diff options
Diffstat (limited to 'arch/mips')
41 files changed, 1735 insertions, 102 deletions
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index 6ecdc690f733..2811ecc1f3c7 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -46,6 +46,7 @@ config MIPS select HAVE_ARCH_TRACEHOOK select HAVE_ARCH_TRANSPARENT_HUGEPAGE if CPU_SUPPORTS_HUGEPAGES select HAVE_ASM_MODVERSIONS + select HAVE_CBPF_JIT if !64BIT && !CPU_MICROMIPS select HAVE_EBPF_JIT if 64BIT && !CPU_MICROMIPS && TARGET_ISA_REV >= 2 select HAVE_CONTEXT_TRACKING select HAVE_COPY_THREAD_TLS @@ -293,6 +294,9 @@ config BCM63XX select SYS_SUPPORTS_32BIT_KERNEL select SYS_SUPPORTS_BIG_ENDIAN select SYS_HAS_EARLY_PRINTK + select SYS_HAS_CPU_BMIPS32_3300 + select SYS_HAS_CPU_BMIPS4350 + select SYS_HAS_CPU_BMIPS4380 select SWAP_IO_SPACE select GPIOLIB select HAVE_CLK @@ -1392,6 +1396,7 @@ config CPU_LOONGSON3 select WEAK_REORDERING_BEYOND_LLSC select MIPS_PGD_C0_CONTEXT select MIPS_L1_CACHE_SHIFT_6 + select MIPS_FP_SUPPORT select GPIOLIB select SWIOTLB help @@ -3054,7 +3059,7 @@ config STACKTRACE_SUPPORT config PGTABLE_LEVELS int default 4 if PAGE_SIZE_4KB && MIPS_VA_BITS_48 - default 3 if 64BIT && !PAGE_SIZE_64KB + default 3 if 64BIT && (!PAGE_SIZE_64KB || MIPS_VA_BITS_48) default 2 config MIPS_AUTO_PFN_OFFSET diff --git a/arch/mips/Makefile b/arch/mips/Makefile index 5403a91ce098..9ff2c70763a0 100644 --- a/arch/mips/Makefile +++ b/arch/mips/Makefile @@ -320,7 +320,7 @@ KBUILD_LDFLAGS += -m $(ld-emul) ifdef CONFIG_MIPS CHECKFLAGS += $(shell $(CC) $(KBUILD_CFLAGS) -dM -E -x c /dev/null | \ - egrep -vw '__GNUC_(|MINOR_|PATCHLEVEL_)_' | \ + egrep -vw '__GNUC_(MINOR_|PATCHLEVEL_)?_' | \ sed -e "s/^\#define /-D'/" -e "s/ /'='/" -e "s/$$/'/" -e 's/\$$/&&/g') endif diff --git a/arch/mips/bcm63xx/clk.c b/arch/mips/bcm63xx/clk.c index 164115944a7f..dcfa0ea912fe 100644 --- a/arch/mips/bcm63xx/clk.c +++ b/arch/mips/bcm63xx/clk.c @@ -381,6 +381,18 @@ void clk_disable(struct clk *clk) EXPORT_SYMBOL(clk_disable); +struct clk *clk_get_parent(struct clk *clk) +{ + return NULL; +} +EXPORT_SYMBOL(clk_get_parent); + +int clk_set_parent(struct clk *clk, struct clk *parent) +{ + return 0; +} +EXPORT_SYMBOL(clk_set_parent); + unsigned long clk_get_rate(struct clk *clk) { if (!clk) diff --git a/arch/mips/boot/compressed/Makefile b/arch/mips/boot/compressed/Makefile index 378cbfb31ee7..1d6ebbc2a5d0 100644 --- a/arch/mips/boot/compressed/Makefile +++ b/arch/mips/boot/compressed/Makefile @@ -33,7 +33,7 @@ KBUILD_AFLAGS := $(KBUILD_AFLAGS) -D__ASSEMBLY__ \ KCOV_INSTRUMENT := n # decompressor objects (linked with vmlinuz) -vmlinuzobjs-y := $(obj)/head.o $(obj)/decompress.o $(obj)/string.o +vmlinuzobjs-y := $(obj)/head.o $(obj)/decompress.o $(obj)/string.o $(obj)/bswapsi.o ifdef CONFIG_DEBUG_ZBOOT vmlinuzobjs-$(CONFIG_DEBUG_ZBOOT) += $(obj)/dbg.o @@ -47,7 +47,7 @@ extra-y += uart-ath79.c $(obj)/uart-ath79.c: $(srctree)/arch/mips/ath79/early_printk.c $(call cmd,shipped) -vmlinuzobjs-$(CONFIG_KERNEL_XZ) += $(obj)/ashldi3.o $(obj)/bswapsi.o +vmlinuzobjs-$(CONFIG_KERNEL_XZ) += $(obj)/ashldi3.o extra-y += ashldi3.c $(obj)/ashldi3.c: $(obj)/%.c: $(srctree)/lib/%.c FORCE diff --git a/arch/mips/boot/compressed/decompress.c b/arch/mips/boot/compressed/decompress.c index a52e929381ea..281e922e5c65 100644 --- a/arch/mips/boot/compressed/decompress.c +++ b/arch/mips/boot/compressed/decompress.c @@ -7,6 +7,8 @@ * Author: Wu Zhangjin <wuzhangjin@gmail.com> */ +#define DISABLE_BRANCH_PROFILING + #include <linux/types.h> #include <linux/kernel.h> #include <linux/string.h> diff --git a/arch/mips/boot/compressed/string.c b/arch/mips/boot/compressed/string.c index 43beecc3587c..0b593b709228 100644 --- a/arch/mips/boot/compressed/string.c +++ b/arch/mips/boot/compressed/string.c @@ -5,6 +5,7 @@ * Very small subset of simple string routines */ +#include <linux/compiler_attributes.h> #include <linux/types.h> void *memcpy(void *dest, const void *src, size_t n) @@ -27,3 +28,19 @@ void *memset(void *s, int c, size_t n) ss[i] = c; return s; } + +void * __weak memmove(void *dest, const void *src, size_t n) +{ + unsigned int i; + const char *s = src; + char *d = dest; + + if ((uintptr_t)dest < (uintptr_t)src) { + for (i = 0; i < n; i++) + d[i] = s[i]; + } else { + for (i = n; i > 0; i--) + d[i - 1] = s[i - 1]; + } + return dest; +} diff --git a/arch/mips/cavium-octeon/octeon-platform.c b/arch/mips/cavium-octeon/octeon-platform.c index 51685f893eab..c214fe4e678b 100644 --- a/arch/mips/cavium-octeon/octeon-platform.c +++ b/arch/mips/cavium-octeon/octeon-platform.c @@ -328,6 +328,7 @@ static int __init octeon_ehci_device_init(void) pd->dev.platform_data = &octeon_ehci_pdata; octeon_ehci_hw_start(&pd->dev); + put_device(&pd->dev); return ret; } @@ -391,6 +392,7 @@ static int __init octeon_ohci_device_init(void) pd->dev.platform_data = &octeon_ohci_pdata; octeon_ohci_hw_start(&pd->dev); + put_device(&pd->dev); return ret; } diff --git a/arch/mips/cavium-octeon/octeon-usb.c b/arch/mips/cavium-octeon/octeon-usb.c index 4017398519cf..e092d86e6358 100644 --- a/arch/mips/cavium-octeon/octeon-usb.c +++ b/arch/mips/cavium-octeon/octeon-usb.c @@ -544,6 +544,7 @@ static int __init dwc3_octeon_device_init(void) devm_iounmap(&pdev->dev, base); devm_release_mem_region(&pdev->dev, res->start, resource_size(res)); + put_device(&pdev->dev); } } while (node != NULL); diff --git a/arch/mips/dec/int-handler.S b/arch/mips/dec/int-handler.S index a25ef822e725..5ed38e618019 100644 --- a/arch/mips/dec/int-handler.S +++ b/arch/mips/dec/int-handler.S @@ -131,7 +131,7 @@ */ mfc0 t0,CP0_CAUSE # get pending interrupts mfc0 t1,CP0_STATUS -#ifdef CONFIG_32BIT +#if defined(CONFIG_32BIT) && defined(CONFIG_MIPS_FP_SUPPORT) lw t2,cpu_fpu_mask #endif andi t0,ST0_IM # CAUSE.CE may be non-zero! @@ -139,7 +139,7 @@ beqz t0,spurious -#ifdef CONFIG_32BIT +#if defined(CONFIG_32BIT) && defined(CONFIG_MIPS_FP_SUPPORT) and t2,t0 bnez t2,fpu # handle FPU immediately #endif @@ -280,7 +280,7 @@ handle_it: j dec_irq_dispatch nop -#ifdef CONFIG_32BIT +#if defined(CONFIG_32BIT) && defined(CONFIG_MIPS_FP_SUPPORT) fpu: lw t0,fpu_kstat_irq nop diff --git a/arch/mips/dec/prom/Makefile b/arch/mips/dec/prom/Makefile index d95016016b42..2bad87551203 100644 --- a/arch/mips/dec/prom/Makefile +++ b/arch/mips/dec/prom/Makefile @@ -6,4 +6,4 @@ lib-y += init.o memory.o cmdline.o identify.o console.o -lib-$(CONFIG_32BIT) += locore.o +lib-$(CONFIG_CPU_R3000) += locore.o diff --git a/arch/mips/dec/setup.c b/arch/mips/dec/setup.c index 1fc8dffa8d1d..649b50ae5b1e 100644 --- a/arch/mips/dec/setup.c +++ b/arch/mips/dec/setup.c @@ -766,7 +766,8 @@ void __init arch_init_irq(void) dec_interrupt[DEC_IRQ_HALT] = -1; /* Register board interrupts: FPU and cascade. */ - if (dec_interrupt[DEC_IRQ_FPU] >= 0 && cpu_has_fpu) { + if (IS_ENABLED(CONFIG_MIPS_FP_SUPPORT) && + dec_interrupt[DEC_IRQ_FPU] >= 0 && cpu_has_fpu) { struct irq_desc *desc_fpu; int irq_fpu; diff --git a/arch/mips/generic/yamon-dt.c b/arch/mips/generic/yamon-dt.c index a3aa22c77cad..a07a5edbcda7 100644 --- a/arch/mips/generic/yamon-dt.c +++ b/arch/mips/generic/yamon-dt.c @@ -75,7 +75,7 @@ static unsigned int __init gen_fdt_mem_array( __init int yamon_dt_append_memory(void *fdt, const struct yamon_mem_region *regions) { - unsigned long phys_memsize, memsize; + unsigned long phys_memsize = 0, memsize; __be32 mem_array[2 * MAX_MEM_ARRAY_ENTRIES]; unsigned int mem_entries; int i, err, mem_off; diff --git a/arch/mips/include/asm/cmpxchg.h b/arch/mips/include/asm/cmpxchg.h index f6136871561d..9182ce828f54 100644 --- a/arch/mips/include/asm/cmpxchg.h +++ b/arch/mips/include/asm/cmpxchg.h @@ -239,6 +239,7 @@ static inline unsigned long __cmpxchg64(volatile void *ptr, " .set " MIPS_ISA_ARCH_LEVEL " \n" /* Load 64 bits from ptr */ "1: lld %L0, %3 # __cmpxchg64 \n" + " .set pop \n" /* * Split the 64 bit value we loaded into the 2 registers that hold the * ret variable. @@ -266,6 +267,8 @@ static inline unsigned long __cmpxchg64(volatile void *ptr, " or %L1, %L1, $at \n" " .set at \n" # endif + " .set push \n" + " .set " MIPS_ISA_ARCH_LEVEL " \n" /* Attempt to store new at ptr */ " scd %L1, %2 \n" /* If we failed, loop! */ diff --git a/arch/mips/include/asm/dec/prom.h b/arch/mips/include/asm/dec/prom.h index 62c7dfb90e06..1e1247add1cf 100644 --- a/arch/mips/include/asm/dec/prom.h +++ b/arch/mips/include/asm/dec/prom.h @@ -43,16 +43,11 @@ */ #define REX_PROM_MAGIC 0x30464354 -#ifdef CONFIG_64BIT - -#define prom_is_rex(magic) 1 /* KN04 and KN05 are REX PROMs. */ - -#else /* !CONFIG_64BIT */ - -#define prom_is_rex(magic) ((magic) == REX_PROM_MAGIC) - -#endif /* !CONFIG_64BIT */ - +/* KN04 and KN05 are REX PROMs, so only do the check for R3k systems. */ +static inline bool prom_is_rex(u32 magic) +{ + return !IS_ENABLED(CONFIG_CPU_R3000) || magic == REX_PROM_MAGIC; +} /* * 3MIN/MAXINE PROM entry points for DS5000/1xx's, DS5000/xx's and diff --git a/arch/mips/include/asm/highmem.h b/arch/mips/include/asm/highmem.h index 9d84aafc33d0..9a6bf4f011b5 100644 --- a/arch/mips/include/asm/highmem.h +++ b/arch/mips/include/asm/highmem.h @@ -36,7 +36,7 @@ extern pte_t *pkmap_page_table; * easily, subsequent pte tables have to be allocated in one physical * chunk of RAM. */ -#ifdef CONFIG_PHYS_ADDR_T_64BIT +#if defined(CONFIG_PHYS_ADDR_T_64BIT) || defined(CONFIG_MIPS_HUGE_TLB_SUPPORT) #define LAST_PKMAP 512 #else #define LAST_PKMAP 1024 diff --git a/arch/mips/include/asm/hugetlb.h b/arch/mips/include/asm/hugetlb.h index 425bb6fc3bda..bf1bf8c7c332 100644 --- a/arch/mips/include/asm/hugetlb.h +++ b/arch/mips/include/asm/hugetlb.h @@ -53,7 +53,13 @@ static inline pte_t huge_ptep_get_and_clear(struct mm_struct *mm, static inline void huge_ptep_clear_flush(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep) { - flush_tlb_page(vma, addr & huge_page_mask(hstate_vma(vma))); + /* + * clear the huge pte entry firstly, so that the other smp threads will + * not get old pte entry after finishing flush_tlb_page and before + * setting new huge pte entry + */ + huge_ptep_get_and_clear(vma->vm_mm, addr, ptep); + flush_tlb_page(vma, addr); } #define __HAVE_ARCH_HUGE_PTE_NONE diff --git a/arch/mips/include/asm/mips-cm.h b/arch/mips/include/asm/mips-cm.h index aeae2effa123..23c67c0871b1 100644 --- a/arch/mips/include/asm/mips-cm.h +++ b/arch/mips/include/asm/mips-cm.h @@ -11,6 +11,7 @@ #ifndef __MIPS_ASM_MIPS_CM_H__ #define __MIPS_ASM_MIPS_CM_H__ +#include <linux/bitfield.h> #include <linux/bitops.h> #include <linux/errno.h> @@ -153,8 +154,8 @@ GCR_ACCESSOR_RO(32, 0x030, rev) #define CM_GCR_REV_MINOR GENMASK(7, 0) #define CM_ENCODE_REV(major, minor) \ - (((major) << __ffs(CM_GCR_REV_MAJOR)) | \ - ((minor) << __ffs(CM_GCR_REV_MINOR))) + (FIELD_PREP(CM_GCR_REV_MAJOR, major) | \ + FIELD_PREP(CM_GCR_REV_MINOR, minor)) #define CM_REV_CM2 CM_ENCODE_REV(6, 0) #define CM_REV_CM2_5 CM_ENCODE_REV(7, 0) @@ -362,10 +363,10 @@ static inline int mips_cm_revision(void) static inline unsigned int mips_cm_max_vp_width(void) { extern int smp_num_siblings; - uint32_t cfg; if (mips_cm_revision() >= CM_REV_CM3) - return read_gcr_sys_config2() & CM_GCR_SYS_CONFIG2_MAXVPW; + return FIELD_GET(CM_GCR_SYS_CONFIG2_MAXVPW, + read_gcr_sys_config2()); if (mips_cm_present()) { /* @@ -373,8 +374,7 @@ static inline unsigned int mips_cm_max_vp_width(void) * number of VP(E)s, and if that ever changes then this will * need revisiting. */ - cfg = read_gcr_cl_config() & CM_GCR_Cx_CONFIG_PVPE; - return (cfg >> __ffs(CM_GCR_Cx_CONFIG_PVPE)) + 1; + return FIELD_GET(CM_GCR_Cx_CONFIG_PVPE, read_gcr_cl_config()) + 1; } if (IS_ENABLED(CONFIG_SMP)) diff --git a/arch/mips/include/asm/mipsregs.h b/arch/mips/include/asm/mipsregs.h index 3afdb39d092a..c28b892937fe 100644 --- a/arch/mips/include/asm/mipsregs.h +++ b/arch/mips/include/asm/mipsregs.h @@ -2007,7 +2007,7 @@ _ASM_MACRO_0(tlbginvf, _ASM_INSN_IF_MIPS(0x4200000c) ({ int __res; \ __asm__ __volatile__( \ ".set\tpush\n\t" \ - ".set\tmips32r2\n\t" \ + ".set\tmips32r5\n\t" \ _ASM_SET_VIRT \ "mfgc0\t%0, " #source ", %1\n\t" \ ".set\tpop" \ @@ -2020,7 +2020,7 @@ _ASM_MACRO_0(tlbginvf, _ASM_INSN_IF_MIPS(0x4200000c) ({ unsigned long long __res; \ __asm__ __volatile__( \ ".set\tpush\n\t" \ - ".set\tmips64r2\n\t" \ + ".set\tmips64r5\n\t" \ _ASM_SET_VIRT \ "dmfgc0\t%0, " #source ", %1\n\t" \ ".set\tpop" \ @@ -2033,7 +2033,7 @@ _ASM_MACRO_0(tlbginvf, _ASM_INSN_IF_MIPS(0x4200000c) do { \ __asm__ __volatile__( \ ".set\tpush\n\t" \ - ".set\tmips32r2\n\t" \ + ".set\tmips32r5\n\t" \ _ASM_SET_VIRT \ "mtgc0\t%z0, " #register ", %1\n\t" \ ".set\tpop" \ @@ -2045,7 +2045,7 @@ do { \ do { \ __asm__ __volatile__( \ ".set\tpush\n\t" \ - ".set\tmips64r2\n\t" \ + ".set\tmips64r5\n\t" \ _ASM_SET_VIRT \ "dmtgc0\t%z0, " #register ", %1\n\t" \ ".set\tpop" \ diff --git a/arch/mips/include/asm/octeon/cvmx-bootinfo.h b/arch/mips/include/asm/octeon/cvmx-bootinfo.h index 62787765575e..ce6e5fddce0b 100644 --- a/arch/mips/include/asm/octeon/cvmx-bootinfo.h +++ b/arch/mips/include/asm/octeon/cvmx-bootinfo.h @@ -315,7 +315,7 @@ enum cvmx_chip_types_enum { /* Functions to return string based on type */ #define ENUM_BRD_TYPE_CASE(x) \ - case x: return(#x + 16); /* Skip CVMX_BOARD_TYPE_ */ + case x: return (&#x[16]); /* Skip CVMX_BOARD_TYPE_ */ static inline const char *cvmx_board_type_to_string(enum cvmx_board_types_enum type) { @@ -404,7 +404,7 @@ static inline const char *cvmx_board_type_to_string(enum } #define ENUM_CHIP_TYPE_CASE(x) \ - case x: return(#x + 15); /* Skip CVMX_CHIP_TYPE */ + case x: return (&#x[15]); /* Skip CVMX_CHIP_TYPE */ static inline const char *cvmx_chip_type_to_string(enum cvmx_chip_types_enum type) { diff --git a/arch/mips/include/asm/setup.h b/arch/mips/include/asm/setup.h index bb36a400203d..8c56b862fd9c 100644 --- a/arch/mips/include/asm/setup.h +++ b/arch/mips/include/asm/setup.h @@ -16,7 +16,7 @@ static inline void setup_8250_early_printk_port(unsigned long base, unsigned int reg_shift, unsigned int timeout) {} #endif -extern void set_handler(unsigned long offset, void *addr, unsigned long len); +void set_handler(unsigned long offset, const void *addr, unsigned long len); extern void set_uncached_handler(unsigned long offset, void *addr, unsigned long len); typedef void (*vi_handler_t)(void); diff --git a/arch/mips/include/asm/timex.h b/arch/mips/include/asm/timex.h index b05bb70a2e46..8026baf46e72 100644 --- a/arch/mips/include/asm/timex.h +++ b/arch/mips/include/asm/timex.h @@ -40,9 +40,9 @@ typedef unsigned int cycles_t; /* - * On R4000/R4400 before version 5.0 an erratum exists such that if the - * cycle counter is read in the exact moment that it is matching the - * compare register, no interrupt will be generated. + * On R4000/R4400 an erratum exists such that if the cycle counter is + * read in the exact moment that it is matching the compare register, + * no interrupt will be generated. * * There is a suggested workaround and also the erratum can't strike if * the compare interrupt isn't being used as the clock source device. @@ -63,7 +63,7 @@ static inline int can_use_mips_counter(unsigned int prid) if (!__builtin_constant_p(cpu_has_counter)) asm volatile("" : "=m" (cpu_data[0].options)); if (likely(cpu_has_counter && - prid >= (PRID_IMP_R4000 | PRID_REV_ENCODE_44(5, 0)))) + prid > (PRID_IMP_R4000 | PRID_REV_ENCODE_44(15, 15)))) return 1; else return 0; diff --git a/arch/mips/include/asm/vdso/vdso.h b/arch/mips/include/asm/vdso/vdso.h index 737ddfc3411c..a327ca21270e 100644 --- a/arch/mips/include/asm/vdso/vdso.h +++ b/arch/mips/include/asm/vdso/vdso.h @@ -67,7 +67,7 @@ static inline const struct vdso_data *get_vdso_data(void) static inline void __iomem *get_gic(const struct vdso_data *data) { - return (void __iomem *)data - PAGE_SIZE; + return (void __iomem *)((unsigned long)data & PAGE_MASK) - PAGE_SIZE; } #endif /* CONFIG_CLKSRC_MIPS_GIC */ diff --git a/arch/mips/kernel/cacheinfo.c b/arch/mips/kernel/cacheinfo.c index 47312c529410..529dab855aac 100644 --- a/arch/mips/kernel/cacheinfo.c +++ b/arch/mips/kernel/cacheinfo.c @@ -17,7 +17,7 @@ do { \ leaf++; \ } while (0) -static int __init_cache_level(unsigned int cpu) +int init_cache_level(unsigned int cpu) { struct cpuinfo_mips *c = ¤t_cpu_data; struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu); @@ -69,7 +69,7 @@ static void fill_cpumask_cluster(int cpu, cpumask_t *cpu_map) cpumask_set_cpu(cpu1, cpu_map); } -static int __populate_cache_leaves(unsigned int cpu) +int populate_cache_leaves(unsigned int cpu) { struct cpuinfo_mips *c = ¤t_cpu_data; struct cpu_cacheinfo *this_cpu_ci = get_cpu_cacheinfo(cpu); @@ -98,6 +98,3 @@ static int __populate_cache_leaves(unsigned int cpu) return 0; } - -DEFINE_SMP_CALL_CACHE_FUNCTION(init_cache_level) -DEFINE_SMP_CALL_CACHE_FUNCTION(populate_cache_leaves) diff --git a/arch/mips/kernel/mips-cm.c b/arch/mips/kernel/mips-cm.c index a9eab83d9148..611ef512c0b8 100644 --- a/arch/mips/kernel/mips-cm.c +++ b/arch/mips/kernel/mips-cm.c @@ -179,8 +179,7 @@ static void mips_cm_probe_l2sync(void) phys_addr_t addr; /* L2-only sync was introduced with CM major revision 6 */ - major_rev = (read_gcr_rev() & CM_GCR_REV_MAJOR) >> - __ffs(CM_GCR_REV_MAJOR); + major_rev = FIELD_GET(CM_GCR_REV_MAJOR, read_gcr_rev()); if (major_rev < 6) return; @@ -263,13 +262,13 @@ void mips_cm_lock_other(unsigned int cluster, unsigned int core, preempt_disable(); if (cm_rev >= CM_REV_CM3) { - val = core << __ffs(CM3_GCR_Cx_OTHER_CORE); - val |= vp << __ffs(CM3_GCR_Cx_OTHER_VP); + val = FIELD_PREP(CM3_GCR_Cx_OTHER_CORE, core) | + FIELD_PREP(CM3_GCR_Cx_OTHER_VP, vp); if (cm_rev >= CM_REV_CM3_5) { val |= CM_GCR_Cx_OTHER_CLUSTER_EN; - val |= cluster << __ffs(CM_GCR_Cx_OTHER_CLUSTER); - val |= block << __ffs(CM_GCR_Cx_OTHER_BLOCK); + val |= FIELD_PREP(CM_GCR_Cx_OTHER_CLUSTER, cluster); + val |= FIELD_PREP(CM_GCR_Cx_OTHER_BLOCK, block); } else { WARN_ON(cluster != 0); WARN_ON(block != CM_GCR_Cx_OTHER_BLOCK_LOCAL); @@ -299,7 +298,7 @@ void mips_cm_lock_other(unsigned int cluster, unsigned int core, spin_lock_irqsave(&per_cpu(cm_core_lock, curr_core), per_cpu(cm_core_lock_flags, curr_core)); - val = core << __ffs(CM_GCR_Cx_OTHER_CORENUM); + val = FIELD_PREP(CM_GCR_Cx_OTHER_CORENUM, core); } write_gcr_cl_other(val); @@ -343,8 +342,8 @@ void mips_cm_error_report(void) cm_other = read_gcr_error_mult(); if (revision < CM_REV_CM3) { /* CM2 */ - cause = cm_error >> __ffs(CM_GCR_ERROR_CAUSE_ERRTYPE); - ocause = cm_other >> __ffs(CM_GCR_ERROR_MULT_ERR2ND); + cause = FIELD_GET(CM_GCR_ERROR_CAUSE_ERRTYPE, cm_error); + ocause = FIELD_GET(CM_GCR_ERROR_MULT_ERR2ND, cm_other); if (!cause) return; @@ -386,8 +385,8 @@ void mips_cm_error_report(void) ulong core_id_bits, vp_id_bits, cmd_bits, cmd_group_bits; ulong cm3_cca_bits, mcp_bits, cm3_tr_bits, sched_bit; - cause = cm_error >> __ffs64(CM3_GCR_ERROR_CAUSE_ERRTYPE); - ocause = cm_other >> __ffs(CM_GCR_ERROR_MULT_ERR2ND); + cause = FIELD_GET(CM3_GCR_ERROR_CAUSE_ERRTYPE, cm_error); + ocause = FIELD_GET(CM_GCR_ERROR_MULT_ERR2ND, cm_other); if (!cause) return; diff --git a/arch/mips/kernel/r2300_fpu.S b/arch/mips/kernel/r2300_fpu.S index 12e58053544f..cbf6db98cfb3 100644 --- a/arch/mips/kernel/r2300_fpu.S +++ b/arch/mips/kernel/r2300_fpu.S @@ -29,8 +29,8 @@ #define EX2(a,b) \ 9: a,##b; \ .section __ex_table,"a"; \ - PTR 9b,bad_stack; \ - PTR 9b+4,bad_stack; \ + PTR 9b,fault; \ + PTR 9b+4,fault; \ .previous .set mips1 diff --git a/arch/mips/kernel/smp.c b/arch/mips/kernel/smp.c index f510c00bda88..c563b03bdccc 100644 --- a/arch/mips/kernel/smp.c +++ b/arch/mips/kernel/smp.c @@ -361,6 +361,9 @@ asmlinkage void start_secondary(void) cpu = smp_processor_id(); cpu_data[cpu].udelay_val = loops_per_jiffy; + set_cpu_sibling_map(cpu); + set_cpu_core_map(cpu); + cpumask_set_cpu(cpu, &cpu_coherent_mask); notify_cpu_starting(cpu); @@ -372,9 +375,6 @@ asmlinkage void start_secondary(void) /* The CPU is running and counters synchronised, now mark it online */ set_cpu_online(cpu, true); - set_cpu_sibling_map(cpu); - set_cpu_core_map(cpu); - calculate_cpu_foreign_map(); /* diff --git a/arch/mips/kernel/syscall.c b/arch/mips/kernel/syscall.c index 3f16f3823031..9a28e79f2e66 100644 --- a/arch/mips/kernel/syscall.c +++ b/arch/mips/kernel/syscall.c @@ -239,12 +239,3 @@ SYSCALL_DEFINE3(cachectl, char *, addr, int, nbytes, int, op) { return -ENOSYS; } - -/* - * If we ever come here the user sp is bad. Zap the process right away. - * Due to the bad stack signaling wouldn't work. - */ -asmlinkage void bad_stack(void) -{ - do_exit(SIGSEGV); -} diff --git a/arch/mips/kernel/time.c b/arch/mips/kernel/time.c index caa01457dce6..ed339d7979f3 100644 --- a/arch/mips/kernel/time.c +++ b/arch/mips/kernel/time.c @@ -141,15 +141,10 @@ static __init int cpu_has_mfc0_count_bug(void) case CPU_R4400MC: /* * The published errata for the R4400 up to 3.0 say the CPU - * has the mfc0 from count bug. + * has the mfc0 from count bug. This seems the last version + * produced. */ - if ((current_cpu_data.processor_id & 0xff) <= 0x30) - return 1; - - /* - * we assume newer revisions are ok - */ - return 0; + return 1; } return 0; diff --git a/arch/mips/kernel/traps.c b/arch/mips/kernel/traps.c index 8282d0feb0b2..749089c25d5e 100644 --- a/arch/mips/kernel/traps.c +++ b/arch/mips/kernel/traps.c @@ -2020,19 +2020,19 @@ static void *set_vi_srs_handler(int n, vi_handler_t addr, int srs) * If no shadow set is selected then use the default handler * that does normal register saving and standard interrupt exit */ - extern char except_vec_vi, except_vec_vi_lui; - extern char except_vec_vi_ori, except_vec_vi_end; - extern char rollback_except_vec_vi; - char *vec_start = using_rollback_handler() ? - &rollback_except_vec_vi : &except_vec_vi; + extern const u8 except_vec_vi[], except_vec_vi_lui[]; + extern const u8 except_vec_vi_ori[], except_vec_vi_end[]; + extern const u8 rollback_except_vec_vi[]; + const u8 *vec_start = using_rollback_handler() ? + rollback_except_vec_vi : except_vec_vi; #if defined(CONFIG_CPU_MICROMIPS) || defined(CONFIG_CPU_BIG_ENDIAN) - const int lui_offset = &except_vec_vi_lui - vec_start + 2; - const int ori_offset = &except_vec_vi_ori - vec_start + 2; + const int lui_offset = except_vec_vi_lui - vec_start + 2; + const int ori_offset = except_vec_vi_ori - vec_start + 2; #else - const int lui_offset = &except_vec_vi_lui - vec_start; - const int ori_offset = &except_vec_vi_ori - vec_start; + const int lui_offset = except_vec_vi_lui - vec_start; + const int ori_offset = except_vec_vi_ori - vec_start; #endif - const int handler_len = &except_vec_vi_end - vec_start; + const int handler_len = except_vec_vi_end - vec_start; if (handler_len > VECTORSPACING) { /* @@ -2240,7 +2240,7 @@ void per_cpu_trap_init(bool is_boot_cpu) } /* Install CPU exception handler */ -void set_handler(unsigned long offset, void *addr, unsigned long size) +void set_handler(unsigned long offset, const void *addr, unsigned long size) { #ifdef CONFIG_CPU_MICROMIPS memcpy((void *)(ebase + offset), ((unsigned char *)addr - 1), size); diff --git a/arch/mips/lantiq/clk.c b/arch/mips/lantiq/clk.c index dd819e31fcbb..7a623684d9b5 100644 --- a/arch/mips/lantiq/clk.c +++ b/arch/mips/lantiq/clk.c @@ -158,6 +158,18 @@ void clk_deactivate(struct clk *clk) } EXPORT_SYMBOL(clk_deactivate); +struct clk *clk_get_parent(struct clk *clk) +{ + return NULL; +} +EXPORT_SYMBOL(clk_get_parent); + +int clk_set_parent(struct clk *clk, struct clk *parent) +{ + return 0; +} +EXPORT_SYMBOL(clk_set_parent); + static inline u32 get_counter_resolution(void) { u32 res; diff --git a/arch/mips/lantiq/xway/dma.c b/arch/mips/lantiq/xway/dma.c index aeb1b989cd4e..e45077aecf83 100644 --- a/arch/mips/lantiq/xway/dma.c +++ b/arch/mips/lantiq/xway/dma.c @@ -11,6 +11,7 @@ #include <linux/export.h> #include <linux/spinlock.h> #include <linux/clk.h> +#include <linux/delay.h> #include <linux/err.h> #include <lantiq_soc.h> @@ -29,6 +30,7 @@ #define LTQ_DMA_PCTRL 0x44 #define LTQ_DMA_IRNEN 0xf4 +#define DMA_ID_CHNR GENMASK(26, 20) /* channel number */ #define DMA_DESCPT BIT(3) /* descriptor complete irq */ #define DMA_TX BIT(8) /* TX channel direction */ #define DMA_CHAN_ON BIT(0) /* channel on / off bit */ @@ -39,7 +41,6 @@ #define DMA_POLL BIT(31) /* turn on channel polling */ #define DMA_CLK_DIV4 BIT(6) /* polling clock divider */ #define DMA_2W_BURST BIT(1) /* 2 word burst length */ -#define DMA_MAX_CHANNEL 20 /* the soc has 20 channels */ #define DMA_ETOP_ENDIANNESS (0xf << 8) /* endianness swap etop channels */ #define DMA_WEIGHT (BIT(17) | BIT(16)) /* default channel wheight */ @@ -205,7 +206,7 @@ ltq_dma_init(struct platform_device *pdev) { struct clk *clk; struct resource *res; - unsigned id; + unsigned int id, nchannels; int i; res = platform_get_resource(pdev, IORESOURCE_MEM, 0); @@ -221,21 +222,24 @@ ltq_dma_init(struct platform_device *pdev) clk_enable(clk); ltq_dma_w32_mask(0, DMA_RESET, LTQ_DMA_CTRL); + usleep_range(1, 10); + /* disable all interrupts */ ltq_dma_w32(0, LTQ_DMA_IRNEN); /* reset/configure each channel */ - for (i = 0; i < DMA_MAX_CHANNEL; i++) { + id = ltq_dma_r32(LTQ_DMA_ID); + nchannels = ((id & DMA_ID_CHNR) >> 20); + for (i = 0; i < nchannels; i++) { ltq_dma_w32(i, LTQ_DMA_CS); ltq_dma_w32(DMA_CHAN_RST, LTQ_DMA_CCTRL); ltq_dma_w32(DMA_POLL | DMA_CLK_DIV4, LTQ_DMA_CPOLL); ltq_dma_w32_mask(DMA_CHAN_ON, 0, LTQ_DMA_CCTRL); } - id = ltq_dma_r32(LTQ_DMA_ID); dev_info(&pdev->dev, "Init done - hw rev: %X, ports: %d, channels: %d\n", - id & 0x1f, (id >> 16) & 0xf, id >> 20); + id & 0x1f, (id >> 16) & 0xf, nchannels); return 0; } diff --git a/arch/mips/loongson64/loongson-3/numa.c b/arch/mips/loongson64/loongson-3/numa.c index 8f20d2cb3767..7e7376cc94b1 100644 --- a/arch/mips/loongson64/loongson-3/numa.c +++ b/arch/mips/loongson64/loongson-3/numa.c @@ -200,6 +200,9 @@ static void __init node_mem_init(unsigned int node) if (node_end_pfn(0) >= (0xffffffff >> PAGE_SHIFT)) memblock_reserve((node_addrspace_offset | 0xfe000000), 32 << 20); + + /* Reserve pfn range 0~node[0]->node_start_pfn */ + memblock_reserve(0, PAGE_SIZE * start_pfn); } } diff --git a/arch/mips/mti-malta/malta-dtshim.c b/arch/mips/mti-malta/malta-dtshim.c index 98a063093b69..0be28adff557 100644 --- a/arch/mips/mti-malta/malta-dtshim.c +++ b/arch/mips/mti-malta/malta-dtshim.c @@ -22,7 +22,7 @@ #define ROCIT_CONFIG_GEN1_MEMMAP_SHIFT 8 #define ROCIT_CONFIG_GEN1_MEMMAP_MASK (0xf << 8) -static unsigned char fdt_buf[16 << 10] __initdata; +static unsigned char fdt_buf[16 << 10] __initdata __aligned(8); /* determined physical memory size, not overridden by command line args */ extern unsigned long physical_memsize; diff --git a/arch/mips/mti-malta/malta-platform.c b/arch/mips/mti-malta/malta-platform.c index 11e9527c6e44..62ffac500eb5 100644 --- a/arch/mips/mti-malta/malta-platform.c +++ b/arch/mips/mti-malta/malta-platform.c @@ -47,7 +47,8 @@ static struct plat_serial8250_port uart8250_data[] = { .mapbase = 0x1f000900, /* The CBUS UART */ .irq = MIPS_CPU_IRQ_BASE + MIPSCPU_INT_MB2, .uartclk = 3686400, /* Twice the usual clk! */ - .iotype = UPIO_MEM32, + .iotype = IS_ENABLED(CONFIG_CPU_BIG_ENDIAN) ? + UPIO_MEM32BE : UPIO_MEM32, .flags = CBUS_UART_FLAGS, .regshift = 3, }, diff --git a/arch/mips/net/Makefile b/arch/mips/net/Makefile index 2d03af7d6b19..d55912349039 100644 --- a/arch/mips/net/Makefile +++ b/arch/mips/net/Makefile @@ -1,4 +1,5 @@ # SPDX-License-Identifier: GPL-2.0-only # MIPS networking code +obj-$(CONFIG_MIPS_CBPF_JIT) += bpf_jit.o bpf_jit_asm.o obj-$(CONFIG_MIPS_EBPF_JIT) += ebpf_jit.o diff --git a/arch/mips/net/bpf_jit.c b/arch/mips/net/bpf_jit.c new file mode 100644 index 000000000000..29a288ff4f18 --- /dev/null +++ b/arch/mips/net/bpf_jit.c @@ -0,0 +1,1299 @@ +/* + * Just-In-Time compiler for BPF filters on MIPS + * + * Copyright (c) 2014 Imagination Technologies Ltd. + * Author: Markos Chandras <markos.chandras@imgtec.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; version 2 of the License. + */ + +#include <linux/bitops.h> +#include <linux/compiler.h> +#include <linux/errno.h> +#include <linux/filter.h> +#include <linux/if_vlan.h> +#include <linux/moduleloader.h> +#include <linux/netdevice.h> +#include <linux/string.h> +#include <linux/slab.h> +#include <linux/types.h> +#include <asm/asm.h> +#include <asm/bitops.h> +#include <asm/cacheflush.h> +#include <asm/cpu-features.h> +#include <asm/uasm.h> + +#include "bpf_jit.h" + +/* ABI + * r_skb_hl SKB header length + * r_data SKB data pointer + * r_off Offset + * r_A BPF register A + * r_X BPF register X + * r_skb *skb + * r_M *scratch memory + * r_skb_len SKB length + * + * On entry (*bpf_func)(*skb, *filter) + * a0 = MIPS_R_A0 = skb; + * a1 = MIPS_R_A1 = filter; + * + * Stack + * ... + * M[15] + * M[14] + * M[13] + * ... + * M[0] <-- r_M + * saved reg k-1 + * saved reg k-2 + * ... + * saved reg 0 <-- r_sp + * <no argument area> + * + * Packet layout + * + * <--------------------- len ------------------------> + * <--skb-len(r_skb_hl)-->< ----- skb->data_len ------> + * ---------------------------------------------------- + * | skb->data | + * ---------------------------------------------------- + */ + +#define ptr typeof(unsigned long) + +#define SCRATCH_OFF(k) (4 * (k)) + +/* JIT flags */ +#define SEEN_CALL (1 << BPF_MEMWORDS) +#define SEEN_SREG_SFT (BPF_MEMWORDS + 1) +#define SEEN_SREG_BASE (1 << SEEN_SREG_SFT) +#define SEEN_SREG(x) (SEEN_SREG_BASE << (x)) +#define SEEN_OFF SEEN_SREG(2) +#define SEEN_A SEEN_SREG(3) +#define SEEN_X SEEN_SREG(4) +#define SEEN_SKB SEEN_SREG(5) +#define SEEN_MEM SEEN_SREG(6) +/* SEEN_SK_DATA also implies skb_hl an skb_len */ +#define SEEN_SKB_DATA (SEEN_SREG(7) | SEEN_SREG(1) | SEEN_SREG(0)) + +/* Arguments used by JIT */ +#define ARGS_USED_BY_JIT 2 /* only applicable to 64-bit */ + +#define SBIT(x) (1 << (x)) /* Signed version of BIT() */ + +/** + * struct jit_ctx - JIT context + * @skf: The sk_filter + * @prologue_bytes: Number of bytes for prologue + * @idx: Instruction index + * @flags: JIT flags + * @offsets: Instruction offsets + * @target: Memory location for the compiled filter + */ +struct jit_ctx { + const struct bpf_prog *skf; + unsigned int prologue_bytes; + u32 idx; + u32 flags; + u32 *offsets; + u32 *target; +}; + + +static inline int optimize_div(u32 *k) +{ + /* power of 2 divides can be implemented with right shift */ + if (!(*k & (*k-1))) { + *k = ilog2(*k); + return 1; + } + + return 0; +} + +static inline void emit_jit_reg_move(ptr dst, ptr src, struct jit_ctx *ctx); + +/* Simply emit the instruction if the JIT memory space has been allocated */ +#define emit_instr(ctx, func, ...) \ +do { \ + if ((ctx)->target != NULL) { \ + u32 *p = &(ctx)->target[ctx->idx]; \ + uasm_i_##func(&p, ##__VA_ARGS__); \ + } \ + (ctx)->idx++; \ +} while (0) + +/* + * Similar to emit_instr but it must be used when we need to emit + * 32-bit or 64-bit instructions + */ +#define emit_long_instr(ctx, func, ...) \ +do { \ + if ((ctx)->target != NULL) { \ + u32 *p = &(ctx)->target[ctx->idx]; \ + UASM_i_##func(&p, ##__VA_ARGS__); \ + } \ + (ctx)->idx++; \ +} while (0) + +/* Determine if immediate is within the 16-bit signed range */ +static inline bool is_range16(s32 imm) +{ + return !(imm >= SBIT(15) || imm < -SBIT(15)); +} + +static inline void emit_addu(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, addu, dst, src1, src2); +} + +static inline void emit_nop(struct jit_ctx *ctx) +{ + emit_instr(ctx, nop); +} + +/* Load a u32 immediate to a register */ +static inline void emit_load_imm(unsigned int dst, u32 imm, struct jit_ctx *ctx) +{ + if (ctx->target != NULL) { + /* addiu can only handle s16 */ + if (!is_range16(imm)) { + u32 *p = &ctx->target[ctx->idx]; + uasm_i_lui(&p, r_tmp_imm, (s32)imm >> 16); + p = &ctx->target[ctx->idx + 1]; + uasm_i_ori(&p, dst, r_tmp_imm, imm & 0xffff); + } else { + u32 *p = &ctx->target[ctx->idx]; + uasm_i_addiu(&p, dst, r_zero, imm); + } + } + ctx->idx++; + + if (!is_range16(imm)) + ctx->idx++; +} + +static inline void emit_or(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, or, dst, src1, src2); +} + +static inline void emit_ori(unsigned int dst, unsigned src, u32 imm, + struct jit_ctx *ctx) +{ + if (imm >= BIT(16)) { + emit_load_imm(r_tmp, imm, ctx); + emit_or(dst, src, r_tmp, ctx); + } else { + emit_instr(ctx, ori, dst, src, imm); + } +} + +static inline void emit_daddiu(unsigned int dst, unsigned int src, + int imm, struct jit_ctx *ctx) +{ + /* + * Only used for stack, so the imm is relatively small + * and it fits in 15-bits + */ + emit_instr(ctx, daddiu, dst, src, imm); +} + +static inline void emit_addiu(unsigned int dst, unsigned int src, + u32 imm, struct jit_ctx *ctx) +{ + if (!is_range16(imm)) { + emit_load_imm(r_tmp, imm, ctx); + emit_addu(dst, r_tmp, src, ctx); + } else { + emit_instr(ctx, addiu, dst, src, imm); + } +} + +static inline void emit_and(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, and, dst, src1, src2); +} + +static inline void emit_andi(unsigned int dst, unsigned int src, + u32 imm, struct jit_ctx *ctx) +{ + /* If imm does not fit in u16 then load it to register */ + if (imm >= BIT(16)) { + emit_load_imm(r_tmp, imm, ctx); + emit_and(dst, src, r_tmp, ctx); + } else { + emit_instr(ctx, andi, dst, src, imm); + } +} + +static inline void emit_xor(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, xor, dst, src1, src2); +} + +static inline void emit_xori(ptr dst, ptr src, u32 imm, struct jit_ctx *ctx) +{ + /* If imm does not fit in u16 then load it to register */ + if (imm >= BIT(16)) { + emit_load_imm(r_tmp, imm, ctx); + emit_xor(dst, src, r_tmp, ctx); + } else { + emit_instr(ctx, xori, dst, src, imm); + } +} + +static inline void emit_stack_offset(int offset, struct jit_ctx *ctx) +{ + emit_long_instr(ctx, ADDIU, r_sp, r_sp, offset); +} + +static inline void emit_subu(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, subu, dst, src1, src2); +} + +static inline void emit_neg(unsigned int reg, struct jit_ctx *ctx) +{ + emit_subu(reg, r_zero, reg, ctx); +} + +static inline void emit_sllv(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + emit_instr(ctx, sllv, dst, src, sa); +} + +static inline void emit_sll(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + /* sa is 5-bits long */ + if (sa >= BIT(5)) + /* Shifting >= 32 results in zero */ + emit_jit_reg_move(dst, r_zero, ctx); + else + emit_instr(ctx, sll, dst, src, sa); +} + +static inline void emit_srlv(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + emit_instr(ctx, srlv, dst, src, sa); +} + +static inline void emit_srl(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + /* sa is 5-bits long */ + if (sa >= BIT(5)) + /* Shifting >= 32 results in zero */ + emit_jit_reg_move(dst, r_zero, ctx); + else + emit_instr(ctx, srl, dst, src, sa); +} + +static inline void emit_slt(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, slt, dst, src1, src2); +} + +static inline void emit_sltu(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, sltu, dst, src1, src2); +} + +static inline void emit_sltiu(unsigned dst, unsigned int src, + unsigned int imm, struct jit_ctx *ctx) +{ + /* 16 bit immediate */ + if (!is_range16((s32)imm)) { + emit_load_imm(r_tmp, imm, ctx); + emit_sltu(dst, src, r_tmp, ctx); + } else { + emit_instr(ctx, sltiu, dst, src, imm); + } + +} + +/* Store register on the stack */ +static inline void emit_store_stack_reg(ptr reg, ptr base, + unsigned int offset, + struct jit_ctx *ctx) +{ + emit_long_instr(ctx, SW, reg, offset, base); +} + +static inline void emit_store(ptr reg, ptr base, unsigned int offset, + struct jit_ctx *ctx) +{ + emit_instr(ctx, sw, reg, offset, base); +} + +static inline void emit_load_stack_reg(ptr reg, ptr base, + unsigned int offset, + struct jit_ctx *ctx) +{ + emit_long_instr(ctx, LW, reg, offset, base); +} + +static inline void emit_load(unsigned int reg, unsigned int base, + unsigned int offset, struct jit_ctx *ctx) +{ + emit_instr(ctx, lw, reg, offset, base); +} + +static inline void emit_load_byte(unsigned int reg, unsigned int base, + unsigned int offset, struct jit_ctx *ctx) +{ + emit_instr(ctx, lb, reg, offset, base); +} + +static inline void emit_half_load(unsigned int reg, unsigned int base, + unsigned int offset, struct jit_ctx *ctx) +{ + emit_instr(ctx, lh, reg, offset, base); +} + +static inline void emit_half_load_unsigned(unsigned int reg, unsigned int base, + unsigned int offset, struct jit_ctx *ctx) +{ + emit_instr(ctx, lhu, reg, offset, base); +} + +static inline void emit_mul(unsigned int dst, unsigned int src1, + unsigned int src2, struct jit_ctx *ctx) +{ + emit_instr(ctx, mul, dst, src1, src2); +} + +static inline void emit_div(unsigned int dst, unsigned int src, + struct jit_ctx *ctx) +{ + if (ctx->target != NULL) { + u32 *p = &ctx->target[ctx->idx]; + uasm_i_divu(&p, dst, src); + p = &ctx->target[ctx->idx + 1]; + uasm_i_mflo(&p, dst); + } + ctx->idx += 2; /* 2 insts */ +} + +static inline void emit_mod(unsigned int dst, unsigned int src, + struct jit_ctx *ctx) +{ + if (ctx->target != NULL) { + u32 *p = &ctx->target[ctx->idx]; + uasm_i_divu(&p, dst, src); + p = &ctx->target[ctx->idx + 1]; + uasm_i_mfhi(&p, dst); + } + ctx->idx += 2; /* 2 insts */ +} + +static inline void emit_dsll(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + emit_instr(ctx, dsll, dst, src, sa); +} + +static inline void emit_dsrl32(unsigned int dst, unsigned int src, + unsigned int sa, struct jit_ctx *ctx) +{ + emit_instr(ctx, dsrl32, dst, src, sa); +} + +static inline void emit_wsbh(unsigned int dst, unsigned int src, + struct jit_ctx *ctx) +{ + emit_instr(ctx, wsbh, dst, src); +} + +/* load pointer to register */ +static inline void emit_load_ptr(unsigned int dst, unsigned int src, + int imm, struct jit_ctx *ctx) +{ + /* src contains the base addr of the 32/64-pointer */ + emit_long_instr(ctx, LW, dst, imm, src); +} + +/* load a function pointer to register */ +static inline void emit_load_func(unsigned int reg, ptr imm, + struct jit_ctx *ctx) +{ + if (IS_ENABLED(CONFIG_64BIT)) { + /* At this point imm is always 64-bit */ + emit_load_imm(r_tmp, (u64)imm >> 32, ctx); + emit_dsll(r_tmp_imm, r_tmp, 16, ctx); /* left shift by 16 */ + emit_ori(r_tmp, r_tmp_imm, (imm >> 16) & 0xffff, ctx); + emit_dsll(r_tmp_imm, r_tmp, 16, ctx); /* left shift by 16 */ + emit_ori(reg, r_tmp_imm, imm & 0xffff, ctx); + } else { + emit_load_imm(reg, imm, ctx); + } +} + +/* Move to real MIPS register */ +static inline void emit_reg_move(ptr dst, ptr src, struct jit_ctx *ctx) +{ + emit_long_instr(ctx, ADDU, dst, src, r_zero); +} + +/* Move to JIT (32-bit) register */ +static inline void emit_jit_reg_move(ptr dst, ptr src, struct jit_ctx *ctx) +{ + emit_addu(dst, src, r_zero, ctx); +} + +/* Compute the immediate value for PC-relative branches. */ +static inline u32 b_imm(unsigned int tgt, struct jit_ctx *ctx) +{ + if (ctx->target == NULL) + return 0; + + /* + * We want a pc-relative branch. We only do forward branches + * so tgt is always after pc. tgt is the instruction offset + * we want to jump to. + + * Branch on MIPS: + * I: target_offset <- sign_extend(offset) + * I+1: PC += target_offset (delay slot) + * + * ctx->idx currently points to the branch instruction + * but the offset is added to the delay slot so we need + * to subtract 4. + */ + return ctx->offsets[tgt] - + (ctx->idx * 4 - ctx->prologue_bytes) - 4; +} + +static inline void emit_bcond(int cond, unsigned int reg1, unsigned int reg2, + unsigned int imm, struct jit_ctx *ctx) +{ + if (ctx->target != NULL) { + u32 *p = &ctx->target[ctx->idx]; + + switch (cond) { + case MIPS_COND_EQ: + uasm_i_beq(&p, reg1, reg2, imm); + break; + case MIPS_COND_NE: + uasm_i_bne(&p, reg1, reg2, imm); + break; + case MIPS_COND_ALL: + uasm_i_b(&p, imm); + break; + default: + pr_warn("%s: Unhandled branch conditional: %d\n", + __func__, cond); + } + } + ctx->idx++; +} + +static inline void emit_b(unsigned int imm, struct jit_ctx *ctx) +{ + emit_bcond(MIPS_COND_ALL, r_zero, r_zero, imm, ctx); +} + +static inline void emit_jalr(unsigned int link, unsigned int reg, + struct jit_ctx *ctx) +{ + emit_instr(ctx, jalr, link, reg); +} + +static inline void emit_jr(unsigned int reg, struct jit_ctx *ctx) +{ + emit_instr(ctx, jr, reg); +} + +static inline u16 align_sp(unsigned int num) +{ + /* Double word alignment for 32-bit, quadword for 64-bit */ + unsigned int align = IS_ENABLED(CONFIG_64BIT) ? 16 : 8; + num = (num + (align - 1)) & -align; + return num; +} + +static void save_bpf_jit_regs(struct jit_ctx *ctx, unsigned offset) +{ + int i = 0, real_off = 0; + u32 sflags, tmp_flags; + + /* Adjust the stack pointer */ + if (offset) + emit_stack_offset(-align_sp(offset), ctx); + + tmp_flags = sflags = ctx->flags >> SEEN_SREG_SFT; + /* sflags is essentially a bitmap */ + while (tmp_flags) { + if ((sflags >> i) & 0x1) { + emit_store_stack_reg(MIPS_R_S0 + i, r_sp, real_off, + ctx); + real_off += SZREG; + } + i++; + tmp_flags >>= 1; + } + + /* save return address */ + if (ctx->flags & SEEN_CALL) { + emit_store_stack_reg(r_ra, r_sp, real_off, ctx); + real_off += SZREG; + } + + /* Setup r_M leaving the alignment gap if necessary */ + if (ctx->flags & SEEN_MEM) { + if (real_off % (SZREG * 2)) + real_off += SZREG; + emit_long_instr(ctx, ADDIU, r_M, r_sp, real_off); + } +} + +static void restore_bpf_jit_regs(struct jit_ctx *ctx, + unsigned int offset) +{ + int i, real_off = 0; + u32 sflags, tmp_flags; + + tmp_flags = sflags = ctx->flags >> SEEN_SREG_SFT; + /* sflags is a bitmap */ + i = 0; + while (tmp_flags) { + if ((sflags >> i) & 0x1) { + emit_load_stack_reg(MIPS_R_S0 + i, r_sp, real_off, + ctx); + real_off += SZREG; + } + i++; + tmp_flags >>= 1; + } + + /* restore return address */ + if (ctx->flags & SEEN_CALL) + emit_load_stack_reg(r_ra, r_sp, real_off, ctx); + + /* Restore the sp and discard the scrach memory */ + if (offset) + emit_stack_offset(align_sp(offset), ctx); +} + +static unsigned int get_stack_depth(struct jit_ctx *ctx) +{ + int sp_off = 0; + + + /* How may s* regs do we need to preserved? */ + sp_off += hweight32(ctx->flags >> SEEN_SREG_SFT) * SZREG; + + if (ctx->flags & SEEN_MEM) + sp_off += 4 * BPF_MEMWORDS; /* BPF_MEMWORDS are 32-bit */ + + if (ctx->flags & SEEN_CALL) + sp_off += SZREG; /* Space for our ra register */ + + return sp_off; +} + +static void build_prologue(struct jit_ctx *ctx) +{ + int sp_off; + + /* Calculate the total offset for the stack pointer */ + sp_off = get_stack_depth(ctx); + save_bpf_jit_regs(ctx, sp_off); + + if (ctx->flags & SEEN_SKB) + emit_reg_move(r_skb, MIPS_R_A0, ctx); + + if (ctx->flags & SEEN_SKB_DATA) { + /* Load packet length */ + emit_load(r_skb_len, r_skb, offsetof(struct sk_buff, len), + ctx); + emit_load(r_tmp, r_skb, offsetof(struct sk_buff, data_len), + ctx); + /* Load the data pointer */ + emit_load_ptr(r_skb_data, r_skb, + offsetof(struct sk_buff, data), ctx); + /* Load the header length */ + emit_subu(r_skb_hl, r_skb_len, r_tmp, ctx); + } + + if (ctx->flags & SEEN_X) + emit_jit_reg_move(r_X, r_zero, ctx); + + /* + * Do not leak kernel data to userspace, we only need to clear + * r_A if it is ever used. In fact if it is never used, we + * will not save/restore it, so clearing it in this case would + * corrupt the state of the caller. + */ + if (bpf_needs_clear_a(&ctx->skf->insns[0]) && + (ctx->flags & SEEN_A)) + emit_jit_reg_move(r_A, r_zero, ctx); +} + +static void build_epilogue(struct jit_ctx *ctx) +{ + unsigned int sp_off; + + /* Calculate the total offset for the stack pointer */ + + sp_off = get_stack_depth(ctx); + restore_bpf_jit_regs(ctx, sp_off); + + /* Return */ + emit_jr(r_ra, ctx); + emit_nop(ctx); +} + +#define CHOOSE_LOAD_FUNC(K, func) \ + ((int)K < 0 ? ((int)K >= SKF_LL_OFF ? func##_negative : func) : \ + func##_positive) + +static bool is_bad_offset(int b_off) +{ + return b_off > 0x1ffff || b_off < -0x20000; +} + +static int build_body(struct jit_ctx *ctx) +{ + const struct bpf_prog *prog = ctx->skf; + const struct sock_filter *inst; + unsigned int i, off, condt; + u32 k, b_off __maybe_unused; + u8 (*sk_load_func)(unsigned long *skb, int offset); + + for (i = 0; i < prog->len; i++) { + u16 code; + + inst = &(prog->insns[i]); + pr_debug("%s: code->0x%02x, jt->0x%x, jf->0x%x, k->0x%x\n", + __func__, inst->code, inst->jt, inst->jf, inst->k); + k = inst->k; + code = bpf_anc_helper(inst); + + if (ctx->target == NULL) + ctx->offsets[i] = ctx->idx * 4; + + switch (code) { + case BPF_LD | BPF_IMM: + /* A <- k ==> li r_A, k */ + ctx->flags |= SEEN_A; + emit_load_imm(r_A, k, ctx); + break; + case BPF_LD | BPF_W | BPF_LEN: + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, len) != 4); + /* A <- len ==> lw r_A, offset(skb) */ + ctx->flags |= SEEN_SKB | SEEN_A; + off = offsetof(struct sk_buff, len); + emit_load(r_A, r_skb, off, ctx); + break; + case BPF_LD | BPF_MEM: + /* A <- M[k] ==> lw r_A, offset(M) */ + ctx->flags |= SEEN_MEM | SEEN_A; + emit_load(r_A, r_M, SCRATCH_OFF(k), ctx); + break; + case BPF_LD | BPF_W | BPF_ABS: + /* A <- P[k:4] */ + sk_load_func = CHOOSE_LOAD_FUNC(k, sk_load_word); + goto load; + case BPF_LD | BPF_H | BPF_ABS: + /* A <- P[k:2] */ + sk_load_func = CHOOSE_LOAD_FUNC(k, sk_load_half); + goto load; + case BPF_LD | BPF_B | BPF_ABS: + /* A <- P[k:1] */ + sk_load_func = CHOOSE_LOAD_FUNC(k, sk_load_byte); +load: + emit_load_imm(r_off, k, ctx); +load_common: + ctx->flags |= SEEN_CALL | SEEN_OFF | + SEEN_SKB | SEEN_A | SEEN_SKB_DATA; + + emit_load_func(r_s0, (ptr)sk_load_func, ctx); + emit_reg_move(MIPS_R_A0, r_skb, ctx); + emit_jalr(MIPS_R_RA, r_s0, ctx); + /* Load second argument to delay slot */ + emit_reg_move(MIPS_R_A1, r_off, ctx); + /* Check the error value */ + emit_bcond(MIPS_COND_EQ, r_ret, 0, b_imm(i + 1, ctx), + ctx); + /* Load return register on DS for failures */ + emit_reg_move(r_ret, r_zero, ctx); + /* Return with error */ + b_off = b_imm(prog->len, ctx); + if (is_bad_offset(b_off)) + return -E2BIG; + emit_b(b_off, ctx); + emit_nop(ctx); + break; + case BPF_LD | BPF_W | BPF_IND: + /* A <- P[X + k:4] */ + sk_load_func = sk_load_word; + goto load_ind; + case BPF_LD | BPF_H | BPF_IND: + /* A <- P[X + k:2] */ + sk_load_func = sk_load_half; + goto load_ind; + case BPF_LD | BPF_B | BPF_IND: + /* A <- P[X + k:1] */ + sk_load_func = sk_load_byte; +load_ind: + ctx->flags |= SEEN_OFF | SEEN_X; + emit_addiu(r_off, r_X, k, ctx); + goto load_common; + case BPF_LDX | BPF_IMM: + /* X <- k */ + ctx->flags |= SEEN_X; + emit_load_imm(r_X, k, ctx); + break; + case BPF_LDX | BPF_MEM: + /* X <- M[k] */ + ctx->flags |= SEEN_X | SEEN_MEM; + emit_load(r_X, r_M, SCRATCH_OFF(k), ctx); + break; + case BPF_LDX | BPF_W | BPF_LEN: + /* X <- len */ + ctx->flags |= SEEN_X | SEEN_SKB; + off = offsetof(struct sk_buff, len); + emit_load(r_X, r_skb, off, ctx); + break; + case BPF_LDX | BPF_B | BPF_MSH: + /* X <- 4 * (P[k:1] & 0xf) */ + ctx->flags |= SEEN_X | SEEN_CALL | SEEN_SKB; + /* Load offset to a1 */ + emit_load_func(r_s0, (ptr)sk_load_byte, ctx); + /* + * This may emit two instructions so it may not fit + * in the delay slot. So use a0 in the delay slot. + */ + emit_load_imm(MIPS_R_A1, k, ctx); + emit_jalr(MIPS_R_RA, r_s0, ctx); + emit_reg_move(MIPS_R_A0, r_skb, ctx); /* delay slot */ + /* Check the error value */ + b_off = b_imm(prog->len, ctx); + if (is_bad_offset(b_off)) + return -E2BIG; + emit_bcond(MIPS_COND_NE, r_ret, 0, b_off, ctx); + emit_reg_move(r_ret, r_zero, ctx); + /* We are good */ + /* X <- P[1:K] & 0xf */ + emit_andi(r_X, r_A, 0xf, ctx); + /* X << 2 */ + emit_b(b_imm(i + 1, ctx), ctx); + emit_sll(r_X, r_X, 2, ctx); /* delay slot */ + break; + case BPF_ST: + /* M[k] <- A */ + ctx->flags |= SEEN_MEM | SEEN_A; + emit_store(r_A, r_M, SCRATCH_OFF(k), ctx); + break; + case BPF_STX: + /* M[k] <- X */ + ctx->flags |= SEEN_MEM | SEEN_X; + emit_store(r_X, r_M, SCRATCH_OFF(k), ctx); + break; + case BPF_ALU | BPF_ADD | BPF_K: + /* A += K */ + ctx->flags |= SEEN_A; + emit_addiu(r_A, r_A, k, ctx); + break; + case BPF_ALU | BPF_ADD | BPF_X: + /* A += X */ + ctx->flags |= SEEN_A | SEEN_X; + emit_addu(r_A, r_A, r_X, ctx); + break; + case BPF_ALU | BPF_SUB | BPF_K: + /* A -= K */ + ctx->flags |= SEEN_A; + emit_addiu(r_A, r_A, -k, ctx); + break; + case BPF_ALU | BPF_SUB | BPF_X: + /* A -= X */ + ctx->flags |= SEEN_A | SEEN_X; + emit_subu(r_A, r_A, r_X, ctx); + break; + case BPF_ALU | BPF_MUL | BPF_K: + /* A *= K */ + /* Load K to scratch register before MUL */ + ctx->flags |= SEEN_A; + emit_load_imm(r_s0, k, ctx); + emit_mul(r_A, r_A, r_s0, ctx); + break; + case BPF_ALU | BPF_MUL | BPF_X: + /* A *= X */ + ctx->flags |= SEEN_A | SEEN_X; + emit_mul(r_A, r_A, r_X, ctx); + break; + case BPF_ALU | BPF_DIV | BPF_K: + /* A /= k */ + if (k == 1) + break; + if (optimize_div(&k)) { + ctx->flags |= SEEN_A; + emit_srl(r_A, r_A, k, ctx); + break; + } + ctx->flags |= SEEN_A; + emit_load_imm(r_s0, k, ctx); + emit_div(r_A, r_s0, ctx); + break; + case BPF_ALU | BPF_MOD | BPF_K: + /* A %= k */ + if (k == 1) { + ctx->flags |= SEEN_A; + emit_jit_reg_move(r_A, r_zero, ctx); + } else { + ctx->flags |= SEEN_A; + emit_load_imm(r_s0, k, ctx); + emit_mod(r_A, r_s0, ctx); + } + break; + case BPF_ALU | BPF_DIV | BPF_X: + /* A /= X */ + ctx->flags |= SEEN_X | SEEN_A; + /* Check if r_X is zero */ + b_off = b_imm(prog->len, ctx); + if (is_bad_offset(b_off)) + return -E2BIG; + emit_bcond(MIPS_COND_EQ, r_X, r_zero, b_off, ctx); + emit_load_imm(r_ret, 0, ctx); /* delay slot */ + emit_div(r_A, r_X, ctx); + break; + case BPF_ALU | BPF_MOD | BPF_X: + /* A %= X */ + ctx->flags |= SEEN_X | SEEN_A; + /* Check if r_X is zero */ + b_off = b_imm(prog->len, ctx); + if (is_bad_offset(b_off)) + return -E2BIG; + emit_bcond(MIPS_COND_EQ, r_X, r_zero, b_off, ctx); + emit_load_imm(r_ret, 0, ctx); /* delay slot */ + emit_mod(r_A, r_X, ctx); + break; + case BPF_ALU | BPF_OR | BPF_K: + /* A |= K */ + ctx->flags |= SEEN_A; + emit_ori(r_A, r_A, k, ctx); + break; + case BPF_ALU | BPF_OR | BPF_X: + /* A |= X */ + ctx->flags |= SEEN_A; + emit_ori(r_A, r_A, r_X, ctx); + break; + case BPF_ALU | BPF_XOR | BPF_K: + /* A ^= k */ + ctx->flags |= SEEN_A; + emit_xori(r_A, r_A, k, ctx); + break; + case BPF_ANC | SKF_AD_ALU_XOR_X: + case BPF_ALU | BPF_XOR | BPF_X: + /* A ^= X */ + ctx->flags |= SEEN_A; + emit_xor(r_A, r_A, r_X, ctx); + break; + case BPF_ALU | BPF_AND | BPF_K: + /* A &= K */ + ctx->flags |= SEEN_A; + emit_andi(r_A, r_A, k, ctx); + break; + case BPF_ALU | BPF_AND | BPF_X: + /* A &= X */ + ctx->flags |= SEEN_A | SEEN_X; + emit_and(r_A, r_A, r_X, ctx); + break; + case BPF_ALU | BPF_LSH | BPF_K: + /* A <<= K */ + ctx->flags |= SEEN_A; + emit_sll(r_A, r_A, k, ctx); + break; + case BPF_ALU | BPF_LSH | BPF_X: + /* A <<= X */ + ctx->flags |= SEEN_A | SEEN_X; + emit_sllv(r_A, r_A, r_X, ctx); + break; + case BPF_ALU | BPF_RSH | BPF_K: + /* A >>= K */ + ctx->flags |= SEEN_A; + emit_srl(r_A, r_A, k, ctx); + break; + case BPF_ALU | BPF_RSH | BPF_X: + ctx->flags |= SEEN_A | SEEN_X; + emit_srlv(r_A, r_A, r_X, ctx); + break; + case BPF_ALU | BPF_NEG: + /* A = -A */ + ctx->flags |= SEEN_A; + emit_neg(r_A, ctx); + break; + case BPF_JMP | BPF_JA: + /* pc += K */ + b_off = b_imm(i + k + 1, ctx); + if (is_bad_offset(b_off)) + return -E2BIG; + emit_b(b_off, ctx); + emit_nop(ctx); + break; + case BPF_JMP | BPF_JEQ | BPF_K: + /* pc += ( A == K ) ? pc->jt : pc->jf */ + condt = MIPS_COND_EQ | MIPS_COND_K; + goto jmp_cmp; + case BPF_JMP | BPF_JEQ | BPF_X: + ctx->flags |= SEEN_X; + /* pc += ( A == X ) ? pc->jt : pc->jf */ + condt = MIPS_COND_EQ | MIPS_COND_X; + goto jmp_cmp; + case BPF_JMP | BPF_JGE | BPF_K: + /* pc += ( A >= K ) ? pc->jt : pc->jf */ + condt = MIPS_COND_GE | MIPS_COND_K; + goto jmp_cmp; + case BPF_JMP | BPF_JGE | BPF_X: + ctx->flags |= SEEN_X; + /* pc += ( A >= X ) ? pc->jt : pc->jf */ + condt = MIPS_COND_GE | MIPS_COND_X; + goto jmp_cmp; + case BPF_JMP | BPF_JGT | BPF_K: + /* pc += ( A > K ) ? pc->jt : pc->jf */ + condt = MIPS_COND_GT | MIPS_COND_K; + goto jmp_cmp; + case BPF_JMP | BPF_JGT | BPF_X: + ctx->flags |= SEEN_X; + /* pc += ( A > X ) ? pc->jt : pc->jf */ + condt = MIPS_COND_GT | MIPS_COND_X; +jmp_cmp: + /* Greater or Equal */ + if ((condt & MIPS_COND_GE) || + (condt & MIPS_COND_GT)) { + if (condt & MIPS_COND_K) { /* K */ + ctx->flags |= SEEN_A; + emit_sltiu(r_s0, r_A, k, ctx); + } else { /* X */ + ctx->flags |= SEEN_A | + SEEN_X; + emit_sltu(r_s0, r_A, r_X, ctx); + } + /* A < (K|X) ? r_scrach = 1 */ + b_off = b_imm(i + inst->jf + 1, ctx); + emit_bcond(MIPS_COND_NE, r_s0, r_zero, b_off, + ctx); + emit_nop(ctx); + /* A > (K|X) ? scratch = 0 */ + if (condt & MIPS_COND_GT) { + /* Checking for equality */ + ctx->flags |= SEEN_A | SEEN_X; + if (condt & MIPS_COND_K) + emit_load_imm(r_s0, k, ctx); + else + emit_jit_reg_move(r_s0, r_X, + ctx); + b_off = b_imm(i + inst->jf + 1, ctx); + emit_bcond(MIPS_COND_EQ, r_A, r_s0, + b_off, ctx); + emit_nop(ctx); + /* Finally, A > K|X */ + b_off = b_imm(i + inst->jt + 1, ctx); + emit_b(b_off, ctx); + emit_nop(ctx); + } else { + /* A >= (K|X) so jump */ + b_off = b_imm(i + inst->jt + 1, ctx); + emit_b(b_off, ctx); + emit_nop(ctx); + } + } else { + /* A == K|X */ + if (condt & MIPS_COND_K) { /* K */ + ctx->flags |= SEEN_A; + emit_load_imm(r_s0, k, ctx); + /* jump true */ + b_off = b_imm(i + inst->jt + 1, ctx); + emit_bcond(MIPS_COND_EQ, r_A, r_s0, + b_off, ctx); + emit_nop(ctx); + /* jump false */ + b_off = b_imm(i + inst->jf + 1, + ctx); + emit_bcond(MIPS_COND_NE, r_A, r_s0, + b_off, ctx); + emit_nop(ctx); + } else { /* X */ + /* jump true */ + ctx->flags |= SEEN_A | SEEN_X; + b_off = b_imm(i + inst->jt + 1, + ctx); + emit_bcond(MIPS_COND_EQ, r_A, r_X, + b_off, ctx); + emit_nop(ctx); + /* jump false */ + b_off = b_imm(i + inst->jf + 1, ctx); + emit_bcond(MIPS_COND_NE, r_A, r_X, + b_off, ctx); + emit_nop(ctx); + } + } + break; + case BPF_JMP | BPF_JSET | BPF_K: + ctx->flags |= SEEN_A; + /* pc += (A & K) ? pc -> jt : pc -> jf */ + emit_load_imm(r_s1, k, ctx); + emit_and(r_s0, r_A, r_s1, ctx); + /* jump true */ + b_off = b_imm(i + inst->jt + 1, ctx); + emit_bcond(MIPS_COND_NE, r_s0, r_zero, b_off, ctx); + emit_nop(ctx); + /* jump false */ + b_off = b_imm(i + inst->jf + 1, ctx); + emit_b(b_off, ctx); + emit_nop(ctx); + break; + case BPF_JMP | BPF_JSET | BPF_X: + ctx->flags |= SEEN_X | SEEN_A; + /* pc += (A & X) ? pc -> jt : pc -> jf */ + emit_and(r_s0, r_A, r_X, ctx); + /* jump true */ + b_off = b_imm(i + inst->jt + 1, ctx); + emit_bcond(MIPS_COND_NE, r_s0, r_zero, b_off, ctx); + emit_nop(ctx); + /* jump false */ + b_off = b_imm(i + inst->jf + 1, ctx); + emit_b(b_off, ctx); + emit_nop(ctx); + break; + case BPF_RET | BPF_A: + ctx->flags |= SEEN_A; + if (i != prog->len - 1) { + /* + * If this is not the last instruction + * then jump to the epilogue + */ + b_off = b_imm(prog->len, ctx); + if (is_bad_offset(b_off)) + return -E2BIG; + emit_b(b_off, ctx); + } + emit_reg_move(r_ret, r_A, ctx); /* delay slot */ + break; + case BPF_RET | BPF_K: + /* + * It can emit two instructions so it does not fit on + * the delay slot. + */ + emit_load_imm(r_ret, k, ctx); + if (i != prog->len - 1) { + /* + * If this is not the last instruction + * then jump to the epilogue + */ + b_off = b_imm(prog->len, ctx); + if (is_bad_offset(b_off)) + return -E2BIG; + emit_b(b_off, ctx); + emit_nop(ctx); + } + break; + case BPF_MISC | BPF_TAX: + /* X = A */ + ctx->flags |= SEEN_X | SEEN_A; + emit_jit_reg_move(r_X, r_A, ctx); + break; + case BPF_MISC | BPF_TXA: + /* A = X */ + ctx->flags |= SEEN_A | SEEN_X; + emit_jit_reg_move(r_A, r_X, ctx); + break; + /* AUX */ + case BPF_ANC | SKF_AD_PROTOCOL: + /* A = ntohs(skb->protocol */ + ctx->flags |= SEEN_SKB | SEEN_OFF | SEEN_A; + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, + protocol) != 2); + off = offsetof(struct sk_buff, protocol); + emit_half_load(r_A, r_skb, off, ctx); +#ifdef CONFIG_CPU_LITTLE_ENDIAN + /* This needs little endian fixup */ + if (cpu_has_wsbh) { + /* R2 and later have the wsbh instruction */ + emit_wsbh(r_A, r_A, ctx); + } else { + /* Get first byte */ + emit_andi(r_tmp_imm, r_A, 0xff, ctx); + /* Shift it */ + emit_sll(r_tmp, r_tmp_imm, 8, ctx); + /* Get second byte */ + emit_srl(r_tmp_imm, r_A, 8, ctx); + emit_andi(r_tmp_imm, r_tmp_imm, 0xff, ctx); + /* Put everyting together in r_A */ + emit_or(r_A, r_tmp, r_tmp_imm, ctx); + } +#endif + break; + case BPF_ANC | SKF_AD_CPU: + ctx->flags |= SEEN_A | SEEN_OFF; + /* A = current_thread_info()->cpu */ + BUILD_BUG_ON(FIELD_SIZEOF(struct thread_info, + cpu) != 4); + off = offsetof(struct thread_info, cpu); + /* $28/gp points to the thread_info struct */ + emit_load(r_A, 28, off, ctx); + break; + case BPF_ANC | SKF_AD_IFINDEX: + /* A = skb->dev->ifindex */ + case BPF_ANC | SKF_AD_HATYPE: + /* A = skb->dev->type */ + ctx->flags |= SEEN_SKB | SEEN_A; + off = offsetof(struct sk_buff, dev); + /* Load *dev pointer */ + emit_load_ptr(r_s0, r_skb, off, ctx); + /* error (0) in the delay slot */ + b_off = b_imm(prog->len, ctx); + if (is_bad_offset(b_off)) + return -E2BIG; + emit_bcond(MIPS_COND_EQ, r_s0, r_zero, b_off, ctx); + emit_reg_move(r_ret, r_zero, ctx); + if (code == (BPF_ANC | SKF_AD_IFINDEX)) { + BUILD_BUG_ON(FIELD_SIZEOF(struct net_device, ifindex) != 4); + off = offsetof(struct net_device, ifindex); + emit_load(r_A, r_s0, off, ctx); + } else { /* (code == (BPF_ANC | SKF_AD_HATYPE) */ + BUILD_BUG_ON(FIELD_SIZEOF(struct net_device, type) != 2); + off = offsetof(struct net_device, type); + emit_half_load_unsigned(r_A, r_s0, off, ctx); + } + break; + case BPF_ANC | SKF_AD_MARK: + ctx->flags |= SEEN_SKB | SEEN_A; + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, mark) != 4); + off = offsetof(struct sk_buff, mark); + emit_load(r_A, r_skb, off, ctx); + break; + case BPF_ANC | SKF_AD_RXHASH: + ctx->flags |= SEEN_SKB | SEEN_A; + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, hash) != 4); + off = offsetof(struct sk_buff, hash); + emit_load(r_A, r_skb, off, ctx); + break; + case BPF_ANC | SKF_AD_VLAN_TAG: + ctx->flags |= SEEN_SKB | SEEN_A; + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, + vlan_tci) != 2); + off = offsetof(struct sk_buff, vlan_tci); + emit_half_load_unsigned(r_A, r_skb, off, ctx); + break; + case BPF_ANC | SKF_AD_VLAN_TAG_PRESENT: + ctx->flags |= SEEN_SKB | SEEN_A; + emit_load_byte(r_A, r_skb, PKT_VLAN_PRESENT_OFFSET(), ctx); + if (PKT_VLAN_PRESENT_BIT) + emit_srl(r_A, r_A, PKT_VLAN_PRESENT_BIT, ctx); + if (PKT_VLAN_PRESENT_BIT < 7) + emit_andi(r_A, r_A, 1, ctx); + break; + case BPF_ANC | SKF_AD_PKTTYPE: + ctx->flags |= SEEN_SKB; + + emit_load_byte(r_tmp, r_skb, PKT_TYPE_OFFSET(), ctx); + /* Keep only the last 3 bits */ + emit_andi(r_A, r_tmp, PKT_TYPE_MAX, ctx); +#ifdef __BIG_ENDIAN_BITFIELD + /* Get the actual packet type to the lower 3 bits */ + emit_srl(r_A, r_A, 5, ctx); +#endif + break; + case BPF_ANC | SKF_AD_QUEUE: + ctx->flags |= SEEN_SKB | SEEN_A; + BUILD_BUG_ON(FIELD_SIZEOF(struct sk_buff, + queue_mapping) != 2); + BUILD_BUG_ON(offsetof(struct sk_buff, + queue_mapping) > 0xff); + off = offsetof(struct sk_buff, queue_mapping); + emit_half_load_unsigned(r_A, r_skb, off, ctx); + break; + default: + pr_debug("%s: Unhandled opcode: 0x%02x\n", __FILE__, + inst->code); + return -1; + } + } + + /* compute offsets only during the first pass */ + if (ctx->target == NULL) + ctx->offsets[i] = ctx->idx * 4; + + return 0; +} + +void bpf_jit_compile(struct bpf_prog *fp) +{ + struct jit_ctx ctx; + unsigned int alloc_size, tmp_idx; + + if (!bpf_jit_enable) + return; + + memset(&ctx, 0, sizeof(ctx)); + + ctx.offsets = kcalloc(fp->len + 1, sizeof(*ctx.offsets), GFP_KERNEL); + if (ctx.offsets == NULL) + return; + + ctx.skf = fp; + + if (build_body(&ctx)) + goto out; + + tmp_idx = ctx.idx; + build_prologue(&ctx); + ctx.prologue_bytes = (ctx.idx - tmp_idx) * 4; + /* just to complete the ctx.idx count */ + build_epilogue(&ctx); + + alloc_size = 4 * ctx.idx; + ctx.target = module_alloc(alloc_size); + if (ctx.target == NULL) + goto out; + + /* Clean it */ + memset(ctx.target, 0, alloc_size); + + ctx.idx = 0; + + /* Generate the actual JIT code */ + build_prologue(&ctx); + if (build_body(&ctx)) { + module_memfree(ctx.target); + goto out; + } + build_epilogue(&ctx); + + /* Update the icache */ + flush_icache_range((ptr)ctx.target, (ptr)(ctx.target + ctx.idx)); + + if (bpf_jit_enable > 1) + /* Dump JIT code */ + bpf_jit_dump(fp->len, alloc_size, 2, ctx.target); + + fp->bpf_func = (void *)ctx.target; + fp->jited = 1; + +out: + kfree(ctx.offsets); +} + +void bpf_jit_free(struct bpf_prog *fp) +{ + if (fp->jited) + module_memfree(fp->bpf_func); + + bpf_prog_unlock_free(fp); +} diff --git a/arch/mips/net/bpf_jit_asm.S b/arch/mips/net/bpf_jit_asm.S new file mode 100644 index 000000000000..57154c5883b6 --- /dev/null +++ b/arch/mips/net/bpf_jit_asm.S @@ -0,0 +1,285 @@ +/* + * bpf_jib_asm.S: Packet/header access helper functions for MIPS/MIPS64 BPF + * compiler. + * + * Copyright (C) 2015 Imagination Technologies Ltd. + * Author: Markos Chandras <markos.chandras@imgtec.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; version 2 of the License. + */ + +#include <asm/asm.h> +#include <asm/isa-rev.h> +#include <asm/regdef.h> +#include "bpf_jit.h" + +/* ABI + * + * r_skb_hl skb header length + * r_skb_data skb data + * r_off(a1) offset register + * r_A BPF register A + * r_X PF register X + * r_skb(a0) *skb + * r_M *scratch memory + * r_skb_le skb length + * r_s0 Scratch register 0 + * r_s1 Scratch register 1 + * + * On entry: + * a0: *skb + * a1: offset (imm or imm + X) + * + * All non-BPF-ABI registers are free for use. On return, we only + * care about r_ret. The BPF-ABI registers are assumed to remain + * unmodified during the entire filter operation. + */ + +#define skb a0 +#define offset a1 +#define SKF_LL_OFF (-0x200000) /* Can't include linux/filter.h in assembly */ + + /* We know better :) so prevent assembler reordering etc */ + .set noreorder + +#define is_offset_negative(TYPE) \ + /* If offset is negative we have more work to do */ \ + slti t0, offset, 0; \ + bgtz t0, bpf_slow_path_##TYPE##_neg; \ + /* Be careful what follows in DS. */ + +#define is_offset_in_header(SIZE, TYPE) \ + /* Reading from header? */ \ + addiu $r_s0, $r_skb_hl, -SIZE; \ + slt t0, $r_s0, offset; \ + bgtz t0, bpf_slow_path_##TYPE; \ + +LEAF(sk_load_word) + is_offset_negative(word) +FEXPORT(sk_load_word_positive) + is_offset_in_header(4, word) + /* Offset within header boundaries */ + PTR_ADDU t1, $r_skb_data, offset + .set reorder + lw $r_A, 0(t1) + .set noreorder +#ifdef CONFIG_CPU_LITTLE_ENDIAN +# if MIPS_ISA_REV >= 2 + wsbh t0, $r_A + rotr $r_A, t0, 16 +# else + sll t0, $r_A, 24 + srl t1, $r_A, 24 + srl t2, $r_A, 8 + or t0, t0, t1 + andi t2, t2, 0xff00 + andi t1, $r_A, 0xff00 + or t0, t0, t2 + sll t1, t1, 8 + or $r_A, t0, t1 +# endif +#endif + jr $r_ra + move $r_ret, zero + END(sk_load_word) + +LEAF(sk_load_half) + is_offset_negative(half) +FEXPORT(sk_load_half_positive) + is_offset_in_header(2, half) + /* Offset within header boundaries */ + PTR_ADDU t1, $r_skb_data, offset + lhu $r_A, 0(t1) +#ifdef CONFIG_CPU_LITTLE_ENDIAN +# if MIPS_ISA_REV >= 2 + wsbh $r_A, $r_A +# else + sll t0, $r_A, 8 + srl t1, $r_A, 8 + andi t0, t0, 0xff00 + or $r_A, t0, t1 +# endif +#endif + jr $r_ra + move $r_ret, zero + END(sk_load_half) + +LEAF(sk_load_byte) + is_offset_negative(byte) +FEXPORT(sk_load_byte_positive) + is_offset_in_header(1, byte) + /* Offset within header boundaries */ + PTR_ADDU t1, $r_skb_data, offset + lbu $r_A, 0(t1) + jr $r_ra + move $r_ret, zero + END(sk_load_byte) + +/* + * call skb_copy_bits: + * (prototype in linux/skbuff.h) + * + * int skb_copy_bits(sk_buff *skb, int offset, void *to, int len) + * + * o32 mandates we leave 4 spaces for argument registers in case + * the callee needs to use them. Even though we don't care about + * the argument registers ourselves, we need to allocate that space + * to remain ABI compliant since the callee may want to use that space. + * We also allocate 2 more spaces for $r_ra and our return register (*to). + * + * n64 is a bit different. The *caller* will allocate the space to preserve + * the arguments. So in 64-bit kernels, we allocate the 4-arg space for no + * good reason but it does not matter that much really. + * + * (void *to) is returned in r_s0 + * + */ +#ifdef CONFIG_CPU_LITTLE_ENDIAN +#define DS_OFFSET(SIZE) (4 * SZREG) +#else +#define DS_OFFSET(SIZE) ((4 * SZREG) + (4 - SIZE)) +#endif +#define bpf_slow_path_common(SIZE) \ + /* Quick check. Are we within reasonable boundaries? */ \ + LONG_ADDIU $r_s1, $r_skb_len, -SIZE; \ + sltu $r_s0, offset, $r_s1; \ + beqz $r_s0, fault; \ + /* Load 4th argument in DS */ \ + LONG_ADDIU a3, zero, SIZE; \ + PTR_ADDIU $r_sp, $r_sp, -(6 * SZREG); \ + PTR_LA t0, skb_copy_bits; \ + PTR_S $r_ra, (5 * SZREG)($r_sp); \ + /* Assign low slot to a2 */ \ + PTR_ADDIU a2, $r_sp, DS_OFFSET(SIZE); \ + jalr t0; \ + /* Reset our destination slot (DS but it's ok) */ \ + INT_S zero, (4 * SZREG)($r_sp); \ + /* \ + * skb_copy_bits returns 0 on success and -EFAULT \ + * on error. Our data live in a2. Do not bother with \ + * our data if an error has been returned. \ + */ \ + /* Restore our frame */ \ + PTR_L $r_ra, (5 * SZREG)($r_sp); \ + INT_L $r_s0, (4 * SZREG)($r_sp); \ + bltz v0, fault; \ + PTR_ADDIU $r_sp, $r_sp, 6 * SZREG; \ + move $r_ret, zero; \ + +NESTED(bpf_slow_path_word, (6 * SZREG), $r_sp) + bpf_slow_path_common(4) +#ifdef CONFIG_CPU_LITTLE_ENDIAN +# if MIPS_ISA_REV >= 2 + wsbh t0, $r_s0 + jr $r_ra + rotr $r_A, t0, 16 +# else + sll t0, $r_s0, 24 + srl t1, $r_s0, 24 + srl t2, $r_s0, 8 + or t0, t0, t1 + andi t2, t2, 0xff00 + andi t1, $r_s0, 0xff00 + or t0, t0, t2 + sll t1, t1, 8 + jr $r_ra + or $r_A, t0, t1 +# endif +#else + jr $r_ra + move $r_A, $r_s0 +#endif + + END(bpf_slow_path_word) + +NESTED(bpf_slow_path_half, (6 * SZREG), $r_sp) + bpf_slow_path_common(2) +#ifdef CONFIG_CPU_LITTLE_ENDIAN +# if MIPS_ISA_REV >= 2 + jr $r_ra + wsbh $r_A, $r_s0 +# else + sll t0, $r_s0, 8 + andi t1, $r_s0, 0xff00 + andi t0, t0, 0xff00 + srl t1, t1, 8 + jr $r_ra + or $r_A, t0, t1 +# endif +#else + jr $r_ra + move $r_A, $r_s0 +#endif + + END(bpf_slow_path_half) + +NESTED(bpf_slow_path_byte, (6 * SZREG), $r_sp) + bpf_slow_path_common(1) + jr $r_ra + move $r_A, $r_s0 + + END(bpf_slow_path_byte) + +/* + * Negative entry points + */ + .macro bpf_is_end_of_data + li t0, SKF_LL_OFF + /* Reading link layer data? */ + slt t1, offset, t0 + bgtz t1, fault + /* Be careful what follows in DS. */ + .endm +/* + * call skb_copy_bits: + * (prototype in linux/filter.h) + * + * void *bpf_internal_load_pointer_neg_helper(const struct sk_buff *skb, + * int k, unsigned int size) + * + * see above (bpf_slow_path_common) for ABI restrictions + */ +#define bpf_negative_common(SIZE) \ + PTR_ADDIU $r_sp, $r_sp, -(6 * SZREG); \ + PTR_LA t0, bpf_internal_load_pointer_neg_helper; \ + PTR_S $r_ra, (5 * SZREG)($r_sp); \ + jalr t0; \ + li a2, SIZE; \ + PTR_L $r_ra, (5 * SZREG)($r_sp); \ + /* Check return pointer */ \ + beqz v0, fault; \ + PTR_ADDIU $r_sp, $r_sp, 6 * SZREG; \ + /* Preserve our pointer */ \ + move $r_s0, v0; \ + /* Set return value */ \ + move $r_ret, zero; \ + +bpf_slow_path_word_neg: + bpf_is_end_of_data +NESTED(sk_load_word_negative, (6 * SZREG), $r_sp) + bpf_negative_common(4) + jr $r_ra + lw $r_A, 0($r_s0) + END(sk_load_word_negative) + +bpf_slow_path_half_neg: + bpf_is_end_of_data +NESTED(sk_load_half_negative, (6 * SZREG), $r_sp) + bpf_negative_common(2) + jr $r_ra + lhu $r_A, 0($r_s0) + END(sk_load_half_negative) + +bpf_slow_path_byte_neg: + bpf_is_end_of_data +NESTED(sk_load_byte_negative, (6 * SZREG), $r_sp) + bpf_negative_common(1) + jr $r_ra + lbu $r_A, 0($r_s0) + END(sk_load_byte_negative) + +fault: + jr $r_ra + addiu $r_ret, zero, 1 diff --git a/arch/mips/net/ebpf_jit.c b/arch/mips/net/ebpf_jit.c index 561154cbcc40..b31b91e57c34 100644 --- a/arch/mips/net/ebpf_jit.c +++ b/arch/mips/net/ebpf_jit.c @@ -1355,6 +1355,9 @@ jeq_common: } break; + case BPF_ST | BPF_NOSPEC: /* speculation barrier */ + break; + case BPF_ST | BPF_B | BPF_MEM: case BPF_ST | BPF_H | BPF_MEM: case BPF_ST | BPF_W | BPF_MEM: diff --git a/arch/mips/ralink/ill_acc.c b/arch/mips/ralink/ill_acc.c index 0ddeb31afa93..45ca2b84f096 100644 --- a/arch/mips/ralink/ill_acc.c +++ b/arch/mips/ralink/ill_acc.c @@ -61,6 +61,7 @@ static int __init ill_acc_of_setup(void) pdev = of_find_device_by_node(np); if (!pdev) { pr_err("%pOFn: failed to lookup pdev\n", np); + of_node_put(np); return -EINVAL; } diff --git a/arch/mips/rb532/devices.c b/arch/mips/rb532/devices.c index c9ecf17f8660..74808619fefe 100644 --- a/arch/mips/rb532/devices.c +++ b/arch/mips/rb532/devices.c @@ -310,11 +310,9 @@ static int __init plat_setup_devices(void) static int __init setup_kmac(char *s) { printk(KERN_INFO "korina mac = %s\n", s); - if (!mac_pton(s, korina_dev0_data.mac)) { + if (!mac_pton(s, korina_dev0_data.mac)) printk(KERN_ERR "Invalid mac\n"); - return -EINVAL; - } - return 0; + return 1; } __setup("kmac=", setup_kmac); diff --git a/arch/mips/sni/time.c b/arch/mips/sni/time.c index dbace1f3e1a9..745ceb945fc5 100644 --- a/arch/mips/sni/time.c +++ b/arch/mips/sni/time.c @@ -18,14 +18,14 @@ static int a20r_set_periodic(struct clock_event_device *evt) { *(volatile u8 *)(A20R_PT_CLOCK_BASE + 12) = 0x34; wmb(); - *(volatile u8 *)(A20R_PT_CLOCK_BASE + 0) = SNI_COUNTER0_DIV; + *(volatile u8 *)(A20R_PT_CLOCK_BASE + 0) = SNI_COUNTER0_DIV & 0xff; wmb(); *(volatile u8 *)(A20R_PT_CLOCK_BASE + 0) = SNI_COUNTER0_DIV >> 8; wmb(); *(volatile u8 *)(A20R_PT_CLOCK_BASE + 12) = 0xb4; wmb(); - *(volatile u8 *)(A20R_PT_CLOCK_BASE + 8) = SNI_COUNTER2_DIV; + *(volatile u8 *)(A20R_PT_CLOCK_BASE + 8) = SNI_COUNTER2_DIV & 0xff; wmb(); *(volatile u8 *)(A20R_PT_CLOCK_BASE + 8) = SNI_COUNTER2_DIV >> 8; wmb(); |