From 596b711ed6b5235f8545680ef38ace00f9898c32 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Sun, 28 Mar 2010 19:42:54 -0700 Subject: x86: Make smp_locks end with page alignment Fix: ------------[ cut here ]------------ WARNING: at arch/x86/mm/init.c:342 free_init_pages+0x4c/0xfa() free_init_pages: range [0x40daf000, 0x40db5c24] is not aligned Modules linked in: Pid: 0, comm: swapper Not tainted 2.6.34-rc2-tip-03946-g4f16b23-dirty #50 Call Trace: [<40232e9f>] warn_slowpath_common+0x65/0x7c [<4021c9f0>] ? free_init_pages+0x4c/0xfa [<40881434>] ? _etext+0x0/0x24 [<40232eea>] warn_slowpath_fmt+0x24/0x27 [<4021c9f0>] free_init_pages+0x4c/0xfa [<40881434>] ? _etext+0x0/0x24 [<40d3f4bd>] alternative_instructions+0xf6/0x100 [<40d3fe4f>] check_bugs+0xbd/0xbf [<40d398a7>] start_kernel+0x2d5/0x2e4 [<40d390ce>] i386_start_kernel+0xce/0xd5 ---[ end trace 4eaa2a86a8e2da22 ]--- Comments in vmlinux.lds.S already said: | /* | * smp_locks might be freed after init | * start/end must be page aligned | */ Signed-off-by: Yinghai Lu Acked-by: Johannes Weiner Cc: David Miller Cc: Benjamin Herrenschmidt Cc: Linus Torvalds LKML-Reference: <1269830604-26214-2-git-send-email-yinghai@kernel.org> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/vmlinux.lds.S b/arch/x86/kernel/vmlinux.lds.S index 44879df..2cc2497 100644 --- a/arch/x86/kernel/vmlinux.lds.S +++ b/arch/x86/kernel/vmlinux.lds.S @@ -291,8 +291,8 @@ SECTIONS .smp_locks : AT(ADDR(.smp_locks) - LOAD_OFFSET) { __smp_locks = .; *(.smp_locks) - __smp_locks_end = .; . = ALIGN(PAGE_SIZE); + __smp_locks_end = .; } #ifdef CONFIG_X86_64 -- cgit v0.10.2 From c967da6a0ba837f762042e931d4afcf72045547c Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Sun, 28 Mar 2010 19:42:55 -0700 Subject: x86: Make sure free_init_pages() frees pages on page boundary When CONFIG_NO_BOOTMEM=y, it could use memory more effiently, or in a more compact fashion. Example: Allocated new RAMDISK: 00ec2000 - 0248ce57 Move RAMDISK from 000000002ea04000 - 000000002ffcee56 to 00ec2000 - 0248ce56 The new RAMDISK's end is not page aligned. Last page could be shared with other users. When free_init_pages are called for initrd or .init, the page could be freed and we could corrupt other data. code segment in free_init_pages(): | for (; addr < end; addr += PAGE_SIZE) { | ClearPageReserved(virt_to_page(addr)); | init_page_count(virt_to_page(addr)); | memset((void *)(addr & ~(PAGE_SIZE-1)), | POISON_FREE_INITMEM, PAGE_SIZE); | free_page(addr); | totalram_pages++; | } last half page could be used as one whole free page. So page align the boundaries. -v2: make the original initramdisk to be aligned, according to Johannes, otherwise we have the chance to lose one page. we still need to keep initrd_end not aligned, otherwise it could confuse decompressor. -v3: change to WARN_ON instead, suggested by Johannes. -v4: use PAGE_ALIGN, suggested by Johannes. We may fix that macro name later to PAGE_ALIGN_UP, and PAGE_ALIGN_DOWN Add comments about assuming ramdisk start is aligned in relocate_initrd(), change to re get ramdisk_image instead of save it to make diff smaller. Add warning for wrong range, suggested by Johannes. -v6: remove one WARN() We need to align beginning in free_init_pages() do not copy more than ramdisk_size, noticed by Johannes Reported-by: Stanislaw Gruszka Tested-by: Stanislaw Gruszka Signed-off-by: Yinghai Lu Acked-by: Johannes Weiner Cc: David Miller Cc: Benjamin Herrenschmidt Cc: Linus Torvalds LKML-Reference: <1269830604-26214-3-git-send-email-yinghai@kernel.org> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/head32.c b/arch/x86/kernel/head32.c index adedeef..b2e2460 100644 --- a/arch/x86/kernel/head32.c +++ b/arch/x86/kernel/head32.c @@ -7,6 +7,7 @@ #include #include +#include #include #include @@ -44,9 +45,10 @@ void __init i386_start_kernel(void) #ifdef CONFIG_BLK_DEV_INITRD /* Reserve INITRD */ if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) { + /* Assume only end is not page aligned */ u64 ramdisk_image = boot_params.hdr.ramdisk_image; u64 ramdisk_size = boot_params.hdr.ramdisk_size; - u64 ramdisk_end = ramdisk_image + ramdisk_size; + u64 ramdisk_end = PAGE_ALIGN(ramdisk_image + ramdisk_size); reserve_early(ramdisk_image, ramdisk_end, "RAMDISK"); } #endif diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c index b5a9896..7147143 100644 --- a/arch/x86/kernel/head64.c +++ b/arch/x86/kernel/head64.c @@ -103,9 +103,10 @@ void __init x86_64_start_reservations(char *real_mode_data) #ifdef CONFIG_BLK_DEV_INITRD /* Reserve INITRD */ if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) { + /* Assume only end is not page aligned */ unsigned long ramdisk_image = boot_params.hdr.ramdisk_image; unsigned long ramdisk_size = boot_params.hdr.ramdisk_size; - unsigned long ramdisk_end = ramdisk_image + ramdisk_size; + unsigned long ramdisk_end = PAGE_ALIGN(ramdisk_image + ramdisk_size); reserve_early(ramdisk_image, ramdisk_end, "RAMDISK"); } #endif diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index 5d7ba1a..d76e185 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -314,16 +314,17 @@ static void __init reserve_brk(void) #define MAX_MAP_CHUNK (NR_FIX_BTMAPS << PAGE_SHIFT) static void __init relocate_initrd(void) { - + /* Assume only end is not page aligned */ u64 ramdisk_image = boot_params.hdr.ramdisk_image; u64 ramdisk_size = boot_params.hdr.ramdisk_size; + u64 area_size = PAGE_ALIGN(ramdisk_size); u64 end_of_lowmem = max_low_pfn_mapped << PAGE_SHIFT; u64 ramdisk_here; unsigned long slop, clen, mapaddr; char *p, *q; /* We need to move the initrd down into lowmem */ - ramdisk_here = find_e820_area(0, end_of_lowmem, ramdisk_size, + ramdisk_here = find_e820_area(0, end_of_lowmem, area_size, PAGE_SIZE); if (ramdisk_here == -1ULL) @@ -332,7 +333,7 @@ static void __init relocate_initrd(void) /* Note: this includes all the lowmem currently occupied by the initrd, we rely on that fact to keep the data intact. */ - reserve_early(ramdisk_here, ramdisk_here + ramdisk_size, + reserve_early(ramdisk_here, ramdisk_here + area_size, "NEW RAMDISK"); initrd_start = ramdisk_here + PAGE_OFFSET; initrd_end = initrd_start + ramdisk_size; @@ -376,9 +377,10 @@ static void __init relocate_initrd(void) static void __init reserve_initrd(void) { + /* Assume only end is not page aligned */ u64 ramdisk_image = boot_params.hdr.ramdisk_image; u64 ramdisk_size = boot_params.hdr.ramdisk_size; - u64 ramdisk_end = ramdisk_image + ramdisk_size; + u64 ramdisk_end = PAGE_ALIGN(ramdisk_image + ramdisk_size); u64 end_of_lowmem = max_low_pfn_mapped << PAGE_SHIFT; if (!boot_params.hdr.type_of_loader || diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c index e71c5cb..452ee5b 100644 --- a/arch/x86/mm/init.c +++ b/arch/x86/mm/init.c @@ -331,11 +331,23 @@ int devmem_is_allowed(unsigned long pagenr) void free_init_pages(char *what, unsigned long begin, unsigned long end) { - unsigned long addr = begin; + unsigned long addr; + unsigned long begin_aligned, end_aligned; - if (addr >= end) + /* Make sure boundaries are page aligned */ + begin_aligned = PAGE_ALIGN(begin); + end_aligned = end & PAGE_MASK; + + if (WARN_ON(begin_aligned != begin || end_aligned != end)) { + begin = begin_aligned; + end = end_aligned; + } + + if (begin >= end) return; + addr = begin; + /* * If debugging page accesses then do not free this memory but * mark them not present - any buggy init-section access will @@ -343,7 +355,7 @@ void free_init_pages(char *what, unsigned long begin, unsigned long end) */ #ifdef CONFIG_DEBUG_PAGEALLOC printk(KERN_INFO "debug: unmapping init memory %08lx..%08lx\n", - begin, PAGE_ALIGN(end)); + begin, end); set_memory_np(begin, (end - begin) >> PAGE_SHIFT); #else /* @@ -358,8 +370,7 @@ void free_init_pages(char *what, unsigned long begin, unsigned long end) for (; addr < end; addr += PAGE_SIZE) { ClearPageReserved(virt_to_page(addr)); init_page_count(virt_to_page(addr)); - memset((void *)(addr & ~(PAGE_SIZE-1)), - POISON_FREE_INITMEM, PAGE_SIZE); + memset((void *)addr, POISON_FREE_INITMEM, PAGE_SIZE); free_page(addr); totalram_pages++; } @@ -376,6 +387,15 @@ void free_initmem(void) #ifdef CONFIG_BLK_DEV_INITRD void free_initrd_mem(unsigned long start, unsigned long end) { - free_init_pages("initrd memory", start, end); + /* + * end could be not aligned, and We can not align that, + * decompresser could be confused by aligned initrd_end + * We already reserve the end partial page before in + * - i386_start_kernel() + * - x86_64_start_kernel() + * - relocate_initrd() + * So here We can do PAGE_ALIGN() safely to get partial page to be freed + */ + free_init_pages("initrd memory", start, PAGE_ALIGN(end)); } #endif -- cgit v0.10.2 From eed63519e3e74d515d2007ecd895338d0ba2a85c Mon Sep 17 00:00:00 2001 From: Ian Campbell Date: Sun, 28 Mar 2010 19:42:56 -0700 Subject: x86: Do not free zero sized per cpu areas This avoids an infinite loop in free_early_partial(). Add a warning to free_early_partial() to catch future problems. -v5: put back start > end back into WARN_ONCE() -v6: use one line for warning, suggested by Linus -v7: more tests -v8: remove the function name as suggested by Johannes WARN_ONCE() will print out that function name. Signed-off-by: Ian Campbell Signed-off-by: Yinghai Lu Tested-by: Konrad Rzeszutek Wilk Tested-by: Joel Becker Tested-by: Stanislaw Gruszka Acked-by: Johannes Weiner Cc: Peter Zijlstra Cc: David Miller Cc: Benjamin Herrenschmidt Cc: Linus Torvalds LKML-Reference: <1269830604-26214-4-git-send-email-yinghai@kernel.org> Signed-off-by: Ingo Molnar diff --git a/kernel/early_res.c b/kernel/early_res.c index 3cb2c66..31aa933 100644 --- a/kernel/early_res.c +++ b/kernel/early_res.c @@ -333,6 +333,12 @@ void __init free_early_partial(u64 start, u64 end) struct early_res *r; int i; + if (start == end) + return; + + if (WARN_ONCE(start > end, " wrong range [%#llx, %#llx]\n", start, end)) + return; + try_next: i = find_overlapped_early(start, end); if (i >= max_early_res) -- cgit v0.10.2