crev) { case 1: - tmp.full = rfixed_const(100); - rdev->pm.igp_sideport_mclk.full = rfixed_const(info->info.ulBootUpMemoryClock); - rdev->pm.igp_sideport_mclk.full = rfixed_div(rdev->pm.igp_sideport_mclk, tmp); - rdev->pm.igp_system_mclk.full = rfixed_const(le16_to_cpu(info->info.usK8MemoryClock)); - rdev->pm.igp_ht_link_clk.full = rfixed_const(le16_to_cpu(info->info.usFSBClock)); - rdev->pm.igp_ht_link_width.full = rfixed_const(info->info.ucHTLinkWidth); + tmp.full = dfixed_const(100); + rdev->pm.igp_sideport_mclk.full = dfixed_const(info->info.ulBootUpMemoryClock); + rdev->pm.igp_sideport_mclk.full = dfixed_div(rdev->pm.igp_sideport_mclk, tmp); + if (info->info.usK8MemoryClock) + rdev->pm.igp_system_mclk.full = dfixed_const(le16_to_cpu(info->info.usK8MemoryClock)); + else if (rdev->clock.default_mclk) { + rdev->pm.igp_system_mclk.full = dfixed_const(rdev->clock.default_mclk); + rdev->pm.igp_system_mclk.full = dfixed_div(rdev->pm.igp_system_mclk, tmp); + } else + rdev->pm.igp_system_mclk.full = dfixed_const(400); + rdev->pm.igp_ht_link_clk.full = dfixed_const(le16_to_cpu(info->info.usFSBClock)); + rdev->pm.igp_ht_link_width.full = dfixed_const(info->info.ucHTLinkWidth); break; case 2: - tmp.full = rfixed_const(100); - rdev->pm.igp_sideport_mclk.full = rfixed_const(info->info_v2.ulBootUpSidePortClock); - rdev->pm.igp_sideport_mclk.full = rfixed_div(rdev->pm.igp_sideport_mclk, tmp); - rdev->pm.igp_system_mclk.full = rfixed_const(info->info_v2.ulBootUpUMAClock); - rdev->pm.igp_system_mclk.full = rfixed_div(rdev->pm.igp_system_mclk, tmp); - rdev->pm.igp_ht_link_clk.full = rfixed_const(info->info_v2.ulHTLinkFreq); - rdev->pm.igp_ht_link_clk.full = rfixed_div(rdev->pm.igp_ht_link_clk, tmp); - rdev->pm.igp_ht_link_width.full = rfixed_const(le16_to_cpu(info->info_v2.usMinHTLinkWidth)); + tmp.full = dfixed_const(100); + rdev->pm.igp_sideport_mclk.full = dfixed_const(info->info_v2.ulBootUpSidePortClock); + rdev->pm.igp_sideport_mclk.full = dfixed_div(rdev->pm.igp_sideport_mclk, tmp); + if (info->info_v2.ulBootUpUMAClock) + rdev->pm.igp_system_mclk.full = dfixed_const(info->info_v2.ulBootUpUMAClock); + else if (rdev->clock.default_mclk) + rdev->pm.igp_system_mclk.full = dfixed_const(rdev->clock.default_mclk); + else + rdev->pm.igp_system_mclk.full = dfixed_const(66700); + rdev->pm.igp_system_mclk.full = dfixed_div(rdev->pm.igp_system_mclk, tmp); + rdev->pm.igp_ht_link_clk.full = dfixed_const(info->info_v2.ulHTLinkFreq); + rdev->pm.igp_ht_link_clk.full = dfixed_div(rdev->pm.igp_ht_link_clk, tmp); + rdev->pm.igp_ht_link_width.full = dfixed_const(le16_to_cpu(info->info_v2.usMinHTLinkWidth)); break; default: - tmp.full = rfixed_const(100); /* We assume the slower possible clock ie worst case */ - /* DDR 333Mhz */ - rdev->pm.igp_sideport_mclk.full = rfixed_const(333); - /* FIXME: system clock ? */ - rdev->pm.igp_system_mclk.full = rfixed_const(100); - rdev->pm.igp_system_mclk.full = rfixed_div(rdev->pm.igp_system_mclk, tmp); - rdev->pm.igp_ht_link_clk.full = rfixed_const(200); - rdev->pm.igp_ht_link_width.full = rfixed_const(8); + rdev->pm.igp_sideport_mclk.full = dfixed_const(200); + rdev->pm.igp_system_mclk.full = dfixed_const(200); + rdev->pm.igp_ht_link_clk.full = dfixed_const(1000); + rdev->pm.igp_ht_link_width.full = dfixed_const(8); DRM_ERROR("No integrated system info for your GPU, using safe default\n"); break; } } else { - tmp.full = rfixed_const(100); /* We assume the slower possible clock ie worst case */ - /* DDR 333Mhz */ - rdev->pm.igp_sideport_mclk.full = rfixed_const(333); - /* FIXME: system clock ? */ - rdev->pm.igp_system_mclk.full = rfixed_const(100); - rdev->pm.igp_system_mclk.full = rfixed_div(rdev->pm.igp_system_mclk, tmp); - rdev->pm.igp_ht_link_clk.full = rfixed_const(200); - rdev->pm.igp_ht_link_width.full = rfixed_const(8); + rdev->pm.igp_sideport_mclk.full = dfixed_const(200); + rdev->pm.igp_system_mclk.full = dfixed_const(200); + rdev->pm.igp_ht_link_clk.full = dfixed_const(1000); + rdev->pm.igp_ht_link_width.full = dfixed_const(8); DRM_ERROR("No integrated system info for your GPU, using safe default\n"); } /* Compute various bandwidth */ /* k8_bandwidth = (memory_clk / 2) * 2 * 8 * 0.5 = memory_clk * 4 */ - tmp.full = rfixed_const(4); - rdev->pm.k8_bandwidth.full = rfixed_mul(rdev->pm.igp_system_mclk, tmp); + tmp.full = dfixed_const(4); + rdev->pm.k8_bandwidth.full = dfixed_mul(rdev->pm.igp_system_mclk, tmp); /* ht_bandwidth = ht_clk * 2 * ht_width / 8 * 0.8 * = ht_clk * ht_width / 5 */ - tmp.full = rfixed_const(5); - rdev->pm.ht_bandwidth.full = rfixed_mul(rdev->pm.igp_ht_link_clk, + tmp.full = dfixed_const(5); + rdev->pm.ht_bandwidth.full = dfixed_mul(rdev->pm.igp_ht_link_clk, rdev->pm.igp_ht_link_width); - rdev->pm.ht_bandwidth.full = rfixed_div(rdev->pm.ht_bandwidth, tmp); + rdev->pm.ht_bandwidth.full = dfixed_div(rdev->pm.ht_bandwidth, tmp); if (tmp.full < rdev->pm.max_bandwidth.full) { /* HT link is a limiting factor */ rdev->pm.max_bandwidth.full = tmp.full; @@ -138,10 +139,10 @@ void rs690_pm_info(struct radeon_device *rdev) /* sideport_bandwidth = (sideport_clk / 2) * 2 * 2 * 0.7 * = (sideport_clk * 14) / 10 */ - tmp.full = rfixed_const(14); - rdev->pm.sideport_bandwidth.full = rfixed_mul(rdev->pm.igp_sideport_mclk, tmp); - tmp.full = rfixed_const(10); - rdev->pm.sideport_bandwidth.full = rfixed_div(rdev->pm.sideport_bandwidth, tmp); + tmp.full = dfixed_const(14); + rdev->pm.sideport_bandwidth.full = dfixed_mul(rdev->pm.igp_sideport_mclk, tmp); + tmp.full = dfixed_const(10); + rdev->pm.sideport_bandwidth.full = dfixed_div(rdev->pm.sideport_bandwidth, tmp); } void rs690_mc_init(struct radeon_device *rdev) @@ -161,6 +162,7 @@ void rs690_mc_init(struct radeon_device *rdev) rs690_pm_info(rdev); rdev->mc.igp_sideport_enabled = radeon_atombios_sideport_present(rdev); radeon_vram_location(rdev, &rdev->mc, base); + rdev->mc.gtt_base_align = rdev->mc.gtt_size - 1; radeon_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); } @@ -230,10 +232,6 @@ void rs690_crtc_bandwidth_compute(struct radeon_device *rdev, fixed20_12 a, b, c; fixed20_12 pclk, request_fifo_depth, tolerable_latency, estimated_width; fixed20_12 consumption_time, line_time, chunk_time, read_delay_latency; - /* FIXME: detect IGP with sideport memory, i don't think there is any - * such product available - */ - bool sideport = false; if (!crtc->base.enabled) { /* FIXME: wouldn't it better to set priority mark to maximum */ @@ -241,20 +239,20 @@ void rs690_crtc_bandwidth_compute(struct radeon_device *rdev, return; } - if (crtc->vsc.full > rfixed_const(2)) - wm->num_line_pair.full = rfixed_const(2); + if (crtc->vsc.full > dfixed_const(2)) + wm->num_line_pair.full = dfixed_const(2); else - wm->num_line_pair.full = rfixed_const(1); - - b.full = rfixed_const(mode->crtc_hdisplay); - c.full = rfixed_const(256); - a.full = rfixed_div(b, c); - request_fifo_depth.full = rfixed_mul(a, wm->num_line_pair); - request_fifo_depth.full = rfixed_ceil(request_fifo_depth); - if (a.full < rfixed_const(4)) { + wm->num_line_pair.full = dfixed_const(1); + + b.full = dfixed_const(mode->crtc_hdisplay); + c.full = dfixed_const(256); + a.full = dfixed_div(b, c); + request_fifo_depth.full = dfixed_mul(a, wm->num_line_pair); + request_fifo_depth.full = dfixed_ceil(request_fifo_depth); + if (a.full < dfixed_const(4)) { wm->lb_request_fifo_depth = 4; } else { - wm->lb_request_fifo_depth = rfixed_trunc(request_fifo_depth); + wm->lb_request_fifo_depth = dfixed_trunc(request_fifo_depth); } /* Determine consumption rate @@ -263,23 +261,23 @@ void rs690_crtc_bandwidth_compute(struct radeon_device *rdev, * vsc = vertical scaling ratio, defined as source/destination * hsc = horizontal scaling ration, defined as source/destination */ - a.full = rfixed_const(mode->clock); - b.full = rfixed_const(1000); - a.full = rfixed_div(a, b); - pclk.full = rfixed_div(b, a); + a.full = dfixed_const(mode->clock); + b.full = dfixed_const(1000); + a.full = dfixed_div(a, b); + pclk.full = dfixed_div(b, a); if (crtc->rmx_type != RMX_OFF) { - b.full = rfixed_const(2); + b.full = dfixed_const(2); if (crtc->vsc.full > b.full) b.full = crtc->vsc.full; - b.full = rfixed_mul(b, crtc->hsc); - c.full = rfixed_const(2); - b.full = rfixed_div(b, c); - consumption_time.full = rfixed_div(pclk, b); + b.full = dfixed_mul(b, crtc->hsc); + c.full = dfixed_const(2); + b.full = dfixed_div(b, c); + consumption_time.full = dfixed_div(pclk, b); } else { consumption_time.full = pclk.full; } - a.full = rfixed_const(1); - wm->consumption_rate.full = rfixed_div(a, consumption_time); + a.full = dfixed_const(1); + wm->consumption_rate.full = dfixed_div(a, consumption_time); /* Determine line time @@ -287,27 +285,27 @@ void rs690_crtc_bandwidth_compute(struct radeon_device *rdev, * LineTime = total number of horizontal pixels * pclk = pixel clock period(ns) */ - a.full = rfixed_const(crtc->base.mode.crtc_htotal); - line_time.full = rfixed_mul(a, pclk); + a.full = dfixed_const(crtc->base.mode.crtc_htotal); + line_time.full = dfixed_mul(a, pclk); /* Determine active time * ActiveTime = time of active region of display within one line, * hactive = total number of horizontal active pixels * htotal = total number of horizontal pixels */ - a.full = rfixed_const(crtc->base.mode.crtc_htotal); - b.full = rfixed_const(crtc->base.mode.crtc_hdisplay); - wm->active_time.full = rfixed_mul(line_time, b); - wm->active_time.full = rfixed_div(wm->active_time, a); + a.full = dfixed_const(crtc->base.mode.crtc_htotal); + b.full = dfixed_const(crtc->base.mode.crtc_hdisplay); + wm->active_time.full = dfixed_mul(line_time, b); + wm->active_time.full = dfixed_div(wm->active_time, a); /* Maximun bandwidth is the minimun bandwidth of all component */ rdev->pm.max_bandwidth = rdev->pm.core_bandwidth; - if (sideport) { + if (rdev->mc.igp_sideport_enabled) { if (rdev->pm.max_bandwidth.full > rdev->pm.sideport_bandwidth.full && rdev->pm.sideport_bandwidth.full) rdev->pm.max_bandwidth = rdev->pm.sideport_bandwidth; - read_delay_latency.full = rfixed_const(370 * 800 * 1000); - read_delay_latency.full = rfixed_div(read_delay_latency, + read_delay_latency.full = dfixed_const(370 * 800 * 1000); + read_delay_latency.full = dfixed_div(read_delay_latency, rdev->pm.igp_sideport_mclk); } else { if (rdev->pm.max_bandwidth.full > rdev->pm.k8_bandwidth.full && @@ -316,23 +314,23 @@ void rs690_crtc_bandwidth_compute(struct radeon_device *rdev, if (rdev->pm.max_bandwidth.full > rdev->pm.ht_bandwidth.full && rdev->pm.ht_bandwidth.full) rdev->pm.max_bandwidth = rdev->pm.ht_bandwidth; - read_delay_latency.full = rfixed_const(5000); + read_delay_latency.full = dfixed_const(5000); } /* sclk = system clocks(ns) = 1000 / max_bandwidth / 16 */ - a.full = rfixed_const(16); - rdev->pm.sclk.full = rfixed_mul(rdev->pm.max_bandwidth, a); - a.full = rfixed_const(1000); - rdev->pm.sclk.full = rfixed_div(a, rdev->pm.sclk); + a.full = dfixed_const(16); + rdev->pm.sclk.full = dfixed_mul(rdev->pm.max_bandwidth, a); + a.full = dfixed_const(1000); + rdev->pm.sclk.full = dfixed_div(a, rdev->pm.sclk); /* Determine chunk time * ChunkTime = the time it takes the DCP to send one chunk of data * to the LB which consists of pipeline delay and inter chunk gap * sclk = system clock(ns) */ - a.full = rfixed_const(256 * 13); - chunk_time.full = rfixed_mul(rdev->pm.sclk, a); - a.full = rfixed_const(10); - chunk_time.full = rfixed_div(chunk_time, a); + a.full = dfixed_const(256 * 13); + chunk_time.full = dfixed_mul(rdev->pm.sclk, a); + a.full = dfixed_const(10); + chunk_time.full = dfixed_div(chunk_time, a); /* Determine the worst case latency * NumLinePair = Number of line pairs to request(1=2 lines, 2=4 lines) @@ -342,13 +340,13 @@ void rs690_crtc_bandwidth_compute(struct radeon_device *rdev, * ChunkTime = time it takes the DCP to send one chunk of data to the LB * which consists of pipeline delay and inter chunk gap */ - if (rfixed_trunc(wm->num_line_pair) > 1) { - a.full = rfixed_const(3); - wm->worst_case_latency.full = rfixed_mul(a, chunk_time); + if (dfixed_trunc(wm->num_line_pair) > 1) { + a.full = dfixed_const(3); + wm->worst_case_latency.full = dfixed_mul(a, chunk_time); wm->worst_case_latency.full += read_delay_latency.full; } else { - a.full = rfixed_const(2); - wm->worst_case_latency.full = rfixed_mul(a, chunk_time); + a.full = dfixed_const(2); + wm->worst_case_latency.full = dfixed_mul(a, chunk_time); wm->worst_case_latency.full += read_delay_latency.full; } @@ -362,34 +360,34 @@ void rs690_crtc_bandwidth_compute(struct radeon_device *rdev, * of data to the LB which consists of * pipeline delay and inter chunk gap */ - if ((2+wm->lb_request_fifo_depth) >= rfixed_trunc(request_fifo_depth)) { + if ((2+wm->lb_request_fifo_depth) >= dfixed_trunc(request_fifo_depth)) { tolerable_latency.full = line_time.full; } else { - tolerable_latency.full = rfixed_const(wm->lb_request_fifo_depth - 2); + tolerable_latency.full = dfixed_const(wm->lb_request_fifo_depth - 2); tolerable_latency.full = request_fifo_depth.full - tolerable_latency.full; - tolerable_latency.full = rfixed_mul(tolerable_latency, chunk_time); + tolerable_latency.full = dfixed_mul(tolerable_latency, chunk_time); tolerable_latency.full = line_time.full - tolerable_latency.full; } /* We assume worst case 32bits (4 bytes) */ - wm->dbpp.full = rfixed_const(4 * 8); + wm->dbpp.full = dfixed_const(4 * 8); /* Determine the maximum priority mark * width = viewport width in pixels */ - a.full = rfixed_const(16); - wm->priority_mark_max.full = rfixed_const(crtc->base.mode.crtc_hdisplay); - wm->priority_mark_max.full = rfixed_div(wm->priority_mark_max, a); - wm->priority_mark_max.full = rfixed_ceil(wm->priority_mark_max); + a.full = dfixed_const(16); + wm->priority_mark_max.full = dfixed_const(crtc->base.mode.crtc_hdisplay); + wm->priority_mark_max.full = dfixed_div(wm->priority_mark_max, a); + wm->priority_mark_max.full = dfixed_ceil(wm->priority_mark_max); /* Determine estimated width */ estimated_width.full = tolerable_latency.full - wm->worst_case_latency.full; - estimated_width.full = rfixed_div(estimated_width, consumption_time); - if (rfixed_trunc(estimated_width) > crtc->base.mode.crtc_hdisplay) { - wm->priority_mark.full = rfixed_const(10); + estimated_width.full = dfixed_div(estimated_width, consumption_time); + if (dfixed_trunc(estimated_width) > crtc->base.mode.crtc_hdisplay) { + wm->priority_mark.full = dfixed_const(10); } else { - a.full = rfixed_const(16); - wm->priority_mark.full = rfixed_div(estimated_width, a); - wm->priority_mark.full = rfixed_ceil(wm->priority_mark); + a.full = dfixed_const(16); + wm->priority_mark.full = dfixed_div(estimated_width, a); + wm->priority_mark.full = dfixed_ceil(wm->priority_mark); wm->priority_mark.full = wm->priority_mark_max.full - wm->priority_mark.full; } } @@ -441,58 +439,58 @@ void rs690_bandwidth_update(struct radeon_device *rdev) WREG32(R_006D58_LB_MAX_REQ_OUTSTANDING, tmp); if (mode0 && mode1) { - if (rfixed_trunc(wm0.dbpp) > 64) - a.full = rfixed_mul(wm0.dbpp, wm0.num_line_pair); + if (dfixed_trunc(wm0.dbpp) > 64) + a.full = dfixed_mul(wm0.dbpp, wm0.num_line_pair); else a.full = wm0.num_line_pair.full; - if (rfixed_trunc(wm1.dbpp) > 64) - b.full = rfixed_mul(wm1.dbpp, wm1.num_line_pair); + if (dfixed_trunc(wm1.dbpp) > 64) + b.full = dfixed_mul(wm1.dbpp, wm1.num_line_pair); else b.full = wm1.num_line_pair.full; a.full += b.full; - fill_rate.full = rfixed_div(wm0.sclk, a); + fill_rate.full = dfixed_div(wm0.sclk, a); if (wm0.consumption_rate.full > fill_rate.full) { b.full = wm0.consumption_rate.full - fill_rate.full; - b.full = rfixed_mul(b, wm0.active_time); - a.full = rfixed_mul(wm0.worst_case_latency, + b.full = dfixed_mul(b, wm0.active_time); + a.full = dfixed_mul(wm0.worst_case_latency, wm0.consumption_rate); a.full = a.full + b.full; - b.full = rfixed_const(16 * 1000); - priority_mark02.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark02.full = dfixed_div(a, b); } else { - a.full = rfixed_mul(wm0.worst_case_latency, + a.full = dfixed_mul(wm0.worst_case_latency, wm0.consumption_rate); - b.full = rfixed_const(16 * 1000); - priority_mark02.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark02.full = dfixed_div(a, b); } if (wm1.consumption_rate.full > fill_rate.full) { b.full = wm1.consumption_rate.full - fill_rate.full; - b.full = rfixed_mul(b, wm1.active_time); - a.full = rfixed_mul(wm1.worst_case_latency, + b.full = dfixed_mul(b, wm1.active_time); + a.full = dfixed_mul(wm1.worst_case_latency, wm1.consumption_rate); a.full = a.full + b.full; - b.full = rfixed_const(16 * 1000); - priority_mark12.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark12.full = dfixed_div(a, b); } else { - a.full = rfixed_mul(wm1.worst_case_latency, + a.full = dfixed_mul(wm1.worst_case_latency, wm1.consumption_rate); - b.full = rfixed_const(16 * 1000); - priority_mark12.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark12.full = dfixed_div(a, b); } if (wm0.priority_mark.full > priority_mark02.full) priority_mark02.full = wm0.priority_mark.full; - if (rfixed_trunc(priority_mark02) < 0) + if (dfixed_trunc(priority_mark02) < 0) priority_mark02.full = 0; if (wm0.priority_mark_max.full > priority_mark02.full) priority_mark02.full = wm0.priority_mark_max.full; if (wm1.priority_mark.full > priority_mark12.full) priority_mark12.full = wm1.priority_mark.full; - if (rfixed_trunc(priority_mark12) < 0) + if (dfixed_trunc(priority_mark12) < 0) priority_mark12.full = 0; if (wm1.priority_mark_max.full > priority_mark12.full) priority_mark12.full = wm1.priority_mark_max.full; - d1mode_priority_a_cnt = rfixed_trunc(priority_mark02); - d2mode_priority_a_cnt = rfixed_trunc(priority_mark12); + d1mode_priority_a_cnt = dfixed_trunc(priority_mark02); + d2mode_priority_a_cnt = dfixed_trunc(priority_mark12); if (rdev->disp_priority == 2) { d1mode_priority_a_cnt |= S_006548_D1MODE_PRIORITY_A_ALWAYS_ON(1); d2mode_priority_a_cnt |= S_006D48_D2MODE_PRIORITY_A_ALWAYS_ON(1); @@ -502,32 +500,32 @@ void rs690_bandwidth_update(struct radeon_device *rdev) WREG32(R_006D48_D2MODE_PRIORITY_A_CNT, d2mode_priority_a_cnt); WREG32(R_006D4C_D2MODE_PRIORITY_B_CNT, d2mode_priority_a_cnt); } else if (mode0) { - if (rfixed_trunc(wm0.dbpp) > 64) - a.full = rfixed_mul(wm0.dbpp, wm0.num_line_pair); + if (dfixed_trunc(wm0.dbpp) > 64) + a.full = dfixed_mul(wm0.dbpp, wm0.num_line_pair); else a.full = wm0.num_line_pair.full; - fill_rate.full = rfixed_div(wm0.sclk, a); + fill_rate.full = dfixed_div(wm0.sclk, a); if (wm0.consumption_rate.full > fill_rate.full) { b.full = wm0.consumption_rate.full - fill_rate.full; - b.full = rfixed_mul(b, wm0.active_time); - a.full = rfixed_mul(wm0.worst_case_latency, + b.full = dfixed_mul(b, wm0.active_time); + a.full = dfixed_mul(wm0.worst_case_latency, wm0.consumption_rate); a.full = a.full + b.full; - b.full = rfixed_const(16 * 1000); - priority_mark02.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark02.full = dfixed_div(a, b); } else { - a.full = rfixed_mul(wm0.worst_case_latency, + a.full = dfixed_mul(wm0.worst_case_latency, wm0.consumption_rate); - b.full = rfixed_const(16 * 1000); - priority_mark02.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark02.full = dfixed_div(a, b); } if (wm0.priority_mark.full > priority_mark02.full) priority_mark02.full = wm0.priority_mark.full; - if (rfixed_trunc(priority_mark02) < 0) + if (dfixed_trunc(priority_mark02) < 0) priority_mark02.full = 0; if (wm0.priority_mark_max.full > priority_mark02.full) priority_mark02.full = wm0.priority_mark_max.full; - d1mode_priority_a_cnt = rfixed_trunc(priority_mark02); + d1mode_priority_a_cnt = dfixed_trunc(priority_mark02); if (rdev->disp_priority == 2) d1mode_priority_a_cnt |= S_006548_D1MODE_PRIORITY_A_ALWAYS_ON(1); WREG32(R_006548_D1MODE_PRIORITY_A_CNT, d1mode_priority_a_cnt); @@ -537,32 +535,32 @@ void rs690_bandwidth_update(struct radeon_device *rdev) WREG32(R_006D4C_D2MODE_PRIORITY_B_CNT, S_006D4C_D2MODE_PRIORITY_B_OFF(1)); } else { - if (rfixed_trunc(wm1.dbpp) > 64) - a.full = rfixed_mul(wm1.dbpp, wm1.num_line_pair); + if (dfixed_trunc(wm1.dbpp) > 64) + a.full = dfixed_mul(wm1.dbpp, wm1.num_line_pair); else a.full = wm1.num_line_pair.full; - fill_rate.full = rfixed_div(wm1.sclk, a); + fill_rate.full = dfixed_div(wm1.sclk, a); if (wm1.consumption_rate.full > fill_rate.full) { b.full = wm1.consumption_rate.full - fill_rate.full; - b.full = rfixed_mul(b, wm1.active_time); - a.full = rfixed_mul(wm1.worst_case_latency, + b.full = dfixed_mul(b, wm1.active_time); + a.full = dfixed_mul(wm1.worst_case_latency, wm1.consumption_rate); a.full = a.full + b.full; - b.full = rfixed_const(16 * 1000); - priority_mark12.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark12.full = dfixed_div(a, b); } else { - a.full = rfixed_mul(wm1.worst_case_latency, + a.full = dfixed_mul(wm1.worst_case_latency, wm1.consumption_rate); - b.full = rfixed_const(16 * 1000); - priority_mark12.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark12.full = dfixed_div(a, b); } if (wm1.priority_mark.full > priority_mark12.full) priority_mark12.full = wm1.priority_mark.full; - if (rfixed_trunc(priority_mark12) < 0) + if (dfixed_trunc(priority_mark12) < 0) priority_mark12.full = 0; if (wm1.priority_mark_max.full > priority_mark12.full) priority_mark12.full = wm1.priority_mark_max.full; - d2mode_priority_a_cnt = rfixed_trunc(priority_mark12); + d2mode_priority_a_cnt = dfixed_trunc(priority_mark12); if (rdev->disp_priority == 2) d2mode_priority_a_cnt |= S_006D48_D2MODE_PRIORITY_A_ALWAYS_ON(1); WREG32(R_006548_D1MODE_PRIORITY_A_CNT, @@ -653,7 +651,7 @@ int rs690_resume(struct radeon_device *rdev) /* Resume clock before doing reset */ rv515_clock_startup(rdev); /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), RREG32(R_0007C0_CP_STAT)); @@ -678,7 +676,6 @@ int rs690_suspend(struct radeon_device *rdev) void rs690_fini(struct radeon_device *rdev) { - radeon_pm_fini(rdev); r100_cp_fini(rdev); r100_wb_fini(rdev); r100_ib_fini(rdev); @@ -717,7 +714,7 @@ int rs690_init(struct radeon_device *rdev) return -EINVAL; } /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), @@ -729,8 +726,6 @@ int rs690_init(struct radeon_device *rdev) /* Initialize clocks */ radeon_get_clock_info(rdev->ddev); - /* Initialize power management */ - radeon_pm_init(rdev); /* initialize memory controller */ rs690_mc_init(rdev); rv515_debugfs(rdev); diff --git a/drivers/gpu/drm/radeon/rv515.c b/drivers/gpu/drm/radeon/rv515.c index 9035121f4b58..0c9c169a6852 100644 --- a/drivers/gpu/drm/radeon/rv515.c +++ b/drivers/gpu/drm/radeon/rv515.c @@ -147,16 +147,11 @@ void rv515_gpu_init(struct radeon_device *rdev) { unsigned pipe_select_current, gb_pipe_select, tmp; - r100_hdp_reset(rdev); - r100_rb2d_reset(rdev); - if (r100_gui_wait_for_idle(rdev)) { printk(KERN_WARNING "Failed to wait GUI idle while " "reseting GPU. Bad things might happen.\n"); } - rv515_vga_render_disable(rdev); - r420_pipes_init(rdev); gb_pipe_select = RREG32(0x402C); tmp = RREG32(0x170C); @@ -174,91 +169,6 @@ void rv515_gpu_init(struct radeon_device *rdev) } } -int rv515_ga_reset(struct radeon_device *rdev) -{ - uint32_t tmp; - bool reinit_cp; - int i; - - reinit_cp = rdev->cp.ready; - rdev->cp.ready = false; - for (i = 0; i < rdev->usec_timeout; i++) { - WREG32(CP_CSQ_MODE, 0); - WREG32(CP_CSQ_CNTL, 0); - WREG32(RBBM_SOFT_RESET, 0x32005); - (void)RREG32(RBBM_SOFT_RESET); - udelay(200); - WREG32(RBBM_SOFT_RESET, 0); - /* Wait to prevent race in RBBM_STATUS */ - mdelay(1); - tmp = RREG32(RBBM_STATUS); - if (tmp & ((1 << 20) | (1 << 26))) { - DRM_ERROR("VAP & CP still busy (RBBM_STATUS=0x%08X)\n", tmp); - /* GA still busy soft reset it */ - WREG32(0x429C, 0x200); - WREG32(VAP_PVS_STATE_FLUSH_REG, 0); - WREG32(0x43E0, 0); - WREG32(0x43E4, 0); - WREG32(0x24AC, 0); - } - /* Wait to prevent race in RBBM_STATUS */ - mdelay(1); - tmp = RREG32(RBBM_STATUS); - if (!(tmp & ((1 << 20) | (1 << 26)))) { - break; - } - } - for (i = 0; i < rdev->usec_timeout; i++) { - tmp = RREG32(RBBM_STATUS); - if (!(tmp & ((1 << 20) | (1 << 26)))) { - DRM_INFO("GA reset succeed (RBBM_STATUS=0x%08X)\n", - tmp); - DRM_INFO("GA_IDLE=0x%08X\n", RREG32(0x425C)); - DRM_INFO("RB3D_RESET_STATUS=0x%08X\n", RREG32(0x46f0)); - DRM_INFO("ISYNC_CNTL=0x%08X\n", RREG32(0x1724)); - if (reinit_cp) { - return r100_cp_init(rdev, rdev->cp.ring_size); - } - return 0; - } - DRM_UDELAY(1); - } - tmp = RREG32(RBBM_STATUS); - DRM_ERROR("Failed to reset GA ! (RBBM_STATUS=0x%08X)\n", tmp); - return -1; -} - -int rv515_gpu_reset(struct radeon_device *rdev) -{ - uint32_t status; - - /* reset order likely matter */ - status = RREG32(RBBM_STATUS); - /* reset HDP */ - r100_hdp_reset(rdev); - /* reset rb2d */ - if (status & ((1 << 17) | (1 << 18) | (1 << 27))) { - r100_rb2d_reset(rdev); - } - /* reset GA */ - if (status & ((1 << 20) | (1 << 26))) { - rv515_ga_reset(rdev); - } - /* reset CP */ - status = RREG32(RBBM_STATUS); - if (status & (1 << 16)) { - r100_cp_reset(rdev); - } - /* Check if GPU is idle */ - status = RREG32(RBBM_STATUS); - if (status & (1 << 31)) { - DRM_ERROR("Failed to reset GPU (RBBM_STATUS=0x%08X)\n", status); - return -1; - } - DRM_INFO("GPU reset succeed (RBBM_STATUS=0x%08X)\n", status); - return 0; -} - static void rv515_vram_get_type(struct radeon_device *rdev) { uint32_t tmp; @@ -285,6 +195,7 @@ void rv515_mc_init(struct radeon_device *rdev) rv515_vram_get_type(rdev); r100_vram_init_sizes(rdev); radeon_vram_location(rdev, &rdev->mc, 0); + rdev->mc.gtt_base_align = 0; if (!(rdev->flags & RADEON_IS_AGP)) radeon_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); @@ -335,7 +246,7 @@ static int rv515_debugfs_ga_info(struct seq_file *m, void *data) tmp = RREG32(0x2140); seq_printf(m, "VAP_CNTL_STATUS 0x%08x\n", tmp); - radeon_gpu_reset(rdev); + radeon_asic_reset(rdev); tmp = RREG32(0x425C); seq_printf(m, "GA_IDLE 0x%08x\n", tmp); return 0; @@ -503,7 +414,7 @@ int rv515_resume(struct radeon_device *rdev) /* Resume clock before doing reset */ rv515_clock_startup(rdev); /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), RREG32(R_0007C0_CP_STAT)); @@ -535,7 +446,6 @@ void rv515_set_safe_registers(struct radeon_device *rdev) void rv515_fini(struct radeon_device *rdev) { - radeon_pm_fini(rdev); r100_cp_fini(rdev); r100_wb_fini(rdev); r100_ib_fini(rdev); @@ -573,7 +483,7 @@ int rv515_init(struct radeon_device *rdev) return -EINVAL; } /* Reset gpu before posting otherwise ATOM will enter infinite loop */ - if (radeon_gpu_reset(rdev)) { + if (radeon_asic_reset(rdev)) { dev_warn(rdev->dev, "GPU reset failed ! (0xE40=0x%08X, 0x7C0=0x%08X)\n", RREG32(R_000E40_RBBM_STATUS), @@ -584,8 +494,6 @@ int rv515_init(struct radeon_device *rdev) return -EINVAL; /* Initialize clocks */ radeon_get_clock_info(rdev->ddev); - /* Initialize power management */ - radeon_pm_init(rdev); /* initialize AGP */ if (rdev->flags & RADEON_IS_AGP) { r = radeon_agp_init(rdev); @@ -885,20 +793,20 @@ void rv515_crtc_bandwidth_compute(struct radeon_device *rdev, return; } - if (crtc->vsc.full > rfixed_const(2)) - wm->num_line_pair.full = rfixed_const(2); + if (crtc->vsc.full > dfixed_const(2)) + wm->num_line_pair.full = dfixed_const(2); else - wm->num_line_pair.full = rfixed_const(1); - - b.full = rfixed_const(mode->crtc_hdisplay); - c.full = rfixed_const(256); - a.full = rfixed_div(b, c); - request_fifo_depth.full = rfixed_mul(a, wm->num_line_pair); - request_fifo_depth.full = rfixed_ceil(request_fifo_depth); - if (a.full < rfixed_const(4)) { + wm->num_line_pair.full = dfixed_const(1); + + b.full = dfixed_const(mode->crtc_hdisplay); + c.full = dfixed_const(256); + a.full = dfixed_div(b, c); + request_fifo_depth.full = dfixed_mul(a, wm->num_line_pair); + request_fifo_depth.full = dfixed_ceil(request_fifo_depth); + if (a.full < dfixed_const(4)) { wm->lb_request_fifo_depth = 4; } else { - wm->lb_request_fifo_depth = rfixed_trunc(request_fifo_depth); + wm->lb_request_fifo_depth = dfixed_trunc(request_fifo_depth); } /* Determine consumption rate @@ -907,23 +815,23 @@ void rv515_crtc_bandwidth_compute(struct radeon_device *rdev, * vsc = vertical scaling ratio, defined as source/destination * hsc = horizontal scaling ration, defined as source/destination */ - a.full = rfixed_const(mode->clock); - b.full = rfixed_const(1000); - a.full = rfixed_div(a, b); - pclk.full = rfixed_div(b, a); + a.full = dfixed_const(mode->clock); + b.full = dfixed_const(1000); + a.full = dfixed_div(a, b); + pclk.full = dfixed_div(b, a); if (crtc->rmx_type != RMX_OFF) { - b.full = rfixed_const(2); + b.full = dfixed_const(2); if (crtc->vsc.full > b.full) b.full = crtc->vsc.full; - b.full = rfixed_mul(b, crtc->hsc); - c.full = rfixed_const(2); - b.full = rfixed_div(b, c); - consumption_time.full = rfixed_div(pclk, b); + b.full = dfixed_mul(b, crtc->hsc); + c.full = dfixed_const(2); + b.full = dfixed_div(b, c); + consumption_time.full = dfixed_div(pclk, b); } else { consumption_time.full = pclk.full; } - a.full = rfixed_const(1); - wm->consumption_rate.full = rfixed_div(a, consumption_time); + a.full = dfixed_const(1); + wm->consumption_rate.full = dfixed_div(a, consumption_time); /* Determine line time @@ -931,27 +839,27 @@ void rv515_crtc_bandwidth_compute(struct radeon_device *rdev, * LineTime = total number of horizontal pixels * pclk = pixel clock period(ns) */ - a.full = rfixed_const(crtc->base.mode.crtc_htotal); - line_time.full = rfixed_mul(a, pclk); + a.full = dfixed_const(crtc->base.mode.crtc_htotal); + line_time.full = dfixed_mul(a, pclk); /* Determine active time * ActiveTime = time of active region of display within one line, * hactive = total number of horizontal active pixels * htotal = total number of horizontal pixels */ - a.full = rfixed_const(crtc->base.mode.crtc_htotal); - b.full = rfixed_const(crtc->base.mode.crtc_hdisplay); - wm->active_time.full = rfixed_mul(line_time, b); - wm->active_time.full = rfixed_div(wm->active_time, a); + a.full = dfixed_const(crtc->base.mode.crtc_htotal); + b.full = dfixed_const(crtc->base.mode.crtc_hdisplay); + wm->active_time.full = dfixed_mul(line_time, b); + wm->active_time.full = dfixed_div(wm->active_time, a); /* Determine chunk time * ChunkTime = the time it takes the DCP to send one chunk of data * to the LB which consists of pipeline delay and inter chunk gap * sclk = system clock(Mhz) */ - a.full = rfixed_const(600 * 1000); - chunk_time.full = rfixed_div(a, rdev->pm.sclk); - read_delay_latency.full = rfixed_const(1000); + a.full = dfixed_const(600 * 1000); + chunk_time.full = dfixed_div(a, rdev->pm.sclk); + read_delay_latency.full = dfixed_const(1000); /* Determine the worst case latency * NumLinePair = Number of line pairs to request(1=2 lines, 2=4 lines) @@ -961,9 +869,9 @@ void rv515_crtc_bandwidth_compute(struct radeon_device *rdev, * ChunkTime = time it takes the DCP to send one chunk of data to the LB * which consists of pipeline delay and inter chunk gap */ - if (rfixed_trunc(wm->num_line_pair) > 1) { - a.full = rfixed_const(3); - wm->worst_case_latency.full = rfixed_mul(a, chunk_time); + if (dfixed_trunc(wm->num_line_pair) > 1) { + a.full = dfixed_const(3); + wm->worst_case_latency.full = dfixed_mul(a, chunk_time); wm->worst_case_latency.full += read_delay_latency.full; } else { wm->worst_case_latency.full = chunk_time.full + read_delay_latency.full; @@ -979,34 +887,34 @@ void rv515_crtc_bandwidth_compute(struct radeon_device *rdev, * of data to the LB which consists of * pipeline delay and inter chunk gap */ - if ((2+wm->lb_request_fifo_depth) >= rfixed_trunc(request_fifo_depth)) { + if ((2+wm->lb_request_fifo_depth) >= dfixed_trunc(request_fifo_depth)) { tolerable_latency.full = line_time.full; } else { - tolerable_latency.full = rfixed_const(wm->lb_request_fifo_depth - 2); + tolerable_latency.full = dfixed_const(wm->lb_request_fifo_depth - 2); tolerable_latency.full = request_fifo_depth.full - tolerable_latency.full; - tolerable_latency.full = rfixed_mul(tolerable_latency, chunk_time); + tolerable_latency.full = dfixed_mul(tolerable_latency, chunk_time); tolerable_latency.full = line_time.full - tolerable_latency.full; } /* We assume worst case 32bits (4 bytes) */ - wm->dbpp.full = rfixed_const(2 * 16); + wm->dbpp.full = dfixed_const(2 * 16); /* Determine the maximum priority mark * width = viewport width in pixels */ - a.full = rfixed_const(16); - wm->priority_mark_max.full = rfixed_const(crtc->base.mode.crtc_hdisplay); - wm->priority_mark_max.full = rfixed_div(wm->priority_mark_max, a); - wm->priority_mark_max.full = rfixed_ceil(wm->priority_mark_max); + a.full = dfixed_const(16); + wm->priority_mark_max.full = dfixed_const(crtc->base.mode.crtc_hdisplay); + wm->priority_mark_max.full = dfixed_div(wm->priority_mark_max, a); + wm->priority_mark_max.full = dfixed_ceil(wm->priority_mark_max); /* Determine estimated width */ estimated_width.full = tolerable_latency.full - wm->worst_case_latency.full; - estimated_width.full = rfixed_div(estimated_width, consumption_time); - if (rfixed_trunc(estimated_width) > crtc->base.mode.crtc_hdisplay) { + estimated_width.full = dfixed_div(estimated_width, consumption_time); + if (dfixed_trunc(estimated_width) > crtc->base.mode.crtc_hdisplay) { wm->priority_mark.full = wm->priority_mark_max.full; } else { - a.full = rfixed_const(16); - wm->priority_mark.full = rfixed_div(estimated_width, a); - wm->priority_mark.full = rfixed_ceil(wm->priority_mark); + a.full = dfixed_const(16); + wm->priority_mark.full = dfixed_div(estimated_width, a); + wm->priority_mark.full = dfixed_ceil(wm->priority_mark); wm->priority_mark.full = wm->priority_mark_max.full - wm->priority_mark.full; } } @@ -1035,58 +943,58 @@ void rv515_bandwidth_avivo_update(struct radeon_device *rdev) WREG32(LB_MAX_REQ_OUTSTANDING, tmp); if (mode0 && mode1) { - if (rfixed_trunc(wm0.dbpp) > 64) - a.full = rfixed_div(wm0.dbpp, wm0.num_line_pair); + if (dfixed_trunc(wm0.dbpp) > 64) + a.full = dfixed_div(wm0.dbpp, wm0.num_line_pair); else a.full = wm0.num_line_pair.full; - if (rfixed_trunc(wm1.dbpp) > 64) - b.full = rfixed_div(wm1.dbpp, wm1.num_line_pair); + if (dfixed_trunc(wm1.dbpp) > 64) + b.full = dfixed_div(wm1.dbpp, wm1.num_line_pair); else b.full = wm1.num_line_pair.full; a.full += b.full; - fill_rate.full = rfixed_div(wm0.sclk, a); + fill_rate.full = dfixed_div(wm0.sclk, a); if (wm0.consumption_rate.full > fill_rate.full) { b.full = wm0.consumption_rate.full - fill_rate.full; - b.full = rfixed_mul(b, wm0.active_time); - a.full = rfixed_const(16); - b.full = rfixed_div(b, a); - a.full = rfixed_mul(wm0.worst_case_latency, + b.full = dfixed_mul(b, wm0.active_time); + a.full = dfixed_const(16); + b.full = dfixed_div(b, a); + a.full = dfixed_mul(wm0.worst_case_latency, wm0.consumption_rate); priority_mark02.full = a.full + b.full; } else { - a.full = rfixed_mul(wm0.worst_case_latency, + a.full = dfixed_mul(wm0.worst_case_latency, wm0.consumption_rate); - b.full = rfixed_const(16 * 1000); - priority_mark02.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark02.full = dfixed_div(a, b); } if (wm1.consumption_rate.full > fill_rate.full) { b.full = wm1.consumption_rate.full - fill_rate.full; - b.full = rfixed_mul(b, wm1.active_time); - a.full = rfixed_const(16); - b.full = rfixed_div(b, a); - a.full = rfixed_mul(wm1.worst_case_latency, + b.full = dfixed_mul(b, wm1.active_time); + a.full = dfixed_const(16); + b.full = dfixed_div(b, a); + a.full = dfixed_mul(wm1.worst_case_latency, wm1.consumption_rate); priority_mark12.full = a.full + b.full; } else { - a.full = rfixed_mul(wm1.worst_case_latency, + a.full = dfixed_mul(wm1.worst_case_latency, wm1.consumption_rate); - b.full = rfixed_const(16 * 1000); - priority_mark12.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark12.full = dfixed_div(a, b); } if (wm0.priority_mark.full > priority_mark02.full) priority_mark02.full = wm0.priority_mark.full; - if (rfixed_trunc(priority_mark02) < 0) + if (dfixed_trunc(priority_mark02) < 0) priority_mark02.full = 0; if (wm0.priority_mark_max.full > priority_mark02.full) priority_mark02.full = wm0.priority_mark_max.full; if (wm1.priority_mark.full > priority_mark12.full) priority_mark12.full = wm1.priority_mark.full; - if (rfixed_trunc(priority_mark12) < 0) + if (dfixed_trunc(priority_mark12) < 0) priority_mark12.full = 0; if (wm1.priority_mark_max.full > priority_mark12.full) priority_mark12.full = wm1.priority_mark_max.full; - d1mode_priority_a_cnt = rfixed_trunc(priority_mark02); - d2mode_priority_a_cnt = rfixed_trunc(priority_mark12); + d1mode_priority_a_cnt = dfixed_trunc(priority_mark02); + d2mode_priority_a_cnt = dfixed_trunc(priority_mark12); if (rdev->disp_priority == 2) { d1mode_priority_a_cnt |= MODE_PRIORITY_ALWAYS_ON; d2mode_priority_a_cnt |= MODE_PRIORITY_ALWAYS_ON; @@ -1096,32 +1004,32 @@ void rv515_bandwidth_avivo_update(struct radeon_device *rdev) WREG32(D2MODE_PRIORITY_A_CNT, d2mode_priority_a_cnt); WREG32(D2MODE_PRIORITY_B_CNT, d2mode_priority_a_cnt); } else if (mode0) { - if (rfixed_trunc(wm0.dbpp) > 64) - a.full = rfixed_div(wm0.dbpp, wm0.num_line_pair); + if (dfixed_trunc(wm0.dbpp) > 64) + a.full = dfixed_div(wm0.dbpp, wm0.num_line_pair); else a.full = wm0.num_line_pair.full; - fill_rate.full = rfixed_div(wm0.sclk, a); + fill_rate.full = dfixed_div(wm0.sclk, a); if (wm0.consumption_rate.full > fill_rate.full) { b.full = wm0.consumption_rate.full - fill_rate.full; - b.full = rfixed_mul(b, wm0.active_time); - a.full = rfixed_const(16); - b.full = rfixed_div(b, a); - a.full = rfixed_mul(wm0.worst_case_latency, + b.full = dfixed_mul(b, wm0.active_time); + a.full = dfixed_const(16); + b.full = dfixed_div(b, a); + a.full = dfixed_mul(wm0.worst_case_latency, wm0.consumption_rate); priority_mark02.full = a.full + b.full; } else { - a.full = rfixed_mul(wm0.worst_case_latency, + a.full = dfixed_mul(wm0.worst_case_latency, wm0.consumption_rate); - b.full = rfixed_const(16); - priority_mark02.full = rfixed_div(a, b); + b.full = dfixed_const(16); + priority_mark02.full = dfixed_div(a, b); } if (wm0.priority_mark.full > priority_mark02.full) priority_mark02.full = wm0.priority_mark.full; - if (rfixed_trunc(priority_mark02) < 0) + if (dfixed_trunc(priority_mark02) < 0) priority_mark02.full = 0; if (wm0.priority_mark_max.full > priority_mark02.full) priority_mark02.full = wm0.priority_mark_max.full; - d1mode_priority_a_cnt = rfixed_trunc(priority_mark02); + d1mode_priority_a_cnt = dfixed_trunc(priority_mark02); if (rdev->disp_priority == 2) d1mode_priority_a_cnt |= MODE_PRIORITY_ALWAYS_ON; WREG32(D1MODE_PRIORITY_A_CNT, d1mode_priority_a_cnt); @@ -1129,32 +1037,32 @@ void rv515_bandwidth_avivo_update(struct radeon_device *rdev) WREG32(D2MODE_PRIORITY_A_CNT, MODE_PRIORITY_OFF); WREG32(D2MODE_PRIORITY_B_CNT, MODE_PRIORITY_OFF); } else { - if (rfixed_trunc(wm1.dbpp) > 64) - a.full = rfixed_div(wm1.dbpp, wm1.num_line_pair); + if (dfixed_trunc(wm1.dbpp) > 64) + a.full = dfixed_div(wm1.dbpp, wm1.num_line_pair); else a.full = wm1.num_line_pair.full; - fill_rate.full = rfixed_div(wm1.sclk, a); + fill_rate.full = dfixed_div(wm1.sclk, a); if (wm1.consumption_rate.full > fill_rate.full) { b.full = wm1.consumption_rate.full - fill_rate.full; - b.full = rfixed_mul(b, wm1.active_time); - a.full = rfixed_const(16); - b.full = rfixed_div(b, a); - a.full = rfixed_mul(wm1.worst_case_latency, + b.full = dfixed_mul(b, wm1.active_time); + a.full = dfixed_const(16); + b.full = dfixed_div(b, a); + a.full = dfixed_mul(wm1.worst_case_latency, wm1.consumption_rate); priority_mark12.full = a.full + b.full; } else { - a.full = rfixed_mul(wm1.worst_case_latency, + a.full = dfixed_mul(wm1.worst_case_latency, wm1.consumption_rate); - b.full = rfixed_const(16 * 1000); - priority_mark12.full = rfixed_div(a, b); + b.full = dfixed_const(16 * 1000); + priority_mark12.full = dfixed_div(a, b); } if (wm1.priority_mark.full > priority_mark12.full) priority_mark12.full = wm1.priority_mark.full; - if (rfixed_trunc(priority_mark12) < 0) + if (dfixed_trunc(priority_mark12) < 0) priority_mark12.full = 0; if (wm1.priority_mark_max.full > priority_mark12.full) priority_mark12.full = wm1.priority_mark_max.full; - d2mode_priority_a_cnt = rfixed_trunc(priority_mark12); + d2mode_priority_a_cnt = dfixed_trunc(priority_mark12); if (rdev->disp_priority == 2) d2mode_priority_a_cnt |= MODE_PRIORITY_ALWAYS_ON; WREG32(D1MODE_PRIORITY_A_CNT, MODE_PRIORITY_OFF); diff --git a/drivers/gpu/drm/radeon/rv515d.h b/drivers/gpu/drm/radeon/rv515d.h index fc216e49384d..590309a710b1 100644 --- a/drivers/gpu/drm/radeon/rv515d.h +++ b/drivers/gpu/drm/radeon/rv515d.h @@ -217,6 +217,52 @@ #define CP_PACKET3_GET_OPCODE(h) (((h) >> 8) & 0xFF) /* Registers */ +#define R_0000F0_RBBM_SOFT_RESET 0x0000F0 +#define S_0000F0_SOFT_RESET_CP(x) (((x) & 0x1) << 0) +#define G_0000F0_SOFT_RESET_CP(x) (((x) >> 0) & 0x1) +#define C_0000F0_SOFT_RESET_CP 0xFFFFFFFE +#define S_0000F0_SOFT_RESET_HI(x) (((x) & 0x1) << 1) +#define G_0000F0_SOFT_RESET_HI(x) (((x) >> 1) & 0x1) +#define C_0000F0_SOFT_RESET_HI 0xFFFFFFFD +#define S_0000F0_SOFT_RESET_VAP(x) (((x) & 0x1) << 2) +#define G_0000F0_SOFT_RESET_VAP(x) (((x) >> 2) & 0x1) +#define C_0000F0_SOFT_RESET_VAP 0xFFFFFFFB +#define S_0000F0_SOFT_RESET_RE(x) (((x) & 0x1) << 3) +#define G_0000F0_SOFT_RESET_RE(x) (((x) >> 3) & 0x1) +#define C_0000F0_SOFT_RESET_RE 0xFFFFFFF7 +#define S_0000F0_SOFT_RESET_PP(x) (((x) & 0x1) << 4) +#define G_0000F0_SOFT_RESET_PP(x) (((x) >> 4) & 0x1) +#define C_0000F0_SOFT_RESET_PP 0xFFFFFFEF +#define S_0000F0_SOFT_RESET_E2(x) (((x) & 0x1) << 5) +#define G_0000F0_SOFT_RESET_E2(x) (((x) >> 5) & 0x1) +#define C_0000F0_SOFT_RESET_E2 0xFFFFFFDF +#define S_0000F0_SOFT_RESET_RB(x) (((x) & 0x1) << 6) +#define G_0000F0_SOFT_RESET_RB(x) (((x) >> 6) & 0x1) +#define C_0000F0_SOFT_RESET_RB 0xFFFFFFBF +#define S_0000F0_SOFT_RESET_HDP(x) (((x) & 0x1) << 7) +#define G_0000F0_SOFT_RESET_HDP(x) (((x) >> 7) & 0x1) +#define C_0000F0_SOFT_RESET_HDP 0xFFFFFF7F +#define S_0000F0_SOFT_RESET_MC(x) (((x) & 0x1) << 8) +#define G_0000F0_SOFT_RESET_MC(x) (((x) >> 8) & 0x1) +#define C_0000F0_SOFT_RESET_MC 0xFFFFFEFF +#define S_0000F0_SOFT_RESET_AIC(x) (((x) & 0x1) << 9) +#define G_0000F0_SOFT_RESET_AIC(x) (((x) >> 9) & 0x1) +#define C_0000F0_SOFT_RESET_AIC 0xFFFFFDFF +#define S_0000F0_SOFT_RESET_VIP(x) (((x) & 0x1) << 10) +#define G_0000F0_SOFT_RESET_VIP(x) (((x) >> 10) & 0x1) +#define C_0000F0_SOFT_RESET_VIP 0xFFFFFBFF +#define S_0000F0_SOFT_RESET_DISP(x) (((x) & 0x1) << 11) +#define G_0000F0_SOFT_RESET_DISP(x) (((x) >> 11) & 0x1) +#define C_0000F0_SOFT_RESET_DISP 0xFFFFF7FF +#define S_0000F0_SOFT_RESET_CG(x) (((x) & 0x1) << 12) +#define G_0000F0_SOFT_RESET_CG(x) (((x) >> 12) & 0x1) +#define C_0000F0_SOFT_RESET_CG 0xFFFFEFFF +#define S_0000F0_SOFT_RESET_GA(x) (((x) & 0x1) << 13) +#define G_0000F0_SOFT_RESET_GA(x) (((x) >> 13) & 0x1) +#define C_0000F0_SOFT_RESET_GA 0xFFFFDFFF +#define S_0000F0_SOFT_RESET_IDCT(x) (((x) & 0x1) << 14) +#define G_0000F0_SOFT_RESET_IDCT(x) (((x) >> 14) & 0x1) +#define C_0000F0_SOFT_RESET_IDCT 0xFFFFBFFF #define R_0000F8_CONFIG_MEMSIZE 0x0000F8 #define S_0000F8_CONFIG_MEMSIZE(x) (((x) & 0xFFFFFFFF) << 0) #define G_0000F8_CONFIG_MEMSIZE(x) (((x) >> 0) & 0xFFFFFFFF) diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c index 97958a64df1a..b7fd82064922 100644 --- a/drivers/gpu/drm/radeon/rv770.c +++ b/drivers/gpu/drm/radeon/rv770.c @@ -42,6 +42,21 @@ static void rv770_gpu_init(struct radeon_device *rdev); void rv770_fini(struct radeon_device *rdev); +void rv770_pm_misc(struct radeon_device *rdev) +{ + int req_ps_idx = rdev->pm.requested_power_state_index; + int req_cm_idx = rdev->pm.requested_clock_mode_index; + struct radeon_power_state *ps = &rdev->pm.power_state[req_ps_idx]; + struct radeon_voltage *voltage = &ps->clock_info[req_cm_idx].voltage; + + if ((voltage->type == VOLTAGE_SW) && voltage->voltage) { + if (voltage->voltage != rdev->pm.current_vddc) { + radeon_atom_set_voltage(rdev, voltage->voltage); + rdev->pm.current_vddc = voltage->voltage; + DRM_DEBUG("Setting: v: %d\n", voltage->voltage); + } + } +} /* * GART @@ -209,7 +224,7 @@ static void rv770_mc_program(struct radeon_device *rdev) WREG32(MC_VM_FB_LOCATION, tmp); WREG32(HDP_NONSURFACE_BASE, (rdev->mc.vram_start >> 8)); WREG32(HDP_NONSURFACE_INFO, (2 << 7)); - WREG32(HDP_NONSURFACE_SIZE, (rdev->mc.mc_vram_size - 1) | 0x3FF); + WREG32(HDP_NONSURFACE_SIZE, 0x3FFFFFFF); if (rdev->flags & RADEON_IS_AGP) { WREG32(MC_VM_AGP_TOP, rdev->mc.gtt_end >> 16); WREG32(MC_VM_AGP_BOT, rdev->mc.gtt_start >> 16); @@ -237,7 +252,6 @@ void r700_cp_stop(struct radeon_device *rdev) WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT)); } - static int rv770_cp_load_microcode(struct radeon_device *rdev) { const __be32 *fw_data; @@ -272,6 +286,11 @@ static int rv770_cp_load_microcode(struct radeon_device *rdev) return 0; } +void r700_cp_fini(struct radeon_device *rdev) +{ + r700_cp_stop(rdev); + radeon_ring_fini(rdev); +} /* * Core functions @@ -906,23 +925,12 @@ int rv770_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.visible_vram_size = rdev->mc.aper_size; - /* FIXME remove this once we support unmappable VRAM */ - if (rdev->mc.mc_vram_size > rdev->mc.aper_size) { - rdev->mc.mc_vram_size = rdev->mc.aper_size; - rdev->mc.real_vram_size = rdev->mc.aper_size; - } r600_vram_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); return 0; } -int rv770_gpu_reset(struct radeon_device *rdev) -{ - /* FIXME: implement any rv770 specific bits */ - return r600_gpu_reset(rdev); -} - static int rv770_startup(struct radeon_device *rdev) { int r; @@ -1094,8 +1102,6 @@ int rv770_init(struct radeon_device *rdev) r = radeon_clocks_init(rdev); if (r) return r; - /* Initialize power management */ - radeon_pm_init(rdev); /* Fence driver */ r = radeon_fence_driver_init(rdev); if (r) @@ -1132,7 +1138,7 @@ int rv770_init(struct radeon_device *rdev) r = rv770_startup(rdev); if (r) { dev_err(rdev->dev, "disabling GPU acceleration\n"); - r600_cp_fini(rdev); + r700_cp_fini(rdev); r600_wb_fini(rdev); r600_irq_fini(rdev); radeon_irq_kms_fini(rdev); @@ -1164,9 +1170,8 @@ int rv770_init(struct radeon_device *rdev) void rv770_fini(struct radeon_device *rdev) { - radeon_pm_fini(rdev); r600_blit_fini(rdev); - r600_cp_fini(rdev); + r700_cp_fini(rdev); r600_wb_fini(rdev); r600_irq_fini(rdev); radeon_irq_kms_fini(rdev); |