diff options
Diffstat (limited to 'arch/powerpc/mm/hash_utils_64.c')
| -rw-r--r-- | arch/powerpc/mm/hash_utils_64.c | 924 |
1 files changed, 669 insertions, 255 deletions
diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c index a83dfa3cf40..88fdd9d2507 100644 --- a/arch/powerpc/mm/hash_utils_64.c +++ b/arch/powerpc/mm/hash_utils_64.c @@ -27,10 +27,13 @@ #include <linux/proc_fs.h> #include <linux/stat.h> #include <linux/sysctl.h> +#include <linux/export.h> #include <linux/ctype.h> #include <linux/cache.h> #include <linux/init.h> #include <linux/signal.h> +#include <linux/memblock.h> +#include <linux/context_tracking.h> #include <asm/processor.h> #include <asm/pgtable.h> @@ -38,11 +41,9 @@ #include <asm/mmu_context.h> #include <asm/page.h> #include <asm/types.h> -#include <asm/system.h> #include <asm/uaccess.h> #include <asm/machdep.h> -#include <asm/lmb.h> -#include <asm/abs_addr.h> +#include <asm/prom.h> #include <asm/tlbflush.h> #include <asm/io.h> #include <asm/eeh.h> @@ -52,6 +53,10 @@ #include <asm/sections.h> #include <asm/spu.h> #include <asm/udbg.h> +#include <asm/code-patching.h> +#include <asm/fadump.h> +#include <asm/firmware.h> +#include <asm/tm.h> #ifdef DEBUG #define DBG(fmt...) udbg_printf(fmt) @@ -67,6 +72,7 @@ #define KB (1024) #define MB (1024*KB) +#define GB (1024L*MB) /* * Note: pte --> Linux PTE @@ -90,17 +96,18 @@ struct mmu_psize_def mmu_psize_defs[MMU_PAGE_COUNT]; struct hash_pte *htab_address; unsigned long htab_size_bytes; unsigned long htab_hash_mask; +EXPORT_SYMBOL_GPL(htab_hash_mask); int mmu_linear_psize = MMU_PAGE_4K; int mmu_virtual_psize = MMU_PAGE_4K; int mmu_vmalloc_psize = MMU_PAGE_4K; +#ifdef CONFIG_SPARSEMEM_VMEMMAP +int mmu_vmemmap_psize = MMU_PAGE_4K; +#endif int mmu_io_psize = MMU_PAGE_4K; int mmu_kernel_ssize = MMU_SEGSIZE_256M; int mmu_highuser_ssize = MMU_SEGSIZE_256M; u16 mmu_slb_size = 64; -#ifdef CONFIG_HUGETLB_PAGE -int mmu_huge_psize = MMU_PAGE_16M; -unsigned int HPAGE_SHIFT; -#endif +EXPORT_SYMBOL_GPL(mmu_slb_size); #ifdef CONFIG_PPC_64K_PAGES int mmu_ci_restrictions; #endif @@ -116,11 +123,11 @@ static DEFINE_SPINLOCK(linear_map_hash_lock); /* Pre-POWER4 CPUs (4k pages only) */ -struct mmu_psize_def mmu_psize_defaults_old[] = { +static struct mmu_psize_def mmu_psize_defaults_old[] = { [MMU_PAGE_4K] = { .shift = 12, .sllp = 0, - .penc = 0, + .penc = {[MMU_PAGE_4K] = 0, [1 ... MMU_PAGE_COUNT - 1] = -1}, .avpnm = 0, .tlbiel = 0, }, @@ -130,56 +137,100 @@ struct mmu_psize_def mmu_psize_defaults_old[] = { * * Support for 16Mb large pages */ -struct mmu_psize_def mmu_psize_defaults_gp[] = { +static struct mmu_psize_def mmu_psize_defaults_gp[] = { [MMU_PAGE_4K] = { .shift = 12, .sllp = 0, - .penc = 0, + .penc = {[MMU_PAGE_4K] = 0, [1 ... MMU_PAGE_COUNT - 1] = -1}, .avpnm = 0, .tlbiel = 1, }, [MMU_PAGE_16M] = { .shift = 24, .sllp = SLB_VSID_L, - .penc = 0, + .penc = {[0 ... MMU_PAGE_16M - 1] = -1, [MMU_PAGE_16M] = 0, + [MMU_PAGE_16M + 1 ... MMU_PAGE_COUNT - 1] = -1 }, .avpnm = 0x1UL, .tlbiel = 0, }, }; +static unsigned long htab_convert_pte_flags(unsigned long pteflags) +{ + unsigned long rflags = pteflags & 0x1fa; + + /* _PAGE_EXEC -> NOEXEC */ + if ((pteflags & _PAGE_EXEC) == 0) + rflags |= HPTE_R_N; + + /* PP bits. PAGE_USER is already PP bit 0x2, so we only + * need to add in 0x1 if it's a read-only user page + */ + if ((pteflags & _PAGE_USER) && !((pteflags & _PAGE_RW) && + (pteflags & _PAGE_DIRTY))) + rflags |= 1; + /* + * Always add "C" bit for perf. Memory coherence is always enabled + */ + return rflags | HPTE_R_C | HPTE_R_M; +} int htab_bolt_mapping(unsigned long vstart, unsigned long vend, - unsigned long pstart, unsigned long mode, + unsigned long pstart, unsigned long prot, int psize, int ssize) { unsigned long vaddr, paddr; unsigned int step, shift; - unsigned long tmp_mode; int ret = 0; shift = mmu_psize_defs[psize].shift; step = 1 << shift; + prot = htab_convert_pte_flags(prot); + + DBG("htab_bolt_mapping(%lx..%lx -> %lx (%lx,%d,%d)\n", + vstart, vend, pstart, prot, psize, ssize); + for (vaddr = vstart, paddr = pstart; vaddr < vend; vaddr += step, paddr += step) { unsigned long hash, hpteg; unsigned long vsid = get_kernel_vsid(vaddr, ssize); - unsigned long va = hpt_va(vaddr, vsid, ssize); + unsigned long vpn = hpt_vpn(vaddr, vsid, ssize); + unsigned long tprot = prot; - tmp_mode = mode; - - /* Make non-kernel text non-executable */ - if (!in_kernel_text(vaddr)) - tmp_mode = mode | HPTE_R_N; + /* + * If we hit a bad address return error. + */ + if (!vsid) + return -1; + /* Make kernel text executable */ + if (overlaps_kernel_text(vaddr, vaddr + step)) + tprot &= ~HPTE_R_N; + + /* Make kvm guest trampolines executable */ + if (overlaps_kvm_tmp(vaddr, vaddr + step)) + tprot &= ~HPTE_R_N; + + /* + * If relocatable, check if it overlaps interrupt vectors that + * are copied down to real 0. For relocatable kernel + * (e.g. kdump case) we copy interrupt vectors down to real + * address 0. Mark that region as executable. This is + * because on p8 system with relocation on exception feature + * enabled, exceptions are raised with MMU (IR=DR=1) ON. Hence + * in order to execute the interrupt handlers in virtual + * mode the vector region need to be marked as executable. + */ + if ((PHYSICAL_START > MEMORY_START) && + overlaps_interrupt_vector_text(vaddr, vaddr + step)) + tprot &= ~HPTE_R_N; - hash = hpt_hash(va, shift, ssize); + hash = hpt_hash(vpn, shift, ssize); hpteg = ((hash & htab_hash_mask) * HPTES_PER_GROUP); - DBG("htab_bolt_mapping: calling %p\n", ppc_md.hpte_insert); - BUG_ON(!ppc_md.hpte_insert); - ret = ppc_md.hpte_insert(hpteg, va, paddr, - tmp_mode, HPTE_V_BOLTED, psize, ssize); + ret = ppc_md.hpte_insert(hpteg, vpn, paddr, tprot, + HPTE_V_BOLTED, psize, psize, ssize); if (ret < 0) break; @@ -191,30 +242,52 @@ int htab_bolt_mapping(unsigned long vstart, unsigned long vend, return ret < 0 ? ret : 0; } +#ifdef CONFIG_MEMORY_HOTPLUG +static int htab_remove_mapping(unsigned long vstart, unsigned long vend, + int psize, int ssize) +{ + unsigned long vaddr; + unsigned int step, shift; + + shift = mmu_psize_defs[psize].shift; + step = 1 << shift; + + if (!ppc_md.hpte_removebolted) { + printk(KERN_WARNING "Platform doesn't implement " + "hpte_removebolted\n"); + return -EINVAL; + } + + for (vaddr = vstart; vaddr < vend; vaddr += step) + ppc_md.hpte_removebolted(vaddr, psize, ssize); + + return 0; +} +#endif /* CONFIG_MEMORY_HOTPLUG */ + static int __init htab_dt_scan_seg_sizes(unsigned long node, const char *uname, int depth, void *data) { - char *type = of_get_flat_dt_prop(node, "device_type", NULL); - u32 *prop; - unsigned long size = 0; + const char *type = of_get_flat_dt_prop(node, "device_type", NULL); + const __be32 *prop; + int size = 0; /* We are scanning "cpu" nodes only */ if (type == NULL || strcmp(type, "cpu") != 0) return 0; - prop = (u32 *)of_get_flat_dt_prop(node, "ibm,processor-segment-sizes", - &size); + prop = of_get_flat_dt_prop(node, "ibm,processor-segment-sizes", &size); if (prop == NULL) return 0; for (; size >= 4; size -= 4, ++prop) { - if (prop[0] == 40) { + if (be32_to_cpu(prop[0]) == 40) { DBG("1T segment support detected\n"); - cur_cpu_spec->cpu_features |= CPU_FTR_1T_SEGMENT; + cur_cpu_spec->mmu_features |= MMU_FTR_1T_SEGMENT; return 1; } } - cur_cpu_spec->cpu_features &= ~CPU_FTR_NO_SLBIE_B; + cur_cpu_spec->mmu_features &= ~MMU_FTR_NO_SLBIE_B; return 0; } @@ -223,89 +296,184 @@ static void __init htab_init_seg_sizes(void) of_scan_flat_dt(htab_dt_scan_seg_sizes, NULL); } +static int __init get_idx_from_shift(unsigned int shift) +{ + int idx = -1; + + switch (shift) { + case 0xc: + idx = MMU_PAGE_4K; + break; + case 0x10: + idx = MMU_PAGE_64K; + break; + case 0x14: + idx = MMU_PAGE_1M; + break; + case 0x18: + idx = MMU_PAGE_16M; + break; + case 0x22: + idx = MMU_PAGE_16G; + break; + } + return idx; +} + static int __init htab_dt_scan_page_sizes(unsigned long node, const char *uname, int depth, void *data) { - char *type = of_get_flat_dt_prop(node, "device_type", NULL); - u32 *prop; - unsigned long size = 0; + const char *type = of_get_flat_dt_prop(node, "device_type", NULL); + const __be32 *prop; + int size = 0; /* We are scanning "cpu" nodes only */ if (type == NULL || strcmp(type, "cpu") != 0) return 0; - prop = (u32 *)of_get_flat_dt_prop(node, - "ibm,segment-page-sizes", &size); + prop = of_get_flat_dt_prop(node, "ibm,segment-page-sizes", &size); if (prop != NULL) { - DBG("Page sizes from device-tree:\n"); + pr_info("Page sizes from device-tree:\n"); size /= 4; - cur_cpu_spec->cpu_features &= ~(CPU_FTR_16M_PAGE); + cur_cpu_spec->mmu_features &= ~(MMU_FTR_16M_PAGE); while(size > 0) { - unsigned int shift = prop[0]; - unsigned int slbenc = prop[1]; - unsigned int lpnum = prop[2]; - unsigned int lpenc = 0; + unsigned int base_shift = be32_to_cpu(prop[0]); + unsigned int slbenc = be32_to_cpu(prop[1]); + unsigned int lpnum = be32_to_cpu(prop[2]); struct mmu_psize_def *def; - int idx = -1; + int idx, base_idx; size -= 3; prop += 3; - while(size > 0 && lpnum) { - if (prop[0] == shift) - lpenc = prop[1]; - prop += 2; size -= 2; - lpnum--; - } - switch(shift) { - case 0xc: - idx = MMU_PAGE_4K; - break; - case 0x10: - idx = MMU_PAGE_64K; - break; - case 0x14: - idx = MMU_PAGE_1M; - break; - case 0x18: - idx = MMU_PAGE_16M; - cur_cpu_spec->cpu_features |= CPU_FTR_16M_PAGE; - break; - case 0x22: - idx = MMU_PAGE_16G; - break; - } - if (idx < 0) + base_idx = get_idx_from_shift(base_shift); + if (base_idx < 0) { + /* + * skip the pte encoding also + */ + prop += lpnum * 2; size -= lpnum * 2; continue; - def = &mmu_psize_defs[idx]; - def->shift = shift; - if (shift <= 23) + } + def = &mmu_psize_defs[base_idx]; + if (base_idx == MMU_PAGE_16M) + cur_cpu_spec->mmu_features |= MMU_FTR_16M_PAGE; + + def->shift = base_shift; + if (base_shift <= 23) def->avpnm = 0; else - def->avpnm = (1 << (shift - 23)) - 1; + def->avpnm = (1 << (base_shift - 23)) - 1; def->sllp = slbenc; - def->penc = lpenc; - /* We don't know for sure what's up with tlbiel, so + /* + * We don't know for sure what's up with tlbiel, so * for now we only set it for 4K and 64K pages */ - if (idx == MMU_PAGE_4K || idx == MMU_PAGE_64K) + if (base_idx == MMU_PAGE_4K || base_idx == MMU_PAGE_64K) def->tlbiel = 1; else def->tlbiel = 0; - DBG(" %d: shift=%02x, sllp=%04x, avpnm=%08x, " - "tlbiel=%d, penc=%d\n", - idx, shift, def->sllp, def->avpnm, def->tlbiel, - def->penc); + while (size > 0 && lpnum) { + unsigned int shift = be32_to_cpu(prop[0]); + int penc = be32_to_cpu(prop[1]); + + prop += 2; size -= 2; + lpnum--; + + idx = get_idx_from_shift(shift); + if (idx < 0) + continue; + + if (penc == -1) + pr_err("Invalid penc for base_shift=%d " + "shift=%d\n", base_shift, shift); + + def->penc[idx] = penc; + pr_info("base_shift=%d: shift=%d, sllp=0x%04lx," + " avpnm=0x%08lx, tlbiel=%d, penc=%d\n", + base_shift, shift, def->sllp, + def->avpnm, def->tlbiel, def->penc[idx]); + } } return 1; } return 0; } +#ifdef CONFIG_HUGETLB_PAGE +/* Scan for 16G memory blocks that have been set aside for huge pages + * and reserve those blocks for 16G huge pages. + */ +static int __init htab_dt_scan_hugepage_blocks(unsigned long node, + const char *uname, int depth, + void *data) { + const char *type = of_get_flat_dt_prop(node, "device_type", NULL); + const __be64 *addr_prop; + const __be32 *page_count_prop; + unsigned int expected_pages; + long unsigned int phys_addr; + long unsigned int block_size; + + /* We are scanning "memory" nodes only */ + if (type == NULL || strcmp(type, "memory") != 0) + return 0; + + /* This property is the log base 2 of the number of virtual pages that + * will represent this memory block. */ + page_count_prop = of_get_flat_dt_prop(node, "ibm,expected#pages", NULL); + if (page_count_prop == NULL) + return 0; + expected_pages = (1 << be32_to_cpu(page_count_prop[0])); + addr_prop = of_get_flat_dt_prop(node, "reg", NULL); + if (addr_prop == NULL) + return 0; + phys_addr = be64_to_cpu(addr_prop[0]); + block_size = be64_to_cpu(addr_prop[1]); + if (block_size != (16 * GB)) + return 0; + printk(KERN_INFO "Huge page(16GB) memory: " + "addr = 0x%lX size = 0x%lX pages = %d\n", + phys_addr, block_size, expected_pages); + if (phys_addr + (16 * GB) <= memblock_end_of_DRAM()) { + memblock_reserve(phys_addr, block_size * expected_pages); + add_gpage(phys_addr, block_size, expected_pages); + } + return 0; +} +#endif /* CONFIG_HUGETLB_PAGE */ + +static void mmu_psize_set_default_penc(void) +{ + int bpsize, apsize; + for (bpsize = 0; bpsize < MMU_PAGE_COUNT; bpsize++) + for (apsize = 0; apsize < MMU_PAGE_COUNT; apsize++) + mmu_psize_defs[bpsize].penc[apsize] = -1; +} + +#ifdef CONFIG_PPC_64K_PAGES + +static bool might_have_hea(void) +{ + /* + * The HEA ethernet adapter requires awareness of the + * GX bus. Without that awareness we can easily assume + * we will never see an HEA ethernet device. + */ +#ifdef CONFIG_IBMEBUS + return !cpu_has_feature(CPU_FTR_ARCH_207S); +#else + return false; +#endif +} + +#endif /* #ifdef CONFIG_PPC_64K_PAGES */ + static void __init htab_init_page_sizes(void) { int rc; + /* se the invalid penc to -1 */ + mmu_psize_set_default_penc(); + /* Default to 4K pages only */ memcpy(mmu_psize_defs, mmu_psize_defaults_old, sizeof(mmu_psize_defaults_old)); @@ -321,7 +489,7 @@ static void __init htab_init_page_sizes(void) * Not in the device-tree, let's fallback on known size * list for 16M capable GP & GR */ - if (cpu_has_feature(CPU_FTR_16M_PAGE)) + if (mmu_has_feature(MMU_FTR_16M_PAGE)) memcpy(mmu_psize_defs, mmu_psize_defaults_gp, sizeof(mmu_psize_defaults_gp)); found: @@ -351,34 +519,49 @@ static void __init htab_init_page_sizes(void) mmu_vmalloc_psize = MMU_PAGE_64K; if (mmu_linear_psize == MMU_PAGE_4K) mmu_linear_psize = MMU_PAGE_64K; - if (cpu_has_feature(CPU_FTR_CI_LARGE_PAGE)) { + if (mmu_has_feature(MMU_FTR_CI_LARGE_PAGE)) { /* - * Don't use 64k pages for ioremap on pSeries, since - * that would stop us accessing the HEA ethernet. + * When running on pSeries using 64k pages for ioremap + * would stop us accessing the HEA ethernet. So if we + * have the chance of ever seeing one, stay at 4k. */ - if (!machine_is(pseries)) + if (!might_have_hea() || !machine_is(pseries)) mmu_io_psize = MMU_PAGE_64K; } else mmu_ci_restrictions = 1; } #endif /* CONFIG_PPC_64K_PAGES */ +#ifdef CONFIG_SPARSEMEM_VMEMMAP + /* We try to use 16M pages for vmemmap if that is supported + * and we have at least 1G of RAM at boot + */ + if (mmu_psize_defs[MMU_PAGE_16M].shift && + memblock_phys_mem_size() >= 0x40000000) + mmu_vmemmap_psize = MMU_PAGE_16M; + else if (mmu_psize_defs[MMU_PAGE_64K].shift) + mmu_vmemmap_psize = MMU_PAGE_64K; + else + mmu_vmemmap_psize = MMU_PAGE_4K; +#endif /* CONFIG_SPARSEMEM_VMEMMAP */ + printk(KERN_DEBUG "Page orders: linear mapping = %d, " - "virtual = %d, io = %d\n", + "virtual = %d, io = %d" +#ifdef CONFIG_SPARSEMEM_VMEMMAP + ", vmemmap = %d" +#endif + "\n", mmu_psize_defs[mmu_linear_psize].shift, mmu_psize_defs[mmu_virtual_psize].shift, - mmu_psize_defs[mmu_io_psize].shift); + mmu_psize_defs[mmu_io_psize].shift +#ifdef CONFIG_SPARSEMEM_VMEMMAP + ,mmu_psize_defs[mmu_vmemmap_psize].shift +#endif + ); #ifdef CONFIG_HUGETLB_PAGE - /* Init large page size. Currently, we pick 16M or 1M depending - * on what is available - */ - if (mmu_psize_defs[MMU_PAGE_16M].shift) - set_huge_psize(MMU_PAGE_16M); - /* With 4k/4level pagetables, we can't (for now) cope with a - * huge page size < PMD_SIZE */ - else if (mmu_psize_defs[MMU_PAGE_1M].shift) - set_huge_psize(MMU_PAGE_1M); + /* Reserve 16G huge page memory sections for huge pages */ + of_scan_flat_dt(htab_dt_scan_hugepage_blocks, NULL); #endif /* CONFIG_HUGETLB_PAGE */ } @@ -386,17 +569,17 @@ static int __init htab_dt_scan_pftsize(unsigned long node, const char *uname, int depth, void *data) { - char *type = of_get_flat_dt_prop(node, "device_type", NULL); - u32 *prop; + const char *type = of_get_flat_dt_prop(node, "device_type", NULL); + const __be32 *prop; /* We are scanning "cpu" nodes only */ if (type == NULL || strcmp(type, "cpu") != 0) return 0; - prop = (u32 *)of_get_flat_dt_prop(node, "ibm,pft-size", NULL); + prop = of_get_flat_dt_prop(node, "ibm,pft-size", NULL); if (prop != NULL) { /* pft_size[0] is the NUMA CEC cookie */ - ppc64_pft_size = prop[1]; + ppc64_pft_size = be32_to_cpu(prop[1]); return 1; } return 0; @@ -404,7 +587,7 @@ static int __init htab_dt_scan_pftsize(unsigned long node, static unsigned long __init htab_get_table_size(void) { - unsigned long mem_size, rnd_mem_size, pteg_count; + unsigned long mem_size, rnd_mem_size, pteg_count, psize; /* If hash size isn't already provided by the platform, we try to * retrieve it from the device-tree. If it's not there neither, we @@ -416,70 +599,80 @@ static unsigned long __init htab_get_table_size(void) return 1UL << ppc64_pft_size; /* round mem_size up to next power of 2 */ - mem_size = lmb_phys_mem_size(); + mem_size = memblock_phys_mem_size(); rnd_mem_size = 1UL << __ilog2(mem_size); if (rnd_mem_size < mem_size) rnd_mem_size <<= 1; /* # pages / 2 */ - pteg_count = max(rnd_mem_size >> (12 + 1), 1UL << 11); + psize = mmu_psize_defs[mmu_virtual_psize].shift; + pteg_count = max(rnd_mem_size >> (psize + 1), 1UL << 11); return pteg_count << 7; } #ifdef CONFIG_MEMORY_HOTPLUG -void create_section_mapping(unsigned long start, unsigned long end) +int create_section_mapping(unsigned long start, unsigned long end) { - BUG_ON(htab_bolt_mapping(start, end, __pa(start), - _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_COHERENT | PP_RWXX, - mmu_linear_psize, mmu_kernel_ssize)); + return htab_bolt_mapping(start, end, __pa(start), + pgprot_val(PAGE_KERNEL), mmu_linear_psize, + mmu_kernel_ssize); } -#endif /* CONFIG_MEMORY_HOTPLUG */ -static inline void make_bl(unsigned int *insn_addr, void *func) +int remove_section_mapping(unsigned long start, unsigned long end) { - unsigned long funcp = *((unsigned long *)func); - int offset = funcp - (unsigned long)insn_addr; - - *insn_addr = (unsigned int)(0x48000001 | (offset & 0x03fffffc)); - flush_icache_range((unsigned long)insn_addr, 4+ - (unsigned long)insn_addr); + return htab_remove_mapping(start, end, mmu_linear_psize, + mmu_kernel_ssize); } +#endif /* CONFIG_MEMORY_HOTPLUG */ + +extern u32 htab_call_hpte_insert1[]; +extern u32 htab_call_hpte_insert2[]; +extern u32 htab_call_hpte_remove[]; +extern u32 htab_call_hpte_updatepp[]; +extern u32 ht64_call_hpte_insert1[]; +extern u32 ht64_call_hpte_insert2[]; +extern u32 ht64_call_hpte_remove[]; +extern u32 ht64_call_hpte_updatepp[]; static void __init htab_finish_init(void) { - extern unsigned int *htab_call_hpte_insert1; - extern unsigned int *htab_call_hpte_insert2; - extern unsigned int *htab_call_hpte_remove; - extern unsigned int *htab_call_hpte_updatepp; - #ifdef CONFIG_PPC_HAS_HASH_64K - extern unsigned int *ht64_call_hpte_insert1; - extern unsigned int *ht64_call_hpte_insert2; - extern unsigned int *ht64_call_hpte_remove; - extern unsigned int *ht64_call_hpte_updatepp; - - make_bl(ht64_call_hpte_insert1, ppc_md.hpte_insert); - make_bl(ht64_call_hpte_insert2, ppc_md.hpte_insert); - make_bl(ht64_call_hpte_remove, ppc_md.hpte_remove); - make_bl(ht64_call_hpte_updatepp, ppc_md.hpte_updatepp); + patch_branch(ht64_call_hpte_insert1, + ppc_function_entry(ppc_md.hpte_insert), + BRANCH_SET_LINK); + patch_branch(ht64_call_hpte_insert2, + ppc_function_entry(ppc_md.hpte_insert), + BRANCH_SET_LINK); + patch_branch(ht64_call_hpte_remove, + ppc_function_entry(ppc_md.hpte_remove), + BRANCH_SET_LINK); + patch_branch(ht64_call_hpte_updatepp, + ppc_function_entry(ppc_md.hpte_updatepp), + BRANCH_SET_LINK); #endif /* CONFIG_PPC_HAS_HASH_64K */ - make_bl(htab_call_hpte_insert1, ppc_md.hpte_insert); - make_bl(htab_call_hpte_insert2, ppc_md.hpte_insert); - make_bl(htab_call_hpte_remove, ppc_md.hpte_remove); - make_bl(htab_call_hpte_updatepp, ppc_md.hpte_updatepp); + patch_branch(htab_call_hpte_insert1, + ppc_function_entry(ppc_md.hpte_insert), + BRANCH_SET_LINK); + patch_branch(htab_call_hpte_insert2, + ppc_function_entry(ppc_md.hpte_insert), + BRANCH_SET_LINK); + patch_branch(htab_call_hpte_remove, + ppc_function_entry(ppc_md.hpte_remove), + BRANCH_SET_LINK); + patch_branch(htab_call_hpte_updatepp, + ppc_function_entry(ppc_md.hpte_updatepp), + BRANCH_SET_LINK); } -void __init htab_initialize(void) +static void __init htab_initialize(void) { unsigned long table; unsigned long pteg_count; - unsigned long mode_rw; + unsigned long prot; unsigned long base = 0, size = 0, limit; - int i; - - extern unsigned long tce_alloc_start, tce_alloc_end; + struct memblock_region *reg; DBG(" -> htab_initialize()\n"); @@ -489,7 +682,7 @@ void __init htab_initialize(void) /* Initialize page sizes */ htab_init_page_sizes(); - if (cpu_has_feature(CPU_FTR_1T_SEGMENT)) { + if (mmu_has_feature(MMU_FTR_1T_SEGMENT)) { mmu_kernel_ssize = MMU_SEGSIZE_1T; mmu_highuser_ssize = MMU_SEGSIZE_1T; printk(KERN_INFO "Using 1TB segments\n"); @@ -508,6 +701,16 @@ void __init htab_initialize(void) /* Using a hypervisor which owns the htab */ htab_address = NULL; _SDR1 = 0; +#ifdef CONFIG_FA_DUMP + /* + * If firmware assisted dump is active firmware preserves + * the contents of htab along with entire partition memory. + * Clear the htab if firmware assisted dump is active so + * that we dont end up using old mappings. + */ + if (is_fadump_active() && ppc_md.hpte_clear_all) + ppc_md.hpte_clear_all(); +#endif } else { /* Find storage for the HPT. Must be contiguous in * the absolute address space. On cell we want it to be @@ -516,14 +719,14 @@ void __init htab_initialize(void) if (machine_is(cell)) limit = 0x80000000; else - limit = 0; + limit = MEMBLOCK_ALLOC_ANYWHERE; - table = lmb_alloc_base(htab_size_bytes, htab_size_bytes, limit); + table = memblock_alloc_base(htab_size_bytes, htab_size_bytes, limit); DBG("Hash table allocated at %lx, size: %lx\n", table, htab_size_bytes); - htab_address = abs_to_virt(table); + htab_address = __va(table); /* htab absolute addr + encoded htabsize */ _SDR1 = table + __ilog2(pteg_count) - 11; @@ -535,12 +738,12 @@ void __init htab_initialize(void) mtspr(SPRN_SDR1, _SDR1); } - mode_rw = _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_COHERENT | PP_RWXX; + prot = pgprot_val(PAGE_KERNEL); #ifdef CONFIG_DEBUG_PAGEALLOC - linear_map_hash_count = lmb_end_of_DRAM() >> PAGE_SHIFT; - linear_map_hash_slots = __va(lmb_alloc_base(linear_map_hash_count, - 1, lmb.rmo_size)); + linear_map_hash_count = memblock_end_of_DRAM() >> PAGE_SHIFT; + linear_map_hash_slots = __va(memblock_alloc_base(linear_map_hash_count, + 1, ppc64_rma_size)); memset(linear_map_hash_slots, 0, linear_map_hash_count); #endif /* CONFIG_DEBUG_PAGEALLOC */ @@ -550,15 +753,16 @@ void __init htab_initialize(void) */ /* create bolted the linear mapping in the hash table */ - for (i=0; i < lmb.memory.cnt; i++) { - base = (unsigned long)__va(lmb.memory.region[i].base); - size = lmb.memory.region[i].size; + for_each_memblock(memory, reg) { + base = (unsigned long)__va(reg->base); + size = reg->size; - DBG("creating mapping for region: %lx : %lx\n", base, size); + DBG("creating mapping for region: %lx..%lx (prot: %lx)\n", + base, size, prot); #ifdef CONFIG_U3_DART /* Do not map the DART space. Fortunately, it will be aligned - * in such a way that it will not cross two lmb regions and + * in such a way that it will not cross two memblock regions and * will fit within a single 16Mb page. * The DART space is assumed to be a full 16Mb region even if * we only use 2Mb of that space. We will use more of it later @@ -571,22 +775,23 @@ void __init htab_initialize(void) unsigned long dart_table_end = dart_tablebase + 16 * MB; if (base != dart_tablebase) BUG_ON(htab_bolt_mapping(base, dart_tablebase, - __pa(base), mode_rw, + __pa(base), prot, mmu_linear_psize, mmu_kernel_ssize)); if ((base + size) > dart_table_end) BUG_ON(htab_bolt_mapping(dart_tablebase+16*MB, base + size, __pa(dart_table_end), - mode_rw, + prot, mmu_linear_psize, mmu_kernel_ssize)); continue; } #endif /* CONFIG_U3_DART */ BUG_ON(htab_bolt_mapping(base, base + size, __pa(base), - mode_rw, mmu_linear_psize, mmu_kernel_ssize)); - } + prot, mmu_linear_psize, mmu_kernel_ssize)); + } + memblock_set_current_limit(MEMBLOCK_ALLOC_ANYWHERE); /* * If we have a memory_limit and we've allocated TCEs then we need to @@ -603,7 +808,7 @@ void __init htab_initialize(void) tce_alloc_start = base + size + 1; BUG_ON(htab_bolt_mapping(tce_alloc_start, tce_alloc_end, - __pa(tce_alloc_start), mode_rw, + __pa(tce_alloc_start), prot, mmu_linear_psize, mmu_kernel_ssize)); } @@ -614,11 +819,41 @@ void __init htab_initialize(void) #undef KB #undef MB -void htab_initialize_secondary(void) +void __init early_init_mmu(void) { + /* Setup initial STAB address in the PACA */ + get_paca()->stab_real = __pa((u64)&initial_stab); + get_paca()->stab_addr = (u64)&initial_stab; + + /* Initialize the MMU Hash table and create the linear mapping + * of memory. Has to be done before stab/slb initialization as + * this is currently where the page size encoding is obtained + */ + htab_initialize(); + + /* Initialize stab / SLB management */ + if (mmu_has_feature(MMU_FTR_SLB)) + slb_initialize(); + else + stab_initialize(get_paca()->stab_real); +} + +#ifdef CONFIG_SMP +void early_init_mmu_secondary(void) +{ + /* Initialize hash table for that CPU */ if (!firmware_has_feature(FW_FEATURE_LPAR)) mtspr(SPRN_SDR1, _SDR1); + + /* Initialize STAB/SLB. We use a virtual address as it works + * in real mode on pSeries. + */ + if (mmu_has_feature(MMU_FTR_SLB)) + slb_initialize(); + else + stab_initialize(get_paca()->stab_addr); } +#endif /* CONFIG_SMP */ /* * Called by asm hashtable.S for doing lazy icache flush @@ -635,7 +870,7 @@ unsigned int hash_page_do_lazy_icache(unsigned int pp, pte_t pte, int trap) /* page is dirty */ if (!test_bit(PG_arch_1, &page->flags) && !PageReserved(page)) { if (trap == 0x400) { - __flush_dcache_icache(page_address(page)); + flush_dcache_icache_page(page); set_bit(PG_arch_1, &page->flags); } else pp |= HPTE_R_N; @@ -643,6 +878,31 @@ unsigned int hash_page_do_lazy_icache(unsigned int pp, pte_t pte, int trap) return pp; } +#ifdef CONFIG_PPC_MM_SLICES +unsigned int get_paca_psize(unsigned long addr) +{ + u64 lpsizes; + unsigned char *hpsizes; + unsigned long index, mask_index; + + if (addr < SLICE_LOW_TOP) { + lpsizes = get_paca()->context.low_slices_psize; + index = GET_LOW_SLICE_INDEX(addr); + return (lpsizes >> (index * 4)) & 0xF; + } + hpsizes = get_paca()->context.high_slices_psize; + index = GET_HIGH_SLICE_INDEX(addr); + mask_index = index & 0x1; + return (hpsizes[index >> 1] >> (mask_index * 4)) & 0xF; +} + +#else +unsigned int get_paca_psize(unsigned long addr) +{ + return get_paca()->context.user_psize; +} +#endif + /* * Demote a segment to using 4k pages. * For now this makes the whole process use 4k pages. @@ -650,13 +910,13 @@ unsigned int hash_page_do_lazy_icache(unsigned int pp, pte_t pte, int trap) #ifdef CONFIG_PPC_64K_PAGES void demote_segment_4k(struct mm_struct *mm, unsigned long addr) { - if (mm->context.user_psize == MMU_PAGE_4K) + if (get_slice_psize(mm, addr) == MMU_PAGE_4K) return; - slice_set_user_psize(mm, MMU_PAGE_4K); + slice_set_range_psize(mm, addr, 1, MMU_PAGE_4K); #ifdef CONFIG_SPU_BASE spu_flush_all_slbs(mm); #endif - if (get_paca()->context.user_psize != MMU_PAGE_4K) { + if (get_paca_psize(addr) != MMU_PAGE_4K) { get_paca()->context = mm->context; slb_flush_and_rebolt(); } @@ -671,15 +931,15 @@ void demote_segment_4k(struct mm_struct *mm, unsigned long addr) * Result is 0: full permissions, _PAGE_RW: read-only, * _PAGE_USER or _PAGE_USER|_PAGE_RW: no access. */ -static int subpage_protection(pgd_t *pgdir, unsigned long ea) +static int subpage_protection(struct mm_struct *mm, unsigned long ea) { - struct subpage_prot_table *spt = pgd_subpage_prot(pgdir); + struct subpage_prot_table *spt = &mm->context.spt; u32 spp = 0; u32 **sbpm, *sbpp; if (ea >= spt->maxaddr) return 0; - if (ea < 0x100000000) { + if (ea < 0x100000000UL) { /* addresses below 4GB use spt->low_prot */ sbpm = spt->low_prot; } else { @@ -701,12 +961,40 @@ static int subpage_protection(pgd_t *pgdir, unsigned long ea) } #else /* CONFIG_PPC_SUBPAGE_PROT */ -static inline int subpage_protection(pgd_t *pgdir, unsigned long ea) +static inline int subpage_protection(struct mm_struct *mm, unsigned long ea) { return 0; } #endif +void hash_failure_debug(unsigned long ea, unsigned long access, + unsigned long vsid, unsigned long trap, + int ssize, int psize, int lpsize, unsigned long pte) +{ + if (!printk_ratelimit()) + return; + pr_info("mm: Hashing failure ! EA=0x%lx access=0x%lx current=%s\n", + ea, access, current->comm); + pr_info(" trap=0x%lx vsid=0x%lx ssize=%d base psize=%d psize %d pte=0x%lx\n", + trap, vsid, ssize, psize, lpsize, pte); +} + +static void check_paca_psize(unsigned long ea, struct mm_struct *mm, + int psize, bool user_region) +{ + if (user_region) { + if (psize != get_paca_psize(ea)) { + get_paca()->context = mm->context; + slb_flush_and_rebolt(); + } + } else if (get_paca()->vmalloc_sllp != + mmu_psize_defs[mmu_vmalloc_psize].sllp) { + get_paca()->vmalloc_sllp = + mmu_psize_defs[mmu_vmalloc_psize].sllp; + slb_vmalloc_update(); + } +} + /* Result code is: * 0 - handled * 1 - normal page fault @@ -715,22 +1003,19 @@ static inline int subpage_protection(pgd_t *pgdir, unsigned long ea) */ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) { - void *pgdir; + enum ctx_state prev_state = exception_enter(); + pgd_t *pgdir; unsigned long vsid; struct mm_struct *mm; pte_t *ptep; - cpumask_t tmp; + unsigned hugeshift; + const struct cpumask *tmp; int rc, user_region = 0, local = 0; int psize, ssize; DBG_LOW("hash_page(ea=%016lx, access=%lx, trap=%lx\n", ea, access, trap); - if ((ea & ~REGION_MASK) >= PGTABLE_RANGE) { - DBG_LOW(" out of pgtable range !\n"); - return 1; - } - /* Get region & vsid */ switch (REGION_ID(ea)) { case USER_REGION_ID: @@ -738,13 +1023,10 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) mm = current->mm; if (! mm) { DBG_LOW(" user region with no mm !\n"); - return 1; + rc = 1; + goto bail; } -#ifdef CONFIG_PPC_MM_SLICES psize = get_slice_psize(mm, ea); -#else - psize = mm->context.user_psize; -#endif ssize = user_segment_size(ea); vsid = get_vsid(mm->context.id, ea, ssize); break; @@ -761,62 +1043,94 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) /* Not a valid range * Send the problem up to do_page_fault */ - return 1; + rc = 1; + goto bail; } DBG_LOW(" mm=%p, mm->pgdir=%p, vsid=%016lx\n", mm, mm->pgd, vsid); + /* Bad address. */ + if (!vsid) { + DBG_LOW("Bad address!\n"); + rc = 1; + goto bail; + } /* Get pgdir */ pgdir = mm->pgd; - if (pgdir == NULL) - return 1; + if (pgdir == NULL) { + rc = 1; + goto bail; + } /* Check CPU locality */ - tmp = cpumask_of_cpu(smp_processor_id()); - if (user_region && cpus_equal(mm->cpu_vm_mask, tmp)) + tmp = cpumask_of(smp_processor_id()); + if (user_region && cpumask_equal(mm_cpumask(mm), tmp)) local = 1; -#ifdef CONFIG_HUGETLB_PAGE - /* Handle hugepage regions */ - if (HPAGE_SHIFT && psize == mmu_huge_psize) { - DBG_LOW(" -> huge page !\n"); - return hash_huge_page(mm, access, ea, vsid, local, trap); - } -#endif /* CONFIG_HUGETLB_PAGE */ - #ifndef CONFIG_PPC_64K_PAGES - /* If we use 4K pages and our psize is not 4K, then we are hitting - * a special driver mapping, we need to align the address before - * we fetch the PTE + /* If we use 4K pages and our psize is not 4K, then we might + * be hitting a special driver mapping, and need to align the + * address before we fetch the PTE. + * + * It could also be a hugepage mapping, in which case this is + * not necessary, but it's not harmful, either. */ if (psize != MMU_PAGE_4K) ea &= ~((1ul << mmu_psize_defs[psize].shift) - 1); #endif /* CONFIG_PPC_64K_PAGES */ /* Get PTE and page size from page tables */ - ptep = find_linux_pte(pgdir, ea); + ptep = find_linux_pte_or_hugepte(pgdir, ea, &hugeshift); if (ptep == NULL || !pte_present(*ptep)) { DBG_LOW(" no PTE !\n"); - return 1; + rc = 1; + goto bail; } -#ifndef CONFIG_PPC_64K_PAGES - DBG_LOW(" i-pte: %016lx\n", pte_val(*ptep)); -#else - DBG_LOW(" i-pte: %016lx %016lx\n", pte_val(*ptep), - pte_val(*(ptep + PTRS_PER_PTE))); -#endif + /* Add _PAGE_PRESENT to the required access perm */ + access |= _PAGE_PRESENT; + /* Pre-check access permissions (will be re-checked atomically * in __hash_page_XX but this pre-check is a fast path */ if (access & ~pte_val(*ptep)) { DBG_LOW(" no access !\n"); - return 1; + rc = 1; + goto bail; } + if (hugeshift) { + if (pmd_trans_huge(*(pmd_t *)ptep)) + rc = __hash_page_thp(ea, access, vsid, (pmd_t *)ptep, + trap, local, ssize, psize); +#ifdef CONFIG_HUGETLB_PAGE + else + rc = __hash_page_huge(ea, access, vsid, ptep, trap, + local, ssize, hugeshift, psize); +#else + else { + /* + * if we have hugeshift, and is not transhuge with + * hugetlb disabled, something is really wrong. + */ + rc = 1; + WARN_ON(1); + } +#endif + check_paca_psize(ea, mm, psize, user_region); + + goto bail; + } + +#ifndef CONFIG_PPC_64K_PAGES + DBG_LOW(" i-pte: %016lx\n", pte_val(*ptep)); +#else + DBG_LOW(" i-pte: %016lx %016lx\n", pte_val(*ptep), + pte_val(*(ptep + PTRS_PER_PTE))); +#endif /* Do actual hashing */ #ifdef CONFIG_PPC_64K_PAGES /* If _PAGE_4K_PFN is set, make sure this is a 4k segment */ - if (pte_val(*ptep) & _PAGE_4K_PFN) { + if ((pte_val(*ptep) & _PAGE_4K_PFN) && psize == MMU_PAGE_64K) { demote_segment_4k(mm, ea); psize = MMU_PAGE_4K; } @@ -844,17 +1158,8 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) #endif } } - if (user_region) { - if (psize != get_paca()->context.user_psize) { - get_paca()->context = mm->context; - slb_flush_and_rebolt(); - } - } else if (get_paca()->vmalloc_sllp != - mmu_psize_defs[mmu_vmalloc_psize].sllp) { - get_paca()->vmalloc_sllp = - mmu_psize_defs[mmu_vmalloc_psize].sllp; - slb_vmalloc_update(); - } + + check_paca_psize(ea, mm, psize, user_region); #endif /* CONFIG_PPC_64K_PAGES */ #ifdef CONFIG_PPC_HAS_HASH_64K @@ -863,7 +1168,7 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) else #endif /* CONFIG_PPC_HAS_HASH_64K */ { - int spp = subpage_protection(pgdir, ea); + int spp = subpage_protection(mm, ea); if (access & spp) rc = -2; else @@ -871,6 +1176,12 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) local, ssize, spp); } + /* Dump some info in case of hash insertion failure, they should + * never happen so it is really useful to know if/when they do + */ + if (rc == -1) + hash_failure_debug(ea, access, vsid, trap, ssize, psize, + psize, pte_val(*ptep)); #ifndef CONFIG_PPC_64K_PAGES DBG_LOW(" o-pte: %016lx\n", pte_val(*ptep)); #else @@ -878,6 +1189,9 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap) pte_val(*(ptep + PTRS_PER_PTE))); #endif DBG_LOW(" -> rc=%d\n", rc); + +bail: + exception_exit(prev_state); return rc; } EXPORT_SYMBOL_GPL(hash_page); @@ -885,13 +1199,12 @@ EXPORT_SYMBOL_GPL(hash_page); void hash_preload(struct mm_struct *mm, unsigned long ea, unsigned long access, unsigned long trap) { + int hugepage_shift; unsigned long vsid; - void *pgdir; + pgd_t *pgdir; pte_t *ptep; - cpumask_t mask; unsigned long flags; - int local = 0; - int ssize; + int rc, ssize, local = 0; BUG_ON(REGION_ID(ea) != USER_REGION_ID); @@ -908,10 +1221,27 @@ void hash_preload(struct mm_struct *mm, unsigned long ea, pgdir = mm->pgd; if (pgdir == NULL) return; - ptep = find_linux_pte(pgdir, ea); - if (!ptep) + + /* Get VSID */ + ssize = user_segment_size(ea); + vsid = get_vsid(mm->context.id, ea, ssize); + if (!vsid) return; + /* + * Hash doesn't like irqs. Walking linux page table with irq disabled + * saves us from holding multiple locks. + */ + local_irq_save(flags); + /* + * THP pages use update_mmu_cache_pmd. We don't do + * hash preload there. Hence can ignore THP here + */ + ptep = find_linux_pte_or_hugepte(pgdir, ea, &hugepage_shift); + if (!ptep) + goto out_exit; + + WARN_ON(hugepage_shift); #ifdef CONFIG_PPC_64K_PAGES /* If either _PAGE_4K_PFN or _PAGE_NO_CACHE is set (and we are on * a 64K kernel), then we don't preload, hash_page() will take @@ -920,52 +1250,73 @@ void hash_preload(struct mm_struct *mm, unsigned long ea, * page size demotion here */ if (pte_val(*ptep) & (_PAGE_4K_PFN | _PAGE_NO_CACHE)) - return; + goto out_exit; #endif /* CONFIG_PPC_64K_PAGES */ - /* Get VSID */ - ssize = user_segment_size(ea); - vsid = get_vsid(mm->context.id, ea, ssize); - - /* Hash doesn't like irqs */ - local_irq_save(flags); - /* Is that local to this CPU ? */ - mask = cpumask_of_cpu(smp_processor_id()); - if (cpus_equal(mm->cpu_vm_mask, mask)) + if (cpumask_equal(mm_cpumask(mm), cpumask_of(smp_processor_id()))) local = 1; /* Hash it in */ #ifdef CONFIG_PPC_HAS_HASH_64K if (mm->context.user_psize == MMU_PAGE_64K) - __hash_page_64K(ea, access, vsid, ptep, trap, local, ssize); + rc = __hash_page_64K(ea, access, vsid, ptep, trap, local, ssize); else #endif /* CONFIG_PPC_HAS_HASH_64K */ - __hash_page_4K(ea, access, vsid, ptep, trap, local, ssize, - subpage_protection(pgdir, ea)); + rc = __hash_page_4K(ea, access, vsid, ptep, trap, local, ssize, + subpage_protection(mm, ea)); + /* Dump some info in case of hash insertion failure, they should + * never happen so it is really useful to know if/when they do + */ + if (rc == -1) + hash_failure_debug(ea, access, vsid, trap, ssize, + mm->context.user_psize, + mm->context.user_psize, + pte_val(*ptep)); +out_exit: local_irq_restore(flags); } /* WARNING: This is called from hash_low_64.S, if you change this prototype, * do not forget to update the assembly call site ! */ -void flush_hash_page(unsigned long va, real_pte_t pte, int psize, int ssize, +void flush_hash_page(unsigned long vpn, real_pte_t pte, int psize, int ssize, int local) { unsigned long hash, index, shift, hidx, slot; - DBG_LOW("flush_hash_page(va=%016x)\n", va); - pte_iterate_hashed_subpages(pte, psize, va, index, shift) { - hash = hpt_hash(va, shift, ssize); + DBG_LOW("flush_hash_page(vpn=%016lx)\n", vpn); + pte_iterate_hashed_subpages(pte, psize, vpn, index, shift) { + hash = hpt_hash(vpn, shift, ssize); hidx = __rpte_to_hidx(pte, index); if (hidx & _PTEIDX_SECONDARY) hash = ~hash; slot = (hash & htab_hash_mask) * HPTES_PER_GROUP; slot += hidx & _PTEIDX_GROUP_IX; - DBG_LOW(" sub %d: hash=%x, hidx=%x\n", index, slot, hidx); - ppc_md.hpte_invalidate(slot, va, psize, ssize, local); + DBG_LOW(" sub %ld: hash=%lx, hidx=%lx\n", index, slot, hidx); + /* + * We use same base page size and actual psize, because we don't + * use these functions for hugepage + */ + ppc_md.hpte_invalidate(slot, vpn, psize, psize, ssize, local); } pte_iterate_hashed_end(); + +#ifdef CONFIG_PPC_TRANSACTIONAL_MEM + /* Transactions are not aborted by tlbiel, only tlbie. + * Without, syncing a page back to a block device w/ PIO could pick up + * transactional data (bad!) so we force an abort here. Before the + * sync the page will be made read-only, which will flush_hash_page. + * BIG ISSUE here: if the kernel uses a page from userspace without + * unmapping it first, it may see the speculated version. + */ + if (local && cpu_has_feature(CPU_FTR_TM) && + current->thread.regs && + MSR_TM_ACTIVE(current->thread.regs->msr)) { + tm_enable(); + tm_abort(TM_CAUSE_TLBI); + } +#endif } void flush_hash_range(unsigned long number, int local) @@ -978,7 +1329,7 @@ void flush_hash_range(unsigned long number, int local) &__get_cpu_var(ppc64_tlb_batch); for (i = 0; i < number; i++) - flush_hash_page(batch->vaddr[i], batch->pte[i], + flush_hash_page(batch->vpn[i], batch->pte[i], batch->psize, batch->ssize, local); } } @@ -989,6 +1340,8 @@ void flush_hash_range(unsigned long number, int local) */ void low_hash_fault(struct pt_regs *regs, unsigned long address, int rc) { + enum ctx_state prev_state = exception_enter(); + if (user_mode(regs)) { #ifdef CONFIG_PPC_SUBPAGE_PROT if (rc == -2) @@ -998,24 +1351,64 @@ void low_hash_fault(struct pt_regs *regs, unsigned long address, int rc) _exception(SIGBUS, regs, BUS_ADRERR, address); } else bad_page_fault(regs, address, SIGBUS); + + exception_exit(prev_state); +} + +long hpte_insert_repeating(unsigned long hash, unsigned long vpn, + unsigned long pa, unsigned long rflags, + unsigned long vflags, int psize, int ssize) +{ + unsigned long hpte_group; + long slot; + +repeat: + hpte_group = ((hash & htab_hash_mask) * + HPTES_PER_GROUP) & ~0x7UL; + + /* Insert into the hash table, primary slot */ + slot = ppc_md.hpte_insert(hpte_group, vpn, pa, rflags, vflags, + psize, psize, ssize); + + /* Primary is full, try the secondary */ + if (unlikely(slot == -1)) { + hpte_group = ((~hash & htab_hash_mask) * + HPTES_PER_GROUP) & ~0x7UL; + slot = ppc_md.hpte_insert(hpte_group, vpn, pa, rflags, + vflags | HPTE_V_SECONDARY, + psize, psize, ssize); + if (slot == -1) { + if (mftb() & 0x1) + hpte_group = ((hash & htab_hash_mask) * + HPTES_PER_GROUP)&~0x7UL; + + ppc_md.hpte_remove(hpte_group); + goto repeat; + } + } + + return slot; } #ifdef CONFIG_DEBUG_PAGEALLOC static void kernel_map_linear_page(unsigned long vaddr, unsigned long lmi) { - unsigned long hash, hpteg; + unsigned long hash; unsigned long vsid = get_kernel_vsid(vaddr, mmu_kernel_ssize); - unsigned long va = hpt_va(vaddr, vsid, mmu_kernel_ssize); - unsigned long mode = _PAGE_ACCESSED | _PAGE_DIRTY | - _PAGE_COHERENT | PP_RWXX | HPTE_R_N; - int ret; + unsigned long vpn = hpt_vpn(vaddr, vsid, mmu_kernel_ssize); + unsigned long mode = htab_convert_pte_flags(PAGE_KERNEL); + long ret; + + hash = hpt_hash(vpn, PAGE_SHIFT, mmu_kernel_ssize); + + /* Don't create HPTE entries for bad address */ + if (!vsid) + return; - hash = hpt_hash(va, PAGE_SHIFT, mmu_kernel_ssize); - hpteg = ((hash & htab_hash_mask) * HPTES_PER_GROUP); + ret = hpte_insert_repeating(hash, vpn, __pa(vaddr), mode, + HPTE_V_BOLTED, + mmu_linear_psize, mmu_kernel_ssize); - ret = ppc_md.hpte_insert(hpteg, va, __pa(vaddr), - mode, HPTE_V_BOLTED, - mmu_linear_psize, mmu_kernel_ssize); BUG_ON (ret < 0); spin_lock(&linear_map_hash_lock); BUG_ON(linear_map_hash_slots[lmi] & 0x80); @@ -1027,9 +1420,9 @@ static void kernel_unmap_linear_page(unsigned long vaddr, unsigned long lmi) { unsigned long hash, hidx, slot; unsigned long vsid = get_kernel_vsid(vaddr, mmu_kernel_ssize); - unsigned long va = hpt_va(vaddr, vsid, mmu_kernel_ssize); + unsigned long vpn = hpt_vpn(vaddr, vsid, mmu_kernel_ssize); - hash = hpt_hash(va, PAGE_SHIFT, mmu_kernel_ssize); + hash = hpt_hash(vpn, PAGE_SHIFT, mmu_kernel_ssize); spin_lock(&linear_map_hash_lock); BUG_ON(!(linear_map_hash_slots[lmi] & 0x80)); hidx = linear_map_hash_slots[lmi] & 0x7f; @@ -1039,7 +1432,8 @@ static void kernel_unmap_linear_page(unsigned long vaddr, unsigned long lmi) hash = ~hash; slot = (hash & htab_hash_mask) * HPTES_PER_GROUP; slot += hidx & _PTEIDX_GROUP_IX; - ppc_md.hpte_invalidate(slot, va, mmu_linear_psize, mmu_kernel_ssize, 0); + ppc_md.hpte_invalidate(slot, vpn, mmu_linear_psize, mmu_linear_psize, + mmu_kernel_ssize, 0); } void kernel_map_pages(struct page *page, int numpages, int enable) @@ -1061,3 +1455,23 @@ void kernel_map_pages(struct page *page, int numpages, int enable) local_irq_restore(flags); } #endif /* CONFIG_DEBUG_PAGEALLOC */ + +void setup_initial_memory_limit(phys_addr_t first_memblock_base, + phys_addr_t first_memblock_size) +{ + /* We don't currently support the first MEMBLOCK not mapping 0 + * physical on those processors + */ + BUG_ON(first_memblock_base != 0); + + /* On LPAR systems, the first entry is our RMA region, + * non-LPAR 64-bit hash MMU systems don't have a limitation + * on real mode access, but using the first entry works well + * enough. We also clamp it to 1G to avoid some funky things + * such as RTAS bugs etc... + */ + ppc64_rma_size = min_t(u64, first_memblock_size, 0x40000000); + + /* Finally limit subsequent allocations */ + memblock_set_current_limit(ppc64_rma_size); +} |