Age | Commit message (Collapse) | Author |
|
Apply proper barriers to suspend/resume operations.
The dmb() in suspend is required to prevent a race
between the next ENTRY_ADDR write and the previous
ENTRY_DATA read. The dsb+outer_sync() in resume()
is needed to ensure that the writes and reads occur
as expected.
Change-Id: I2e66ecce6ceb4f4647bde82cb706b739affeb858
Reviewed-on: http://git-master/r/1291
Reviewed-by: Gary King <gking@nvidia.com>
Reviewed-by: Mayuresh Kulkarni <mkulkarni@nvidia.com>
Tested-by: Mayuresh Kulkarni <mkulkarni@nvidia.com>
|
|
Enabled USB-VBUS detection through pmu on whistler for turning off USB
power rail to reduce the USB power when there is no cable connection.
Bug 667912: AVDD_USB_Power is consuming 3.82mW of power in OSIdle and ULP
audio playback case.
Tested on Android/whistler and USB power goes to 0mW.
Change-Id: Ia4bdc327b6d1b86921c296cadc48c3a4a2c35e73
Reviewed-on: http://git-master/r/1245
Tested-by: Dara Ramesh <dramesh@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
|
|
Added functionality to detect the VBUS and turn on/off the USB power rail.
Bug 667912: AVDD_USB_Power is consuming 3.82mW of power in OSIdle and ULP
audio playback case.
Change-Id: Ia64e5a04eff16c1ebec2afe1c7bc1e8f72f9da30
Reviewed-on: http://git-master/r/1224
Reviewed-by: Aleksandr Frid <afrid@nvidia.com>
Tested-by: Dara Ramesh <dramesh@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
|
|
Removed helper thread and replaced it with Worker Queues in udc and ehci
drivers to handle usbphy power up/down sequence. Made changes to turn off
usb power rail based on vbus detection mechanism is selected as PMU.
Fixed usb host LP0 exit sequence.
Bug 667912: AVDD_USB_Power is consuming 3.82mW of power in OSIdle and ULP
audio playback case.
Change-Id: I3a77d0ecb4f0b81dafe705100451c42641f0bfb9
Reviewed-on: http://git-master/r/1221
Tested-by: Hanumanth Venkateswa Moganty <vmoganty@nvidia.com>
Tested-by: Dara Ramesh <dramesh@nvidia.com>
Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
Tested-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
|
|
Checking ODM Bosch accelerometer driver for BMA150 part.
This is not a complete driver. The interrupts from BMA150
are not reliably received. The polling mode works fine.
This driver need to be updated for changing config settings
through sysfs interface.
Change-Id: I1492c5661509ba23c3967ed3ea1908bfb4336948
Reviewed-on: http://git-master/r/1180
Reviewed-by: Gary King <gking@nvidia.com>
Reviewed-by: Hoang Pham <hopham@nvidia.com>
Tested-by: Hoang Pham <hopham@nvidia.com>
|
|
Correcting the spelling error in enum definition TEGRA_DMA_REQ_ERROR_ABOTRED
to TEGRA_DMA_REQ_ERROR_ABORTED.
Tested on harmony.
Change-Id: Ie28aea9f87c1a0a2d3f27ce6132fb3eb48eb59e6
Reviewed-on: http://git-master/r/1225
Reviewed-by: Ramachandrudu Kandhala <rkandhala@nvidia.com>
Reviewed-by: Venkata (Muni) Anda <vanda@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
Tested-by: Ramachandrudu Kandhala <rkandhala@nvidia.com>
|
|
Default value of 2MB for DMA allocations is not large enough
to cover all the DMA allocation use cases of Tegra drivers.
4MB is a safe and working value.
Change-Id: I1f1e9db64ece21e9eef94c909a36cb1a2f50ed68
Reviewed-on: http://git-master/r/1247
Reviewed-by: Venkata (Muni) Anda <vanda@nvidia.com>
Tested-by: Venkata (Muni) Anda <vanda@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
|
|
This change should help in finding which drivers are sending busy hints at run time.
Bug 475616: All DFS clients should specify client tags while registering with DFS
Change-Id: Ic46cebba485dfc1e1b546276658bbcf04fe25d6f
Reviewed-on: http://git-master/r/1205
Reviewed-by: Sharad Ranjan <shranjan@nvidia.com>
Tested-by: Sharad Ranjan <shranjan@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Aleksandr Frid <afrid@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
|
|
When this code was written, we were not aware of how to turn on harmony's
display using a single key press. Also, after resume if none of the wake-locks
is acquired then system immediately goes to suspend.
To avoid this, we added this WAR to acquire a timed wake lock of 10 sec.
This gives the user a chance to either press a
key/insert-remove usb cable/insert-remove SD card after
which the display turns on.
Now that we have the way to turn on display by any press, this WAR is not
needed.
Tested on android + harmony (A02, R04 EC firmware) for LP0 and LP1.
For bug 680652 - [SW] Harmony specific suspend_ops implementation
Change-Id: I26492f01ed8cf07c5cf448c0e1f67dc99251aaf4
Reviewed-on: http://git-master/r/1236
Reviewed-by: Mayuresh Kulkarni <mkulkarni@nvidia.com>
Tested-by: Mayuresh Kulkarni <mkulkarni@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
|
|
- Tuned CPU non real-time starvation (NRT) parameters to increase
DVFS responsiveness: reduced boost degradation rate to 12%/sample
(from 50%/sample); increased initial boost step to 10MHz.
- Updated low corner (LC) detection: dropped activity margin from
corner hit requirements; cut off NRT degradation tail (this should
keep LP2 duty cycle unchanged despite NRT boost increase)
Change-Id: I747e7342cb25d8d103577bd2ad01c92446340f42
Reviewed-on: http://git-master/r/1218
Tested-by: Aleksandr Frid <afrid@nvidia.com>
Reviewed-by: Sharad Ranjan <shranjan@nvidia.com>
Tested-by: Sharad Ranjan <shranjan@nvidia.com>
Reviewed-by: Niranjan Wartikar <nwartikar@nvidia.com>
Tested-by: Niranjan Wartikar <nwartikar@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
|
|
For Bug 679393 to enable selective suspend/resume of USB.
Change-Id: I4013f9a3a3a7779f7537b75b419bdb81fcee3878
Reviewed-on: http://git-master/r/1231
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
Tested-by: Yu-Huan Hsu <yhsu@nvidia.com>
|
|
The CPU clock was previously boosted to 300 MHz for USB device on cable
connect for performance optimization. Following issues were noted with
this setting:
1. it is above low limit (250MHz) used to turn off slave CPU. Hence it
will never be turned off when cable is connected.
2. it is also above PLLP 216MHz output. Hence high power PLLX will be
kept alive.
Reducing the CPU boost frequency to 200 MHz to address the above issues
with minimal impact on the performance.
Bug 654486: [whistler/android] - Large difference in USB MSD perf with
DFS OFF.
Change-Id: Ia57db137b493dd8fe8436ede7942df0661dad67b
Reviewed-on: http://git-master/r/1228
Tested-by: Abhishek Aggarwal <aaggarwal@nvidia.com>
Reviewed-by: Aleksandr Frid <afrid@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
|
|
Change-Id: I883ea945f7bac85d6b02530246fc2e0f23937dbe
Reviewed-on: http://git-master/r/1219
Reviewed-by: Antti Hatala <ahatala@nvidia.com>
Tested-by: Antti Hatala <ahatala@nvidia.com>
|
|
lid close code was partially checked in. This is causing the screen
orientation to change. Disabling the lid support temporarily.
Bug 679243 : Android Harmony LVDS rotated into portrait
Bug 635301 : [Android\ Whistler \ Harmony] Power off/ Lid (close/open)
Button support in Android
Tested on : harmony boot orientation restored.
Change-Id: I9a5807af5860d2d9c602e9cc3b5ea2216063fa5f
Reviewed-on: http://git-master/r/1210
Reviewed-by: Mayuresh Kulkarni <mkulkarni@nvidia.com>
Tested-by: Mayuresh Kulkarni <mkulkarni@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
|
|
Change-Id: I274c8b80a5734e2b017e8fa9c85e994f435ed0ad
Reviewed-on: http://git-master/r/1208
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
|
|
... to match expected config name from new korg l2x0 code
Change-Id: Ia54867363b2087e8c68784d1a5553bb11a18113e
Reviewed-on: http://git-master/r/1207
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
|
|
Whistler Android: enable PPP, OUTER_CACHE_SYNC, UINPUT
Whistler LDK: enable OTG_UTILS, OUTER_CACHE_SYNC
Harmony Android: enable OUTER_CACHE_SYNC, UINPUT
Harmony LDK: enable OUTER_CACHE_SYNC, USB_ETH, USB_ETH_RNDIS
bug 635519 (UINPUT)
bug 636848 (USB OTG)
bug 673802 (OUTER_CACHE_SYNC)
Change-Id: I35531fc1fbb440fdf5173e580cca8b51ea3c6423
Reviewed-on: http://git-master/r/1206
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
|
|
Get rid of some unneeded variables, indented and cleaned things up.
Change-Id: I06bf8df81de7f1e68f7175f5f507836a2a88a608
Reviewed-on: http://git-master/r/1145
Reviewed-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Tested-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
The compiler was changing the literal ldr instructions
to mov instructions, as a result of which things worked
normally. Issues were seen when the compiler stopped automatically
optimizing these instructions.
Change-Id: I58f1bb018e6aa7a67e1a9ba177182e160dc64e9a
Reviewed-on: http://git-master/r/1203
Reviewed-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Tested-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Change-Id: I35cce6b67b2ad0611c742c31ac8e4f6c5b271e3d
Reviewed-on: http://git-master/r/1182
Reviewed-by: Acorn Pooley <apooley@nvidia.com>
Tested-by: Acorn Pooley <apooley@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Summary of changes to AP20 USB PHY:
1. UTMIP_XCVR_CFG0:
a. UTMIP_XCVR_SETUP field is being set to 0x9. Its programming through
fuses is disabled.
b. UTMIP_XCVR_HSSLEW_MSB field is programmed to 0 for non-host mode.
Its default value on reset is 1.
2. UTMIP_HSRX_CFG1: UTMIP_HS_SYNC_START_DLY field is now set to 0.
Previously it was 9.
3. UTMIP_TX_CFG0: Removed programming of UTMIP_FS_PREAMBLE_J field to 1.
It should be 0 which is its default value.
4. UTMIP_SPARE_CFG0: Removed programming of this reg since now it is not
required to read fuses to obtain value for UTMIP_XCVR_SETUP field.
5. UTMIP_XCVR_CFG1: Programming UTMIP_XCVR_TERM_RANGE_ADJ field to 0x6.
Bug 672008: AP20: USB3_UTMIP electrical parameters cause physical-level
failures
Change-Id: Idadff030d49584b25678179b876ab200f670af21
Reviewed-on: http://git-master/r/1198
Tested-by: Abhishek Aggarwal <aaggarwal@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
The GMC group is used by UARTD in the harmony and by power on reset, this
group's pullupdown is set to normal. So during kernel boot it throws some
junk character in the serial terminal.
Setting this group pullupdown to pull up resolve this issue.
Change-Id: I28868cd87c88d542c5b29016b7c27ae29cfbc189
Reviewed-on: http://git-master/r/1159
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
|
|
Bug 676157: Kernel builds in 9.12.8, 9.12.9 cause section mismatch warnings
Change-Id: Ic288f077d24b03f9d6af2d20dcb01a1bfce25599
Reviewed-on: http://git-master/r/1195
Reviewed-by: Mayuresh Kulkarni <mkulkarni@nvidia.com>
Tested-by: Mayuresh Kulkarni <mkulkarni@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Checking in odm panjit touch screen driver. This is required for
new harmony tablet board. This is disabled by default as of now.
Also checking updated tegra-gpio.c to have the correct gpio lies
info for harmony tablet. This is also disabled by default.
Added the acceleremeter properties in the odm kit. The driver
for accelerometer will be checked in later.
updated copyright headers for panjit.
Change-Id: I7089ddb11670670dc0aa683dce190bc619fb07e2
Reviewed-on: http://git-master/r/1139
Reviewed-by: Gary King <gking@nvidia.com>
Reviewed-by: Krishna Reddy <vdumpa@nvidia.com>
Tested-by: Krishna Reddy <vdumpa@nvidia.com>
|
|
Following improvements are done in the uart rx and tx path:
- Separate locks for rx and tx paths.
- Implementing the continuous double buffering with same buffer for
uart rx.
- Deactivating the rts line only incase of if dma does not read the
fifo i.e. nonmultiple of 4.
- Handling the modem control signal chnage properly.
- Created the separate workqueue for the receive to handle dma thresold
callback as well as the rx uart interrupts.
Tested on harmony and whistler with uart testcase and BT on harmony.
we tested the simulatenous file transfer and the A2DP on BT-harmony
and did not observe any glitches.
Change-Id: Id451ca2f42e246b524a7c9e4bbf3aa832cd0d535
Reviewed-on: http://git-master/r/1172
Reviewed-by: Gary King <gking@nvidia.com>
Reviewed-by: Anantha Idapalapati <aidapalapati@nvidia.com>
Tested-by: Anantha Idapalapati <aidapalapati@nvidia.com>
|
|
Separated EMC DFS parameters for DDR2 and LPDDR2 so they can be tuned
independently. Kept all parameters for both SDRAM types unchanged with the
exception of LPDDR2 EMC activity margin - set at 100% (2x); this change is
a partial revert of commit I8467a52d that caused perf regression on LPDDR2
platforms (no perf regression is observed for DDR2).
Change-Id: I193732a376e0a477aef3faf2348b4187d8aca40f
Reviewed-on: http://git-master/r/1165
Tested-by: Aleksandr Frid <afrid@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Used EV_SW and SW_LID input events, and modified Android policy manager
to respect lid events as wakeups.
1) kernel changes - added driver registration using gpio_keys library
Bug 635301: [Android\Whistler\Harmony] Power off/ Lid (close/open) Button
support in Android
Tested on: harmony. lp1 suspend/wakeup using slider S4 right/left respectively.
Change-Id: I694117d64e7319f2efda8251aa8b708ea200e279
Reviewed-on: http://git-master/r/1125
Reviewed-by: Gary King <gking@nvidia.com>
Reviewed-by: Varun Wadekar <vwadekar@nvidia.com>
Tested-by: Mallikarjun Kasoju <mkasoju@nvidia.com>
|
|
Also disabling sd card detect as a wake source for harmony
Change-Id: If826e916847b44086c531b63938861a47718b7da
Reviewed-on: http://git-master/r/1183
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Upadted CPU clock control in LP2 state:
- Disabled PLLX on entry to LP2
- Forced CPU divider 1:1 setting on entry to LP2, and restore divider
on exit (speed up LP2 entry/exit)
- Removed PLLC and PLLM from wake source consideration (commonly these
PLLs are disabled in LP2 anyway, but using them in rare case when they
are available may create dangerous over-clocking condition)
Change-Id: Ied51ebee553766e66d6007e1149270e243df0543
Reviewed-on: http://git-master/r/1155
Tested-by: Aleksandr Frid <afrid@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Bug 673802 - [T20/Harmony/Warm boot] Warm boot stress test failed on Harmony.
Backport the changes related to explicit cpu_down and L2 cache shutdown
from experimental kernel branch to main line
Tested on: Harmony
Change-Id: I8615c7e5010ea7133941857cd400a9fb73c0bcc5
Reviewed-on: http://git-master/r/1151
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
|
|
Bug 651676, this is causing failure while entering system suspend when lp0 is lowest power state
Change-Id: I4337fc00571e21f9a6d39706b74ec62522cb6d46
Change-Id: I4337fc00571e21f9a6d39706b74ec62522cb6d46
Reviewed-on: http://git-master/r/1166
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
CPU power request line needs to be explicitly enabled so that PMUs with
separate request lines for CPU and core rails will shutdown the CPU power
rail as expected.
Change-Id: I95bc048b6e1df9029e900d33c4c0f5aa63c008a7
Reviewed-on: http://git-master/r/1156
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Venkata (Muni) Anda <vanda@nvidia.com>
|
|
Change-Id: I50cb153fd240818ca876dda53a58809740a272d2
Reviewed-on: http://git-master/r/1141
Tested-by: Aleksandr Frid <afrid@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Bug 671189, 663161
Tested on Nand + android.
Change-Id: I518d8a1b57a9766dc69a115b1189789f054c2d0e
Reviewed-on: http://git-master/r/1088
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
|
|
It seems that after writing SYSCLK_OE bit in APBDEV_PMC_CNTRL_0,
a delay needs to be added for stabilization.
Reviewed by Bharat. Tested on Harmony (A02, R04 EC firmware) for
multiple flash and reboot.
Bug 676490 - [T20/Harmony] Bootup failure on harmony-inconsistent (3/5 times)
Change-Id: I1f45a86aa23ac00b43f0103285cceda970ee5b39
Reviewed-on: http://git-master/r/1160
Reviewed-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
|
|
The interrupt flag in the 32KHz domain might not be getting cleared correctly.
Clear the kbc interrupt explicitly prior to LP0 to fix this.
Change-Id: I68c93cb3130fd123e28f1c697b9314b1bbcf4c6e
Reviewed-on: http://git-master/r/1154
Reviewed-by: Gary King <gking@nvidia.com>
Reviewed-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Tested-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
|
|
Bug 663285
Enable nvmap to reclaim unpinned I/O virtual memory after it has been
unpinned and re-use it for other objects.
Checking in on behalf of yhsu.
Change-Id: I27304b45a80ae0d2fcad9f126672fe6ff5ded2af
Reviewed-on: http://git-master/r/1144
Reviewed-by: Krishna Reddy <vdumpa@nvidia.com>
Tested-by: Krishna Reddy <vdumpa@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>
Tested-by: Yu-Huan Hsu <yhsu@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Change-Id: I6969716342df0b28db1167df9bca50a5c03ffb2f
Reviewed-on: http://git-master/r/1138
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
LP0 turns off core voltage. As a result, the L2 rams
need to be flushed prior to shutdown. The L2 cache
will be re-enabled after returning from LP0.
Change-Id: Ie6e69a3e5615393c5c6e7189b4d7e3559304e7f9
Reviewed-on: http://git-master/r/1036
Reviewed-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Tested-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
To enhance the USB MSD performance, the busy hints for cpu clock have
been enabled on cable connect and its frequency is being boosted to
300 MHz.
Bug 654486: [whistler/android] - Large difference in USB MSD perf with
DFS OFF
Change-Id: I4471665bd02b7b49b368aec4aac5f1b89038c309
Reviewed-on: http://git-master/r/1114
Tested-by: Abhishek Aggarwal <aaggarwal@nvidia.com>
Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
Tested-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Set the GPIO in a reasonable state before calling NvRmGpioConfigPins(Output) to
avoid 5us~1ms low-pulse, since the low-pulse causes some EC firmwares to crash.
reviewed by Artiste-Hsu
tested on Harmony
Change-Id: Ia30dfd10dfd16eec979f2a8e6ffc8286121a2eeb
Reviewed-on: http://git-master/r/1126
Tested-by: Varun Wadekar <vwadekar@nvidia.com>
Reviewed-by: Phillip Smith <psmith@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Change-Id: Ied68fde9cc9556088837ccaabfdd93c396777218
Reviewed-on: http://git-master/r/1130
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Aleksandr Frid <afrid@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
On Whistler enabled CONFIG_CGROUP_SCHED instead of CONFIG_USER_SCHED
to be in sync with Harmony kernel config. Enables cgroup file system.
Bug 669840
Change-Id: I4366b0ad0af5cb57fb7adc373347e081811d0590
Reviewed-on: http://git-master/r/1111
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
|
|
Added PMU low power state configuration to RM kernel suspend/resume.
Included PMU interrupt control, and core rail control on platforms
with combined cpu/core power request.
Restricted core power groups gating to LP1 entry (core domain is down
in LP0, anyway).
Change-Id: If37ddfd42bd861b2cbc31767775583bd13549da8
Reviewed-on: http://git-master/r/1086
Reviewed-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Tested-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Reviewed-by: Narendra Damahe <ndamahe@nvidia.com>
Tested-by: Narendra Damahe <ndamahe@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Updated EMC DVFS table on Whistler:
- made sure RFC settings are above 5 for all entries
- disabled DQS_PULLD for 150MHz entry and below
- lowered operational voltage for 18MHz and 27MHz entries from 1.0V to 0.95V
Change-Id: I682ed68bb5bfd4497a214065007159ab2fe042b7
Reviewed-on: http://git-master/r/1089
Reviewed-by: Sharad Ranjan <shranjan@nvidia.com>
Tested-by: Sharad Ranjan <shranjan@nvidia.com>
Tested-by: Aleksandr Frid <afrid@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Disabled NOR clock in RM ("orphan" boot clock). If/when NOR driver is
added it will re-enable it as necessary.
Change-Id: I5f9b15a5be86c594b6638c3110d9f1d8ba5db694
Reviewed-on: http://git-master/r/1116
Tested-by: Aleksandr Frid <afrid@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Following issue has been fixed:
- Blocking write was returning immediatly if data request is multiple of 4.
- Blocking write was not able to complete if data length is nonmultiple of 4
and more than 4.
- The close was taking too much time because proper timeout and fifo size was
not configured.
- Tx dma path optimized to fill more data to dma buffer if there is more
pending chars in the buffer.
- Tx path is fixed to properly signal the wakup event to tty layer.
- RTS flow control is not getting set from second open even cflag is
requested for that.
- Rx dma was not receiving the correct data after second open. The multiple
request was
getting queued for the receive path at the time of closing.
- Rx dma was started before uart controller is configured and it is creating
to misbehave the dma.
- Transfer count was not getting calculated in the dma driver.
Pending issue:
- Loosing the data id more than 32K of data was sennt is single shot. Debugging
this.
Tested on harmony with different testcase developed for testing the linux serial
driver.
Change-Id: I6ed9095dd6340d2b5e7ef036823d2e4e5a61abcc
Reviewed-on: http://git-master/r/1065
Tested-by: Suresh Mangipudi <smangipudi@nvidia.com>
Reviewed-by: Udaykumar Rameshchan Raval <uraval@nvidia.com>
Tested-by: Udaykumar Rameshchan Raval <uraval@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
patch: removing extra spaces
Change-Id: Icc7fa7d5744134c0846b99c77b90cd76f9f938fc
Reviewed-on: http://git-master/r/1068
Tested-by: Sharad Ranjan <shranjan@nvidia.com>
Reviewed-by: Aleksandr Frid <afrid@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
|
|
Adjusted LP2 time padding factor to reduce LP2 entry threshold back to 5ms
(commit ad159fa937e9d917737956d12d530b723cba2b13 set it to 12.5ms as a side
effect of adding power good delay to LP2 turnaround time)
Change-Id: I1f00803e5d043cc6e99aeb7b373c0bdbda56e0ed
Reviewed-on: http://git-master/r/1087
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
|
|
by the time that the slave processors' page table pointers are written
to memory, the L1 and L2 caches may already be enabled on the master,
so the writes need to be forcefully cleaned to memory to ensure that
the slave will be able to read it.
patch originally submitted to 2.6.34 by RMK
Change-Id: Id9952149ce17c693775ec2b21e9a0425465fb770
Reviewed-on: http://git-master/r/1067
Reviewed-by: Trivikram Kasivajhula <tkasivajhula@nvidia.com>
Reviewed-by: Scott Williams <scwilliams@nvidia.com>
Reviewed-by: Gary King <gking@nvidia.com>
Tested-by: Gary King <gking@nvidia.com>
|