diff options
Diffstat (limited to 'arch/mips/mm')
| -rw-r--r-- | arch/mips/mm/Makefile | 2 | ||||
| -rw-r--r-- | arch/mips/mm/c-octeon.c | 29 | ||||
| -rw-r--r-- | arch/mips/mm/c-r3k.c | 9 | ||||
| -rw-r--r-- | arch/mips/mm/c-r4k.c | 391 | ||||
| -rw-r--r-- | arch/mips/mm/c-tx39.c | 2 | ||||
| -rw-r--r-- | arch/mips/mm/cache.c | 12 | ||||
| -rw-r--r-- | arch/mips/mm/cerr-sb1.c | 4 | ||||
| -rw-r--r-- | arch/mips/mm/cex-sb1.S | 5 | ||||
| -rw-r--r-- | arch/mips/mm/dma-default.c | 38 | ||||
| -rw-r--r-- | arch/mips/mm/fault.c | 23 | ||||
| -rw-r--r-- | arch/mips/mm/gup.c | 3 | ||||
| -rw-r--r-- | arch/mips/mm/hugetlbpage.c | 1 | ||||
| -rw-r--r-- | arch/mips/mm/init.c | 162 | ||||
| -rw-r--r-- | arch/mips/mm/mmap.c | 2 | ||||
| -rw-r--r-- | arch/mips/mm/page.c | 48 | ||||
| -rw-r--r-- | arch/mips/mm/sc-ip22.c | 2 | ||||
| -rw-r--r-- | arch/mips/mm/sc-mips.c | 9 | ||||
| -rw-r--r-- | arch/mips/mm/sc-r5k.c | 2 | ||||
| -rw-r--r-- | arch/mips/mm/sc-rm7k.c | 13 | ||||
| -rw-r--r-- | arch/mips/mm/tlb-funcs.S | 39 | ||||
| -rw-r--r-- | arch/mips/mm/tlb-r3k.c | 3 | ||||
| -rw-r--r-- | arch/mips/mm/tlb-r4k.c | 179 | ||||
| -rw-r--r-- | arch/mips/mm/tlb-r8k.c | 5 | ||||
| -rw-r--r-- | arch/mips/mm/tlbex.c | 638 | ||||
| -rw-r--r-- | arch/mips/mm/uasm-micromips.c | 26 | ||||
| -rw-r--r-- | arch/mips/mm/uasm-mips.c | 27 | ||||
| -rw-r--r-- | arch/mips/mm/uasm.c | 165 |
27 files changed, 1104 insertions, 735 deletions
diff --git a/arch/mips/mm/Makefile b/arch/mips/mm/Makefile index e87aae1f2e8..7f4f93ab22b 100644 --- a/arch/mips/mm/Makefile +++ b/arch/mips/mm/Makefile @@ -4,7 +4,7 @@ obj-y += cache.o dma-default.o extable.o fault.o \ gup.o init.o mmap.o page.o page-funcs.o \ - tlbex.o tlbex-fault.o uasm-mips.o + tlbex.o tlbex-fault.o tlb-funcs.o uasm-mips.o obj-$(CONFIG_32BIT) += ioremap.o pgtable-32.o obj-$(CONFIG_64BIT) += pgtable-64.o diff --git a/arch/mips/mm/c-octeon.c b/arch/mips/mm/c-octeon.c index 8557fb55286..05b1d7cf951 100644 --- a/arch/mips/mm/c-octeon.c +++ b/arch/mips/mm/c-octeon.c @@ -6,7 +6,6 @@ * Copyright (C) 2005-2007 Cavium Networks */ #include <linux/export.h> -#include <linux/init.h> #include <linux/kernel.h> #include <linux/sched.h> #include <linux/smp.h> @@ -19,6 +18,7 @@ #include <asm/bootinfo.h> #include <asm/cacheops.h> #include <asm/cpu-features.h> +#include <asm/cpu-type.h> #include <asm/page.h> #include <asm/pgtable.h> #include <asm/r4kcache.h> @@ -137,8 +137,10 @@ static void octeon_flush_cache_sigtramp(unsigned long addr) { struct vm_area_struct *vma; + down_read(¤t->mm->mmap_sem); vma = find_vma(current->mm, addr); octeon_flush_icache_all_cores(vma); + up_read(¤t->mm->mmap_sem); } @@ -180,15 +182,16 @@ static void octeon_flush_kernel_vmap_range(unsigned long vaddr, int size) * Probe Octeon's caches * */ -static void __cpuinit probe_octeon(void) +static void probe_octeon(void) { unsigned long icache_size; unsigned long dcache_size; unsigned int config1; struct cpuinfo_mips *c = ¤t_cpu_data; + int cputype = current_cpu_type(); config1 = read_c0_config1(); - switch (c->cputype) { + switch (cputype) { case CPU_CAVIUM_OCTEON: case CPU_CAVIUM_OCTEON_PLUS: c->icache.linesz = 2 << ((config1 >> 19) & 7); @@ -199,7 +202,7 @@ static void __cpuinit probe_octeon(void) c->icache.sets * c->icache.ways * c->icache.linesz; c->icache.waybit = ffs(icache_size / c->icache.ways) - 1; c->dcache.linesz = 128; - if (c->cputype == CPU_CAVIUM_OCTEON_PLUS) + if (cputype == CPU_CAVIUM_OCTEON_PLUS) c->dcache.sets = 2; /* CN5XXX has two Dcache sets */ else c->dcache.sets = 1; /* CN3XXX has one Dcache set */ @@ -224,6 +227,20 @@ static void __cpuinit probe_octeon(void) c->options |= MIPS_CPU_PREFETCH; break; + case CPU_CAVIUM_OCTEON3: + c->icache.linesz = 128; + c->icache.sets = 16; + c->icache.ways = 39; + c->icache.flags |= MIPS_CACHE_VTAG; + icache_size = c->icache.sets * c->icache.ways * c->icache.linesz; + + c->dcache.linesz = 128; + c->dcache.ways = 32; + c->dcache.sets = 8; + dcache_size = c->dcache.sets * c->dcache.ways * c->dcache.linesz; + c->options |= MIPS_CPU_PREFETCH; + break; + default: panic("Unsupported Cavium Networks CPU type"); break; @@ -251,7 +268,7 @@ static void __cpuinit probe_octeon(void) } } -static void __cpuinit octeon_cache_error_setup(void) +static void octeon_cache_error_setup(void) { extern char except_vec2_octeon; set_handler(0x100, &except_vec2_octeon, 0x80); @@ -261,7 +278,7 @@ static void __cpuinit octeon_cache_error_setup(void) * Setup the Octeon cache flush routines * */ -void __cpuinit octeon_cache_init(void) +void octeon_cache_init(void) { probe_octeon(); diff --git a/arch/mips/mm/c-r3k.c b/arch/mips/mm/c-r3k.c index 704dc735a59..135ec313c1f 100644 --- a/arch/mips/mm/c-r3k.c +++ b/arch/mips/mm/c-r3k.c @@ -9,7 +9,6 @@ * Copyright (C) 1998 Gleb Raiko & Vladimir Roganov * Copyright (C) 2001, 2004, 2007 Maciej W. Rozycki */ -#include <linux/init.h> #include <linux/kernel.h> #include <linux/sched.h> #include <linux/smp.h> @@ -26,7 +25,7 @@ static unsigned long icache_size, dcache_size; /* Size in bytes */ static unsigned long icache_lsize, dcache_lsize; /* Size in bytes */ -unsigned long __cpuinit r3k_cache_size(unsigned long ca_flags) +unsigned long r3k_cache_size(unsigned long ca_flags) { unsigned long flags, status, dummy, size; volatile unsigned long *p; @@ -61,7 +60,7 @@ unsigned long __cpuinit r3k_cache_size(unsigned long ca_flags) return size * sizeof(*p); } -unsigned long __cpuinit r3k_cache_lsize(unsigned long ca_flags) +unsigned long r3k_cache_lsize(unsigned long ca_flags) { unsigned long flags, status, lsize, i; volatile unsigned long *p; @@ -90,7 +89,7 @@ unsigned long __cpuinit r3k_cache_lsize(unsigned long ca_flags) return lsize * sizeof(*p); } -static void __cpuinit r3k_probe_cache(void) +static void r3k_probe_cache(void) { dcache_size = r3k_cache_size(ST0_ISC); if (dcache_size) @@ -312,7 +311,7 @@ static void r3k_dma_cache_wback_inv(unsigned long start, unsigned long size) r3k_flush_dcache_range(start, start + size); } -void __cpuinit r3k_cache_init(void) +void r3k_cache_init(void) { extern void build_clear_page(void); extern void build_copy_page(void); diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c index 21813beec7a..f2e8302fa70 100644 --- a/arch/mips/mm/c-r4k.c +++ b/arch/mips/mm/c-r4k.c @@ -7,11 +7,13 @@ * Copyright (C) 1997, 1998, 1999, 2000, 2001, 2002 Ralf Baechle (ralf@gnu.org) * Copyright (C) 1999, 2000 Silicon Graphics, Inc. */ +#include <linux/cpu_pm.h> #include <linux/hardirq.h> #include <linux/init.h> #include <linux/highmem.h> #include <linux/kernel.h> #include <linux/linkage.h> +#include <linux/preempt.h> #include <linux/sched.h> #include <linux/smp.h> #include <linux/mm.h> @@ -24,6 +26,7 @@ #include <asm/cacheops.h> #include <asm/cpu.h> #include <asm/cpu-features.h> +#include <asm/cpu-type.h> #include <asm/io.h> #include <asm/page.h> #include <asm/pgtable.h> @@ -48,14 +51,14 @@ static inline void r4k_on_each_cpu(void (*func) (void *info), void *info) { preempt_disable(); -#if !defined(CONFIG_MIPS_MT_SMP) && !defined(CONFIG_MIPS_MT_SMTC) +#ifndef CONFIG_MIPS_MT_SMP smp_call_function(func, info, 1); #endif func(info); preempt_enable(); } -#if defined(CONFIG_MIPS_CMP) +#if defined(CONFIG_MIPS_CMP) || defined(CONFIG_MIPS_CPS) #define cpu_has_safe_index_cacheops 0 #else #define cpu_has_safe_index_cacheops 1 @@ -103,27 +106,64 @@ static inline void r4k_blast_dcache_page_dc32(unsigned long addr) static inline void r4k_blast_dcache_page_dc64(unsigned long addr) { - R4600_HIT_CACHEOP_WAR_IMPL; blast_dcache64_page(addr); } -static void __cpuinit r4k_blast_dcache_page_setup(void) +static inline void r4k_blast_dcache_page_dc128(unsigned long addr) +{ + blast_dcache128_page(addr); +} + +static void r4k_blast_dcache_page_setup(void) { unsigned long dc_lsize = cpu_dcache_line_size(); - if (dc_lsize == 0) + switch (dc_lsize) { + case 0: r4k_blast_dcache_page = (void *)cache_noop; - else if (dc_lsize == 16) + break; + case 16: r4k_blast_dcache_page = blast_dcache16_page; - else if (dc_lsize == 32) + break; + case 32: r4k_blast_dcache_page = r4k_blast_dcache_page_dc32; - else if (dc_lsize == 64) + break; + case 64: r4k_blast_dcache_page = r4k_blast_dcache_page_dc64; + break; + case 128: + r4k_blast_dcache_page = r4k_blast_dcache_page_dc128; + break; + default: + break; + } +} + +#ifndef CONFIG_EVA +#define r4k_blast_dcache_user_page r4k_blast_dcache_page +#else + +static void (*r4k_blast_dcache_user_page)(unsigned long addr); + +static void r4k_blast_dcache_user_page_setup(void) +{ + unsigned long dc_lsize = cpu_dcache_line_size(); + + if (dc_lsize == 0) + r4k_blast_dcache_user_page = (void *)cache_noop; + else if (dc_lsize == 16) + r4k_blast_dcache_user_page = blast_dcache16_user_page; + else if (dc_lsize == 32) + r4k_blast_dcache_user_page = blast_dcache32_user_page; + else if (dc_lsize == 64) + r4k_blast_dcache_user_page = blast_dcache64_user_page; } +#endif + static void (* r4k_blast_dcache_page_indexed)(unsigned long addr); -static void __cpuinit r4k_blast_dcache_page_indexed_setup(void) +static void r4k_blast_dcache_page_indexed_setup(void) { unsigned long dc_lsize = cpu_dcache_line_size(); @@ -135,12 +175,14 @@ static void __cpuinit r4k_blast_dcache_page_indexed_setup(void) r4k_blast_dcache_page_indexed = blast_dcache32_page_indexed; else if (dc_lsize == 64) r4k_blast_dcache_page_indexed = blast_dcache64_page_indexed; + else if (dc_lsize == 128) + r4k_blast_dcache_page_indexed = blast_dcache128_page_indexed; } void (* r4k_blast_dcache)(void); EXPORT_SYMBOL(r4k_blast_dcache); -static void __cpuinit r4k_blast_dcache_setup(void) +static void r4k_blast_dcache_setup(void) { unsigned long dc_lsize = cpu_dcache_line_size(); @@ -152,6 +194,8 @@ static void __cpuinit r4k_blast_dcache_setup(void) r4k_blast_dcache = blast_dcache32; else if (dc_lsize == 64) r4k_blast_dcache = blast_dcache64; + else if (dc_lsize == 128) + r4k_blast_dcache = blast_dcache128; } /* force code alignment (used for TX49XX_ICACHE_INDEX_INV_WAR) */ @@ -227,7 +271,7 @@ static inline void tx49_blast_icache32_page_indexed(unsigned long page) static void (* r4k_blast_icache_page)(unsigned long addr); -static void __cpuinit r4k_blast_icache_page_setup(void) +static void r4k_blast_icache_page_setup(void) { unsigned long ic_lsize = cpu_icache_line_size(); @@ -235,16 +279,41 @@ static void __cpuinit r4k_blast_icache_page_setup(void) r4k_blast_icache_page = (void *)cache_noop; else if (ic_lsize == 16) r4k_blast_icache_page = blast_icache16_page; + else if (ic_lsize == 32 && current_cpu_type() == CPU_LOONGSON2) + r4k_blast_icache_page = loongson2_blast_icache32_page; else if (ic_lsize == 32) r4k_blast_icache_page = blast_icache32_page; else if (ic_lsize == 64) r4k_blast_icache_page = blast_icache64_page; + else if (ic_lsize == 128) + r4k_blast_icache_page = blast_icache128_page; } +#ifndef CONFIG_EVA +#define r4k_blast_icache_user_page r4k_blast_icache_page +#else + +static void (*r4k_blast_icache_user_page)(unsigned long addr); + +static void __cpuinit r4k_blast_icache_user_page_setup(void) +{ + unsigned long ic_lsize = cpu_icache_line_size(); + + if (ic_lsize == 0) + r4k_blast_icache_user_page = (void *)cache_noop; + else if (ic_lsize == 16) + r4k_blast_icache_user_page = blast_icache16_user_page; + else if (ic_lsize == 32) + r4k_blast_icache_user_page = blast_icache32_user_page; + else if (ic_lsize == 64) + r4k_blast_icache_user_page = blast_icache64_user_page; +} + +#endif static void (* r4k_blast_icache_page_indexed)(unsigned long addr); -static void __cpuinit r4k_blast_icache_page_indexed_setup(void) +static void r4k_blast_icache_page_indexed_setup(void) { unsigned long ic_lsize = cpu_icache_line_size(); @@ -259,6 +328,9 @@ static void __cpuinit r4k_blast_icache_page_indexed_setup(void) else if (TX49XX_ICACHE_INDEX_INV_WAR) r4k_blast_icache_page_indexed = tx49_blast_icache32_page_indexed; + else if (current_cpu_type() == CPU_LOONGSON2) + r4k_blast_icache_page_indexed = + loongson2_blast_icache32_page_indexed; else r4k_blast_icache_page_indexed = blast_icache32_page_indexed; @@ -269,7 +341,7 @@ static void __cpuinit r4k_blast_icache_page_indexed_setup(void) void (* r4k_blast_icache)(void); EXPORT_SYMBOL(r4k_blast_icache); -static void __cpuinit r4k_blast_icache_setup(void) +static void r4k_blast_icache_setup(void) { unsigned long ic_lsize = cpu_icache_line_size(); @@ -282,15 +354,19 @@ static void __cpuinit r4k_blast_icache_setup(void) r4k_blast_icache = blast_r4600_v1_icache32; else if (TX49XX_ICACHE_INDEX_INV_WAR) r4k_blast_icache = tx49_blast_icache32; + else if (current_cpu_type() == CPU_LOONGSON2) + r4k_blast_icache = loongson2_blast_icache32; else r4k_blast_icache = blast_icache32; } else if (ic_lsize == 64) r4k_blast_icache = blast_icache64; + else if (ic_lsize == 128) + r4k_blast_icache = blast_icache128; } static void (* r4k_blast_scache_page)(unsigned long addr); -static void __cpuinit r4k_blast_scache_page_setup(void) +static void r4k_blast_scache_page_setup(void) { unsigned long sc_lsize = cpu_scache_line_size(); @@ -308,7 +384,7 @@ static void __cpuinit r4k_blast_scache_page_setup(void) static void (* r4k_blast_scache_page_indexed)(unsigned long addr); -static void __cpuinit r4k_blast_scache_page_indexed_setup(void) +static void r4k_blast_scache_page_indexed_setup(void) { unsigned long sc_lsize = cpu_scache_line_size(); @@ -326,7 +402,7 @@ static void __cpuinit r4k_blast_scache_page_indexed_setup(void) static void (* r4k_blast_scache)(void); -static void __cpuinit r4k_blast_scache_setup(void) +static void r4k_blast_scache_setup(void) { unsigned long sc_lsize = cpu_scache_line_size(); @@ -344,14 +420,9 @@ static void __cpuinit r4k_blast_scache_setup(void) static inline void local_r4k___flush_cache_all(void * args) { -#if defined(CONFIG_CPU_LOONGSON2) - r4k_blast_scache(); - return; -#endif - r4k_blast_dcache(); - r4k_blast_icache(); - switch (current_cpu_type()) { + case CPU_LOONGSON2: + case CPU_LOONGSON3: case CPU_R4000SC: case CPU_R4000MC: case CPU_R4400SC: @@ -359,7 +430,18 @@ static inline void local_r4k___flush_cache_all(void * args) case CPU_R10000: case CPU_R12000: case CPU_R14000: + /* + * These caches are inclusive caches, that is, if something + * is not cached in the S-cache, we know it also won't be + * in one of the primary caches. + */ r4k_blast_scache(); + break; + + default: + r4k_blast_dcache(); + r4k_blast_icache(); + break; } } @@ -370,7 +452,7 @@ static void r4k___flush_cache_all(void) static inline int has_valid_asid(const struct mm_struct *mm) { -#if defined(CONFIG_MIPS_MT_SMP) || defined(CONFIG_MIPS_MT_SMTC) +#ifdef CONFIG_MIPS_MT_SMP int i; for_each_online_cpu(i) @@ -505,7 +587,8 @@ static inline void local_r4k_flush_cache_page(void *args) } if (cpu_has_dc_aliases || (exec && !cpu_has_ic_fills_f_dc)) { - r4k_blast_dcache_page(addr); + vaddr ? r4k_blast_dcache_page(addr) : + r4k_blast_dcache_user_page(addr); if (exec && !cpu_icache_snoops_remote_store) r4k_blast_scache_page(addr); } @@ -516,7 +599,8 @@ static inline void local_r4k_flush_cache_page(void *args) if (cpu_context(cpu, mm) != 0) drop_mmu_context(mm, cpu); } else - r4k_blast_icache_page(addr); + vaddr ? r4k_blast_icache_page(addr) : + r4k_blast_icache_user_page(addr); } if (vaddr) { @@ -570,8 +654,28 @@ static inline void local_r4k_flush_icache_range(unsigned long start, unsigned lo if (end - start > icache_size) r4k_blast_icache(); - else - protected_blast_icache_range(start, end); + else { + switch (boot_cpu_type()) { + case CPU_LOONGSON2: + protected_loongson2_blast_icache_range(start, end); + break; + + default: + protected_blast_icache_range(start, end); + break; + } + } +#ifdef CONFIG_EVA + /* + * Due to all possible segment mappings, there might cache aliases + * caused by the bootloader being in non-EVA mode, and the CPU switching + * to EVA during early kernel init. It's best to flush the scache + * to avoid having secondary cores fetching stale data and lead to + * kernel crashes. + */ + bc_wback_inv(start, (end - start)); + __sync(); +#endif } static inline void local_r4k_flush_icache_range_ipi(void *args) @@ -594,18 +698,20 @@ static void r4k_flush_icache_range(unsigned long start, unsigned long end) instruction_hazard(); } -#ifdef CONFIG_DMA_NONCOHERENT +#if defined(CONFIG_DMA_NONCOHERENT) || defined(CONFIG_DMA_MAYBE_COHERENT) static void r4k_dma_cache_wback_inv(unsigned long addr, unsigned long size) { /* Catch bad driver code */ BUG_ON(size == 0); + preempt_disable(); if (cpu_has_inclusive_pcaches) { if (size >= scache_size) r4k_blast_scache(); else blast_scache_range(addr, addr + size); + preempt_enable(); __sync(); return; } @@ -621,6 +727,7 @@ static void r4k_dma_cache_wback_inv(unsigned long addr, unsigned long size) R4600_HIT_CACHEOP_WAR_IMPL; blast_dcache_range(addr, addr + size); } + preempt_enable(); bc_wback_inv(addr, size); __sync(); @@ -631,6 +738,7 @@ static void r4k_dma_cache_inv(unsigned long addr, unsigned long size) /* Catch bad driver code */ BUG_ON(size == 0); + preempt_disable(); if (cpu_has_inclusive_pcaches) { if (size >= scache_size) r4k_blast_scache(); @@ -645,6 +753,7 @@ static void r4k_dma_cache_inv(unsigned long addr, unsigned long size) */ blast_inv_scache_range(addr, addr + size); } + preempt_enable(); __sync(); return; } @@ -655,11 +764,12 @@ static void r4k_dma_cache_inv(unsigned long addr, unsigned long size) R4600_HIT_CACHEOP_WAR_IMPL; blast_inv_dcache_range(addr, addr + size); } + preempt_enable(); bc_inv(addr, size); __sync(); } -#endif /* CONFIG_DMA_NONCOHERENT */ +#endif /* CONFIG_DMA_NONCOHERENT || CONFIG_DMA_MAYBE_COHERENT */ /* * While we're protected against bad userland addresses we don't care @@ -780,28 +890,38 @@ static inline void rm7k_erratum31(void) static inline void alias_74k_erratum(struct cpuinfo_mips *c) { + unsigned int imp = c->processor_id & PRID_IMP_MASK; + unsigned int rev = c->processor_id & PRID_REV_MASK; + /* * Early versions of the 74K do not update the cache tags on a * vtag miss/ptag hit which can occur in the case of KSEG0/KUSEG * aliases. In this case it is better to treat the cache as always * having aliases. */ - if ((c->processor_id & 0xff) <= PRID_REV_ENCODE_332(2, 4, 0)) - c->dcache.flags |= MIPS_CACHE_VTAG; - if ((c->processor_id & 0xff) == PRID_REV_ENCODE_332(2, 4, 0)) - write_c0_config6(read_c0_config6() | MIPS_CONF6_SYND); - if (((c->processor_id & 0xff00) == PRID_IMP_1074K) && - ((c->processor_id & 0xff) <= PRID_REV_ENCODE_332(1, 1, 0))) { - c->dcache.flags |= MIPS_CACHE_VTAG; - write_c0_config6(read_c0_config6() | MIPS_CONF6_SYND); + switch (imp) { + case PRID_IMP_74K: + if (rev <= PRID_REV_ENCODE_332(2, 4, 0)) + c->dcache.flags |= MIPS_CACHE_VTAG; + if (rev == PRID_REV_ENCODE_332(2, 4, 0)) + write_c0_config6(read_c0_config6() | MIPS_CONF6_SYND); + break; + case PRID_IMP_1074K: + if (rev <= PRID_REV_ENCODE_332(1, 1, 0)) { + c->dcache.flags |= MIPS_CACHE_VTAG; + write_c0_config6(read_c0_config6() | MIPS_CONF6_SYND); + } + break; + default: + BUG(); } } -static char *way_string[] __cpuinitdata = { NULL, "direct mapped", "2-way", +static char *way_string[] = { NULL, "direct mapped", "2-way", "3-way", "4-way", "5-way", "6-way", "7-way", "8-way" }; -static void __cpuinit probe_pcache(void) +static void probe_pcache(void) { struct cpuinfo_mips *c = ¤t_cpu_data; unsigned int config = read_c0_config(); @@ -809,7 +929,7 @@ static void __cpuinit probe_pcache(void) unsigned long config1; unsigned int lsize; - switch (c->cputype) { + switch (current_cpu_type()) { case CPU_R4600: /* QED style two way caches? */ case CPU_R4700: case CPU_R5000: @@ -971,6 +1091,48 @@ static void __cpuinit probe_pcache(void) c->dcache.waybit = 0; break; + case CPU_LOONGSON3: + config1 = read_c0_config1(); + lsize = (config1 >> 19) & 7; + if (lsize) + c->icache.linesz = 2 << lsize; + else + c->icache.linesz = 0; + c->icache.sets = 64 << ((config1 >> 22) & 7); + c->icache.ways = 1 + ((config1 >> 16) & 7); + icache_size = c->icache.sets * + c->icache.ways * + c->icache.linesz; + c->icache.waybit = 0; + + lsize = (config1 >> 10) & 7; + if (lsize) + c->dcache.linesz = 2 << lsize; + else + c->dcache.linesz = 0; + c->dcache.sets = 64 << ((config1 >> 13) & 7); + c->dcache.ways = 1 + ((config1 >> 7) & 7); + dcache_size = c->dcache.sets * + c->dcache.ways * + c->dcache.linesz; + c->dcache.waybit = 0; + break; + + case CPU_CAVIUM_OCTEON3: + /* For now lie about the number of ways. */ + c->icache.linesz = 128; + c->icache.sets = 16; + c->icache.ways = 8; + c->icache.flags |= MIPS_CACHE_VTAG; + icache_size = c->icache.sets * c->icache.ways * c->icache.linesz; + + c->dcache.linesz = 128; + c->dcache.ways = 8; + c->dcache.sets = 8; + dcache_size = c->dcache.sets * c->dcache.ways * c->dcache.linesz; + c->options |= MIPS_CPU_PREFETCH; + break; + default: if (!(config & MIPS_CONF_M)) panic("Don't know how to probe P-caches on this cpu."); @@ -981,10 +1143,14 @@ static void __cpuinit probe_pcache(void) */ config1 = read_c0_config1(); - if ((lsize = ((config1 >> 19) & 7))) - c->icache.linesz = 2 << lsize; - else - c->icache.linesz = lsize; + lsize = (config1 >> 19) & 7; + + /* IL == 7 is reserved */ + if (lsize == 7) + panic("Invalid icache line size"); + + c->icache.linesz = lsize ? 2 << lsize : 0; + c->icache.sets = 32 << (((config1 >> 22) + 1) & 7); c->icache.ways = 1 + ((config1 >> 16) & 7); @@ -1001,10 +1167,14 @@ static void __cpuinit probe_pcache(void) */ c->dcache.flags = 0; - if ((lsize = ((config1 >> 10) & 7))) - c->dcache.linesz = 2 << lsize; - else - c->dcache.linesz= lsize; + lsize = (config1 >> 10) & 7; + + /* DL == 7 is reserved */ + if (lsize == 7) + panic("Invalid dcache line size"); + + c->dcache.linesz = lsize ? 2 << lsize : 0; + c->dcache.sets = 32 << (((config1 >> 13) + 1) & 7); c->dcache.ways = 1 + ((config1 >> 7) & 7); @@ -1025,7 +1195,8 @@ static void __cpuinit probe_pcache(void) * presumably no vendor is shipping his hardware in the "bad" * configuration. */ - if ((prid & 0xff00) == PRID_IMP_R4000 && (prid & 0xff) < 0x40 && + if ((prid & PRID_IMP_MASK) == PRID_IMP_R4000 && + (prid & PRID_REV_MASK) < PRID_REV_R4400 && !(config & CONF_SC) && c->icache.linesz != 16 && PAGE_SIZE <= 0x8000) panic("Improper R4000SC processor configuration detected"); @@ -1045,7 +1216,7 @@ static void __cpuinit probe_pcache(void) * normally they'd suffer from aliases but magic in the hardware deals * with that for us so we don't need to take care ourselves. */ - switch (c->cputype) { + switch (current_cpu_type()) { case CPU_20KC: case CPU_25KF: case CPU_SB1: @@ -1065,11 +1236,21 @@ static void __cpuinit probe_pcache(void) case CPU_34K: case CPU_74K: case CPU_1004K: - if (c->cputype == CPU_74K) + case CPU_1074K: + case CPU_INTERAPTIV: + case CPU_P5600: + case CPU_PROAPTIV: + case CPU_M5150: + if ((c->cputype == CPU_74K) || (c->cputype == CPU_1074K)) alias_74k_erratum(c); - if ((read_c0_config7() & (1 << 16))) { - /* effectively physically indexed dcache, - thus no virtual aliases. */ + if (!(read_c0_config7() & MIPS_CONF7_IAR) && + (c->icache.waysize > PAGE_SIZE)) + c->icache.flags |= MIPS_CACHE_ALIASES; + if (read_c0_config7() & MIPS_CONF7_AR) { + /* + * Effectively physically indexed dcache, + * thus no virtual aliases. + */ c->dcache.flags |= MIPS_CACHE_PINDEX; break; } @@ -1078,7 +1259,7 @@ static void __cpuinit probe_pcache(void) c->dcache.flags |= MIPS_CACHE_ALIASES; } - switch (c->cputype) { + switch (current_cpu_type()) { case CPU_20KC: /* * Some older 20Kc chips doesn't have the 'VI' bit in @@ -1090,15 +1271,14 @@ static void __cpuinit probe_pcache(void) case CPU_ALCHEMY: c->icache.flags |= MIPS_CACHE_IC_F_DC; break; - } -#ifdef CONFIG_CPU_LOONGSON2 - /* - * LOONGSON2 has 4 way icache, but when using indexed cache op, - * one op will act on all 4 ways - */ - c->icache.ways = 1; -#endif + case CPU_LOONGSON2: + /* + * LOONGSON2 has 4 way icache, but when using indexed cache op, + * one op will act on all 4 ways + */ + c->icache.ways = 1; + } printk("Primary instruction cache %ldkB, %s, %s, linesize %d bytes.\n", icache_size >> 10, @@ -1119,7 +1299,7 @@ static void __cpuinit probe_pcache(void) * executes in KSEG1 space or else you will crash and burn badly. You have * been warned. */ -static int __cpuinit probe_scache(void) +static int probe_scache(void) { unsigned long flags, addr, begin, end, pow2; unsigned int config = read_c0_config(); @@ -1174,7 +1354,6 @@ static int __cpuinit probe_scache(void) return 1; } -#if defined(CONFIG_CPU_LOONGSON2) static void __init loongson2_sc_init(void) { struct cpuinfo_mips *c = ¤t_cpu_data; @@ -1190,13 +1369,39 @@ static void __init loongson2_sc_init(void) c->options |= MIPS_CPU_INCLUSIVE_CACHES; } -#endif + +static void __init loongson3_sc_init(void) +{ + struct cpuinfo_mips *c = ¤t_cpu_data; + unsigned int config2, lsize; + + config2 = read_c0_config2(); + lsize = (config2 >> 4) & 15; + if (lsize) + c->scache.linesz = 2 << lsize; + else + c->scache.linesz = 0; + c->scache.sets = 64 << ((config2 >> 8) & 15); + c->scache.ways = 1 + (config2 & 15); + + scache_size = c->scache.sets * + c->scache.ways * + c->scache.linesz; + /* Loongson-3 has 4 cores, 1MB scache for each. scaches are shared */ + scache_size *= 4; + c->scache.waybit = 0; + pr_info("Unified secondary cache %ldkB %s, linesize %d bytes.\n", + scache_size >> 10, way_string[c->scache.ways], c->scache.linesz); + if (scache_size) + c->options |= MIPS_CPU_INCLUSIVE_CACHES; + return; +} extern int r5k_sc_init(void); extern int rm7k_sc_init(void); extern int mips_sc_init(void); -static void __cpuinit setup_scache(void) +static void setup_scache(void) { struct cpuinfo_mips *c = ¤t_cpu_data; unsigned int config = read_c0_config(); @@ -1207,7 +1412,7 @@ static void __cpuinit setup_scache(void) * processors don't have a S-cache that would be relevant to the * Linux memory management. */ - switch (c->cputype) { + switch (current_cpu_type()) { case CPU_R4000SC: case CPU_R4000MC: case CPU_R4400SC: @@ -1240,11 +1445,15 @@ static void __cpuinit setup_scache(void) #endif return; -#if defined(CONFIG_CPU_LOONGSON2) case CPU_LOONGSON2: loongson2_sc_init(); return; -#endif + + case CPU_LOONGSON3: + loongson3_sc_init(); + return; + + case CPU_CAVIUM_OCTEON3: case CPU_XLP: /* don't need to worry about L2, fully coherent */ return; @@ -1329,7 +1538,7 @@ static void nxp_pr4450_fixup_config(void) NXP_BARRIER(); } -static int __cpuinitdata cca = -1; +static int cca = -1; static int __init cca_setup(char *str) { @@ -1340,7 +1549,7 @@ static int __init cca_setup(char *str) early_param("cca", cca_setup); -static void __cpuinit coherency_setup(void) +static void coherency_setup(void) { if (cca < 0 || cca > 7) cca = read_c0_config() & CONF_CM_CMASK; @@ -1380,13 +1589,12 @@ static void __cpuinit coherency_setup(void) } } -static void __cpuinit r4k_cache_error_setup(void) +static void r4k_cache_error_setup(void) { extern char __weak except_vec2_generic; extern char __weak except_vec2_sb1; - struct cpuinfo_mips *c = ¤t_cpu_data; - switch (c->cputype) { + switch (current_cpu_type()) { case CPU_SB1: case CPU_SB1A: set_uncached_handler(0x100, &except_vec2_sb1, 0x80); @@ -1398,7 +1606,7 @@ static void __cpuinit r4k_cache_error_setup(void) } } -void __cpuinit r4k_cache_init(void) +void r4k_cache_init(void) { extern void build_clear_page(void); extern void build_copy_page(void); @@ -1416,6 +1624,10 @@ void __cpuinit r4k_cache_init(void) r4k_blast_scache_page_setup(); r4k_blast_scache_page_indexed_setup(); r4k_blast_scache_setup(); +#ifdef CONFIG_EVA + r4k_blast_dcache_user_page_setup(); + r4k_blast_icache_user_page_setup(); +#endif /* * Some MIPS32 and MIPS64 processors have physically indexed caches. @@ -1447,7 +1659,7 @@ void __cpuinit r4k_cache_init(void) flush_icache_range = r4k_flush_icache_range; local_flush_icache_range = local_r4k_flush_icache_range; -#if defined(CONFIG_DMA_NONCOHERENT) +#if defined(CONFIG_DMA_NONCOHERENT) || defined(CONFIG_DMA_MAYBE_COHERENT) if (coherentio) { _dma_cache_wback_inv = (void *)cache_noop; _dma_cache_wback = (void *)cache_noop; @@ -1472,3 +1684,26 @@ void __cpuinit r4k_cache_init(void) coherency_setup(); board_cache_error_setup = r4k_cache_error_setup; } + +static int r4k_cache_pm_notifier(struct notifier_block *self, unsigned long cmd, + void *v) +{ + switch (cmd) { + case CPU_PM_ENTER_FAILED: + case CPU_PM_EXIT: + coherency_setup(); + break; + } + + return NOTIFY_OK; +} + +static struct notifier_block r4k_cache_pm_notifier_block = { + .notifier_call = r4k_cache_pm_notifier, +}; + +int __init r4k_cache_init_pm(void) +{ + return cpu_pm_register_notifier(&r4k_cache_pm_notifier_block); +} +arch_initcall(r4k_cache_init_pm); diff --git a/arch/mips/mm/c-tx39.c b/arch/mips/mm/c-tx39.c index ba9da270289..8d909dbbf37 100644 --- a/arch/mips/mm/c-tx39.c +++ b/arch/mips/mm/c-tx39.c @@ -344,7 +344,7 @@ static __init void tx39_probe_cache(void) } } -void __cpuinit tx39_cache_init(void) +void tx39_cache_init(void) { extern void build_clear_page(void); extern void build_copy_page(void); diff --git a/arch/mips/mm/cache.c b/arch/mips/mm/cache.c index 5aeb3eb0b72..f7b91d3a371 100644 --- a/arch/mips/mm/cache.c +++ b/arch/mips/mm/cache.c @@ -8,7 +8,6 @@ */ #include <linux/fs.h> #include <linux/fcntl.h> -#include <linux/init.h> #include <linux/kernel.h> #include <linux/linkage.h> #include <linux/module.h> @@ -30,15 +29,16 @@ void (*flush_cache_range)(struct vm_area_struct *vma, unsigned long start, void (*flush_cache_page)(struct vm_area_struct *vma, unsigned long page, unsigned long pfn); void (*flush_icache_range)(unsigned long start, unsigned long end); +EXPORT_SYMBOL_GPL(flush_icache_range); void (*local_flush_icache_range)(unsigned long start, unsigned long end); +EXPORT_SYMBOL_GPL(local_flush_icache_range); void (*__flush_cache_vmap)(void); void (*__flush_cache_vunmap)(void); void (*__flush_kernel_vmap_range)(unsigned long vaddr, int size); -void (*__invalidate_kernel_vmap_range)(unsigned long vaddr, int size); - EXPORT_SYMBOL_GPL(__flush_kernel_vmap_range); +void (*__invalidate_kernel_vmap_range)(unsigned long vaddr, int size); /* MIPS specific cache operations */ void (*flush_cache_sigtramp)(unsigned long addr); @@ -50,7 +50,7 @@ EXPORT_SYMBOL_GPL(local_flush_data_cache_page); EXPORT_SYMBOL(flush_data_cache_page); EXPORT_SYMBOL(flush_icache_all); -#ifdef CONFIG_DMA_NONCOHERENT +#if defined(CONFIG_DMA_NONCOHERENT) || defined(CONFIG_DMA_MAYBE_COHERENT) /* DMA cache operations. */ void (*_dma_cache_wback_inv)(unsigned long start, unsigned long size); @@ -59,7 +59,7 @@ void (*_dma_cache_inv)(unsigned long start, unsigned long size); EXPORT_SYMBOL(_dma_cache_wback_inv); -#endif /* CONFIG_DMA_NONCOHERENT */ +#endif /* CONFIG_DMA_NONCOHERENT || CONFIG_DMA_MAYBE_COHERENT */ /* * We could optimize the case where the cache argument is not BCACHE but @@ -182,7 +182,7 @@ static inline void setup_protection_map(void) } } -void __cpuinit cpu_cache_init(void) +void cpu_cache_init(void) { if (cpu_has_3k_cache) { extern void __weak r3k_cache_init(void); diff --git a/arch/mips/mm/cerr-sb1.c b/arch/mips/mm/cerr-sb1.c index 576add33bf5..ee5c1ff861a 100644 --- a/arch/mips/mm/cerr-sb1.c +++ b/arch/mips/mm/cerr-sb1.c @@ -182,11 +182,7 @@ asmlinkage void sb1_cache_error(void) #ifdef CONFIG_SIBYTE_BW_TRACE /* Freeze the trace buffer now */ -#if defined(CONFIG_SIBYTE_BCM1x55) || defined(CONFIG_SIBYTE_BCM1x80) - csr_out32(M_BCM1480_SCD_TRACE_CFG_FREEZE, IOADDR(A_SCD_TRACE_CFG)); -#else csr_out32(M_SCD_TRACE_CFG_FREEZE, IOADDR(A_SCD_TRACE_CFG)); -#endif printk("Trace buffer frozen\n"); #endif diff --git a/arch/mips/mm/cex-sb1.S b/arch/mips/mm/cex-sb1.S index fe1d887e8d7..5d5f29681a2 100644 --- a/arch/mips/mm/cex-sb1.S +++ b/arch/mips/mm/cex-sb1.S @@ -15,7 +15,6 @@ * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ -#include <linux/init.h> #include <asm/asm.h> #include <asm/regdef.h> @@ -49,8 +48,6 @@ * (0x170-0x17f) are used to preserve k0, k1, and ra. */ - __CPUINIT - LEAF(except_vec2_sb1) /* * If this error is recoverable, we need to exit the handler @@ -142,8 +139,6 @@ unrecoverable: END(except_vec2_sb1) - __FINIT - LEAF(handle_vec2_sb1) mfc0 k0,CP0_CONFIG li k1,~CONF_CM_CMASK diff --git a/arch/mips/mm/dma-default.c b/arch/mips/mm/dma-default.c index caf92ecb37d..44b6dff5aba 100644 --- a/arch/mips/mm/dma-default.c +++ b/arch/mips/mm/dma-default.c @@ -18,10 +18,12 @@ #include <linux/highmem.h> #include <asm/cache.h> +#include <asm/cpu-type.h> #include <asm/io.h> #include <dma-coherence.h> +#ifdef CONFIG_DMA_MAYBE_COHERENT int coherentio = 0; /* User defined DMA coherency from command line. */ EXPORT_SYMBOL_GPL(coherentio); int hw_coherentio = 0; /* Actual hardware supported DMA coherency setting. */ @@ -41,6 +43,7 @@ static int __init setnocoherentio(char *str) return 0; } early_param("nocoherentio", setnocoherentio); +#endif static inline struct page *dma_addr_to_page(struct device *dev, dma_addr_t dma_addr) @@ -50,16 +53,20 @@ static inline struct page *dma_addr_to_page(struct device *dev, } /* + * The affected CPUs below in 'cpu_needs_post_dma_flush()' can + * speculatively fill random cachelines with stale data at any time, + * requiring an extra flush post-DMA. + * * Warning on the terminology - Linux calls an uncached area coherent; * MIPS terminology calls memory areas with hardware maintained coherency * coherent. */ - -static inline int cpu_is_noncoherent_r10000(struct device *dev) +static inline int cpu_needs_post_dma_flush(struct device *dev) { return !plat_device_is_coherent(dev) && - (current_cpu_type() == CPU_R10000 || - current_cpu_type() == CPU_R12000); + (boot_cpu_type() == CPU_R10000 || + boot_cpu_type() == CPU_R12000 || + boot_cpu_type() == CPU_BMIPS5000); } static gfp_t massage_gfp_flags(const struct device *dev, gfp_t gfp) @@ -230,7 +237,7 @@ static inline void __dma_sync(struct page *page, static void mips_dma_unmap_page(struct device *dev, dma_addr_t dma_addr, size_t size, enum dma_data_direction direction, struct dma_attrs *attrs) { - if (cpu_is_noncoherent_r10000(dev)) + if (cpu_needs_post_dma_flush(dev)) __dma_sync(dma_addr_to_page(dev, dma_addr), dma_addr & ~PAGE_MASK, size, direction); @@ -246,6 +253,9 @@ static int mips_dma_map_sg(struct device *dev, struct scatterlist *sg, if (!plat_device_is_coherent(dev)) __dma_sync(sg_page(sg), sg->offset, sg->length, direction); +#ifdef CONFIG_NEED_SG_DMA_LENGTH + sg->dma_length = sg->length; +#endif sg->dma_address = plat_map_dma_mem_page(dev, sg_page(sg)) + sg->offset; } @@ -281,7 +291,7 @@ static void mips_dma_unmap_sg(struct device *dev, struct scatterlist *sg, static void mips_dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, size_t size, enum dma_data_direction direction) { - if (cpu_is_noncoherent_r10000(dev)) + if (cpu_needs_post_dma_flush(dev)) __dma_sync(dma_addr_to_page(dev, dma_handle), dma_handle & ~PAGE_MASK, size, direction); } @@ -289,7 +299,6 @@ static void mips_dma_sync_single_for_cpu(struct device *dev, static void mips_dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle, size_t size, enum dma_data_direction direction) { - plat_extra_sync_for_device(dev); if (!plat_device_is_coherent(dev)) __dma_sync(dma_addr_to_page(dev, dma_handle), dma_handle & ~PAGE_MASK, size, direction); @@ -300,12 +309,10 @@ static void mips_dma_sync_sg_for_cpu(struct device *dev, { int i; - /* Make sure that gcc doesn't leave the empty loop body. */ - for (i = 0; i < nelems; i++, sg++) { - if (cpu_is_noncoherent_r10000(dev)) + if (cpu_needs_post_dma_flush(dev)) + for (i = 0; i < nelems; i++, sg++) __dma_sync(sg_page(sg), sg->offset, sg->length, direction); - } } static void mips_dma_sync_sg_for_device(struct device *dev, @@ -313,17 +320,15 @@ static void mips_dma_sync_sg_for_device(struct device *dev, { int i; - /* Make sure that gcc doesn't leave the empty loop body. */ - for (i = 0; i < nelems; i++, sg++) { - if (!plat_device_is_coherent(dev)) + if (!plat_device_is_coherent(dev)) + for (i = 0; i < nelems; i++, sg++) __dma_sync(sg_page(sg), sg->offset, sg->length, direction); - } } int mips_dma_mapping_error(struct device *dev, dma_addr_t dma_addr) { - return plat_dma_mapping_error(dev, dma_addr); + return 0; } int mips_dma_supported(struct device *dev, u64 mask) @@ -336,7 +341,6 @@ void dma_cache_sync(struct device *dev, void *vaddr, size_t size, { BUG_ON(direction == DMA_NONE); - plat_extra_sync_for_device(dev); if (!plat_device_is_coherent(dev)) __dma_sync_virtual(vaddr, size, direction); } diff --git a/arch/mips/mm/fault.c b/arch/mips/mm/fault.c index 0fead53d1c2..becc42bb184 100644 --- a/arch/mips/mm/fault.c +++ b/arch/mips/mm/fault.c @@ -5,6 +5,7 @@ * * Copyright (C) 1995 - 2000 by Ralf Baechle */ +#include <linux/context_tracking.h> #include <linux/signal.h> #include <linux/sched.h> #include <linux/interrupt.h> @@ -32,8 +33,8 @@ * and the problem, and then passes it off to one of the appropriate * routines. */ -asmlinkage void __kprobes do_page_fault(struct pt_regs *regs, unsigned long write, - unsigned long address) +static void __kprobes __do_page_fault(struct pt_regs *regs, unsigned long write, + unsigned long address) { struct vm_area_struct * vma = NULL; struct task_struct *tsk = current; @@ -41,8 +42,7 @@ asmlinkage void __kprobes do_page_fault(struct pt_regs *regs, unsigned long writ const int field = sizeof(unsigned long) * 2; siginfo_t info; int fault; - unsigned int flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE | - (write ? FAULT_FLAG_WRITE : 0); + unsigned int flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE; #if 0 printk("Cpu%d[%s:%d:%0*lx:%ld:%0*lx]\n", raw_smp_processor_id(), @@ -92,6 +92,8 @@ asmlinkage void __kprobes do_page_fault(struct pt_regs *regs, unsigned long writ if (in_atomic() || !mm) goto bad_area_nosemaphore; + if (user_mode(regs)) + flags |= FAULT_FLAG_USER; retry: down_read(&mm->mmap_sem); vma = find_vma(mm, address); @@ -113,6 +115,7 @@ good_area: if (write) { if (!(vma->vm_flags & VM_WRITE)) goto bad_area; + flags |= FAULT_FLAG_WRITE; } else { if (cpu_has_rixi) { if (address == regs->cp0_epc && !(vma->vm_flags & VM_EXEC)) { @@ -240,6 +243,8 @@ out_of_memory: * (which will retry the fault, or kill us if we got oom-killed). */ up_read(&mm->mmap_sem); + if (!user_mode(regs)) + goto no_context; pagefault_out_of_memory(); return; @@ -312,3 +317,13 @@ vmalloc_fault: } #endif } + +asmlinkage void __kprobes do_page_fault(struct pt_regs *regs, + unsigned long write, unsigned long address) +{ + enum ctx_state prev_state; + + prev_state = exception_enter(); + __do_page_fault(regs, write, address); + exception_exit(prev_state); +} diff --git a/arch/mips/mm/gup.c b/arch/mips/mm/gup.c index d4ea5c9c4a9..06ce17c2a90 100644 --- a/arch/mips/mm/gup.c +++ b/arch/mips/mm/gup.c @@ -12,6 +12,7 @@ #include <linux/swap.h> #include <linux/hugetlb.h> +#include <asm/cpu-features.h> #include <asm/pgtable.h> static inline pte_t gup_get_pte(pte_t *ptep) @@ -273,7 +274,7 @@ int get_user_pages_fast(unsigned long start, int nr_pages, int write, len = (unsigned long) nr_pages << PAGE_SHIFT; end = start + len; - if (end < start) + if (end < start || cpu_has_dc_aliases) goto slow_irqon; /* XXX: batch / limit 'nr' */ diff --git a/arch/mips/mm/hugetlbpage.c b/arch/mips/mm/hugetlbpage.c index a7fee0dfb7a..4ec8ee10d37 100644 --- a/arch/mips/mm/hugetlbpage.c +++ b/arch/mips/mm/hugetlbpage.c @@ -11,7 +11,6 @@ * Copyright (C) 2008, 2009 Cavium Networks, Inc. */ -#include <linux/init.h> #include <linux/fs.h> #include <linux/mm.h> #include <linux/hugetlb.h> diff --git a/arch/mips/mm/init.c b/arch/mips/mm/init.c index 9b973e0af9c..6e4413330e3 100644 --- a/arch/mips/mm/init.c +++ b/arch/mips/mm/init.c @@ -44,27 +44,6 @@ #include <asm/tlb.h> #include <asm/fixmap.h> -/* Atomicity and interruptability */ -#ifdef CONFIG_MIPS_MT_SMTC - -#include <asm/mipsmtregs.h> - -#define ENTER_CRITICAL(flags) \ - { \ - unsigned int mvpflags; \ - local_irq_save(flags);\ - mvpflags = dvpe() -#define EXIT_CRITICAL(flags) \ - evpe(mvpflags); \ - local_irq_restore(flags); \ - } -#else - -#define ENTER_CRITICAL(flags) local_irq_save(flags) -#define EXIT_CRITICAL(flags) local_irq_restore(flags) - -#endif /* CONFIG_MIPS_MT_SMTC */ - /* * We have up to 8 empty zeroed pages so we can map one of the right colour * when needed. This is necessary only on R4000 / R4400 SC and MC versions @@ -100,21 +79,7 @@ void setup_zero_pages(void) zero_page_mask = ((PAGE_SIZE << order) - 1) & PAGE_MASK; } -#ifdef CONFIG_MIPS_MT_SMTC -static pte_t *kmap_coherent_pte; -static void __init kmap_coherent_init(void) -{ - unsigned long vaddr; - - /* cache the first coherent kmap pte */ - vaddr = __fix_to_virt(FIX_CMAP_BEGIN); - kmap_coherent_pte = kmap_get_fixmap_pte(vaddr); -} -#else -static inline void kmap_coherent_init(void) {} -#endif - -void *kmap_coherent(struct page *page, unsigned long addr) +static void *__kmap_pgprot(struct page *page, unsigned long addr, pgprot_t prot) { enum fixed_addresses idx; unsigned long vaddr, flags, entrylo; @@ -124,62 +89,50 @@ void *kmap_coherent(struct page *page, unsigned long addr) BUG_ON(Page_dcache_dirty(page)); - inc_preempt_count(); + pagefault_disable(); idx = (addr >> PAGE_SHIFT) & (FIX_N_COLOURS - 1); -#ifdef CONFIG_MIPS_MT_SMTC - idx += FIX_N_COLOURS * smp_processor_id() + - (in_interrupt() ? (FIX_N_COLOURS * NR_CPUS) : 0); -#else idx += in_interrupt() ? FIX_N_COLOURS : 0; -#endif vaddr = __fix_to_virt(FIX_CMAP_END - idx); - pte = mk_pte(page, PAGE_KERNEL); + pte = mk_pte(page, prot); #if defined(CONFIG_64BIT_PHYS_ADDR) && defined(CONFIG_CPU_MIPS32) entrylo = pte.pte_high; #else entrylo = pte_to_entrylo(pte_val(pte)); #endif - ENTER_CRITICAL(flags); + local_irq_save(flags); old_ctx = read_c0_entryhi(); write_c0_entryhi(vaddr & (PAGE_MASK << 1)); write_c0_entrylo0(entrylo); write_c0_entrylo1(entrylo); -#ifdef CONFIG_MIPS_MT_SMTC - set_pte(kmap_coherent_pte - (FIX_CMAP_END - idx), pte); - /* preload TLB instead of local_flush_tlb_one() */ - mtc0_tlbw_hazard(); - tlb_probe(); - tlb_probe_hazard(); - tlbidx = read_c0_index(); - mtc0_tlbw_hazard(); - if (tlbidx < 0) - tlb_write_random(); - else - tlb_write_indexed(); -#else tlbidx = read_c0_wired(); write_c0_wired(tlbidx + 1); write_c0_index(tlbidx); mtc0_tlbw_hazard(); tlb_write_indexed(); -#endif tlbw_use_hazard(); write_c0_entryhi(old_ctx); - EXIT_CRITICAL(flags); + local_irq_restore(flags); return (void*) vaddr; } -#define UNIQUE_ENTRYHI(idx) (CKSEG0 + ((idx) << (PAGE_SHIFT + 1))) +void *kmap_coherent(struct page *page, unsigned long addr) +{ + return __kmap_pgprot(page, addr, PAGE_KERNEL); +} + +void *kmap_noncoherent(struct page *page, unsigned long addr) +{ + return __kmap_pgprot(page, addr, PAGE_KERNEL_NC); +} void kunmap_coherent(void) { -#ifndef CONFIG_MIPS_MT_SMTC unsigned int wired; unsigned long flags, old_ctx; - ENTER_CRITICAL(flags); + local_irq_save(flags); old_ctx = read_c0_entryhi(); wired = read_c0_wired() - 1; write_c0_wired(wired); @@ -191,10 +144,8 @@ void kunmap_coherent(void) tlb_write_indexed(); tlbw_use_hazard(); write_c0_entryhi(old_ctx); - EXIT_CRITICAL(flags); -#endif - dec_preempt_count(); - preempt_check_resched(); + local_irq_restore(flags); + pagefault_enable(); } void copy_user_highpage(struct page *to, struct page *from, @@ -254,11 +205,12 @@ void copy_from_user_page(struct vm_area_struct *vma, SetPageDcacheDirty(page); } } +EXPORT_SYMBOL_GPL(copy_from_user_page); void __init fixrange_init(unsigned long start, unsigned long end, pgd_t *pgd_base) { -#if defined(CONFIG_HIGHMEM) || defined(CONFIG_MIPS_MT_SMTC) +#ifdef CONFIG_HIGHMEM pgd_t *pgd; pud_t *pud; pmd_t *pmd; @@ -329,8 +281,6 @@ void __init paging_init(void) #ifdef CONFIG_HIGHMEM kmap_init(); #endif - kmap_coherent_init(); - #ifdef CONFIG_ZONE_DMA max_zone_pfns[ZONE_DMA] = MAX_DMA_PFN; #endif @@ -359,11 +309,24 @@ void __init paging_init(void) static struct kcore_list kcore_kseg0; #endif -void __init mem_init(void) +static inline void mem_init_free_highmem(void) { - unsigned long codesize, reservedpages, datasize, initsize; - unsigned long tmp, ram; +#ifdef CONFIG_HIGHMEM + unsigned long tmp; + + for (tmp = highstart_pfn; tmp < highend_pfn; tmp++) { + struct page *page = pfn_to_page(tmp); + + if (!page_is_ram(tmp)) + SetPageReserved(page); + else + free_highmem_page(page); + } +#endif +} +void __init mem_init(void) +{ #ifdef CONFIG_HIGHMEM #ifdef CONFIG_DISCONTIGMEM #error "CONFIG_HIGHMEM and CONFIG_DISCONTIGMEM dont work together yet" @@ -374,34 +337,10 @@ void __init mem_init(void) #endif high_memory = (void *) __va(max_low_pfn << PAGE_SHIFT); - totalram_pages += free_all_bootmem(); + free_all_bootmem(); setup_zero_pages(); /* Setup zeroed pages. */ - - reservedpages = ram = 0; - for (tmp = 0; tmp < max_low_pfn; tmp++) - if (page_is_ram(tmp) && pfn_valid(tmp)) { - ram++; - if (PageReserved(pfn_to_page(tmp))) - reservedpages++; - } - num_physpages = ram; - -#ifdef CONFIG_HIGHMEM - for (tmp = highstart_pfn; tmp < highend_pfn; tmp++) { - struct page *page = pfn_to_page(tmp); - - if (!page_is_ram(tmp)) { - SetPageReserved(page); - continue; - } - free_highmem_page(page); - } - num_physpages += totalhigh_pages; -#endif - - codesize = (unsigned long) &_etext - (unsigned long) &_text; - datasize = (unsigned long) &_edata - (unsigned long) &_etext; - initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin; + mem_init_free_highmem(); + mem_init_print_info(NULL); #ifdef CONFIG_64BIT if ((unsigned long) &_text > (unsigned long) CKSEG0) @@ -410,16 +349,6 @@ void __init mem_init(void) kclist_add(&kcore_kseg0, (void *) CKSEG0, 0x80000000 - 4, KCORE_TEXT); #endif - - printk(KERN_INFO "Memory: %luk/%luk available (%ldk kernel code, " - "%ldk reserved, %ldk data, %ldk init, %ldk highmem)\n", - nr_free_pages() << (PAGE_SHIFT-10), - ram << (PAGE_SHIFT-10), - codesize >> 10, - reservedpages << (PAGE_SHIFT-10), - datasize >> 10, - initsize >> 10, - totalhigh_pages << (PAGE_SHIFT-10)); } #endif /* !CONFIG_NEED_MULTIPLE_NODES */ @@ -440,14 +369,25 @@ void free_init_pages(const char *what, unsigned long begin, unsigned long end) #ifdef CONFIG_BLK_DEV_INITRD void free_initrd_mem(unsigned long start, unsigned long end) { - free_reserved_area(start, end, POISON_FREE_INITMEM, "initrd"); + free_reserved_area((void *)start, (void *)end, POISON_FREE_INITMEM, + "initrd"); } #endif +void (*free_init_pages_eva)(void *begin, void *end) = NULL; + void __init_refok free_initmem(void) { prom_free_prom_memory(); - free_initmem_default(POISON_FREE_INITMEM); + /* + * Let the platform define a specific function to free the + * init section since EVA may have used any possible mapping + * between virtual and physical addresses. + */ + if (free_init_pages_eva) + free_init_pages_eva((void *)&__init_begin, (void *)&__init_end); + else + free_initmem_default(POISON_FREE_INITMEM); } #ifndef CONFIG_MIPS_PGD_C0_CONTEXT diff --git a/arch/mips/mm/mmap.c b/arch/mips/mm/mmap.c index 7e5fe2790d8..f1baadd56e8 100644 --- a/arch/mips/mm/mmap.c +++ b/arch/mips/mm/mmap.c @@ -158,11 +158,9 @@ void arch_pick_mmap_layout(struct mm_struct *mm) if (mmap_is_legacy()) { mm->mmap_base = TASK_UNMAPPED_BASE + random_factor; mm->get_unmapped_area = arch_get_unmapped_area; - mm->unmap_area = arch_unmap_area; } else { mm->mmap_base = mmap_base(random_factor); mm->get_unmapped_area = arch_get_unmapped_area_topdown; - mm->unmap_area = arch_unmap_area_topdown; } } diff --git a/arch/mips/mm/page.c b/arch/mips/mm/page.c index 4eb8dcfaf1c..b611102e23b 100644 --- a/arch/mips/mm/page.c +++ b/arch/mips/mm/page.c @@ -8,7 +8,6 @@ * Copyright (C) 2008 Thiemo Seufer * Copyright (C) 2012 MIPS Technologies, Inc. */ -#include <linux/init.h> #include <linux/kernel.h> #include <linux/sched.h> #include <linux/smp.h> @@ -18,6 +17,7 @@ #include <asm/bugs.h> #include <asm/cacheops.h> +#include <asm/cpu-type.h> #include <asm/inst.h> #include <asm/io.h> #include <asm/page.h> @@ -66,29 +66,29 @@ UASM_L_LA(_copy_pref_both) UASM_L_LA(_copy_pref_store) /* We need one branch and therefore one relocation per target label. */ -static struct uasm_label __cpuinitdata labels[5]; -static struct uasm_reloc __cpuinitdata relocs[5]; +static struct uasm_label labels[5]; +static struct uasm_reloc relocs[5]; #define cpu_is_r4600_v1_x() ((read_c0_prid() & 0xfffffff0) == 0x00002010) #define cpu_is_r4600_v2_x() ((read_c0_prid() & 0xfffffff0) == 0x00002020) -static int pref_bias_clear_store __cpuinitdata; -static int pref_bias_copy_load __cpuinitdata; -static int pref_bias_copy_store __cpuinitdata; +static int pref_bias_clear_store; +static int pref_bias_copy_load; +static int pref_bias_copy_store; -static u32 pref_src_mode __cpuinitdata; -static u32 pref_dst_mode __cpuinitdata; +static u32 pref_src_mode; +static u32 pref_dst_mode; -static int clear_word_size __cpuinitdata; -static int copy_word_size __cpuinitdata; +static int clear_word_size; +static int copy_word_size; -static int half_clear_loop_size __cpuinitdata; -static int half_copy_loop_size __cpuinitdata; +static int half_clear_loop_size; +static int half_copy_loop_size; -static int cache_line_size __cpuinitdata; +static int cache_line_size; #define cache_line_mask() (cache_line_size - 1) -static inline void __cpuinit +static inline void pg_addiu(u32 **buf, unsigned int reg1, unsigned int reg2, unsigned int off) { if (cpu_has_64bit_gp_regs && DADDI_WAR && r4k_daddiu_bug()) { @@ -108,7 +108,7 @@ pg_addiu(u32 **buf, unsigned int reg1, unsigned int reg2, unsigned int off) } } -static void __cpuinit set_prefetch_parameters(void) +static void set_prefetch_parameters(void) { if (cpu_has_64bit_gp_regs || cpu_has_64bit_zero_reg) clear_word_size = 8; @@ -199,7 +199,7 @@ static void __cpuinit set_prefetch_parameters(void) 4 * copy_word_size)); } -static void __cpuinit build_clear_store(u32 **buf, int off) +static void build_clear_store(u32 **buf, int off) { if (cpu_has_64bit_gp_regs || cpu_has_64bit_zero_reg) { uasm_i_sd(buf, ZERO, off, A0); @@ -208,7 +208,7 @@ static void __cpuinit build_clear_store(u32 **buf, int off) } } -static inline void __cpuinit build_clear_pref(u32 **buf, int off) +static inline void build_clear_pref(u32 **buf, int off) { if (off & cache_line_mask()) return; @@ -232,7 +232,7 @@ static inline void __cpuinit build_clear_pref(u32 **buf, int off) uasm_i_cache(buf, Create_Dirty_Excl_D, off, A0); } - } + } } extern u32 __clear_page_start; @@ -240,7 +240,7 @@ extern u32 __clear_page_end; extern u32 __copy_page_start; extern u32 __copy_page_end; -void __cpuinit build_clear_page(void) +void build_clear_page(void) { int off; u32 *buf = &__clear_page_start; @@ -273,7 +273,7 @@ void __cpuinit build_clear_page(void) uasm_i_ori(&buf, A2, A0, off); if (R4600_V2_HIT_CACHEOP_WAR && cpu_is_r4600_v2_x()) - uasm_i_lui(&buf, AT, 0xa000); + uasm_i_lui(&buf, AT, uasm_rel_hi(0xa0000000)); off = cache_line_size ? min(8, pref_bias_clear_store / cache_line_size) * cache_line_size : 0; @@ -333,7 +333,7 @@ void __cpuinit build_clear_page(void) pr_debug("\t.set pop\n"); } -static void __cpuinit build_copy_load(u32 **buf, int reg, int off) +static void build_copy_load(u32 **buf, int reg, int off) { if (cpu_has_64bit_gp_regs) { uasm_i_ld(buf, reg, off, A1); @@ -342,7 +342,7 @@ static void __cpuinit build_copy_load(u32 **buf, int reg, int off) } } -static void __cpuinit build_copy_store(u32 **buf, int reg, int off) +static void build_copy_store(u32 **buf, int reg, int off) { if (cpu_has_64bit_gp_regs) { uasm_i_sd(buf, reg, off, A0); @@ -387,7 +387,7 @@ static inline void build_copy_store_pref(u32 **buf, int off) } } -void __cpuinit build_copy_page(void) +void build_copy_page(void) { int off; u32 *buf = &__copy_page_start; @@ -424,7 +424,7 @@ void __cpuinit build_copy_page(void) uasm_i_ori(&buf, A2, A0, off); if (R4600_V2_HIT_CACHEOP_WAR && cpu_is_r4600_v2_x()) - uasm_i_lui(&buf, AT, 0xa000); + uasm_i_lui(&buf, AT, uasm_rel_hi(0xa0000000)); off = cache_line_size ? min(8, pref_bias_copy_load / cache_line_size) * cache_line_size : 0; diff --git a/arch/mips/mm/sc-ip22.c b/arch/mips/mm/sc-ip22.c index c6aaed934d5..dc7c5a5214a 100644 --- a/arch/mips/mm/sc-ip22.c +++ b/arch/mips/mm/sc-ip22.c @@ -167,7 +167,7 @@ static struct bcache_ops indy_sc_ops = { .bc_inv = indy_sc_wback_invalidate }; -void __cpuinit indy_sc_init(void) +void indy_sc_init(void) { if (indy_sc_probe()) { indy_sc_enable(); diff --git a/arch/mips/mm/sc-mips.c b/arch/mips/mm/sc-mips.c index df96da7e939..99eb8fabab6 100644 --- a/arch/mips/mm/sc-mips.c +++ b/arch/mips/mm/sc-mips.c @@ -6,6 +6,7 @@ #include <linux/sched.h> #include <linux/mm.h> +#include <asm/cpu-type.h> #include <asm/mipsregs.h> #include <asm/bcache.h> #include <asm/cacheops.h> @@ -71,10 +72,14 @@ static inline int mips_sc_is_activated(struct cpuinfo_mips *c) unsigned int tmp; /* Check the bypass bit (L2B) */ - switch (c->cputype) { + switch (current_cpu_type()) { case CPU_34K: case CPU_74K: case CPU_1004K: + case CPU_1074K: + case CPU_INTERAPTIV: + case CPU_PROAPTIV: + case CPU_P5600: case CPU_BMIPS5000: if (config2 & (1 << 12)) return 0; @@ -132,7 +137,7 @@ static inline int __init mips_sc_probe(void) return 1; } -int __cpuinit mips_sc_init(void) +int mips_sc_init(void) { int found = mips_sc_probe(); if (found) { diff --git a/arch/mips/mm/sc-r5k.c b/arch/mips/mm/sc-r5k.c index 8bc67720e14..0216ed6eaa2 100644 --- a/arch/mips/mm/sc-r5k.c +++ b/arch/mips/mm/sc-r5k.c @@ -98,7 +98,7 @@ static struct bcache_ops r5k_sc_ops = { .bc_inv = r5k_dma_cache_inv_sc }; -void __cpuinit r5k_sc_init(void) +void r5k_sc_init(void) { if (r5k_sc_probe()) { r5k_sc_enable(); diff --git a/arch/mips/mm/sc-rm7k.c b/arch/mips/mm/sc-rm7k.c index 274af3be144..9ac1efcfbcc 100644 --- a/arch/mips/mm/sc-rm7k.c +++ b/arch/mips/mm/sc-rm7k.c @@ -6,7 +6,6 @@ #undef DEBUG -#include <linux/init.h> #include <linux/kernel.h> #include <linux/mm.h> #include <linux/bitops.h> @@ -104,7 +103,7 @@ static void blast_rm7k_tcache(void) /* * This function is executed in uncached address space. */ -static __cpuinit void __rm7k_tc_enable(void) +static void __rm7k_tc_enable(void) { int i; @@ -117,7 +116,7 @@ static __cpuinit void __rm7k_tc_enable(void) cache_op(Index_Store_Tag_T, CKSEG0ADDR(i)); } -static __cpuinit void rm7k_tc_enable(void) +static void rm7k_tc_enable(void) { if (read_c0_config() & RM7K_CONF_TE) return; @@ -130,7 +129,7 @@ static __cpuinit void rm7k_tc_enable(void) /* * This function is executed in uncached address space. */ -static __cpuinit void __rm7k_sc_enable(void) +static void __rm7k_sc_enable(void) { int i; @@ -143,7 +142,7 @@ static __cpuinit void __rm7k_sc_enable(void) cache_op(Index_Store_Tag_SD, CKSEG0ADDR(i)); } -static __cpuinit void rm7k_sc_enable(void) +static void rm7k_sc_enable(void) { if (read_c0_config() & RM7K_CONF_SE) return; @@ -184,7 +183,7 @@ static struct bcache_ops rm7k_sc_ops = { * This is a probing function like the one found in c-r4k.c, we look for the * wrap around point with different addresses. */ -static __cpuinit void __probe_tcache(void) +static void __probe_tcache(void) { unsigned long flags, addr, begin, end, pow2; @@ -226,7 +225,7 @@ static __cpuinit void __probe_tcache(void) local_irq_restore(flags); } -void __cpuinit rm7k_sc_init(void) +void rm7k_sc_init(void) { struct cpuinfo_mips *c = ¤t_cpu_data; unsigned int config = read_c0_config(); diff --git a/arch/mips/mm/tlb-funcs.S b/arch/mips/mm/tlb-funcs.S new file mode 100644 index 00000000000..a5427c6e975 --- /dev/null +++ b/arch/mips/mm/tlb-funcs.S @@ -0,0 +1,39 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Micro-assembler generated tlb handler functions. + * + * Copyright (C) 2013 Broadcom Corporation. + * + * Based on mm/page-funcs.c + * Copyright (C) 2012 MIPS Technologies, Inc. + * Copyright (C) 2012 Ralf Baechle <ralf@linux-mips.org> + */ +#include <asm/asm.h> +#include <asm/regdef.h> + +#define FASTPATH_SIZE 128 + +EXPORT(tlbmiss_handler_setup_pgd_start) +LEAF(tlbmiss_handler_setup_pgd) +1: j 1b /* Dummy, will be replaced. */ + .space 64 +END(tlbmiss_handler_setup_pgd) +EXPORT(tlbmiss_handler_setup_pgd_end) + +LEAF(handle_tlbm) + .space FASTPATH_SIZE * 4 +END(handle_tlbm) +EXPORT(handle_tlbm_end) + +LEAF(handle_tlbs) + .space FASTPATH_SIZE * 4 +END(handle_tlbs) +EXPORT(handle_tlbs_end) + +LEAF(handle_tlbl) + .space FASTPATH_SIZE * 4 +END(handle_tlbl) +EXPORT(handle_tlbl_end) diff --git a/arch/mips/mm/tlb-r3k.c b/arch/mips/mm/tlb-r3k.c index a63d1ed0827..d657493ef56 100644 --- a/arch/mips/mm/tlb-r3k.c +++ b/arch/mips/mm/tlb-r3k.c @@ -10,7 +10,6 @@ * Copyright (C) 2002 Ralf Baechle * Copyright (C) 2002 Maciej W. Rozycki */ -#include <linux/init.h> #include <linux/kernel.h> #include <linux/sched.h> #include <linux/smp.h> @@ -276,7 +275,7 @@ void add_wired_entry(unsigned long entrylo0, unsigned long entrylo1, } } -void __cpuinit tlb_init(void) +void tlb_init(void) { local_flush_tlb_all(); diff --git a/arch/mips/mm/tlb-r4k.c b/arch/mips/mm/tlb-r4k.c index c643de4c473..3914e27456f 100644 --- a/arch/mips/mm/tlb-r4k.c +++ b/arch/mips/mm/tlb-r4k.c @@ -8,6 +8,7 @@ * Carsten Langgaard, carstenl@mips.com * Copyright (C) 2002 MIPS Technologies, Inc. All rights reserved. */ +#include <linux/cpu_pm.h> #include <linux/init.h> #include <linux/sched.h> #include <linux/smp.h> @@ -16,64 +17,44 @@ #include <linux/module.h> #include <asm/cpu.h> +#include <asm/cpu-type.h> #include <asm/bootinfo.h> #include <asm/mmu_context.h> #include <asm/pgtable.h> +#include <asm/tlb.h> #include <asm/tlbmisc.h> extern void build_tlb_refill_handler(void); /* - * Make sure all entries differ. If they're not different - * MIPS32 will take revenge ... + * LOONGSON2/3 has a 4 entry itlb which is a subset of dtlb, + * unfortunately, itlb is not totally transparent to software. */ -#define UNIQUE_ENTRYHI(idx) (CKSEG0 + ((idx) << (PAGE_SHIFT + 1))) - -/* Atomicity and interruptability */ -#ifdef CONFIG_MIPS_MT_SMTC - -#include <asm/smtc.h> -#include <asm/mipsmtregs.h> - -#define ENTER_CRITICAL(flags) \ - { \ - unsigned int mvpflags; \ - local_irq_save(flags);\ - mvpflags = dvpe() -#define EXIT_CRITICAL(flags) \ - evpe(mvpflags); \ - local_irq_restore(flags); \ +static inline void flush_itlb(void) +{ + switch (current_cpu_type()) { + case CPU_LOONGSON2: + case CPU_LOONGSON3: + write_c0_diag(4); + break; + default: + break; } -#else - -#define ENTER_CRITICAL(flags) local_irq_save(flags) -#define EXIT_CRITICAL(flags) local_irq_restore(flags) - -#endif /* CONFIG_MIPS_MT_SMTC */ - -#if defined(CONFIG_CPU_LOONGSON2) -/* - * LOONGSON2 has a 4 entry itlb which is a subset of dtlb, - * unfortrunately, itlb is not totally transparent to software. - */ -#define FLUSH_ITLB write_c0_diag(4); - -#define FLUSH_ITLB_VM(vma) { if ((vma)->vm_flags & VM_EXEC) write_c0_diag(4); } - -#else - -#define FLUSH_ITLB -#define FLUSH_ITLB_VM(vma) +} -#endif +static inline void flush_itlb_vm(struct vm_area_struct *vma) +{ + if (vma->vm_flags & VM_EXEC) + flush_itlb(); +} void local_flush_tlb_all(void) { unsigned long flags; unsigned long old_ctx; - int entry; + int entry, ftlbhighset; - ENTER_CRITICAL(flags); + local_irq_save(flags); /* Save old context and create impossible VPN2 value */ old_ctx = read_c0_entryhi(); write_c0_entrylo0(0); @@ -82,18 +63,35 @@ void local_flush_tlb_all(void) entry = read_c0_wired(); /* Blast 'em all away. */ - while (entry < current_cpu_data.tlbsize) { - /* Make sure all entries differ. */ - write_c0_entryhi(UNIQUE_ENTRYHI(entry)); - write_c0_index(entry); - mtc0_tlbw_hazard(); - tlb_write_indexed(); - entry++; + if (cpu_has_tlbinv) { + if (current_cpu_data.tlbsizevtlb) { + write_c0_index(0); + mtc0_tlbw_hazard(); + tlbinvf(); /* invalidate VTLB */ + } + ftlbhighset = current_cpu_data.tlbsizevtlb + + current_cpu_data.tlbsizeftlbsets; + for (entry = current_cpu_data.tlbsizevtlb; + entry < ftlbhighset; + entry++) { + write_c0_index(entry); + mtc0_tlbw_hazard(); + tlbinvf(); /* invalidate one FTLB set */ + } + } else { + while (entry < current_cpu_data.tlbsize) { + /* Make sure all entries differ. */ + write_c0_entryhi(UNIQUE_ENTRYHI(entry)); + write_c0_index(entry); + mtc0_tlbw_hazard(); + tlb_write_indexed(); + entry++; + } } tlbw_use_hazard(); write_c0_entryhi(old_ctx); - FLUSH_ITLB; - EXIT_CRITICAL(flags); + flush_itlb(); + local_irq_restore(flags); } EXPORT_SYMBOL(local_flush_tlb_all); @@ -123,11 +121,13 @@ void local_flush_tlb_range(struct vm_area_struct *vma, unsigned long start, if (cpu_context(cpu, mm) != 0) { unsigned long size, flags; - ENTER_CRITICAL(flags); + local_irq_save(flags); start = round_down(start, PAGE_SIZE << 1); end = round_up(end, PAGE_SIZE << 1); size = (end - start) >> (PAGE_SHIFT + 1); - if (size <= current_cpu_data.tlbsize/2) { + if (size <= (current_cpu_data.tlbsizeftlbsets ? + current_cpu_data.tlbsize / 8 : + current_cpu_data.tlbsize / 2)) { int oldpid = read_c0_entryhi(); int newpid = cpu_asid(cpu, mm); @@ -154,8 +154,8 @@ void local_flush_tlb_range(struct vm_area_struct *vma, unsigned long start, } else { drop_mmu_context(mm, cpu); } - FLUSH_ITLB; - EXIT_CRITICAL(flags); + flush_itlb(); + local_irq_restore(flags); } } @@ -163,10 +163,12 @@ void local_flush_tlb_kernel_range(unsigned long start, unsigned long end) { unsigned long size, flags; - ENTER_CRITICAL(flags); + local_irq_save(flags); size = (end - start + (PAGE_SIZE - 1)) >> PAGE_SHIFT; size = (size + 1) >> 1; - if (size <= current_cpu_data.tlbsize / 2) { + if (size <= (current_cpu_data.tlbsizeftlbsets ? + current_cpu_data.tlbsize / 8 : + current_cpu_data.tlbsize / 2)) { int pid = read_c0_entryhi(); start &= (PAGE_MASK << 1); @@ -196,8 +198,8 @@ void local_flush_tlb_kernel_range(unsigned long start, unsigned long end) } else { local_flush_tlb_all(); } - FLUSH_ITLB; - EXIT_CRITICAL(flags); + flush_itlb(); + local_irq_restore(flags); } void local_flush_tlb_page(struct vm_area_struct *vma, unsigned long page) @@ -210,7 +212,7 @@ void local_flush_tlb_page(struct vm_area_struct *vma, unsigned long page) newpid = cpu_asid(cpu, vma->vm_mm); page &= (PAGE_MASK << 1); - ENTER_CRITICAL(flags); + local_irq_save(flags); oldpid = read_c0_entryhi(); write_c0_entryhi(page | newpid); mtc0_tlbw_hazard(); @@ -229,8 +231,8 @@ void local_flush_tlb_page(struct vm_area_struct *vma, unsigned long page) finish: write_c0_entryhi(oldpid); - FLUSH_ITLB_VM(vma); - EXIT_CRITICAL(flags); + flush_itlb_vm(vma); + local_irq_restore(flags); } } @@ -243,7 +245,7 @@ void local_flush_tlb_one(unsigned long page) unsigned long flags; int oldpid, idx; - ENTER_CRITICAL(flags); + local_irq_save(flags); oldpid = read_c0_entryhi(); page &= (PAGE_MASK << 1); write_c0_entryhi(page); @@ -261,8 +263,8 @@ void local_flush_tlb_one(unsigned long page) tlbw_use_hazard(); } write_c0_entryhi(oldpid); - FLUSH_ITLB; - EXIT_CRITICAL(flags); + flush_itlb(); + local_irq_restore(flags); } /* @@ -285,7 +287,7 @@ void __update_tlb(struct vm_area_struct * vma, unsigned long address, pte_t pte) if (current->active_mm != vma->vm_mm) return; - ENTER_CRITICAL(flags); + local_irq_save(flags); pid = read_c0_entryhi() & ASID_MASK; address &= (PAGE_MASK << 1); @@ -334,8 +336,8 @@ void __update_tlb(struct vm_area_struct * vma, unsigned long address, pte_t pte) tlb_write_indexed(); } tlbw_use_hazard(); - FLUSH_ITLB_VM(vma); - EXIT_CRITICAL(flags); + flush_itlb_vm(vma); + local_irq_restore(flags); } void add_wired_entry(unsigned long entrylo0, unsigned long entrylo1, @@ -346,7 +348,7 @@ void add_wired_entry(unsigned long entrylo0, unsigned long entrylo1, unsigned long old_pagemask; unsigned long old_ctx; - ENTER_CRITICAL(flags); + local_irq_save(flags); /* Save old context and create impossible VPN2 value */ old_ctx = read_c0_entryhi(); old_pagemask = read_c0_pagemask(); @@ -366,7 +368,7 @@ void add_wired_entry(unsigned long entrylo0, unsigned long entrylo1, tlbw_use_hazard(); /* What is the hazard here? */ write_c0_pagemask(old_pagemask); local_flush_tlb_all(); - EXIT_CRITICAL(flags); + local_irq_restore(flags); } #ifdef CONFIG_TRANSPARENT_HUGEPAGE @@ -376,20 +378,20 @@ int __init has_transparent_hugepage(void) unsigned int mask; unsigned long flags; - ENTER_CRITICAL(flags); + local_irq_save(flags); write_c0_pagemask(PM_HUGE_MASK); back_to_back_c0_hazard(); mask = read_c0_pagemask(); write_c0_pagemask(PM_DEFAULT_MASK); - EXIT_CRITICAL(flags); + local_irq_restore(flags); return mask == PM_HUGE_MASK; } #endif /* CONFIG_TRANSPARENT_HUGEPAGE */ -static int __cpuinitdata ntlb; +static int ntlb; static int __init set_ntlb(char *str) { get_option(&str, &ntlb); @@ -398,7 +400,10 @@ static int __init set_ntlb(char *str) __setup("ntlb=", set_ntlb); -void __cpuinit tlb_init(void) +/* + * Configure TLB (for init or after a CPU has been powered off). + */ +static void r4k_tlb_configure(void) { /* * You should never change this register: @@ -430,6 +435,11 @@ void __cpuinit tlb_init(void) local_flush_tlb_all(); /* Did I tell you that ARC SUCKS? */ +} + +void tlb_init(void) +{ + r4k_tlb_configure(); if (ntlb) { if (ntlb > 1 && ntlb <= current_cpu_data.tlbsize) { @@ -443,3 +453,26 @@ void __cpuinit tlb_init(void) build_tlb_refill_handler(); } + +static int r4k_tlb_pm_notifier(struct notifier_block *self, unsigned long cmd, + void *v) +{ + switch (cmd) { + case CPU_PM_ENTER_FAILED: + case CPU_PM_EXIT: + r4k_tlb_configure(); + break; + } + + return NOTIFY_OK; +} + +static struct notifier_block r4k_tlb_pm_notifier_block = { + .notifier_call = r4k_tlb_pm_notifier, +}; + +static int __init r4k_tlb_init_pm(void) +{ + return cpu_pm_register_notifier(&r4k_tlb_pm_notifier_block); +} +arch_initcall(r4k_tlb_init_pm); diff --git a/arch/mips/mm/tlb-r8k.c b/arch/mips/mm/tlb-r8k.c index 91c2499f806..138a2ec7cc6 100644 --- a/arch/mips/mm/tlb-r8k.c +++ b/arch/mips/mm/tlb-r8k.c @@ -8,7 +8,6 @@ * Carsten Langgaard, carstenl@mips.com * Copyright (C) 2002 MIPS Technologies, Inc. All rights reserved. */ -#include <linux/init.h> #include <linux/sched.h> #include <linux/smp.h> #include <linux/mm.h> @@ -213,14 +212,14 @@ void __update_tlb(struct vm_area_struct * vma, unsigned long address, pte_t pte) local_irq_restore(flags); } -static void __cpuinit probe_tlb(unsigned long config) +static void probe_tlb(unsigned long config) { struct cpuinfo_mips *c = ¤t_cpu_data; c->tlbsize = 3 * 128; /* 3 sets each 128 entries */ } -void __cpuinit tlb_init(void) +void tlb_init(void) { unsigned int config = read_c0_config(); unsigned long status; diff --git a/arch/mips/mm/tlbex.c b/arch/mips/mm/tlbex.c index afeef93f81a..e80e10bafc8 100644 --- a/arch/mips/mm/tlbex.c +++ b/arch/mips/mm/tlbex.c @@ -26,10 +26,10 @@ #include <linux/types.h> #include <linux/smp.h> #include <linux/string.h> -#include <linux/init.h> #include <linux/cache.h> #include <asm/cacheflush.h> +#include <asm/cpu-type.h> #include <asm/pgtable.h> #include <asm/war.h> #include <asm/uasm.h> @@ -85,6 +85,7 @@ static int use_bbit_insns(void) case CPU_CAVIUM_OCTEON: case CPU_CAVIUM_OCTEON_PLUS: case CPU_CAVIUM_OCTEON2: + case CPU_CAVIUM_OCTEON3: return 1; default: return 0; @@ -95,6 +96,7 @@ static int use_lwx_insns(void) { switch (current_cpu_type()) { case CPU_CAVIUM_OCTEON2: + case CPU_CAVIUM_OCTEON3: return 1; default: return 0; @@ -136,7 +138,7 @@ static int scratchpad_offset(int i) * why; it's not an issue caused by the core RTL. * */ -static int __cpuinit m4kc_tlbp_war(void) +static int m4kc_tlbp_war(void) { return (current_cpu_data.processor_id & 0xffff00) == (PRID_COMP_MIPS | PRID_IMP_4KC); @@ -181,11 +183,9 @@ UASM_L_LA(_large_segbits_fault) UASM_L_LA(_tlb_huge_update) #endif -static int __cpuinitdata hazard_instance; +static int hazard_instance; -static void __cpuinit uasm_bgezl_hazard(u32 **p, - struct uasm_reloc **r, - int instance) +static void uasm_bgezl_hazard(u32 **p, struct uasm_reloc **r, int instance) { switch (instance) { case 0 ... 7: @@ -196,9 +196,7 @@ static void __cpuinit uasm_bgezl_hazard(u32 **p, } } -static void __cpuinit uasm_bgezl_label(struct uasm_label **l, - u32 **p, - int instance) +static void uasm_bgezl_label(struct uasm_label **l, u32 **p, int instance) { switch (instance) { case 0 ... 7: @@ -295,17 +293,28 @@ static inline void dump_handler(const char *symbol, const u32 *handler, int coun * We deliberately chose a buffer size of 128, so we won't scribble * over anything important on overflow before we panic. */ -static u32 tlb_handler[128] __cpuinitdata; +static u32 tlb_handler[128]; /* simply assume worst case size for labels and relocs */ -static struct uasm_label labels[128] __cpuinitdata; -static struct uasm_reloc relocs[128] __cpuinitdata; +static struct uasm_label labels[128]; +static struct uasm_reloc relocs[128]; -static int check_for_high_segbits __cpuinitdata; +static int check_for_high_segbits; -static unsigned int kscratch_used_mask __cpuinitdata; +static unsigned int kscratch_used_mask; -static int __cpuinit allocate_kscratch(void) +static inline int __maybe_unused c0_kscratch(void) +{ + switch (current_cpu_type()) { + case CPU_XLP: + case CPU_XLR: + return 22; + default: + return 31; + } +} + +static int allocate_kscratch(void) { int r; unsigned int a = cpu_data[0].kscratch_mask & ~kscratch_used_mask; @@ -322,21 +331,17 @@ static int __cpuinit allocate_kscratch(void) return r; } -static int scratch_reg __cpuinitdata; -static int pgd_reg __cpuinitdata; +static int scratch_reg; +static int pgd_reg; enum vmalloc64_mode {not_refill, refill_scratch, refill_noscratch}; -static struct work_registers __cpuinit build_get_work_registers(u32 **p) +static struct work_registers build_get_work_registers(u32 **p) { struct work_registers r; - int smp_processor_id_reg; - int smp_processor_id_sel; - int smp_processor_id_shift; - - if (scratch_reg > 0) { + if (scratch_reg >= 0) { /* Save in CPU local C0_KScratch? */ - UASM_i_MTC0(p, 1, 31, scratch_reg); + UASM_i_MTC0(p, 1, c0_kscratch(), scratch_reg); r.r1 = K0; r.r2 = K1; r.r3 = 1; @@ -344,25 +349,9 @@ static struct work_registers __cpuinit build_get_work_registers(u32 **p) } if (num_possible_cpus() > 1) { -#ifdef CONFIG_MIPS_PGD_C0_CONTEXT - smp_processor_id_shift = 51; - smp_processor_id_reg = 20; /* XContext */ - smp_processor_id_sel = 0; -#else -# ifdef CONFIG_32BIT - smp_processor_id_shift = 25; - smp_processor_id_reg = 4; /* Context */ - smp_processor_id_sel = 0; -# endif -# ifdef CONFIG_64BIT - smp_processor_id_shift = 26; - smp_processor_id_reg = 4; /* Context */ - smp_processor_id_sel = 0; -# endif -#endif /* Get smp_processor_id */ - UASM_i_MFC0(p, K0, smp_processor_id_reg, smp_processor_id_sel); - UASM_i_SRL_SAFE(p, K0, K0, smp_processor_id_shift); + UASM_i_CPUID_MFC0(p, K0, SMP_CPUID_REG); + UASM_i_SRL_SAFE(p, K0, K0, SMP_CPUID_REGSHIFT); /* handler_reg_save index in K0 */ UASM_i_SLL(p, K0, K0, ilog2(sizeof(struct tlb_reg_save))); @@ -382,10 +371,10 @@ static struct work_registers __cpuinit build_get_work_registers(u32 **p) return r; } -static void __cpuinit build_restore_work_registers(u32 **p) +static void build_restore_work_registers(u32 **p) { - if (scratch_reg > 0) { - UASM_i_MFC0(p, 1, 31, scratch_reg); + if (scratch_reg >= 0) { + UASM_i_MFC0(p, 1, c0_kscratch(), scratch_reg); return; } /* K0 already points to save area, restore $1 and $2 */ @@ -407,7 +396,7 @@ extern unsigned long pgd_current[]; /* * The R3000 TLB handler is simple. */ -static void __cpuinit build_r3000_tlb_refill_handler(void) +static void build_r3000_tlb_refill_handler(void) { long pgdc = (long)pgd_current; u32 *p; @@ -452,7 +441,7 @@ static void __cpuinit build_r3000_tlb_refill_handler(void) * other one.To keep things simple, we first assume linear space, * then we relocate it to the final handler layout as needed. */ -static u32 final_handler[64] __cpuinitdata; +static u32 final_handler[64]; /* * Hazards @@ -476,7 +465,7 @@ static u32 final_handler[64] __cpuinitdata; * * As if we MIPS hackers wouldn't know how to nop pipelines happy ... */ -static void __cpuinit __maybe_unused build_tlb_probe_entry(u32 **p) +static void __maybe_unused build_tlb_probe_entry(u32 **p) { switch (current_cpu_type()) { /* Found by experiment: R4600 v2.0/R4700 needs this, too. */ @@ -500,9 +489,9 @@ static void __cpuinit __maybe_unused build_tlb_probe_entry(u32 **p) */ enum tlb_write_entry { tlb_random, tlb_indexed }; -static void __cpuinit build_tlb_write_entry(u32 **p, struct uasm_label **l, - struct uasm_reloc **r, - enum tlb_write_entry wmode) +static void build_tlb_write_entry(u32 **p, struct uasm_label **l, + struct uasm_reloc **r, + enum tlb_write_entry wmode) { void(*tlbw)(u32 **) = NULL; @@ -520,6 +509,10 @@ static void __cpuinit build_tlb_write_entry(u32 **p, struct uasm_label **l, switch (current_cpu_type()) { case CPU_M14KC: case CPU_74K: + case CPU_1074K: + case CPU_PROAPTIV: + case CPU_P5600: + case CPU_M5150: break; default: @@ -589,6 +582,7 @@ static void __cpuinit build_tlb_write_entry(u32 **p, struct uasm_label **l, case CPU_BMIPS4380: case CPU_BMIPS5000: case CPU_LOONGSON2: + case CPU_LOONGSON3: case CPU_R5500: if (m4kc_tlbp_war()) uasm_i_nop(p); @@ -631,13 +625,13 @@ static void __cpuinit build_tlb_write_entry(u32 **p, struct uasm_label **l, default: panic("No TLB refill handler yet (CPU type: %d)", - current_cpu_data.cputype); + current_cpu_type()); break; } } -static __cpuinit __maybe_unused void build_convert_pte_to_entrylo(u32 **p, - unsigned int reg) +static __maybe_unused void build_convert_pte_to_entrylo(u32 **p, + unsigned int reg) { if (cpu_has_rixi) { UASM_i_ROTR(p, reg, reg, ilog2(_PAGE_GLOBAL)); @@ -652,11 +646,9 @@ static __cpuinit __maybe_unused void build_convert_pte_to_entrylo(u32 **p, #ifdef CONFIG_MIPS_HUGE_TLB_SUPPORT -static __cpuinit void build_restore_pagemask(u32 **p, - struct uasm_reloc **r, - unsigned int tmp, - enum label_id lid, - int restore_scratch) +static void build_restore_pagemask(u32 **p, struct uasm_reloc **r, + unsigned int tmp, enum label_id lid, + int restore_scratch) { if (restore_scratch) { /* Reset default page size */ @@ -673,8 +665,8 @@ static __cpuinit void build_restore_pagemask(u32 **p, uasm_i_mtc0(p, 0, C0_PAGEMASK); uasm_il_b(p, r, lid); } - if (scratch_reg > 0) - UASM_i_MFC0(p, 1, 31, scratch_reg); + if (scratch_reg >= 0) + UASM_i_MFC0(p, 1, c0_kscratch(), scratch_reg); else UASM_i_LW(p, 1, scratchpad_offset(0), 0); } else { @@ -695,12 +687,11 @@ static __cpuinit void build_restore_pagemask(u32 **p, } } -static __cpuinit void build_huge_tlb_write_entry(u32 **p, - struct uasm_label **l, - struct uasm_reloc **r, - unsigned int tmp, - enum tlb_write_entry wmode, - int restore_scratch) +static void build_huge_tlb_write_entry(u32 **p, struct uasm_label **l, + struct uasm_reloc **r, + unsigned int tmp, + enum tlb_write_entry wmode, + int restore_scratch) { /* Set huge page tlb entry size */ uasm_i_lui(p, tmp, PM_HUGE_MASK >> 16); @@ -715,9 +706,9 @@ static __cpuinit void build_huge_tlb_write_entry(u32 **p, /* * Check if Huge PTE is present, if so then jump to LABEL. */ -static void __cpuinit +static void build_is_huge_pte(u32 **p, struct uasm_reloc **r, unsigned int tmp, - unsigned int pmd, int lid) + unsigned int pmd, int lid) { UASM_i_LW(p, tmp, 0, pmd); if (use_bbit_insns()) { @@ -728,9 +719,8 @@ build_is_huge_pte(u32 **p, struct uasm_reloc **r, unsigned int tmp, } } -static __cpuinit void build_huge_update_entries(u32 **p, - unsigned int pte, - unsigned int tmp) +static void build_huge_update_entries(u32 **p, unsigned int pte, + unsigned int tmp) { int small_sequence; @@ -760,11 +750,10 @@ static __cpuinit void build_huge_update_entries(u32 **p, UASM_i_MTC0(p, pte, C0_ENTRYLO1); /* load it */ } -static __cpuinit void build_huge_handler_tail(u32 **p, - struct uasm_reloc **r, - struct uasm_label **l, - unsigned int pte, - unsigned int ptr) +static void build_huge_handler_tail(u32 **p, struct uasm_reloc **r, + struct uasm_label **l, + unsigned int pte, + unsigned int ptr) { #ifdef CONFIG_SMP UASM_i_SC(p, pte, 0, ptr); @@ -783,7 +772,7 @@ static __cpuinit void build_huge_handler_tail(u32 **p, * TMP and PTR are scratch. * TMP will be clobbered, PTR will hold the pmd entry. */ -static void __cpuinit +static void build_get_pmde64(u32 **p, struct uasm_label **l, struct uasm_reloc **r, unsigned int tmp, unsigned int ptr) { @@ -814,11 +803,11 @@ build_get_pmde64(u32 **p, struct uasm_label **l, struct uasm_reloc **r, } /* No uasm_i_nop needed here, since the next insn doesn't touch TMP. */ -#ifdef CONFIG_MIPS_PGD_C0_CONTEXT if (pgd_reg != -1) { /* pgd is in pgd_reg */ - UASM_i_MFC0(p, ptr, 31, pgd_reg); + UASM_i_MFC0(p, ptr, c0_kscratch(), pgd_reg); } else { +#if defined(CONFIG_MIPS_PGD_C0_CONTEXT) /* * &pgd << 11 stored in CONTEXT [23..63]. */ @@ -830,30 +819,18 @@ build_get_pmde64(u32 **p, struct uasm_label **l, struct uasm_reloc **r, /* 1 0 1 0 1 << 6 xkphys cached */ uasm_i_ori(p, ptr, ptr, 0x540); uasm_i_drotr(p, ptr, ptr, 11); - } #elif defined(CONFIG_SMP) -# ifdef CONFIG_MIPS_MT_SMTC - /* - * SMTC uses TCBind value as "CPU" index - */ - uasm_i_mfc0(p, ptr, C0_TCBIND); - uasm_i_dsrl_safe(p, ptr, ptr, 19); -# else - /* - * 64 bit SMP running in XKPHYS has smp_processor_id() << 3 - * stored in CONTEXT. - */ - uasm_i_dmfc0(p, ptr, C0_CONTEXT); - uasm_i_dsrl_safe(p, ptr, ptr, 23); -# endif - UASM_i_LA_mostly(p, tmp, pgdc); - uasm_i_daddu(p, ptr, ptr, tmp); - uasm_i_dmfc0(p, tmp, C0_BADVADDR); - uasm_i_ld(p, ptr, uasm_rel_lo(pgdc), ptr); + UASM_i_CPUID_MFC0(p, ptr, SMP_CPUID_REG); + uasm_i_dsrl_safe(p, ptr, ptr, SMP_CPUID_PTRSHIFT); + UASM_i_LA_mostly(p, tmp, pgdc); + uasm_i_daddu(p, ptr, ptr, tmp); + uasm_i_dmfc0(p, tmp, C0_BADVADDR); + uasm_i_ld(p, ptr, uasm_rel_lo(pgdc), ptr); #else - UASM_i_LA_mostly(p, ptr, pgdc); - uasm_i_ld(p, ptr, uasm_rel_lo(pgdc), ptr); + UASM_i_LA_mostly(p, ptr, pgdc); + uasm_i_ld(p, ptr, uasm_rel_lo(pgdc), ptr); #endif + } uasm_l_vmalloc_done(l, *p); @@ -875,7 +852,7 @@ build_get_pmde64(u32 **p, struct uasm_label **l, struct uasm_reloc **r, * BVADDR is the faulting address, PTR is scratch. * PTR will hold the pgd for vmalloc. */ -static void __cpuinit +static void build_get_pgd_vmalloc64(u32 **p, struct uasm_label **l, struct uasm_reloc **r, unsigned int bvaddr, unsigned int ptr, enum vmalloc64_mode mode) @@ -929,8 +906,8 @@ build_get_pgd_vmalloc64(u32 **p, struct uasm_label **l, struct uasm_reloc **r, uasm_i_jr(p, ptr); if (mode == refill_scratch) { - if (scratch_reg > 0) - UASM_i_MFC0(p, 1, 31, scratch_reg); + if (scratch_reg >= 0) + UASM_i_MFC0(p, 1, c0_kscratch(), scratch_reg); else UASM_i_LW(p, 1, scratchpad_offset(0), 0); } else { @@ -945,34 +922,28 @@ build_get_pgd_vmalloc64(u32 **p, struct uasm_label **l, struct uasm_reloc **r, * TMP and PTR are scratch. * TMP will be clobbered, PTR will hold the pgd entry. */ -static void __cpuinit __maybe_unused +static void __maybe_unused build_get_pgde32(u32 **p, unsigned int tmp, unsigned int ptr) { - long pgdc = (long)pgd_current; + if (pgd_reg != -1) { + /* pgd is in pgd_reg */ + uasm_i_mfc0(p, ptr, c0_kscratch(), pgd_reg); + uasm_i_mfc0(p, tmp, C0_BADVADDR); /* get faulting address */ + } else { + long pgdc = (long)pgd_current; - /* 32 bit SMP has smp_processor_id() stored in CONTEXT. */ + /* 32 bit SMP has smp_processor_id() stored in CONTEXT. */ #ifdef CONFIG_SMP -#ifdef CONFIG_MIPS_MT_SMTC - /* - * SMTC uses TCBind value as "CPU" index - */ - uasm_i_mfc0(p, ptr, C0_TCBIND); - UASM_i_LA_mostly(p, tmp, pgdc); - uasm_i_srl(p, ptr, ptr, 19); + uasm_i_mfc0(p, ptr, SMP_CPUID_REG); + UASM_i_LA_mostly(p, tmp, pgdc); + uasm_i_srl(p, ptr, ptr, SMP_CPUID_PTRSHIFT); + uasm_i_addu(p, ptr, tmp, ptr); #else - /* - * smp_processor_id() << 3 is stored in CONTEXT. - */ - uasm_i_mfc0(p, ptr, C0_CONTEXT); - UASM_i_LA_mostly(p, tmp, pgdc); - uasm_i_srl(p, ptr, ptr, 23); -#endif - uasm_i_addu(p, ptr, tmp, ptr); -#else - UASM_i_LA_mostly(p, ptr, pgdc); + UASM_i_LA_mostly(p, ptr, pgdc); #endif - uasm_i_mfc0(p, tmp, C0_BADVADDR); /* get faulting address */ - uasm_i_lw(p, ptr, uasm_rel_lo(pgdc), ptr); + uasm_i_mfc0(p, tmp, C0_BADVADDR); /* get faulting address */ + uasm_i_lw(p, ptr, uasm_rel_lo(pgdc), ptr); + } uasm_i_srl(p, tmp, tmp, PGDIR_SHIFT); /* get pgd only bits */ uasm_i_sll(p, tmp, tmp, PGD_T_LOG2); uasm_i_addu(p, ptr, ptr, tmp); /* add in pgd offset */ @@ -980,7 +951,7 @@ build_get_pgde32(u32 **p, unsigned int tmp, unsigned int ptr) #endif /* !CONFIG_64BIT */ -static void __cpuinit build_adjust_context(u32 **p, unsigned int ctx) +static void build_adjust_context(u32 **p, unsigned int ctx) { unsigned int shift = 4 - (PTE_T_LOG2 + 1) + PAGE_SHIFT - 12; unsigned int mask = (PTRS_PER_PTE / 2 - 1) << (PTE_T_LOG2 + 1); @@ -1006,7 +977,7 @@ static void __cpuinit build_adjust_context(u32 **p, unsigned int ctx) uasm_i_andi(p, ctx, ctx, mask); } -static void __cpuinit build_get_ptep(u32 **p, unsigned int tmp, unsigned int ptr) +static void build_get_ptep(u32 **p, unsigned int tmp, unsigned int ptr) { /* * Bug workaround for the Nevada. It seems as if under certain @@ -1031,8 +1002,7 @@ static void __cpuinit build_get_ptep(u32 **p, unsigned int tmp, unsigned int ptr UASM_i_ADDU(p, ptr, ptr, tmp); /* add in offset */ } -static void __cpuinit build_update_entries(u32 **p, unsigned int tmp, - unsigned int ptep) +static void build_update_entries(u32 **p, unsigned int tmp, unsigned int ptep) { /* * 64bit address support (36bit on a 32bit CPU) in a 32bit @@ -1093,10 +1063,10 @@ struct mips_huge_tlb_info { int restore_scratch; }; -static struct mips_huge_tlb_info __cpuinit +static struct mips_huge_tlb_info build_fast_tlb_refill_handler (u32 **p, struct uasm_label **l, struct uasm_reloc **r, unsigned int tmp, - unsigned int ptr, int c0_scratch) + unsigned int ptr, int c0_scratch_reg) { struct mips_huge_tlb_info rv; unsigned int even, odd; @@ -1110,12 +1080,12 @@ build_fast_tlb_refill_handler (u32 **p, struct uasm_label **l, UASM_i_MFC0(p, tmp, C0_BADVADDR); if (pgd_reg != -1) - UASM_i_MFC0(p, ptr, 31, pgd_reg); + UASM_i_MFC0(p, ptr, c0_kscratch(), pgd_reg); else UASM_i_MFC0(p, ptr, C0_CONTEXT); - if (c0_scratch >= 0) - UASM_i_MTC0(p, scratch, 31, c0_scratch); + if (c0_scratch_reg >= 0) + UASM_i_MTC0(p, scratch, c0_kscratch(), c0_scratch_reg); else UASM_i_SW(p, scratch, scratchpad_offset(0), 0); @@ -1130,14 +1100,14 @@ build_fast_tlb_refill_handler (u32 **p, struct uasm_label **l, } } else { if (pgd_reg != -1) - UASM_i_MFC0(p, ptr, 31, pgd_reg); + UASM_i_MFC0(p, ptr, c0_kscratch(), pgd_reg); else UASM_i_MFC0(p, ptr, C0_CONTEXT); UASM_i_MFC0(p, tmp, C0_BADVADDR); - if (c0_scratch >= 0) - UASM_i_MTC0(p, scratch, 31, c0_scratch); + if (c0_scratch_reg >= 0) + UASM_i_MTC0(p, scratch, c0_kscratch(), c0_scratch_reg); else UASM_i_SW(p, scratch, scratchpad_offset(0), 0); @@ -1242,8 +1212,8 @@ build_fast_tlb_refill_handler (u32 **p, struct uasm_label **l, } UASM_i_MTC0(p, odd, C0_ENTRYLO1); /* load it */ - if (c0_scratch >= 0) { - UASM_i_MFC0(p, scratch, 31, c0_scratch); + if (c0_scratch_reg >= 0) { + UASM_i_MFC0(p, scratch, c0_kscratch(), c0_scratch_reg); build_tlb_write_entry(p, l, r, tlb_random); uasm_l_leave(l, *p); rv.restore_scratch = 1; @@ -1271,7 +1241,7 @@ build_fast_tlb_refill_handler (u32 **p, struct uasm_label **l, */ #define MIPS64_REFILL_INSNS 32 -static void __cpuinit build_r4000_tlb_refill_handler(void) +static void build_r4000_tlb_refill_handler(void) { u32 *p = tlb_handler; struct uasm_label *l = labels; @@ -1286,7 +1256,7 @@ static void __cpuinit build_r4000_tlb_refill_handler(void) memset(relocs, 0, sizeof(relocs)); memset(final_handler, 0, sizeof(final_handler)); - if ((scratch_reg > 0 || scratchpad_available()) && use_bbit_insns()) { + if (IS_ENABLED(CONFIG_64BIT) && (scratch_reg >= 0 || scratchpad_available()) && use_bbit_insns()) { htlb_info = build_fast_tlb_refill_handler(&p, &l, &r, K0, K1, scratch_reg); vmalloc_mode = refill_scratch; @@ -1345,95 +1315,100 @@ static void __cpuinit build_r4000_tlb_refill_handler(void) * need three, with the second nop'ed and the third being * unused. */ - /* Loongson2 ebase is different than r4k, we have more space */ -#if defined(CONFIG_32BIT) || defined(CONFIG_CPU_LOONGSON2) - if ((p - tlb_handler) > 64) - panic("TLB refill handler space exceeded"); -#else - if (((p - tlb_handler) > (MIPS64_REFILL_INSNS * 2) - 1) - || (((p - tlb_handler) > (MIPS64_REFILL_INSNS * 2) - 3) - && uasm_insn_has_bdelay(relocs, - tlb_handler + MIPS64_REFILL_INSNS - 3))) - panic("TLB refill handler space exceeded"); -#endif - - /* - * Now fold the handler in the TLB refill handler space. - */ -#if defined(CONFIG_32BIT) || defined(CONFIG_CPU_LOONGSON2) - f = final_handler; - /* Simplest case, just copy the handler. */ - uasm_copy_handler(relocs, labels, tlb_handler, p, f); - final_len = p - tlb_handler; -#else /* CONFIG_64BIT */ - f = final_handler + MIPS64_REFILL_INSNS; - if ((p - tlb_handler) <= MIPS64_REFILL_INSNS) { - /* Just copy the handler. */ - uasm_copy_handler(relocs, labels, tlb_handler, p, f); - final_len = p - tlb_handler; - } else { -#ifdef CONFIG_MIPS_HUGE_TLB_SUPPORT - const enum label_id ls = label_tlb_huge_update; -#else - const enum label_id ls = label_vmalloc; -#endif - u32 *split; - int ov = 0; - int i; - - for (i = 0; i < ARRAY_SIZE(labels) && labels[i].lab != ls; i++) - ; - BUG_ON(i == ARRAY_SIZE(labels)); - split = labels[i].addr; - - /* - * See if we have overflown one way or the other. - */ - if (split > tlb_handler + MIPS64_REFILL_INSNS || - split < p - MIPS64_REFILL_INSNS) - ov = 1; - - if (ov) { + switch (boot_cpu_type()) { + default: + if (sizeof(long) == 4) { + case CPU_LOONGSON2: + /* Loongson2 ebase is different than r4k, we have more space */ + if ((p - tlb_handler) > 64) + panic("TLB refill handler space exceeded"); /* - * Split two instructions before the end. One - * for the branch and one for the instruction - * in the delay slot. + * Now fold the handler in the TLB refill handler space. */ - split = tlb_handler + MIPS64_REFILL_INSNS - 2; - + f = final_handler; + /* Simplest case, just copy the handler. */ + uasm_copy_handler(relocs, labels, tlb_handler, p, f); + final_len = p - tlb_handler; + break; + } else { + if (((p - tlb_handler) > (MIPS64_REFILL_INSNS * 2) - 1) + || (((p - tlb_handler) > (MIPS64_REFILL_INSNS * 2) - 3) + && uasm_insn_has_bdelay(relocs, + tlb_handler + MIPS64_REFILL_INSNS - 3))) + panic("TLB refill handler space exceeded"); /* - * If the branch would fall in a delay slot, - * we must back up an additional instruction - * so that it is no longer in a delay slot. + * Now fold the handler in the TLB refill handler space. */ - if (uasm_insn_has_bdelay(relocs, split - 1)) - split--; - } - /* Copy first part of the handler. */ - uasm_copy_handler(relocs, labels, tlb_handler, split, f); - f += split - tlb_handler; - - if (ov) { - /* Insert branch. */ - uasm_l_split(&l, final_handler); - uasm_il_b(&f, &r, label_split); - if (uasm_insn_has_bdelay(relocs, split)) - uasm_i_nop(&f); - else { - uasm_copy_handler(relocs, labels, - split, split + 1, f); - uasm_move_labels(labels, f, f + 1, -1); - f++; - split++; + f = final_handler + MIPS64_REFILL_INSNS; + if ((p - tlb_handler) <= MIPS64_REFILL_INSNS) { + /* Just copy the handler. */ + uasm_copy_handler(relocs, labels, tlb_handler, p, f); + final_len = p - tlb_handler; + } else { +#ifdef CONFIG_MIPS_HUGE_TLB_SUPPORT + const enum label_id ls = label_tlb_huge_update; +#else + const enum label_id ls = label_vmalloc; +#endif + u32 *split; + int ov = 0; + int i; + + for (i = 0; i < ARRAY_SIZE(labels) && labels[i].lab != ls; i++) + ; + BUG_ON(i == ARRAY_SIZE(labels)); + split = labels[i].addr; + + /* + * See if we have overflown one way or the other. + */ + if (split > tlb_handler + MIPS64_REFILL_INSNS || + split < p - MIPS64_REFILL_INSNS) + ov = 1; + + if (ov) { + /* + * Split two instructions before the end. One + * for the branch and one for the instruction + * in the delay slot. + */ + split = tlb_handler + MIPS64_REFILL_INSNS - 2; + + /* + * If the branch would fall in a delay slot, + * we must back up an additional instruction + * so that it is no longer in a delay slot. + */ + if (uasm_insn_has_bdelay(relocs, split - 1)) + split--; + } + /* Copy first part of the handler. */ + uasm_copy_handler(relocs, labels, tlb_handler, split, f); + f += split - tlb_handler; + + if (ov) { + /* Insert branch. */ + uasm_l_split(&l, final_handler); + uasm_il_b(&f, &r, label_split); + if (uasm_insn_has_bdelay(relocs, split)) + uasm_i_nop(&f); + else { + uasm_copy_handler(relocs, labels, + split, split + 1, f); + uasm_move_labels(labels, f, f + 1, -1); + f++; + split++; + } + } + + /* Copy the rest of the handler. */ + uasm_copy_handler(relocs, labels, split, p, final_handler); + final_len = (f - (final_handler + MIPS64_REFILL_INSNS)) + + (p - split); } } - - /* Copy the rest of the handler. */ - uasm_copy_handler(relocs, labels, split, p, final_handler); - final_len = (f - (final_handler + MIPS64_REFILL_INSNS)) + - (p - split); + break; } -#endif /* CONFIG_64BIT */ uasm_resolve_relocs(relocs, labels); pr_debug("Wrote TLB refill handler (%u instructions).\n", @@ -1444,33 +1419,34 @@ static void __cpuinit build_r4000_tlb_refill_handler(void) dump_handler("r4000_tlb_refill", (u32 *)ebase, 64); } -/* - * 128 instructions for the fastpath handler is generous and should - * never be exceeded. - */ -#define FASTPATH_SIZE 128 - -u32 handle_tlbl[FASTPATH_SIZE] __cacheline_aligned; -u32 handle_tlbs[FASTPATH_SIZE] __cacheline_aligned; -u32 handle_tlbm[FASTPATH_SIZE] __cacheline_aligned; -#ifdef CONFIG_MIPS_PGD_C0_CONTEXT -u32 tlbmiss_handler_setup_pgd_array[16] __cacheline_aligned; +extern u32 handle_tlbl[], handle_tlbl_end[]; +extern u32 handle_tlbs[], handle_tlbs_end[]; +extern u32 handle_tlbm[], handle_tlbm_end[]; +extern u32 tlbmiss_handler_setup_pgd_start[], tlbmiss_handler_setup_pgd[]; +extern u32 tlbmiss_handler_setup_pgd_end[]; -static void __cpuinit build_r4000_setup_pgd(void) +static void build_setup_pgd(void) { const int a0 = 4; - const int a1 = 5; - u32 *p = tlbmiss_handler_setup_pgd_array; - struct uasm_label *l = labels; - struct uasm_reloc *r = relocs; + const int __maybe_unused a1 = 5; + const int __maybe_unused a2 = 6; + u32 *p = tlbmiss_handler_setup_pgd_start; + const int tlbmiss_handler_setup_pgd_size = + tlbmiss_handler_setup_pgd_end - tlbmiss_handler_setup_pgd_start; +#ifndef CONFIG_MIPS_PGD_C0_CONTEXT + long pgdc = (long)pgd_current; +#endif - memset(tlbmiss_handler_setup_pgd_array, 0, sizeof(tlbmiss_handler_setup_pgd_array)); + memset(tlbmiss_handler_setup_pgd, 0, tlbmiss_handler_setup_pgd_size * + sizeof(tlbmiss_handler_setup_pgd[0])); memset(labels, 0, sizeof(labels)); memset(relocs, 0, sizeof(relocs)); - pgd_reg = allocate_kscratch(); - +#ifdef CONFIG_MIPS_PGD_C0_CONTEXT if (pgd_reg == -1) { + struct uasm_label *l = labels; + struct uasm_reloc *r = relocs; + /* PGD << 11 in c0_Context */ /* * If it is a ckseg0 address, convert to a physical @@ -1490,21 +1466,40 @@ static void __cpuinit build_r4000_setup_pgd(void) } else { /* PGD in c0_KScratch */ uasm_i_jr(&p, 31); - UASM_i_MTC0(&p, a0, 31, pgd_reg); + UASM_i_MTC0(&p, a0, c0_kscratch(), pgd_reg); } - if (p - tlbmiss_handler_setup_pgd_array > ARRAY_SIZE(tlbmiss_handler_setup_pgd_array)) - panic("tlbmiss_handler_setup_pgd_array space exceeded"); +#else +#ifdef CONFIG_SMP + /* Save PGD to pgd_current[smp_processor_id()] */ + UASM_i_CPUID_MFC0(&p, a1, SMP_CPUID_REG); + UASM_i_SRL_SAFE(&p, a1, a1, SMP_CPUID_PTRSHIFT); + UASM_i_LA_mostly(&p, a2, pgdc); + UASM_i_ADDU(&p, a2, a2, a1); + UASM_i_SW(&p, a0, uasm_rel_lo(pgdc), a2); +#else + UASM_i_LA_mostly(&p, a2, pgdc); + UASM_i_SW(&p, a0, uasm_rel_lo(pgdc), a2); +#endif /* SMP */ + uasm_i_jr(&p, 31); + + /* if pgd_reg is allocated, save PGD also to scratch register */ + if (pgd_reg != -1) + UASM_i_MTC0(&p, a0, c0_kscratch(), pgd_reg); + else + uasm_i_nop(&p); +#endif + if (p >= tlbmiss_handler_setup_pgd_end) + panic("tlbmiss_handler_setup_pgd space exceeded"); + uasm_resolve_relocs(relocs, labels); - pr_debug("Wrote tlbmiss_handler_setup_pgd_array (%u instructions).\n", - (unsigned int)(p - tlbmiss_handler_setup_pgd_array)); + pr_debug("Wrote tlbmiss_handler_setup_pgd (%u instructions).\n", + (unsigned int)(p - tlbmiss_handler_setup_pgd)); - dump_handler("tlbmiss_handler", - tlbmiss_handler_setup_pgd_array, - ARRAY_SIZE(tlbmiss_handler_setup_pgd_array)); + dump_handler("tlbmiss_handler", tlbmiss_handler_setup_pgd, + tlbmiss_handler_setup_pgd_size); } -#endif -static void __cpuinit +static void iPTE_LW(u32 **p, unsigned int pte, unsigned int ptr) { #ifdef CONFIG_SMP @@ -1524,7 +1519,7 @@ iPTE_LW(u32 **p, unsigned int pte, unsigned int ptr) #endif } -static void __cpuinit +static void iPTE_SW(u32 **p, struct uasm_reloc **r, unsigned int pte, unsigned int ptr, unsigned int mode) { @@ -1584,7 +1579,7 @@ iPTE_SW(u32 **p, struct uasm_reloc **r, unsigned int pte, unsigned int ptr, * the page table where this PTE is located, PTE will be re-loaded * with it's original value. */ -static void __cpuinit +static void build_pte_present(u32 **p, struct uasm_reloc **r, int pte, int ptr, int scratch, enum label_id lid) { @@ -1612,7 +1607,7 @@ build_pte_present(u32 **p, struct uasm_reloc **r, } /* Make PTE valid, store result in PTR. */ -static void __cpuinit +static void build_make_valid(u32 **p, struct uasm_reloc **r, unsigned int pte, unsigned int ptr) { @@ -1625,7 +1620,7 @@ build_make_valid(u32 **p, struct uasm_reloc **r, unsigned int pte, * Check if PTE can be written to, if not branch to LABEL. Regardless * restore PTE with value from PTR when done. */ -static void __cpuinit +static void build_pte_writable(u32 **p, struct uasm_reloc **r, unsigned int pte, unsigned int ptr, int scratch, enum label_id lid) @@ -1645,7 +1640,7 @@ build_pte_writable(u32 **p, struct uasm_reloc **r, /* Make PTE writable, update software status bits as well, then store * at PTR. */ -static void __cpuinit +static void build_make_write(u32 **p, struct uasm_reloc **r, unsigned int pte, unsigned int ptr) { @@ -1659,7 +1654,7 @@ build_make_write(u32 **p, struct uasm_reloc **r, unsigned int pte, * Check if PTE can be modified, if not branch to LABEL. Regardless * restore PTE with value from PTR when done. */ -static void __cpuinit +static void build_pte_modifiable(u32 **p, struct uasm_reloc **r, unsigned int pte, unsigned int ptr, int scratch, enum label_id lid) @@ -1688,7 +1683,7 @@ build_pte_modifiable(u32 **p, struct uasm_reloc **r, * This places the pte into ENTRYLO0 and writes it with tlbwi. * Then it returns. */ -static void __cpuinit +static void build_r3000_pte_reload_tlbwi(u32 **p, unsigned int pte, unsigned int tmp) { uasm_i_mtc0(p, pte, C0_ENTRYLO0); /* cp0 delay */ @@ -1704,7 +1699,7 @@ build_r3000_pte_reload_tlbwi(u32 **p, unsigned int pte, unsigned int tmp) * may have the probe fail bit set as a result of a trap on a * kseg2 access, i.e. without refill. Then it returns. */ -static void __cpuinit +static void build_r3000_tlb_reload_write(u32 **p, struct uasm_label **l, struct uasm_reloc **r, unsigned int pte, unsigned int tmp) @@ -1722,7 +1717,7 @@ build_r3000_tlb_reload_write(u32 **p, struct uasm_label **l, uasm_i_rfe(p); /* branch delay */ } -static void __cpuinit +static void build_r3000_tlbchange_handler_head(u32 **p, unsigned int pte, unsigned int ptr) { @@ -1742,13 +1737,14 @@ build_r3000_tlbchange_handler_head(u32 **p, unsigned int pte, uasm_i_tlbp(p); /* load delay */ } -static void __cpuinit build_r3000_tlb_load_handler(void) +static void build_r3000_tlb_load_handler(void) { u32 *p = handle_tlbl; + const int handle_tlbl_size = handle_tlbl_end - handle_tlbl; struct uasm_label *l = labels; struct uasm_reloc *r = relocs; - memset(handle_tlbl, 0, sizeof(handle_tlbl)); + memset(handle_tlbl, 0, handle_tlbl_size * sizeof(handle_tlbl[0])); memset(labels, 0, sizeof(labels)); memset(relocs, 0, sizeof(relocs)); @@ -1762,23 +1758,24 @@ static void __cpuinit build_r3000_tlb_load_handler(void) uasm_i_j(&p, (unsigned long)tlb_do_page_fault_0 & 0x0fffffff); uasm_i_nop(&p); - if ((p - handle_tlbl) > FASTPATH_SIZE) + if (p >= handle_tlbl_end) panic("TLB load handler fastpath space exceeded"); uasm_resolve_relocs(relocs, labels); pr_debug("Wrote TLB load handler fastpath (%u instructions).\n", (unsigned int)(p - handle_tlbl)); - dump_handler("r3000_tlb_load", handle_tlbl, ARRAY_SIZE(handle_tlbl)); + dump_handler("r3000_tlb_load", handle_tlbl, handle_tlbl_size); } -static void __cpuinit build_r3000_tlb_store_handler(void) +static void build_r3000_tlb_store_handler(void) { u32 *p = handle_tlbs; + const int handle_tlbs_size = handle_tlbs_end - handle_tlbs; struct uasm_label *l = labels; struct uasm_reloc *r = relocs; - memset(handle_tlbs, 0, sizeof(handle_tlbs)); + memset(handle_tlbs, 0, handle_tlbs_size * sizeof(handle_tlbs[0])); memset(labels, 0, sizeof(labels)); memset(relocs, 0, sizeof(relocs)); @@ -1792,23 +1789,24 @@ static void __cpuinit build_r3000_tlb_store_handler(void) uasm_i_j(&p, (unsigned long)tlb_do_page_fault_1 & 0x0fffffff); uasm_i_nop(&p); - if ((p - handle_tlbs) > FASTPATH_SIZE) + if (p >= handle_tlbs_end) panic("TLB store handler fastpath space exceeded"); uasm_resolve_relocs(relocs, labels); pr_debug("Wrote TLB store handler fastpath (%u instructions).\n", (unsigned int)(p - handle_tlbs)); - dump_handler("r3000_tlb_store", handle_tlbs, ARRAY_SIZE(handle_tlbs)); + dump_handler("r3000_tlb_store", handle_tlbs, handle_tlbs_size); } -static void __cpuinit build_r3000_tlb_modify_handler(void) +static void build_r3000_tlb_modify_handler(void) { u32 *p = handle_tlbm; + const int handle_tlbm_size = handle_tlbm_end - handle_tlbm; struct uasm_label *l = labels; struct uasm_reloc *r = relocs; - memset(handle_tlbm, 0, sizeof(handle_tlbm)); + memset(handle_tlbm, 0, handle_tlbm_size * sizeof(handle_tlbm[0])); memset(labels, 0, sizeof(labels)); memset(relocs, 0, sizeof(relocs)); @@ -1822,21 +1820,21 @@ static void __cpuinit build_r3000_tlb_modify_handler(void) uasm_i_j(&p, (unsigned long)tlb_do_page_fault_1 & 0x0fffffff); uasm_i_nop(&p); - if ((p - handle_tlbm) > FASTPATH_SIZE) + if (p >= handle_tlbm_end) panic("TLB modify handler fastpath space exceeded"); uasm_resolve_relocs(relocs, labels); pr_debug("Wrote TLB modify handler fastpath (%u instructions).\n", (unsigned int)(p - handle_tlbm)); - dump_handler("r3000_tlb_modify", handle_tlbm, ARRAY_SIZE(handle_tlbm)); + dump_handler("r3000_tlb_modify", handle_tlbm, handle_tlbm_size); } #endif /* CONFIG_MIPS_PGD_C0_CONTEXT */ /* * R4000 style TLB load/store/modify handlers. */ -static struct work_registers __cpuinit +static struct work_registers build_r4000_tlbchange_handler_head(u32 **p, struct uasm_label **l, struct uasm_reloc **r) { @@ -1872,7 +1870,7 @@ build_r4000_tlbchange_handler_head(u32 **p, struct uasm_label **l, return wr; } -static void __cpuinit +static void build_r4000_tlbchange_handler_tail(u32 **p, struct uasm_label **l, struct uasm_reloc **r, unsigned int tmp, unsigned int ptr) @@ -1890,14 +1888,15 @@ build_r4000_tlbchange_handler_tail(u32 **p, struct uasm_label **l, #endif } -static void __cpuinit build_r4000_tlb_load_handler(void) +static void build_r4000_tlb_load_handler(void) { u32 *p = handle_tlbl; + const int handle_tlbl_size = handle_tlbl_end - handle_tlbl; struct uasm_label *l = labels; struct uasm_reloc *r = relocs; struct work_registers wr; - memset(handle_tlbl, 0, sizeof(handle_tlbl)); + memset(handle_tlbl, 0, handle_tlbl_size * sizeof(handle_tlbl[0])); memset(labels, 0, sizeof(labels)); memset(relocs, 0, sizeof(relocs)); @@ -1935,6 +1934,19 @@ static void __cpuinit build_r4000_tlb_load_handler(void) uasm_i_nop(&p); uasm_i_tlbr(&p); + + switch (current_cpu_type()) { + default: + if (cpu_has_mips_r2) { + uasm_i_ehb(&p); + + case CPU_CAVIUM_OCTEON: + case CPU_CAVIUM_OCTEON_PLUS: + case CPU_CAVIUM_OCTEON2: + break; + } + } + /* Examine entrylo 0 or 1 based on ptr. */ if (use_bbit_insns()) { uasm_i_bbit0(&p, wr.r2, ilog2(sizeof(pte_t)), 8); @@ -1989,6 +2001,19 @@ static void __cpuinit build_r4000_tlb_load_handler(void) uasm_i_nop(&p); uasm_i_tlbr(&p); + + switch (current_cpu_type()) { + default: + if (cpu_has_mips_r2) { + uasm_i_ehb(&p); + + case CPU_CAVIUM_OCTEON: + case CPU_CAVIUM_OCTEON_PLUS: + case CPU_CAVIUM_OCTEON2: + break; + } + } + /* Examine entrylo 0 or 1 based on ptr. */ if (use_bbit_insns()) { uasm_i_bbit0(&p, wr.r2, ilog2(sizeof(pte_t)), 8); @@ -2036,24 +2061,25 @@ static void __cpuinit build_r4000_tlb_load_handler(void) uasm_i_j(&p, (unsigned long)tlb_do_page_fault_0 & 0x0fffffff); uasm_i_nop(&p); - if ((p - handle_tlbl) > FASTPATH_SIZE) + if (p >= handle_tlbl_end) panic("TLB load handler fastpath space exceeded"); uasm_resolve_relocs(relocs, labels); pr_debug("Wrote TLB load handler fastpath (%u instructions).\n", (unsigned int)(p - handle_tlbl)); - dump_handler("r4000_tlb_load", handle_tlbl, ARRAY_SIZE(handle_tlbl)); + dump_handler("r4000_tlb_load", handle_tlbl, handle_tlbl_size); } -static void __cpuinit build_r4000_tlb_store_handler(void) +static void build_r4000_tlb_store_handler(void) { u32 *p = handle_tlbs; + const int handle_tlbs_size = handle_tlbs_end - handle_tlbs; struct uasm_label *l = labels; struct uasm_reloc *r = relocs; struct work_registers wr; - memset(handle_tlbs, 0, sizeof(handle_tlbs)); + memset(handle_tlbs, 0, handle_tlbs_size * sizeof(handle_tlbs[0])); memset(labels, 0, sizeof(labels)); memset(relocs, 0, sizeof(relocs)); @@ -2090,24 +2116,25 @@ static void __cpuinit build_r4000_tlb_store_handler(void) uasm_i_j(&p, (unsigned long)tlb_do_page_fault_1 & 0x0fffffff); uasm_i_nop(&p); - if ((p - handle_tlbs) > FASTPATH_SIZE) + if (p >= handle_tlbs_end) panic("TLB store handler fastpath space exceeded"); uasm_resolve_relocs(relocs, labels); pr_debug("Wrote TLB store handler fastpath (%u instructions).\n", (unsigned int)(p - handle_tlbs)); - dump_handler("r4000_tlb_store", handle_tlbs, ARRAY_SIZE(handle_tlbs)); + dump_handler("r4000_tlb_store", handle_tlbs, handle_tlbs_size); } -static void __cpuinit build_r4000_tlb_modify_handler(void) +static void build_r4000_tlb_modify_handler(void) { u32 *p = handle_tlbm; + const int handle_tlbm_size = handle_tlbm_end - handle_tlbm; struct uasm_label *l = labels; struct uasm_reloc *r = relocs; struct work_registers wr; - memset(handle_tlbm, 0, sizeof(handle_tlbm)); + memset(handle_tlbm, 0, handle_tlbm_size * sizeof(handle_tlbm[0])); memset(labels, 0, sizeof(labels)); memset(relocs, 0, sizeof(relocs)); @@ -2145,17 +2172,29 @@ static void __cpuinit build_r4000_tlb_modify_handler(void) uasm_i_j(&p, (unsigned long)tlb_do_page_fault_1 & 0x0fffffff); uasm_i_nop(&p); - if ((p - handle_tlbm) > FASTPATH_SIZE) + if (p >= handle_tlbm_end) panic("TLB modify handler fastpath space exceeded"); uasm_resolve_relocs(relocs, labels); pr_debug("Wrote TLB modify handler fastpath (%u instructions).\n", (unsigned int)(p - handle_tlbm)); - dump_handler("r4000_tlb_modify", handle_tlbm, ARRAY_SIZE(handle_tlbm)); + dump_handler("r4000_tlb_modify", handle_tlbm, handle_tlbm_size); +} + +static void flush_tlb_handlers(void) +{ + local_flush_icache_range((unsigned long)handle_tlbl, + (unsigned long)handle_tlbl_end); + local_flush_icache_range((unsigned long)handle_tlbs, + (unsigned long)handle_tlbs_end); + local_flush_icache_range((unsigned long)handle_tlbm, + (unsigned long)handle_tlbm_end); + local_flush_icache_range((unsigned long)tlbmiss_handler_setup_pgd, + (unsigned long)tlbmiss_handler_setup_pgd_end); } -void __cpuinit build_tlb_refill_handler(void) +void build_tlb_refill_handler(void) { /* * The refill handler is generated per-CPU, multi-node systems @@ -2184,9 +2223,11 @@ void __cpuinit build_tlb_refill_handler(void) if (!run_once) { if (!cpu_has_local_ebase) build_r3000_tlb_refill_handler(); + build_setup_pgd(); build_r3000_tlb_load_handler(); build_r3000_tlb_store_handler(); build_r3000_tlb_modify_handler(); + flush_tlb_handlers(); run_once++; } #else @@ -2206,31 +2247,16 @@ void __cpuinit build_tlb_refill_handler(void) default: if (!run_once) { scratch_reg = allocate_kscratch(); -#ifdef CONFIG_MIPS_PGD_C0_CONTEXT - build_r4000_setup_pgd(); -#endif + build_setup_pgd(); build_r4000_tlb_load_handler(); build_r4000_tlb_store_handler(); build_r4000_tlb_modify_handler(); if (!cpu_has_local_ebase) build_r4000_tlb_refill_handler(); + flush_tlb_handlers(); run_once++; } if (cpu_has_local_ebase) build_r4000_tlb_refill_handler(); } } - -void __cpuinit flush_tlb_handlers(void) -{ - local_flush_icache_range((unsigned long)handle_tlbl, - (unsigned long)handle_tlbl + sizeof(handle_tlbl)); - local_flush_icache_range((unsigned long)handle_tlbs, - (unsigned long)handle_tlbs + sizeof(handle_tlbs)); - local_flush_icache_range((unsigned long)handle_tlbm, - (unsigned long)handle_tlbm + sizeof(handle_tlbm)); -#ifdef CONFIG_MIPS_PGD_C0_CONTEXT - local_flush_icache_range((unsigned long)tlbmiss_handler_setup_pgd_array, - (unsigned long)tlbmiss_handler_setup_pgd_array + sizeof(handle_tlbm)); -#endif -} diff --git a/arch/mips/mm/uasm-micromips.c b/arch/mips/mm/uasm-micromips.c index 162ee6d6278..8399ddf03a0 100644 --- a/arch/mips/mm/uasm-micromips.c +++ b/arch/mips/mm/uasm-micromips.c @@ -15,7 +15,6 @@ #include <linux/kernel.h> #include <linux/types.h> -#include <linux/init.h> #include <asm/inst.h> #include <asm/elf.h> @@ -49,7 +48,7 @@ #include "uasm.c" -static struct insn insn_table_MM[] __uasminitdata = { +static struct insn insn_table_MM[] = { { insn_addu, M(mm_pool32a_op, 0, 0, 0, 0, mm_addu32_op), RT | RS | RD }, { insn_addiu, M(mm_addiu32_op, 0, 0, 0, 0, 0), RT | RS | SIMM }, { insn_and, M(mm_pool32a_op, 0, 0, 0, 0, mm_and_op), RT | RS | RD }, @@ -64,6 +63,7 @@ static struct insn insn_table_MM[] __uasminitdata = { { insn_cache, M(mm_pool32b_op, 0, 0, mm_cache_func, 0, 0), RT | RS | SIMM }, { insn_daddu, 0, 0 }, { insn_daddiu, 0, 0 }, + { insn_divu, M(mm_pool32a_op, 0, 0, 0, mm_divu_op, mm_pool32axf_op), RT | RS }, { insn_dmfc0, 0, 0 }, { insn_dmtc0, 0, 0 }, { insn_dsll, 0, 0 }, @@ -79,14 +79,20 @@ static struct insn insn_table_MM[] __uasminitdata = { { insn_ext, M(mm_pool32a_op, 0, 0, 0, 0, mm_ext_op), RT | RS | RD | RE }, { insn_j, M(mm_j32_op, 0, 0, 0, 0, 0), JIMM }, { insn_jal, M(mm_jal32_op, 0, 0, 0, 0, 0), JIMM }, + { insn_jalr, M(mm_pool32a_op, 0, 0, 0, mm_jalr_op, mm_pool32axf_op), RT | RS }, { insn_jr, M(mm_pool32a_op, 0, 0, 0, mm_jalr_op, mm_pool32axf_op), RS }, + { insn_lb, M(mm_lb32_op, 0, 0, 0, 0, 0), RT | RS | SIMM }, { insn_ld, 0, 0 }, + { insn_lh, M(mm_lh32_op, 0, 0, 0, 0, 0), RS | RS | SIMM }, { insn_ll, M(mm_pool32c_op, 0, 0, (mm_ll_func << 1), 0, 0), RS | RT | SIMM }, { insn_lld, 0, 0 }, { insn_lui, M(mm_pool32i_op, mm_lui_op, 0, 0, 0, 0), RS | SIMM }, { insn_lw, M(mm_lw32_op, 0, 0, 0, 0, 0), RT | RS | SIMM }, { insn_mfc0, M(mm_pool32a_op, 0, 0, 0, mm_mfc0_op, mm_pool32axf_op), RT | RS | RD }, + { insn_mfhi, M(mm_pool32a_op, 0, 0, 0, mm_mfhi32_op, mm_pool32axf_op), RS }, + { insn_mflo, M(mm_pool32a_op, 0, 0, 0, mm_mflo32_op, mm_pool32axf_op), RS }, { insn_mtc0, M(mm_pool32a_op, 0, 0, 0, mm_mtc0_op, mm_pool32axf_op), RT | RS | RD }, + { insn_mul, M(mm_pool32a_op, 0, 0, 0, 0, mm_mul_op), RT | RS | RD }, { insn_or, M(mm_pool32a_op, 0, 0, 0, 0, mm_or32_op), RT | RS | RD }, { insn_ori, M(mm_ori32_op, 0, 0, 0, 0, 0), RT | RS | UIMM }, { insn_pref, M(mm_pool32c_op, 0, 0, (mm_pref_func << 1), 0, 0), RT | RS | SIMM }, @@ -95,15 +101,23 @@ static struct insn insn_table_MM[] __uasminitdata = { { insn_scd, 0, 0 }, { insn_sd, 0, 0 }, { insn_sll, M(mm_pool32a_op, 0, 0, 0, 0, mm_sll32_op), RT | RS | RD }, + { insn_sllv, M(mm_pool32a_op, 0, 0, 0, 0, mm_sllv32_op), RT | RS | RD }, + { insn_slt, M(mm_pool32a_op, 0, 0, 0, 0, mm_slt_op), RT | RS | RD }, + { insn_sltiu, M(mm_sltiu32_op, 0, 0, 0, 0, 0), RT | RS | SIMM }, + { insn_sltu, M(mm_pool32a_op, 0, 0, 0, 0, mm_sltu_op), RT | RS | RD }, { insn_sra, M(mm_pool32a_op, 0, 0, 0, 0, mm_sra_op), RT | RS | RD }, { insn_srl, M(mm_pool32a_op, 0, 0, 0, 0, mm_srl32_op), RT | RS | RD }, + { insn_srlv, M(mm_pool32a_op, 0, 0, 0, 0, mm_srlv32_op), RT | RS | RD }, { insn_rotr, M(mm_pool32a_op, 0, 0, 0, 0, mm_rotr_op), RT | RS | RD }, { insn_subu, M(mm_pool32a_op, 0, 0, 0, 0, mm_subu32_op), RT | RS | RD }, { insn_sw, M(mm_sw32_op, 0, 0, 0, 0, 0), RT | RS | SIMM }, + { insn_sync, M(mm_pool32a_op, 0, 0, 0, mm_sync_op, mm_pool32axf_op), RS }, { insn_tlbp, M(mm_pool32a_op, 0, 0, 0, mm_tlbp_op, mm_pool32axf_op), 0 }, { insn_tlbr, M(mm_pool32a_op, 0, 0, 0, mm_tlbr_op, mm_pool32axf_op), 0 }, { insn_tlbwi, M(mm_pool32a_op, 0, 0, 0, mm_tlbwi_op, mm_pool32axf_op), 0 }, { insn_tlbwr, M(mm_pool32a_op, 0, 0, 0, mm_tlbwr_op, mm_pool32axf_op), 0 }, + { insn_wait, M(mm_pool32a_op, 0, 0, 0, mm_wait_op, mm_pool32axf_op), SCIMM }, + { insn_wsbh, M(mm_pool32a_op, 0, 0, 0, mm_wsbh_op, mm_pool32axf_op), RT | RS }, { insn_xor, M(mm_pool32a_op, 0, 0, 0, 0, mm_xor32_op), RT | RS | RD }, { insn_xori, M(mm_xori32_op, 0, 0, 0, 0, 0), RT | RS | UIMM }, { insn_dins, 0, 0 }, @@ -118,7 +132,7 @@ static struct insn insn_table_MM[] __uasminitdata = { #undef M -static inline __uasminit u32 build_bimm(s32 arg) +static inline u32 build_bimm(s32 arg) { WARN(arg > 0xffff || arg < -0x10000, KERN_WARNING "Micro-assembler field overflow\n"); @@ -128,7 +142,7 @@ static inline __uasminit u32 build_bimm(s32 arg) return ((arg < 0) ? (1 << 15) : 0) | ((arg >> 1) & 0x7fff); } -static inline __uasminit u32 build_jimm(u32 arg) +static inline u32 build_jimm(u32 arg) { WARN(arg & ~((JIMM_MASK << 2) | 1), @@ -141,7 +155,7 @@ static inline __uasminit u32 build_jimm(u32 arg) * The order of opcode arguments is implicitly left to right, * starting with RS and ending with FUNC or IMM. */ -static void __uasminit build_insn(u32 **buf, enum opcode opc, ...) +static void build_insn(u32 **buf, enum opcode opc, ...) { struct insn *ip = NULL; unsigned int i; @@ -199,7 +213,7 @@ static void __uasminit build_insn(u32 **buf, enum opcode opc, ...) (*buf)++; } -static inline void __uasminit +static inline void __resolve_relocs(struct uasm_reloc *rel, struct uasm_label *lab) { long laddr = (long)lab->addr; diff --git a/arch/mips/mm/uasm-mips.c b/arch/mips/mm/uasm-mips.c index 5fcdd8fe3e8..6708a2dbf93 100644 --- a/arch/mips/mm/uasm-mips.c +++ b/arch/mips/mm/uasm-mips.c @@ -15,7 +15,6 @@ #include <linux/kernel.h> #include <linux/types.h> -#include <linux/init.h> #include <asm/inst.h> #include <asm/elf.h> @@ -49,7 +48,7 @@ #include "uasm.c" -static struct insn insn_table[] __uasminitdata = { +static struct insn insn_table[] = { { insn_addiu, M(addiu_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_addu, M(spec_op, 0, 0, 0, 0, addu_op), RS | RT | RD }, { insn_andi, M(andi_op, 0, 0, 0, 0, 0), RS | RT | UIMM }, @@ -68,6 +67,7 @@ static struct insn insn_table[] __uasminitdata = { { insn_daddu, M(spec_op, 0, 0, 0, 0, daddu_op), RS | RT | RD }, { insn_dinsm, M(spec3_op, 0, 0, 0, 0, dinsm_op), RS | RT | RD | RE }, { insn_dins, M(spec3_op, 0, 0, 0, 0, dins_op), RS | RT | RD | RE }, + { insn_divu, M(spec_op, 0, 0, 0, 0, divu_op), RS | RT }, { insn_dmfc0, M(cop0_op, dmfc_op, 0, 0, 0, 0), RT | RD | SET}, { insn_dmtc0, M(cop0_op, dmtc_op, 0, 0, 0, 0), RT | RD | SET}, { insn_drotr32, M(spec_op, 1, 0, 0, 0, dsrl32_op), RT | RD | RE }, @@ -83,17 +83,23 @@ static struct insn insn_table[] __uasminitdata = { { insn_ins, M(spec3_op, 0, 0, 0, 0, ins_op), RS | RT | RD | RE }, { insn_j, M(j_op, 0, 0, 0, 0, 0), JIMM }, { insn_jal, M(jal_op, 0, 0, 0, 0, 0), JIMM }, + { insn_jalr, M(spec_op, 0, 0, 0, 0, jalr_op), RS | RD }, { insn_j, M(j_op, 0, 0, 0, 0, 0), JIMM }, { insn_jr, M(spec_op, 0, 0, 0, 0, jr_op), RS }, + { insn_lb, M(lb_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_ld, M(ld_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_ldx, M(spec3_op, 0, 0, 0, ldx_op, lx_op), RS | RT | RD }, + { insn_lh, M(lh_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_lld, M(lld_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_ll, M(ll_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_lui, M(lui_op, 0, 0, 0, 0, 0), RT | SIMM }, { insn_lw, M(lw_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_lwx, M(spec3_op, 0, 0, 0, lwx_op, lx_op), RS | RT | RD }, { insn_mfc0, M(cop0_op, mfc_op, 0, 0, 0, 0), RT | RD | SET}, + { insn_mfhi, M(spec_op, 0, 0, 0, 0, mfhi_op), RD }, + { insn_mflo, M(spec_op, 0, 0, 0, 0, mflo_op), RD }, { insn_mtc0, M(cop0_op, mtc_op, 0, 0, 0, 0), RT | RD | SET}, + { insn_mul, M(spec2_op, 0, 0, 0, 0, mul_op), RS | RT | RD}, { insn_ori, M(ori_op, 0, 0, 0, 0, 0), RS | RT | UIMM }, { insn_or, M(spec_op, 0, 0, 0, 0, or_op), RS | RT | RD }, { insn_pref, M(pref_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, @@ -103,23 +109,32 @@ static struct insn insn_table[] __uasminitdata = { { insn_sc, M(sc_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_sd, M(sd_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_sll, M(spec_op, 0, 0, 0, 0, sll_op), RT | RD | RE }, + { insn_sllv, M(spec_op, 0, 0, 0, 0, sllv_op), RS | RT | RD }, + { insn_slt, M(spec_op, 0, 0, 0, 0, slt_op), RS | RT | RD }, + { insn_sltiu, M(sltiu_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, + { insn_sltu, M(spec_op, 0, 0, 0, 0, sltu_op), RS | RT | RD }, { insn_sra, M(spec_op, 0, 0, 0, 0, sra_op), RT | RD | RE }, { insn_srl, M(spec_op, 0, 0, 0, 0, srl_op), RT | RD | RE }, + { insn_srlv, M(spec_op, 0, 0, 0, 0, srlv_op), RS | RT | RD }, { insn_subu, M(spec_op, 0, 0, 0, 0, subu_op), RS | RT | RD }, { insn_sw, M(sw_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, + { insn_sync, M(spec_op, 0, 0, 0, 0, sync_op), RE }, { insn_syscall, M(spec_op, 0, 0, 0, 0, syscall_op), SCIMM}, { insn_tlbp, M(cop0_op, cop_op, 0, 0, 0, tlbp_op), 0 }, { insn_tlbr, M(cop0_op, cop_op, 0, 0, 0, tlbr_op), 0 }, { insn_tlbwi, M(cop0_op, cop_op, 0, 0, 0, tlbwi_op), 0 }, { insn_tlbwr, M(cop0_op, cop_op, 0, 0, 0, tlbwr_op), 0 }, + { insn_wait, M(cop0_op, cop_op, 0, 0, 0, wait_op), SCIMM }, + { insn_wsbh, M(spec3_op, 0, 0, 0, wsbh_op, bshfl_op), RT | RD }, { insn_xori, M(xori_op, 0, 0, 0, 0, 0), RS | RT | UIMM }, { insn_xor, M(spec_op, 0, 0, 0, 0, xor_op), RS | RT | RD }, + { insn_yield, M(spec3_op, 0, 0, 0, 0, yield_op), RS | RD }, { insn_invalid, 0, 0 } }; #undef M -static inline __uasminit u32 build_bimm(s32 arg) +static inline u32 build_bimm(s32 arg) { WARN(arg > 0x1ffff || arg < -0x20000, KERN_WARNING "Micro-assembler field overflow\n"); @@ -129,7 +144,7 @@ static inline __uasminit u32 build_bimm(s32 arg) return ((arg < 0) ? (1 << 15) : 0) | ((arg >> 2) & 0x7fff); } -static inline __uasminit u32 build_jimm(u32 arg) +static inline u32 build_jimm(u32 arg) { WARN(arg & ~(JIMM_MASK << 2), KERN_WARNING "Micro-assembler field overflow\n"); @@ -141,7 +156,7 @@ static inline __uasminit u32 build_jimm(u32 arg) * The order of opcode arguments is implicitly left to right, * starting with RS and ending with FUNC or IMM. */ -static void __uasminit build_insn(u32 **buf, enum opcode opc, ...) +static void build_insn(u32 **buf, enum opcode opc, ...) { struct insn *ip = NULL; unsigned int i; @@ -187,7 +202,7 @@ static void __uasminit build_insn(u32 **buf, enum opcode opc, ...) (*buf)++; } -static inline void __uasminit +static inline void __resolve_relocs(struct uasm_reloc *rel, struct uasm_label *lab) { long laddr = (long)lab->addr; diff --git a/arch/mips/mm/uasm.c b/arch/mips/mm/uasm.c index 7eb5e4355d2..a01b0d6cedd 100644 --- a/arch/mips/mm/uasm.c +++ b/arch/mips/mm/uasm.c @@ -47,14 +47,16 @@ enum opcode { insn_addiu, insn_addu, insn_and, insn_andi, insn_bbit0, insn_bbit1, insn_beq, insn_beql, insn_bgez, insn_bgezl, insn_bltz, insn_bltzl, insn_bne, insn_cache, insn_daddiu, insn_daddu, insn_dins, insn_dinsm, - insn_dmfc0, insn_dmtc0, insn_drotr, insn_drotr32, insn_dsll, + insn_divu, insn_dmfc0, insn_dmtc0, insn_drotr, insn_drotr32, insn_dsll, insn_dsll32, insn_dsra, insn_dsrl, insn_dsrl32, insn_dsubu, insn_eret, - insn_ext, insn_ins, insn_j, insn_jal, insn_jr, insn_ld, insn_ldx, - insn_ll, insn_lld, insn_lui, insn_lw, insn_lwx, insn_mfc0, insn_mtc0, + insn_ext, insn_ins, insn_j, insn_jal, insn_jalr, insn_jr, insn_lb, + insn_ld, insn_ldx, insn_lh, insn_ll, insn_lld, insn_lui, insn_lw, + insn_lwx, insn_mfc0, insn_mfhi, insn_mflo, insn_mtc0, insn_mul, insn_or, insn_ori, insn_pref, insn_rfe, insn_rotr, insn_sc, insn_scd, - insn_sd, insn_sll, insn_sra, insn_srl, insn_subu, insn_sw, - insn_syscall, insn_tlbp, insn_tlbr, insn_tlbwi, insn_tlbwr, insn_xor, - insn_xori, + insn_sd, insn_sll, insn_sllv, insn_slt, insn_sltiu, insn_sltu, insn_sra, + insn_srl, insn_srlv, insn_subu, insn_sw, insn_sync, insn_syscall, + insn_tlbp, insn_tlbr, insn_tlbwi, insn_tlbwr, insn_wait, insn_wsbh, + insn_xor, insn_xori, insn_yield, }; struct insn { @@ -63,35 +65,35 @@ struct insn { enum fields fields; }; -static inline __uasminit u32 build_rs(u32 arg) +static inline u32 build_rs(u32 arg) { WARN(arg & ~RS_MASK, KERN_WARNING "Micro-assembler field overflow\n"); return (arg & RS_MASK) << RS_SH; } -static inline __uasminit u32 build_rt(u32 arg) +static inline u32 build_rt(u32 arg) { WARN(arg & ~RT_MASK, KERN_WARNING "Micro-assembler field overflow\n"); return (arg & RT_MASK) << RT_SH; } -static inline __uasminit u32 build_rd(u32 arg) +static inline u32 build_rd(u32 arg) { WARN(arg & ~RD_MASK, KERN_WARNING "Micro-assembler field overflow\n"); return (arg & RD_MASK) << RD_SH; } -static inline __uasminit u32 build_re(u32 arg) +static inline u32 build_re(u32 arg) { WARN(arg & ~RE_MASK, KERN_WARNING "Micro-assembler field overflow\n"); return (arg & RE_MASK) << RE_SH; } -static inline __uasminit u32 build_simm(s32 arg) +static inline u32 build_simm(s32 arg) { WARN(arg > 0x7fff || arg < -0x8000, KERN_WARNING "Micro-assembler field overflow\n"); @@ -99,14 +101,14 @@ static inline __uasminit u32 build_simm(s32 arg) return arg & 0xffff; } -static inline __uasminit u32 build_uimm(u32 arg) +static inline u32 build_uimm(u32 arg) { WARN(arg & ~IMM_MASK, KERN_WARNING "Micro-assembler field overflow\n"); return arg & IMM_MASK; } -static inline __uasminit u32 build_scimm(u32 arg) +static inline u32 build_scimm(u32 arg) { WARN(arg & ~SCIMM_MASK, KERN_WARNING "Micro-assembler field overflow\n"); @@ -114,21 +116,21 @@ static inline __uasminit u32 build_scimm(u32 arg) return (arg & SCIMM_MASK) << SCIMM_SH; } -static inline __uasminit u32 build_func(u32 arg) +static inline u32 build_func(u32 arg) { WARN(arg & ~FUNC_MASK, KERN_WARNING "Micro-assembler field overflow\n"); return arg & FUNC_MASK; } -static inline __uasminit u32 build_set(u32 arg) +static inline u32 build_set(u32 arg) { WARN(arg & ~SET_MASK, KERN_WARNING "Micro-assembler field overflow\n"); return arg & SET_MASK; } -static void __uasminit build_insn(u32 **buf, enum opcode opc, ...); +static void build_insn(u32 **buf, enum opcode opc, ...); #define I_u1u2u3(op) \ Ip_u1u2u3(op) \ @@ -137,6 +139,13 @@ Ip_u1u2u3(op) \ } \ UASM_EXPORT_SYMBOL(uasm_i##op); +#define I_s3s1s2(op) \ +Ip_s3s1s2(op) \ +{ \ + build_insn(buf, insn##op, b, c, a); \ +} \ +UASM_EXPORT_SYMBOL(uasm_i##op); + #define I_u2u1u3(op) \ Ip_u2u1u3(op) \ { \ @@ -144,6 +153,13 @@ Ip_u2u1u3(op) \ } \ UASM_EXPORT_SYMBOL(uasm_i##op); +#define I_u3u2u1(op) \ +Ip_u3u2u1(op) \ +{ \ + build_insn(buf, insn##op, c, b, a); \ +} \ +UASM_EXPORT_SYMBOL(uasm_i##op); + #define I_u3u1u2(op) \ Ip_u3u1u2(op) \ { \ @@ -200,6 +216,13 @@ Ip_u1u2(op) \ } \ UASM_EXPORT_SYMBOL(uasm_i##op); +#define I_u2u1(op) \ +Ip_u1u2(op) \ +{ \ + build_insn(buf, insn##op, b, a); \ +} \ +UASM_EXPORT_SYMBOL(uasm_i##op); + #define I_u1s2(op) \ Ip_u1s2(op) \ { \ @@ -237,6 +260,7 @@ I_u1u2u3(_dmfc0) I_u1u2u3(_dmtc0) I_u2u1s3(_daddiu) I_u3u1u2(_daddu) +I_u1u2(_divu) I_u2u1u3(_dsll) I_u2u1u3(_dsll32) I_u2u1u3(_dsra) @@ -250,14 +274,20 @@ I_u2u1msbdu3(_ext) I_u2u1msbu3(_ins) I_u1(_j) I_u1(_jal) +I_u2u1(_jalr) I_u1(_jr) +I_u2s3u1(_lb) I_u2s3u1(_ld) +I_u2s3u1(_lh) I_u2s3u1(_ll) I_u2s3u1(_lld) I_u1s2(_lui) I_u2s3u1(_lw) I_u1u2u3(_mfc0) +I_u1(_mfhi) +I_u1(_mflo) I_u1u2u3(_mtc0) +I_u3u1u2(_mul) I_u2u1u3(_ori) I_u3u1u2(_or) I_0(_rfe) @@ -265,17 +295,26 @@ I_u2s3u1(_sc) I_u2s3u1(_scd) I_u2s3u1(_sd) I_u2u1u3(_sll) +I_u3u2u1(_sllv) +I_s3s1s2(_slt) +I_u2u1s3(_sltiu) +I_u3u1u2(_sltu) I_u2u1u3(_sra) I_u2u1u3(_srl) +I_u3u2u1(_srlv) I_u2u1u3(_rotr) I_u3u1u2(_subu) I_u2s3u1(_sw) +I_u1(_sync) I_0(_tlbp) I_0(_tlbr) I_0(_tlbwi) I_0(_tlbwr) +I_u1(_wait); +I_u2u1(_wsbh) I_u3u1u2(_xor) I_u2u1u3(_xori) +I_u2u1(_yield) I_u2u1msbu3(_dins); I_u2u1msb32u3(_dinsm); I_u1(_syscall); @@ -286,7 +325,7 @@ I_u3u1u2(_ldx) #ifdef CONFIG_CPU_CAVIUM_OCTEON #include <asm/octeon/octeon.h> -void __uasminit ISAFUNC(uasm_i_pref)(u32 **buf, unsigned int a, signed int b, +void ISAFUNC(uasm_i_pref)(u32 **buf, unsigned int a, signed int b, unsigned int c) { if (OCTEON_IS_MODEL(OCTEON_CN63XX_PASS1_X) && a <= 24 && a != 5) @@ -304,7 +343,7 @@ I_u2s3u1(_pref) #endif /* Handle labels. */ -void __uasminit ISAFUNC(uasm_build_label)(struct uasm_label **lab, u32 *addr, int lid) +void ISAFUNC(uasm_build_label)(struct uasm_label **lab, u32 *addr, int lid) { (*lab)->addr = addr; (*lab)->lab = lid; @@ -312,7 +351,7 @@ void __uasminit ISAFUNC(uasm_build_label)(struct uasm_label **lab, u32 *addr, in } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_build_label)); -int __uasminit ISAFUNC(uasm_in_compat_space_p)(long addr) +int ISAFUNC(uasm_in_compat_space_p)(long addr) { /* Is this address in 32bit compat space? */ #ifdef CONFIG_64BIT @@ -323,7 +362,7 @@ int __uasminit ISAFUNC(uasm_in_compat_space_p)(long addr) } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_in_compat_space_p)); -static int __uasminit uasm_rel_highest(long val) +static int uasm_rel_highest(long val) { #ifdef CONFIG_64BIT return ((((val + 0x800080008000L) >> 48) & 0xffff) ^ 0x8000) - 0x8000; @@ -332,7 +371,7 @@ static int __uasminit uasm_rel_highest(long val) #endif } -static int __uasminit uasm_rel_higher(long val) +static int uasm_rel_higher(long val) { #ifdef CONFIG_64BIT return ((((val + 0x80008000L) >> 32) & 0xffff) ^ 0x8000) - 0x8000; @@ -341,19 +380,19 @@ static int __uasminit uasm_rel_higher(long val) #endif } -int __uasminit ISAFUNC(uasm_rel_hi)(long val) +int ISAFUNC(uasm_rel_hi)(long val) { return ((((val + 0x8000L) >> 16) & 0xffff) ^ 0x8000) - 0x8000; } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_rel_hi)); -int __uasminit ISAFUNC(uasm_rel_lo)(long val) +int ISAFUNC(uasm_rel_lo)(long val) { return ((val & 0xffff) ^ 0x8000) - 0x8000; } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_rel_lo)); -void __uasminit ISAFUNC(UASM_i_LA_mostly)(u32 **buf, unsigned int rs, long addr) +void ISAFUNC(UASM_i_LA_mostly)(u32 **buf, unsigned int rs, long addr) { if (!ISAFUNC(uasm_in_compat_space_p)(addr)) { ISAFUNC(uasm_i_lui)(buf, rs, uasm_rel_highest(addr)); @@ -371,7 +410,7 @@ void __uasminit ISAFUNC(UASM_i_LA_mostly)(u32 **buf, unsigned int rs, long addr) } UASM_EXPORT_SYMBOL(ISAFUNC(UASM_i_LA_mostly)); -void __uasminit ISAFUNC(UASM_i_LA)(u32 **buf, unsigned int rs, long addr) +void ISAFUNC(UASM_i_LA)(u32 **buf, unsigned int rs, long addr) { ISAFUNC(UASM_i_LA_mostly)(buf, rs, addr); if (ISAFUNC(uasm_rel_lo(addr))) { @@ -386,8 +425,7 @@ void __uasminit ISAFUNC(UASM_i_LA)(u32 **buf, unsigned int rs, long addr) UASM_EXPORT_SYMBOL(ISAFUNC(UASM_i_LA)); /* Handle relocations. */ -void __uasminit -ISAFUNC(uasm_r_mips_pc16)(struct uasm_reloc **rel, u32 *addr, int lid) +void ISAFUNC(uasm_r_mips_pc16)(struct uasm_reloc **rel, u32 *addr, int lid) { (*rel)->addr = addr; (*rel)->type = R_MIPS_PC16; @@ -396,11 +434,11 @@ ISAFUNC(uasm_r_mips_pc16)(struct uasm_reloc **rel, u32 *addr, int lid) } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_r_mips_pc16)); -static inline void __uasminit -__resolve_relocs(struct uasm_reloc *rel, struct uasm_label *lab); +static inline void __resolve_relocs(struct uasm_reloc *rel, + struct uasm_label *lab); -void __uasminit -ISAFUNC(uasm_resolve_relocs)(struct uasm_reloc *rel, struct uasm_label *lab) +void ISAFUNC(uasm_resolve_relocs)(struct uasm_reloc *rel, + struct uasm_label *lab) { struct uasm_label *l; @@ -411,8 +449,8 @@ ISAFUNC(uasm_resolve_relocs)(struct uasm_reloc *rel, struct uasm_label *lab) } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_resolve_relocs)); -void __uasminit -ISAFUNC(uasm_move_relocs)(struct uasm_reloc *rel, u32 *first, u32 *end, long off) +void ISAFUNC(uasm_move_relocs)(struct uasm_reloc *rel, u32 *first, u32 *end, + long off) { for (; rel->lab != UASM_LABEL_INVALID; rel++) if (rel->addr >= first && rel->addr < end) @@ -420,8 +458,8 @@ ISAFUNC(uasm_move_relocs)(struct uasm_reloc *rel, u32 *first, u32 *end, long off } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_move_relocs)); -void __uasminit -ISAFUNC(uasm_move_labels)(struct uasm_label *lab, u32 *first, u32 *end, long off) +void ISAFUNC(uasm_move_labels)(struct uasm_label *lab, u32 *first, u32 *end, + long off) { for (; lab->lab != UASM_LABEL_INVALID; lab++) if (lab->addr >= first && lab->addr < end) @@ -429,9 +467,8 @@ ISAFUNC(uasm_move_labels)(struct uasm_label *lab, u32 *first, u32 *end, long off } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_move_labels)); -void __uasminit -ISAFUNC(uasm_copy_handler)(struct uasm_reloc *rel, struct uasm_label *lab, u32 *first, - u32 *end, u32 *target) +void ISAFUNC(uasm_copy_handler)(struct uasm_reloc *rel, struct uasm_label *lab, + u32 *first, u32 *end, u32 *target) { long off = (long)(target - first); @@ -442,7 +479,7 @@ ISAFUNC(uasm_copy_handler)(struct uasm_reloc *rel, struct uasm_label *lab, u32 * } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_copy_handler)); -int __uasminit ISAFUNC(uasm_insn_has_bdelay)(struct uasm_reloc *rel, u32 *addr) +int ISAFUNC(uasm_insn_has_bdelay)(struct uasm_reloc *rel, u32 *addr) { for (; rel->lab != UASM_LABEL_INVALID; rel++) { if (rel->addr == addr @@ -456,83 +493,87 @@ int __uasminit ISAFUNC(uasm_insn_has_bdelay)(struct uasm_reloc *rel, u32 *addr) UASM_EXPORT_SYMBOL(ISAFUNC(uasm_insn_has_bdelay)); /* Convenience functions for labeled branches. */ -void __uasminit -ISAFUNC(uasm_il_bltz)(u32 **p, struct uasm_reloc **r, unsigned int reg, int lid) +void ISAFUNC(uasm_il_bltz)(u32 **p, struct uasm_reloc **r, unsigned int reg, + int lid) { uasm_r_mips_pc16(r, *p, lid); ISAFUNC(uasm_i_bltz)(p, reg, 0); } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_bltz)); -void __uasminit -ISAFUNC(uasm_il_b)(u32 **p, struct uasm_reloc **r, int lid) +void ISAFUNC(uasm_il_b)(u32 **p, struct uasm_reloc **r, int lid) { uasm_r_mips_pc16(r, *p, lid); ISAFUNC(uasm_i_b)(p, 0); } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_b)); -void __uasminit -ISAFUNC(uasm_il_beqz)(u32 **p, struct uasm_reloc **r, unsigned int reg, int lid) +void ISAFUNC(uasm_il_beq)(u32 **p, struct uasm_reloc **r, unsigned int r1, + unsigned int r2, int lid) +{ + uasm_r_mips_pc16(r, *p, lid); + ISAFUNC(uasm_i_beq)(p, r1, r2, 0); +} +UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_beq)); + +void ISAFUNC(uasm_il_beqz)(u32 **p, struct uasm_reloc **r, unsigned int reg, + int lid) { uasm_r_mips_pc16(r, *p, lid); ISAFUNC(uasm_i_beqz)(p, reg, 0); } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_beqz)); -void __uasminit -ISAFUNC(uasm_il_beqzl)(u32 **p, struct uasm_reloc **r, unsigned int reg, int lid) +void ISAFUNC(uasm_il_beqzl)(u32 **p, struct uasm_reloc **r, unsigned int reg, + int lid) { uasm_r_mips_pc16(r, *p, lid); ISAFUNC(uasm_i_beqzl)(p, reg, 0); } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_beqzl)); -void __uasminit -ISAFUNC(uasm_il_bne)(u32 **p, struct uasm_reloc **r, unsigned int reg1, - unsigned int reg2, int lid) +void ISAFUNC(uasm_il_bne)(u32 **p, struct uasm_reloc **r, unsigned int reg1, + unsigned int reg2, int lid) { uasm_r_mips_pc16(r, *p, lid); ISAFUNC(uasm_i_bne)(p, reg1, reg2, 0); } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_bne)); -void __uasminit -ISAFUNC(uasm_il_bnez)(u32 **p, struct uasm_reloc **r, unsigned int reg, int lid) +void ISAFUNC(uasm_il_bnez)(u32 **p, struct uasm_reloc **r, unsigned int reg, + int lid) { uasm_r_mips_pc16(r, *p, lid); ISAFUNC(uasm_i_bnez)(p, reg, 0); } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_bnez)); -void __uasminit -ISAFUNC(uasm_il_bgezl)(u32 **p, struct uasm_reloc **r, unsigned int reg, int lid) +void ISAFUNC(uasm_il_bgezl)(u32 **p, struct uasm_reloc **r, unsigned int reg, + int lid) { uasm_r_mips_pc16(r, *p, lid); ISAFUNC(uasm_i_bgezl)(p, reg, 0); } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_bgezl)); -void __uasminit -ISAFUNC(uasm_il_bgez)(u32 **p, struct uasm_reloc **r, unsigned int reg, int lid) +void ISAFUNC(uasm_il_bgez)(u32 **p, struct uasm_reloc **r, unsigned int reg, + int lid) { uasm_r_mips_pc16(r, *p, lid); ISAFUNC(uasm_i_bgez)(p, reg, 0); } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_bgez)); -void __uasminit -ISAFUNC(uasm_il_bbit0)(u32 **p, struct uasm_reloc **r, unsigned int reg, - unsigned int bit, int lid) +void ISAFUNC(uasm_il_bbit0)(u32 **p, struct uasm_reloc **r, unsigned int reg, + unsigned int bit, int lid) { uasm_r_mips_pc16(r, *p, lid); ISAFUNC(uasm_i_bbit0)(p, reg, bit, 0); } UASM_EXPORT_SYMBOL(ISAFUNC(uasm_il_bbit0)); -void __uasminit -ISAFUNC(uasm_il_bbit1)(u32 **p, struct uasm_reloc **r, unsigned int reg, - unsigned int bit, int lid) +void ISAFUNC(uasm_il_bbit1)(u32 **p, struct uasm_reloc **r, unsigned int reg, + unsigned int bit, int lid) { uasm_r_mips_pc16(r, *p, lid); ISAFUNC(uasm_i_bbit1)(p, reg, bit, 0); |