X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;ds=sidebyside;f=arch%2Fx86_64%2Fmm%2Finit.c;h=1d61913d75f4b8778e16879207bf5c6513f43cf7;hb=43bc926fffd92024b46cafaf7350d669ba9ca884;hp=dbe53b4c7e66c0c0f75fbccfb8d404961d7eb10d;hpb=cee37fe97739d85991964371c1f3a745c00dd236;p=linux-2.6.git diff --git a/arch/x86_64/mm/init.c b/arch/x86_64/mm/init.c index dbe53b4c7..1d61913d7 100644 --- a/arch/x86_64/mm/init.c +++ b/arch/x86_64/mm/init.c @@ -7,6 +7,7 @@ */ #include +#include #include #include #include @@ -22,6 +23,10 @@ #include #include #include +#include +#include +#include +#include #include #include @@ -36,16 +41,18 @@ #include #include #include +#include +#include +#include #ifndef Dprintk #define Dprintk(x...) #endif -#ifdef CONFIG_GART_IOMMU -extern int swiotlb; -#endif +struct dma_mapping_ops* dma_ops; +EXPORT_SYMBOL(dma_ops); -extern char _stext[]; +static unsigned long dma_reserve __initdata; DEFINE_PER_CPU(struct mmu_gather, mmu_gathers); @@ -57,41 +64,38 @@ DEFINE_PER_CPU(struct mmu_gather, mmu_gathers); void show_mem(void) { - int i, total = 0, reserved = 0; - int shared = 0, cached = 0; + long i, total = 0, reserved = 0; + long shared = 0, cached = 0; pg_data_t *pgdat; struct page *page; - printk("Mem-info:\n"); + printk(KERN_INFO "Mem-info:\n"); show_free_areas(); - printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); + printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); - for_each_pgdat(pgdat) { + for_each_online_pgdat(pgdat) { for (i = 0; i < pgdat->node_spanned_pages; ++i) { page = pfn_to_page(pgdat->node_start_pfn + i); total++; - if (PageReserved(page)) - reserved++; - else if (PageSwapCache(page)) - cached++; - else if (page_count(page)) - shared += page_count(page) - 1; + if (PageReserved(page)) + reserved++; + else if (PageSwapCache(page)) + cached++; + else if (page_count(page)) + shared += page_count(page) - 1; } } - printk("%d pages of RAM\n", total); - printk("%d reserved pages\n",reserved); - printk("%d pages shared\n",shared); - printk("%d pages swap cached\n",cached); + printk(KERN_INFO "%lu pages of RAM\n", total); + printk(KERN_INFO "%lu reserved pages\n",reserved); + printk(KERN_INFO "%lu pages shared\n",shared); + printk(KERN_INFO "%lu pages swap cached\n",cached); } /* References to section boundaries */ -extern char _text, _etext, _edata, __bss_start, _end[]; -extern char __init_begin, __init_end; - int after_bootmem; -static void *spp_getpage(void) +static __init void *spp_getpage(void) { void *ptr; if (after_bootmem) @@ -105,7 +109,7 @@ static void *spp_getpage(void) return ptr; } -static void set_pte_phys(unsigned long vaddr, +static __init void set_pte_phys(unsigned long vaddr, unsigned long phys, pgprot_t prot) { pgd_t *pgd; @@ -154,7 +158,8 @@ static void set_pte_phys(unsigned long vaddr, } /* NOTE: this is meant to be run only at boot */ -void __set_fixmap (enum fixed_addresses idx, unsigned long phys, pgprot_t prot) +void __init +__set_fixmap (enum fixed_addresses idx, unsigned long phys, pgprot_t prot) { unsigned long address = __fix_to_virt(idx); @@ -179,13 +184,19 @@ static struct temp_map { {} }; -static __init void *alloc_low_page(int *index, unsigned long *phys) +static __meminit void *alloc_low_page(int *index, unsigned long *phys) { struct temp_map *ti; int i; unsigned long pfn = table_end++, paddr; void *adr; + if (after_bootmem) { + adr = (void *)get_zeroed_page(GFP_ATOMIC); + *phys = __pa(adr); + return adr; + } + if (pfn >= end_pfn) panic("alloc_low_page: ran out of memory"); for (i = 0; temp_mappings[i].allocated; i++) { @@ -198,55 +209,113 @@ static __init void *alloc_low_page(int *index, unsigned long *phys) ti->allocated = 1; __flush_tlb(); adr = ti->address + ((pfn << PAGE_SHIFT) & ~PMD_MASK); + memset(adr, 0, PAGE_SIZE); *index = i; *phys = pfn * PAGE_SIZE; return adr; } -static __init void unmap_low_page(int i) +static __meminit void unmap_low_page(int i) { - struct temp_map *ti = &temp_mappings[i]; + struct temp_map *ti; + + if (after_bootmem) + return; + + ti = &temp_mappings[i]; set_pmd(ti->pmd, __pmd(0)); ti->allocated = 0; } -static void __init phys_pud_init(pud_t *pud, unsigned long address, unsigned long end) +/* Must run before zap_low_mappings */ +__init void *early_ioremap(unsigned long addr, unsigned long size) +{ + unsigned long map = round_down(addr, LARGE_PAGE_SIZE); + + /* actually usually some more */ + if (size >= LARGE_PAGE_SIZE) { + printk("SMBIOS area too long %lu\n", size); + return NULL; + } + set_pmd(temp_mappings[0].pmd, __pmd(map | _KERNPG_TABLE | _PAGE_PSE)); + map += LARGE_PAGE_SIZE; + set_pmd(temp_mappings[1].pmd, __pmd(map | _KERNPG_TABLE | _PAGE_PSE)); + __flush_tlb(); + return temp_mappings[0].address + (addr & (LARGE_PAGE_SIZE-1)); +} + +/* To avoid virtual aliases later */ +__init void early_iounmap(void *addr, unsigned long size) +{ + if ((void *)round_down((unsigned long)addr, LARGE_PAGE_SIZE) != temp_mappings[0].address) + printk("early_iounmap: bad address %p\n", addr); + set_pmd(temp_mappings[0].pmd, __pmd(0)); + set_pmd(temp_mappings[1].pmd, __pmd(0)); + __flush_tlb(); +} + +static void __meminit +phys_pmd_init(pmd_t *pmd, unsigned long address, unsigned long end) +{ + int i; + + for (i = 0; i < PTRS_PER_PMD; pmd++, i++, address += PMD_SIZE) { + unsigned long entry; + + if (address > end) { + for (; i < PTRS_PER_PMD; i++, pmd++) + set_pmd(pmd, __pmd(0)); + break; + } + entry = _PAGE_NX|_PAGE_PSE|_KERNPG_TABLE|_PAGE_GLOBAL|address; + entry &= __supported_pte_mask; + set_pmd(pmd, __pmd(entry)); + } +} + +static void __meminit +phys_pmd_update(pud_t *pud, unsigned long address, unsigned long end) +{ + pmd_t *pmd = pmd_offset(pud, (unsigned long)__va(address)); + + if (pmd_none(*pmd)) { + spin_lock(&init_mm.page_table_lock); + phys_pmd_init(pmd, address, end); + spin_unlock(&init_mm.page_table_lock); + __flush_tlb_all(); + } +} + +static void __meminit phys_pud_init(pud_t *pud, unsigned long address, unsigned long end) { - long i, j; + long i = pud_index(address); - i = pud_index(address); pud = pud + i; + + if (after_bootmem && pud_val(*pud)) { + phys_pmd_update(pud, address, end); + return; + } + for (; i < PTRS_PER_PUD; pud++, i++) { int map; unsigned long paddr, pmd_phys; pmd_t *pmd; - paddr = address + i*PUD_SIZE; - if (paddr >= end) { - for (; i < PTRS_PER_PUD; i++, pud++) - set_pud(pud, __pud(0)); + paddr = (address & PGDIR_MASK) + i*PUD_SIZE; + if (paddr >= end) break; - } - if (!e820_mapped(paddr, paddr+PUD_SIZE, 0)) { + if (!after_bootmem && !e820_any_mapped(paddr, paddr+PUD_SIZE, 0)) { set_pud(pud, __pud(0)); continue; } pmd = alloc_low_page(&map, &pmd_phys); + spin_lock(&init_mm.page_table_lock); set_pud(pud, __pud(pmd_phys | _KERNPG_TABLE)); - for (j = 0; j < PTRS_PER_PMD; pmd++, j++, paddr += PMD_SIZE) { - unsigned long pe; - - if (paddr >= end) { - for (; j < PTRS_PER_PMD; j++, pmd++) - set_pmd(pmd, __pmd(0)); - break; - } - pe = _PAGE_NX|_PAGE_PSE | _KERNPG_TABLE | _PAGE_GLOBAL | paddr; - pe &= __supported_pte_mask; - set_pmd(pmd, __pmd(pe)); - } + phys_pmd_init(pmd, paddr, end); + spin_unlock(&init_mm.page_table_lock); unmap_low_page(map); } __flush_tlb(); @@ -254,25 +323,32 @@ static void __init phys_pud_init(pud_t *pud, unsigned long address, unsigned lon static void __init find_early_table_space(unsigned long end) { - unsigned long puds, pmds, tables; + unsigned long puds, pmds, tables, start; puds = (end + PUD_SIZE - 1) >> PUD_SHIFT; pmds = (end + PMD_SIZE - 1) >> PMD_SHIFT; tables = round_up(puds * sizeof(pud_t), PAGE_SIZE) + round_up(pmds * sizeof(pmd_t), PAGE_SIZE); - table_start = find_e820_area(0x8000, __pa_symbol(&_text), tables); + /* RED-PEN putting page tables only on node 0 could + cause a hotspot and fill up ZONE_DMA. The page tables + need roughly 0.5KB per GB. */ + start = 0x8000; + table_start = find_e820_area(start, end, tables); if (table_start == -1UL) panic("Cannot find space for the kernel page tables"); table_start >>= PAGE_SHIFT; table_end = table_start; + + early_printk("kernel direct mapping tables up to %lx @ %lx-%lx\n", + end, table_start << PAGE_SHIFT, table_end << PAGE_SHIFT); } /* Setup the direct mapping of the physical memory at PAGE_OFFSET. This runs before bootmem is initialized and gets pages directly from the physical memory. To access them they are temporarily mapped. */ -void __init init_memory_mapping(unsigned long start, unsigned long end) +void __meminit init_memory_mapping(unsigned long start, unsigned long end) { unsigned long next; @@ -284,7 +360,8 @@ void __init init_memory_mapping(unsigned long start, unsigned long end) * mapped. Unfortunately this is done currently before the nodes are * discovered. */ - find_early_table_space(end); + if (!after_bootmem) + find_early_table_space(end); start = (unsigned long)__va(start); end = (unsigned long)__va(end); @@ -292,50 +369,106 @@ void __init init_memory_mapping(unsigned long start, unsigned long end) for (; start < end; start = next) { int map; unsigned long pud_phys; - pud_t *pud = alloc_low_page(&map, &pud_phys); + pgd_t *pgd = pgd_offset_k(start); + pud_t *pud; + + if (after_bootmem) + pud = pud_offset_k(pgd, start & PGDIR_MASK); + else + pud = alloc_low_page(&map, &pud_phys); + next = start + PGDIR_SIZE; if (next > end) next = end; phys_pud_init(pud, __pa(start), __pa(next)); - set_pgd(pgd_offset_k(start), mk_kernel_pgd(pud_phys)); + if (!after_bootmem) + set_pgd(pgd_offset_k(start), mk_kernel_pgd(pud_phys)); unmap_low_page(map); } - asm volatile("movq %%cr4,%0" : "=r" (mmu_cr4_features)); + if (!after_bootmem) + asm volatile("movq %%cr4,%0" : "=r" (mmu_cr4_features)); __flush_tlb_all(); - early_printk("kernel direct mapping tables upto %lx @ %lx-%lx\n", end, - table_start< end_pfn) + dma32_pfn = end_pfn; + z[ZONE_DMA32] = dma32_pfn - start_pfn; + } + z[ZONE_NORMAL] = end_pfn - start_pfn; + + /* Remove lower zones from higher ones. */ + w = 0; + for (i = 0; i < MAX_NR_ZONES; i++) { + if (z[i]) + z[i] -= w; + w += z[i]; + } - max_dma = virt_to_phys((char *)MAX_DMA_ADDRESS) >> PAGE_SHIFT; + /* Compute holes */ + w = start_pfn; + for (i = 0; i < MAX_NR_ZONES; i++) { + unsigned long s = w; + w += z[i]; + h[i] = e820_hole_size(s, w); + } - if (end_pfn < max_dma) - zones_size[ZONE_DMA] = end_pfn; - else { - zones_size[ZONE_DMA] = max_dma; - zones_size[ZONE_NORMAL] = end_pfn - max_dma; + /* Add the space pace needed for mem_map to the holes too. */ + for (i = 0; i < MAX_NR_ZONES; i++) + h[i] += (z[i] * sizeof(struct page)) / PAGE_SIZE; + + /* The 16MB DMA zone has the kernel and other misc mappings. + Account them too */ + if (h[ZONE_DMA]) { + h[ZONE_DMA] += dma_reserve; + if (h[ZONE_DMA] >= z[ZONE_DMA]) { + printk(KERN_WARNING + "Kernel too large and filling up ZONE_DMA?\n"); + h[ZONE_DMA] = z[ZONE_DMA]; } - free_area_init(zones_size); } - return; +} + +#ifndef CONFIG_NUMA +void __init paging_init(void) +{ + unsigned long zones[MAX_NR_ZONES], holes[MAX_NR_ZONES]; + + memory_present(0, 0, end_pfn); + sparse_init(); + size_zones(zones, holes, 0, end_pfn); + free_area_init_node(0, NODE_DATA(0), zones, + __pa(PAGE_OFFSET) >> PAGE_SHIFT, holes); } #endif @@ -383,9 +516,9 @@ static inline int page_is_ram (unsigned long pagenr) if (e820.map[i].type != E820_RAM) /* not usable memory */ continue; /* - * !!!FIXME!!! Some BIOSen report areas as RAM that - * are not. Notably the 640->1Mb area. We need a sanity - * check here. + * !!!FIXME!!! Some BIOSen report areas as RAM that + * are not. Notably the 640->1Mb area. We need a sanity + * check here. */ addr = (e820.map[i].addr+PAGE_SIZE-1) >> PAGE_SHIFT; end = (e820.map[i].addr+e820.map[i].size) >> PAGE_SHIFT; @@ -395,25 +528,114 @@ static inline int page_is_ram (unsigned long pagenr) return 0; } -extern int swiotlb_force; +/* + * Memory hotplug specific functions + */ +#if defined(CONFIG_ACPI_HOTPLUG_MEMORY) || defined(CONFIG_ACPI_HOTPLUG_MEMORY_MODULE) + +void online_page(struct page *page) +{ + ClearPageReserved(page); + init_page_count(page); + __free_page(page); + totalram_pages++; + num_physpages++; +} + +#ifndef CONFIG_MEMORY_HOTPLUG +/* + * Memory Hotadd without sparsemem. The mem_maps have been allocated in advance, + * just online the pages. + */ +int __add_pages(struct zone *z, unsigned long start_pfn, unsigned long nr_pages) +{ + int err = -EIO; + unsigned long pfn; + unsigned long total = 0, mem = 0; + for (pfn = start_pfn; pfn < start_pfn + nr_pages; pfn++) { + if (pfn_valid(pfn)) { + online_page(pfn_to_page(pfn)); + err = 0; + mem++; + } + total++; + } + if (!err) { + z->spanned_pages += total; + z->present_pages += mem; + z->zone_pgdat->node_spanned_pages += total; + z->zone_pgdat->node_present_pages += mem; + } + return err; +} +#endif + +/* + * Memory is added always to NORMAL zone. This means you will never get + * additional DMA/DMA32 memory. + */ +int add_memory(u64 start, u64 size) +{ + struct pglist_data *pgdat = NODE_DATA(0); + struct zone *zone = pgdat->node_zones + MAX_NR_ZONES-2; + unsigned long start_pfn = start >> PAGE_SHIFT; + unsigned long nr_pages = size >> PAGE_SHIFT; + int ret; + + ret = __add_pages(zone, start_pfn, nr_pages); + if (ret) + goto error; + + init_memory_mapping(start, (start + size -1)); + + return ret; +error: + printk("%s: Problem encountered in __add_pages!\n", __func__); + return ret; +} +EXPORT_SYMBOL_GPL(add_memory); + +int remove_memory(u64 start, u64 size) +{ + return -EINVAL; +} +EXPORT_SYMBOL_GPL(remove_memory); + +#endif + +/* + * devmem_is_allowed() checks to see if /dev/mem access to a certain address is + * valid. The argument is a physical page number. + * + * + * On x86-64, access has to be given to the first megabyte of ram because that area + * contains bios code and data regions used by X and dosemu and similar apps. + * Access has to be given to non-kernel-ram areas as well, these contain the PCI + * mmio resources as well as potential bios/acpi data regions. + */ +int devmem_is_allowed(unsigned long pagenr) +{ + if (pagenr <= 256) + return 1; + if (!page_is_ram(pagenr)) + return 1; + return 0; +} + + +EXPORT_SYMBOL_GPL(page_is_ram); static struct kcore_list kcore_mem, kcore_vmalloc, kcore_kernel, kcore_modules, kcore_vsyscall; void __init mem_init(void) { - int codesize, reservedpages, datasize, initsize; - int tmp; + long codesize, reservedpages, datasize, initsize; #ifdef CONFIG_SWIOTLB - if (swiotlb_force) - swiotlb = 1; - if (!iommu_aperture && - (end_pfn >= 0xffffffff>>PAGE_SHIFT || force_iommu)) - swiotlb = 1; - if (swiotlb) - swiotlb_init(); + pci_swiotlb_init(); #endif + no_iommu_init(); /* How many end-of-memory variables you have, grandma! */ max_low_pfn = end_pfn; @@ -427,23 +649,12 @@ void __init mem_init(void) reservedpages = 0; /* this will put all low memory onto the freelists */ -#ifdef CONFIG_DISCONTIGMEM - totalram_pages += numa_free_all_bootmem(); - tmp = 0; - /* should count reserved pages here for all nodes */ +#ifdef CONFIG_NUMA + totalram_pages = numa_free_all_bootmem(); #else - max_mapnr = end_pfn; - if (!mem_map) BUG(); - - totalram_pages += free_all_bootmem(); - - for (tmp = 0; tmp < end_pfn; tmp++) - /* - * Only count reserved RAM pages - */ - if (page_is_ram(tmp) && PageReserved(pfn_to_page(tmp))) - reservedpages++; + totalram_pages = free_all_bootmem(); #endif + reservedpages = end_pfn - totalram_pages - e820_hole_size(0, end_pfn); after_bootmem = 1; @@ -460,7 +671,7 @@ void __init mem_init(void) kclist_add(&kcore_vsyscall, (void *)VSYSCALL_START, VSYSCALL_END - VSYSCALL_START); - printk("Memory: %luk/%luk available (%dk kernel code, %dk reserved, %dk data, %dk init)\n", + printk("Memory: %luk/%luk available (%ldk kernel code, %ldk reserved, %ldk data, %ldk init)\n", (unsigned long) nr_free_pages() << (PAGE_SHIFT-10), end_pfn << (PAGE_SHIFT-10), codesize >> 10, @@ -468,19 +679,16 @@ void __init mem_init(void) datasize >> 10, initsize >> 10); +#ifdef CONFIG_SMP /* - * Subtle. SMP is doing its boot stuff late (because it has to - * fork idle threads) - but it also needs low mappings for the - * protected-mode entry to work. We zap these entries only after - * the WP-bit has been tested. + * Sync boot_level4_pgt mappings with the init_level4_pgt + * except for the low identity mappings which are already zapped + * in init_level4_pgt. This sync-up is essential for AP's bringup */ -#ifndef CONFIG_SMP - zap_low_mappings(); + memcpy(boot_level4_pgt+1, init_level4_pgt+1, (PTRS_PER_PGD-1)*sizeof(pgd_t)); #endif } -extern char __initdata_begin[], __initdata_end[]; - void free_initmem(void) { unsigned long addr; @@ -488,24 +696,47 @@ void free_initmem(void) addr = (unsigned long)(&__init_begin); for (; addr < (unsigned long)(&__init_end); addr += PAGE_SIZE) { ClearPageReserved(virt_to_page(addr)); - set_page_count(virt_to_page(addr), 1); + init_page_count(virt_to_page(addr)); memset((void *)(addr & ~(PAGE_SIZE-1)), 0xcc, PAGE_SIZE); free_page(addr); totalram_pages++; } memset(__initdata_begin, 0xba, __initdata_end - __initdata_begin); - printk ("Freeing unused kernel memory: %luk freed\n", (&__init_end - &__init_begin) >> 10); + printk ("Freeing unused kernel memory: %luk freed\n", (__init_end - __init_begin) >> 10); +} + +#ifdef CONFIG_DEBUG_RODATA + +extern char __start_rodata, __end_rodata; +void mark_rodata_ro(void) +{ + unsigned long addr = (unsigned long)&__start_rodata; + + for (; addr < (unsigned long)&__end_rodata; addr += PAGE_SIZE) + change_page_attr_addr(addr, 1, PAGE_KERNEL_RO); + + printk ("Write protecting the kernel read-only data: %luk\n", + (&__end_rodata - &__start_rodata) >> 10); + + /* + * change_page_attr_addr() requires a global_flush_tlb() call after it. + * We do this after the printk so that if something went wrong in the + * change, the printk gets out at least to give a better debug hint + * of who is the culprit. + */ + global_flush_tlb(); } +#endif #ifdef CONFIG_BLK_DEV_INITRD void free_initrd_mem(unsigned long start, unsigned long end) { - if (start < (unsigned long)&_end) + if (start >= end) return; printk ("Freeing initrd memory: %ldk freed\n", (end - start) >> 10); for (; start < end; start += PAGE_SIZE) { ClearPageReserved(virt_to_page(start)); - set_page_count(virt_to_page(start), 1); + init_page_count(virt_to_page(start)); free_page(start); totalram_pages++; } @@ -515,12 +746,14 @@ void free_initrd_mem(unsigned long start, unsigned long end) void __init reserve_bootmem_generic(unsigned long phys, unsigned len) { /* Should check here against the e820 map to avoid double free */ -#ifdef CONFIG_DISCONTIGMEM +#ifdef CONFIG_NUMA int nid = phys_to_nid(phys); reserve_bootmem_node(NODE_DATA(nid), phys, len); #else reserve_bootmem(phys, len); #endif + if (phys+len <= MAX_DMA_PFN*PAGE_SIZE) + dma_reserve += len / PAGE_SIZE; } int kern_addr_valid(unsigned long addr) @@ -562,10 +795,6 @@ extern int exception_trace, page_fault_trace; static ctl_table debug_table2[] = { { 99, "exception-trace", &exception_trace, sizeof(int), 0644, NULL, proc_dointvec }, -#ifdef CONFIG_CHECKING - { 100, "page-fault-trace", &page_fault_trace, sizeof(int), 0644, NULL, - proc_dointvec }, -#endif { 0, } };