diff options
Diffstat (limited to 'arch/arm/mm/mmu.c')
-rw-r--r-- | arch/arm/mm/mmu.c | 302 |
1 files changed, 133 insertions, 169 deletions
diff --git a/arch/arm/mm/mmu.c b/arch/arm/mm/mmu.c index 285894171186..72ad3e1f56cf 100644 --- a/arch/arm/mm/mmu.c +++ b/arch/arm/mm/mmu.c @@ -11,13 +11,12 @@ #include <linux/kernel.h> #include <linux/errno.h> #include <linux/init.h> -#include <linux/bootmem.h> #include <linux/mman.h> #include <linux/nodemask.h> -#include <linux/sort.h> +#include <linux/memblock.h> +#include <linux/fs.h> #include <asm/cputype.h> -#include <asm/mach-types.h> #include <asm/sections.h> #include <asm/cachetype.h> #include <asm/setup.h> @@ -247,6 +246,9 @@ static struct mem_type mem_types[] = { .domain = DOMAIN_USER, }, [MT_MEMORY] = { + .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY | + L_PTE_WRITE | L_PTE_EXEC, + .prot_l1 = PMD_TYPE_TABLE, .prot_sect = PMD_TYPE_SECT | PMD_SECT_AP_WRITE, .domain = DOMAIN_KERNEL, }, @@ -255,9 +257,25 @@ static struct mem_type mem_types[] = { .domain = DOMAIN_KERNEL, }, [MT_MEMORY_NONCACHED] = { + .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY | + L_PTE_WRITE | L_PTE_EXEC | L_PTE_MT_BUFFERABLE, + .prot_l1 = PMD_TYPE_TABLE, .prot_sect = PMD_TYPE_SECT | PMD_SECT_AP_WRITE, .domain = DOMAIN_KERNEL, }, + [MT_MEMORY_DTCM] = { + .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY | + L_PTE_WRITE, + .prot_l1 = PMD_TYPE_TABLE, + .prot_sect = PMD_TYPE_SECT | PMD_SECT_XN, + .domain = DOMAIN_KERNEL, + }, + [MT_MEMORY_ITCM] = { + .prot_pte = L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY | + L_PTE_WRITE | L_PTE_EXEC, + .prot_l1 = PMD_TYPE_TABLE, + .domain = DOMAIN_KERNEL, + }, }; const struct mem_type *get_mem_type(unsigned int type) @@ -291,9 +309,8 @@ static void __init build_mem_type_table(void) cachepolicy = CPOLICY_WRITEBACK; ecc_mask = 0; } -#ifdef CONFIG_SMP - cachepolicy = CPOLICY_WRITEALLOC; -#endif + if (is_smp()) + cachepolicy = CPOLICY_WRITEALLOC; /* * Strip out features not present on earlier architectures. @@ -387,21 +404,22 @@ static void __init build_mem_type_table(void) cp = &cache_policies[cachepolicy]; vecs_pgprot = kern_pgprot = user_pgprot = cp->pte; -#ifndef CONFIG_SMP /* * Only use write-through for non-SMP systems */ - if (cpu_arch >= CPU_ARCH_ARMv5 && cachepolicy > CPOLICY_WRITETHROUGH) + if (!is_smp() && cpu_arch >= CPU_ARCH_ARMv5 && cachepolicy > CPOLICY_WRITETHROUGH) vecs_pgprot = cache_policies[CPOLICY_WRITETHROUGH].pte; -#endif /* * Enable CPU-specific coherency if supported. * (Only available on XSC3 at the moment.) */ - if (arch_is_coherent() && cpu_is_xsc3()) + if (arch_is_coherent() && cpu_is_xsc3()) { mem_types[MT_MEMORY].prot_sect |= PMD_SECT_S; - + mem_types[MT_MEMORY].prot_pte |= L_PTE_SHARED; + mem_types[MT_MEMORY_NONCACHED].prot_sect |= PMD_SECT_S; + mem_types[MT_MEMORY_NONCACHED].prot_pte |= L_PTE_SHARED; + } /* * ARMv6 and above have extended page tables. */ @@ -414,20 +432,23 @@ static void __init build_mem_type_table(void) mem_types[MT_MINICLEAN].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; mem_types[MT_CACHECLEAN].prot_sect |= PMD_SECT_APX|PMD_SECT_AP_WRITE; -#ifdef CONFIG_SMP - /* - * Mark memory with the "shared" attribute for SMP systems - */ - user_pgprot |= L_PTE_SHARED; - kern_pgprot |= L_PTE_SHARED; - vecs_pgprot |= L_PTE_SHARED; - mem_types[MT_DEVICE_WC].prot_sect |= PMD_SECT_S; - mem_types[MT_DEVICE_WC].prot_pte |= L_PTE_SHARED; - mem_types[MT_DEVICE_CACHED].prot_sect |= PMD_SECT_S; - mem_types[MT_DEVICE_CACHED].prot_pte |= L_PTE_SHARED; - mem_types[MT_MEMORY].prot_sect |= PMD_SECT_S; - mem_types[MT_MEMORY_NONCACHED].prot_sect |= PMD_SECT_S; -#endif + if (is_smp()) { + /* + * Mark memory with the "shared" attribute + * for SMP systems + */ + user_pgprot |= L_PTE_SHARED; + kern_pgprot |= L_PTE_SHARED; + vecs_pgprot |= L_PTE_SHARED; + mem_types[MT_DEVICE_WC].prot_sect |= PMD_SECT_S; + mem_types[MT_DEVICE_WC].prot_pte |= L_PTE_SHARED; + mem_types[MT_DEVICE_CACHED].prot_sect |= PMD_SECT_S; + mem_types[MT_DEVICE_CACHED].prot_pte |= L_PTE_SHARED; + mem_types[MT_MEMORY].prot_sect |= PMD_SECT_S; + mem_types[MT_MEMORY].prot_pte |= L_PTE_SHARED; + mem_types[MT_MEMORY_NONCACHED].prot_sect |= PMD_SECT_S; + mem_types[MT_MEMORY_NONCACHED].prot_pte |= L_PTE_SHARED; + } } /* @@ -463,6 +484,8 @@ static void __init build_mem_type_table(void) mem_types[MT_LOW_VECTORS].prot_l1 |= ecc_mask; mem_types[MT_HIGH_VECTORS].prot_l1 |= ecc_mask; mem_types[MT_MEMORY].prot_sect |= ecc_mask | cp->pmd; + mem_types[MT_MEMORY].prot_pte |= kern_pgprot; + mem_types[MT_MEMORY_NONCACHED].prot_sect |= ecc_mask; mem_types[MT_ROM].prot_sect |= cp->pmd; switch (cp->pmd) { @@ -486,20 +509,43 @@ static void __init build_mem_type_table(void) } } +#ifdef CONFIG_ARM_DMA_MEM_BUFFERABLE +pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn, + unsigned long size, pgprot_t vma_prot) +{ + if (!pfn_valid(pfn)) + return pgprot_noncached(vma_prot); + else if (file->f_flags & O_SYNC) + return pgprot_writecombine(vma_prot); + return vma_prot; +} +EXPORT_SYMBOL(phys_mem_access_prot); +#endif + #define vectors_base() (vectors_high() ? 0xffff0000 : 0) -static void __init alloc_init_pte(pmd_t *pmd, unsigned long addr, - unsigned long end, unsigned long pfn, - const struct mem_type *type) +static void __init *early_alloc(unsigned long sz) { - pte_t *pte; + void *ptr = __va(memblock_alloc(sz, sz)); + memset(ptr, 0, sz); + return ptr; +} +static pte_t * __init early_pte_alloc(pmd_t *pmd, unsigned long addr, unsigned long prot) +{ if (pmd_none(*pmd)) { - pte = alloc_bootmem_low_pages(2 * PTRS_PER_PTE * sizeof(pte_t)); - __pmd_populate(pmd, __pa(pte) | type->prot_l1); + pte_t *pte = early_alloc(2 * PTRS_PER_PTE * sizeof(pte_t)); + __pmd_populate(pmd, __pa(pte) | prot); } + BUG_ON(pmd_bad(*pmd)); + return pte_offset_kernel(pmd, addr); +} - pte = pte_offset_kernel(pmd, addr); +static void __init alloc_init_pte(pmd_t *pmd, unsigned long addr, + unsigned long end, unsigned long pfn, + const struct mem_type *type) +{ + pte_t *pte = early_pte_alloc(pmd, addr, type->prot_l1); do { set_pte_ext(pte, pfn_pte(pfn, __pgprot(type->prot_pte)), 0); pfn++; @@ -668,7 +714,7 @@ void __init iotable_init(struct map_desc *io_desc, int nr) create_mapping(io_desc + i); } -static unsigned long __initdata vmalloc_reserve = SZ_128M; +static void * __initdata vmalloc_min = (void *)(VMALLOC_END - SZ_128M); /* * vmalloc=size forces the vmalloc area to be exactly 'size' @@ -677,7 +723,7 @@ static unsigned long __initdata vmalloc_reserve = SZ_128M; */ static int __init early_vmalloc(char *arg) { - vmalloc_reserve = memparse(arg, NULL); + unsigned long vmalloc_reserve = memparse(arg, NULL); if (vmalloc_reserve < SZ_16M) { vmalloc_reserve = SZ_16M; @@ -692,22 +738,27 @@ static int __init early_vmalloc(char *arg) "vmalloc area is too big, limiting to %luMB\n", vmalloc_reserve >> 20); } + + vmalloc_min = (void *)(VMALLOC_END - vmalloc_reserve); return 0; } early_param("vmalloc", early_vmalloc); -#define VMALLOC_MIN (void *)(VMALLOC_END - vmalloc_reserve) +static phys_addr_t lowmem_limit __initdata = 0; static void __init sanity_check_meminfo(void) { int i, j, highmem = 0; + lowmem_limit = __pa(vmalloc_min - 1) + 1; + memblock_set_current_limit(lowmem_limit); + for (i = 0, j = 0; i < meminfo.nr_banks; i++) { struct membank *bank = &meminfo.bank[j]; *bank = meminfo.bank[i]; #ifdef CONFIG_HIGHMEM - if (__va(bank->start) > VMALLOC_MIN || + if (__va(bank->start) > vmalloc_min || __va(bank->start) < (void *)PAGE_OFFSET) highmem = 1; @@ -717,8 +768,8 @@ static void __init sanity_check_meminfo(void) * Split those memory banks which are partially overlapping * the vmalloc area greatly simplifying things later. */ - if (__va(bank->start) < VMALLOC_MIN && - bank->size > VMALLOC_MIN - __va(bank->start)) { + if (__va(bank->start) < vmalloc_min && + bank->size > vmalloc_min - __va(bank->start)) { if (meminfo.nr_banks >= NR_BANKS) { printk(KERN_CRIT "NR_BANKS too low, " "ignoring high memory\n"); @@ -727,12 +778,12 @@ static void __init sanity_check_meminfo(void) (meminfo.nr_banks - i) * sizeof(*bank)); meminfo.nr_banks++; i++; - bank[1].size -= VMALLOC_MIN - __va(bank->start); - bank[1].start = __pa(VMALLOC_MIN - 1) + 1; + bank[1].size -= vmalloc_min - __va(bank->start); + bank[1].start = __pa(vmalloc_min - 1) + 1; bank[1].highmem = highmem = 1; j++; } - bank->size = VMALLOC_MIN - __va(bank->start); + bank->size = vmalloc_min - __va(bank->start); } #else bank->highmem = highmem; @@ -741,7 +792,7 @@ static void __init sanity_check_meminfo(void) * Check whether this memory bank would entirely overlap * the vmalloc area. */ - if (__va(bank->start) >= VMALLOC_MIN || + if (__va(bank->start) >= vmalloc_min || __va(bank->start) < (void *)PAGE_OFFSET) { printk(KERN_NOTICE "Ignoring RAM at %.8lx-%.8lx " "(vmalloc region overlap).\n", @@ -753,9 +804,9 @@ static void __init sanity_check_meminfo(void) * Check whether this memory bank would partially overlap * the vmalloc area. */ - if (__va(bank->start + bank->size) > VMALLOC_MIN || + if (__va(bank->start + bank->size) > vmalloc_min || __va(bank->start + bank->size) < __va(bank->start)) { - unsigned long newsize = VMALLOC_MIN - __va(bank->start); + unsigned long newsize = vmalloc_min - __va(bank->start); printk(KERN_NOTICE "Truncating RAM at %.8lx-%.8lx " "to -%.8lx (vmalloc region overlap).\n", bank->start, bank->start + bank->size - 1, @@ -776,8 +827,7 @@ static void __init sanity_check_meminfo(void) * rather difficult. */ reason = "with VIPT aliasing cache"; -#ifdef CONFIG_SMP - } else if (tlb_ops_need_broadcast()) { + } else if (is_smp() && tlb_ops_need_broadcast()) { /* * kmap_high needs to occasionally flush TLB entries, * however, if the TLB entries need to be broadcast @@ -787,7 +837,6 @@ static void __init sanity_check_meminfo(void) * (must not be called with irqs off) */ reason = "without hardware TLB ops broadcasting"; -#endif } if (reason) { printk(KERN_CRIT "HIGHMEM is not supported %s, ignoring high memory\n", @@ -803,6 +852,7 @@ static void __init sanity_check_meminfo(void) static inline void prepare_page_table(void) { unsigned long addr; + phys_addr_t end; /* * Clear out all the mappings below the kernel image. @@ -818,110 +868,39 @@ static inline void prepare_page_table(void) pmd_clear(pmd_off_k(addr)); /* + * Find the end of the first block of lowmem. + */ + end = memblock.memory.regions[0].base + memblock.memory.regions[0].size; + if (end >= lowmem_limit) + end = lowmem_limit; + + /* * Clear out all the kernel space mappings, except for the first * memory bank, up to the end of the vmalloc region. */ - for (addr = __phys_to_virt(bank_phys_end(&meminfo.bank[0])); + for (addr = __phys_to_virt(end); addr < VMALLOC_END; addr += PGDIR_SIZE) pmd_clear(pmd_off_k(addr)); } /* - * Reserve the various regions of node 0 + * Reserve the special regions of memory */ -void __init reserve_node_zero(pg_data_t *pgdat) +void __init arm_mm_memblock_reserve(void) { - unsigned long res_size = 0; - - /* - * Register the kernel text and data with bootmem. - * Note that this can only be in node 0. - */ -#ifdef CONFIG_XIP_KERNEL - reserve_bootmem_node(pgdat, __pa(_data), _end - _data, - BOOTMEM_DEFAULT); -#else - reserve_bootmem_node(pgdat, __pa(_stext), _end - _stext, - BOOTMEM_DEFAULT); -#endif - /* * Reserve the page tables. These are already in use, * and can only be in node 0. */ - reserve_bootmem_node(pgdat, __pa(swapper_pg_dir), - PTRS_PER_PGD * sizeof(pgd_t), BOOTMEM_DEFAULT); - - /* - * Hmm... This should go elsewhere, but we really really need to - * stop things allocating the low memory; ideally we need a better - * implementation of GFP_DMA which does not assume that DMA-able - * memory starts at zero. - */ - if (machine_is_integrator() || machine_is_cintegrator()) - res_size = __pa(swapper_pg_dir) - PHYS_OFFSET; - - /* - * These should likewise go elsewhere. They pre-reserve the - * screen memory region at the start of main system memory. - */ - if (machine_is_edb7211()) - res_size = 0x00020000; - if (machine_is_p720t()) - res_size = 0x00014000; - - /* H1940, RX3715 and RX1950 need to reserve this for suspend */ - - if (machine_is_h1940() || machine_is_rx3715() - || machine_is_rx1950()) { - reserve_bootmem_node(pgdat, 0x30003000, 0x1000, - BOOTMEM_DEFAULT); - reserve_bootmem_node(pgdat, 0x30081000, 0x1000, - BOOTMEM_DEFAULT); - } - - if (machine_is_palmld() || machine_is_palmtx()) { - reserve_bootmem_node(pgdat, 0xa0000000, 0x1000, - BOOTMEM_EXCLUSIVE); - reserve_bootmem_node(pgdat, 0xa0200000, 0x1000, - BOOTMEM_EXCLUSIVE); - } - - if (machine_is_treo680() || machine_is_centro()) { - reserve_bootmem_node(pgdat, 0xa0000000, 0x1000, - BOOTMEM_EXCLUSIVE); - reserve_bootmem_node(pgdat, 0xa2000000, 0x1000, - BOOTMEM_EXCLUSIVE); - } - - if (machine_is_palmt5()) - reserve_bootmem_node(pgdat, 0xa0200000, 0x1000, - BOOTMEM_EXCLUSIVE); - - /* - * U300 - This platform family can share physical memory - * between two ARM cpus, one running Linux and the other - * running another OS. - */ - if (machine_is_u300()) { -#ifdef CONFIG_MACH_U300_SINGLE_RAM -#if ((CONFIG_MACH_U300_ACCESS_MEM_SIZE & 1) == 1) && \ - CONFIG_MACH_U300_2MB_ALIGNMENT_FIX - res_size = 0x00100000; -#endif -#endif - } + memblock_reserve(__pa(swapper_pg_dir), PTRS_PER_PGD * sizeof(pgd_t)); #ifdef CONFIG_SA1111 /* * Because of the SA1111 DMA bug, we want to preserve our * precious DMA-able memory... */ - res_size = __pa(swapper_pg_dir) - PHYS_OFFSET; + memblock_reserve(PHYS_OFFSET, __pa(swapper_pg_dir) - PHYS_OFFSET); #endif - if (res_size) - reserve_bootmem_node(pgdat, PHYS_OFFSET, res_size, - BOOTMEM_DEFAULT); } /* @@ -940,7 +919,7 @@ static void __init devicemaps_init(struct machine_desc *mdesc) /* * Allocate the vector page early. */ - vectors = alloc_bootmem_low_pages(PAGE_SIZE); + vectors = early_alloc(PAGE_SIZE); for (addr = VMALLOC_END; addr; addr += PGDIR_SIZE) pmd_clear(pmd_off_k(addr)); @@ -1011,45 +990,33 @@ static void __init devicemaps_init(struct machine_desc *mdesc) static void __init kmap_init(void) { #ifdef CONFIG_HIGHMEM - pmd_t *pmd = pmd_off_k(PKMAP_BASE); - pte_t *pte = alloc_bootmem_low_pages(2 * PTRS_PER_PTE * sizeof(pte_t)); - BUG_ON(!pmd_none(*pmd) || !pte); - __pmd_populate(pmd, __pa(pte) | _PAGE_KERNEL_TABLE); - pkmap_page_table = pte + PTRS_PER_PTE; + pkmap_page_table = early_pte_alloc(pmd_off_k(PKMAP_BASE), + PKMAP_BASE, _PAGE_KERNEL_TABLE); #endif } -static inline void map_memory_bank(struct membank *bank) -{ - struct map_desc map; - - map.pfn = bank_pfn_start(bank); - map.virtual = __phys_to_virt(bank_phys_start(bank)); - map.length = bank_phys_size(bank); - map.type = MT_MEMORY; - - create_mapping(&map); -} - static void __init map_lowmem(void) { - struct meminfo *mi = &meminfo; - int i; + struct memblock_region *reg; /* Map all the lowmem memory banks. */ - for (i = 0; i < mi->nr_banks; i++) { - struct membank *bank = &mi->bank[i]; + for_each_memblock(memory, reg) { + phys_addr_t start = reg->base; + phys_addr_t end = start + reg->size; + struct map_desc map; + + if (end > lowmem_limit) + end = lowmem_limit; + if (start >= end) + break; - if (!bank->highmem) - map_memory_bank(bank); - } -} + map.pfn = __phys_to_pfn(start); + map.virtual = __phys_to_virt(start); + map.length = end - start; + map.type = MT_MEMORY; -static int __init meminfo_cmp(const void *_a, const void *_b) -{ - const struct membank *a = _a, *b = _b; - long cmp = bank_pfn_start(a) - bank_pfn_start(b); - return cmp < 0 ? -1 : cmp > 0 ? 1 : 0; + create_mapping(&map); + } } /* @@ -1060,23 +1027,20 @@ void __init paging_init(struct machine_desc *mdesc) { void *zero_page; - sort(&meminfo.bank, meminfo.nr_banks, sizeof(meminfo.bank[0]), meminfo_cmp, NULL); - build_mem_type_table(); sanity_check_meminfo(); prepare_page_table(); map_lowmem(); - bootmem_init(); devicemaps_init(mdesc); kmap_init(); top_pmd = pmd_off_k(0xffff0000); - /* - * allocate the zero page. Note that this always succeeds and - * returns a zeroed result. - */ - zero_page = alloc_bootmem_low_pages(PAGE_SIZE); + /* allocate the zero page. */ + zero_page = early_alloc(PAGE_SIZE); + + bootmem_init(); + empty_zero_page = virt_to_page(zero_page); __flush_dcache_page(NULL, empty_zero_page); } |