From e009d5dc0a94a7133e5f1c083732d760bfd038e6 Mon Sep 17 00:00:00 2001 From: Michal Hocko Date: Thu, 12 Mar 2015 16:25:52 -0700 Subject: mm, oom: do not fail __GFP_NOFAIL allocation if oom killer is disabled Tetsuo Handa has pointed out that __GFP_NOFAIL allocations might fail after OOM killer is disabled if the allocation is performed by a kernel thread. This behavior was introduced from the very beginning by 7f33d49a2ed5 ("mm, PM/Freezer: Disable OOM killer when tasks are frozen"). This means that the basic contract for the allocation request is broken and the context requesting such an allocation might blow up unexpectedly. There are basically two ways forward. 1) move oom_killer_disable after kernel threads are frozen. This has a risk that the OOM victim wouldn't be able to finish because it would depend on an already frozen kernel thread. This would be really tricky to debug. 2) do not fail GFP_NOFAIL allocation no matter what and risk a potential Freezable kernel threads will loop and fail the suspend. Incidental allocations after kernel threads are frozen will at least dump a warning - if we are lucky and the serial console is still active of course... This patch implements the later option because it is safer. We would see warning rather than allocation failures for the kernel threads which would blow up otherwise and have a higher chances to identify __GFP_NOFAIL users from deeper pm code. Signed-off-by: Michal Hocko Acked-by: David Rientjes Cc: Johannes Weiner Cc: Tetsuo Handa Cc: "Rafael J. Wysocki" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/page_alloc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'mm') diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 7abfa70cdc1a..40e29429e7b0 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -2373,7 +2373,8 @@ __alloc_pages_may_oom(gfp_t gfp_mask, unsigned int order, goto out; } /* Exhausted what can be done so it's blamo time */ - if (out_of_memory(ac->zonelist, gfp_mask, order, ac->nodemask, false)) + if (out_of_memory(ac->zonelist, gfp_mask, order, ac->nodemask, false) + || WARN_ON_ONCE(gfp_mask & __GFP_NOFAIL)) *did_some_progress = 1; out: oom_zonelist_unlock(ac->zonelist, gfp_mask); -- cgit v1.2.3-58-ga151 From 44fc80573cc760a7154f41fd0a958ee10eba1a81 Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Thu, 12 Mar 2015 16:25:54 -0700 Subject: mm, hugetlb: close race when setting PageTail for gigantic pages Now that gigantic pages are dynamically allocatable, care must be taken to ensure that p->first_page is valid before setting PageTail. If this isn't done, then it is possible to race and have compound_head() return NULL. Signed-off-by: David Rientjes Acked-by: Davidlohr Bueso Cc: Luiz Capitulino Cc: Joonsoo Kim Acked-by: Hillf Danton Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/hugetlb.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'mm') diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 0a9ac6c26832..c41b2a0ee273 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -917,7 +917,6 @@ static void prep_compound_gigantic_page(struct page *page, unsigned long order) __SetPageHead(page); __ClearPageReserved(page); for (i = 1; i < nr_pages; i++, p = mem_map_next(p, page, i)) { - __SetPageTail(p); /* * For gigantic hugepages allocated through bootmem at * boot, it's safer to be consistent with the not-gigantic @@ -933,6 +932,9 @@ static void prep_compound_gigantic_page(struct page *page, unsigned long order) __ClearPageReserved(p); set_page_count(p, 0); p->first_page = page; + /* Make sure p->first_page is always valid for PageTail() */ + smp_wmb(); + __SetPageTail(p); } } -- cgit v1.2.3-58-ga151 From 850fc430f47aad52092deaaeb32b99f97f0e6aca Mon Sep 17 00:00:00 2001 From: Danesh Petigara Date: Thu, 12 Mar 2015 16:25:57 -0700 Subject: mm: cma: fix CMA aligned offset calculation The CMA aligned offset calculation is incorrect for non-zero order_per_bit values. For example, if cma->order_per_bit=1, cma->base_pfn= 0x2f800000 and align_order=12, the function returns a value of 0x17c00 instead of 0x400. This patch fixes the CMA aligned offset calculation. The previous calculation was wrong and would return too-large values for the offset, so that when cma_alloc looks for free pages in the bitmap with the requested alignment > order_per_bit, it starts too far into the bitmap and so CMA allocations will fail despite there actually being plenty of free pages remaining. It will also probably have the wrong alignment. With this change, we will get the correct offset into the bitmap. One affected user is powerpc KVM, which has kvm_cma->order_per_bit set to KVM_CMA_CHUNK_ORDER - PAGE_SHIFT, or 18 - 12 = 6. [gregory.0xf0@gmail.com: changelog additions] Signed-off-by: Danesh Petigara Reviewed-by: Gregory Fong Acked-by: Michal Nazarewicz Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/cma.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) (limited to 'mm') diff --git a/mm/cma.c b/mm/cma.c index 75016fd1de90..68ecb7a42983 100644 --- a/mm/cma.c +++ b/mm/cma.c @@ -64,15 +64,17 @@ static unsigned long cma_bitmap_aligned_mask(struct cma *cma, int align_order) return (1UL << (align_order - cma->order_per_bit)) - 1; } +/* + * Find a PFN aligned to the specified order and return an offset represented in + * order_per_bits. + */ static unsigned long cma_bitmap_aligned_offset(struct cma *cma, int align_order) { - unsigned int alignment; - if (align_order <= cma->order_per_bit) return 0; - alignment = 1UL << (align_order - cma->order_per_bit); - return ALIGN(cma->base_pfn, alignment) - - (cma->base_pfn >> cma->order_per_bit); + + return (ALIGN(cma->base_pfn, (1UL << align_order)) + - cma->base_pfn) >> cma->order_per_bit; } static unsigned long cma_bitmap_maxno(struct cma *cma) -- cgit v1.2.3-58-ga151 From 5b8bf30721980b254be7a07315c353b3a3175b74 Mon Sep 17 00:00:00 2001 From: gchen gchen Date: Thu, 12 Mar 2015 16:26:05 -0700 Subject: mm/nommu.c: export symbol max_mapnr Several modules may need max_mapnr, so export, the related error with allmodconfig under c6x: MODPOST 3327 modules ERROR: "max_mapnr" [fs/pstore/ramoops.ko] undefined! ERROR: "max_mapnr" [drivers/media/v4l2-core/videobuf2-dma-contig.ko] undefined! Signed-off-by: Chen Gang Cc: Mark Salter Cc: Aurelien Jacquiot Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/nommu.c | 1 + 1 file changed, 1 insertion(+) (limited to 'mm') diff --git a/mm/nommu.c b/mm/nommu.c index 3e67e7538ecf..3fba2dc97c44 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -62,6 +62,7 @@ void *high_memory; EXPORT_SYMBOL(high_memory); struct page *mem_map; unsigned long max_mapnr; +EXPORT_SYMBOL(max_mapnr); unsigned long highest_memmap_pfn; struct percpu_counter vm_committed_as; int sysctl_overcommit_memory = OVERCOMMIT_GUESS; /* heuristic overcommit */ -- cgit v1.2.3-58-ga151 From a5af5aa8b67dfdba36c853b70564fd2dfe73d478 Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Thu, 12 Mar 2015 16:26:11 -0700 Subject: kasan, module, vmalloc: rework shadow allocation for modules Current approach in handling shadow memory for modules is broken. Shadow memory could be freed only after memory shadow corresponds it is no longer used. vfree() called from interrupt context could use memory its freeing to store 'struct llist_node' in it: void vfree(const void *addr) { ... if (unlikely(in_interrupt())) { struct vfree_deferred *p = this_cpu_ptr(&vfree_deferred); if (llist_add((struct llist_node *)addr, &p->list)) schedule_work(&p->wq); Later this list node used in free_work() which actually frees memory. Currently module_memfree() called in interrupt context will free shadow before freeing module's memory which could provoke kernel crash. So shadow memory should be freed after module's memory. However, such deallocation order could race with kasan_module_alloc() in module_alloc(). Free shadow right before releasing vm area. At this point vfree()'d memory is not used anymore and yet not available for other allocations. New VM_KASAN flag used to indicate that vm area has dynamically allocated shadow memory so kasan frees shadow only if it was previously allocated. Signed-off-by: Andrey Ryabinin Acked-by: Rusty Russell Cc: Dmitry Vyukov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/kasan.h | 5 +++-- include/linux/vmalloc.h | 1 + kernel/module.c | 2 -- mm/kasan/kasan.c | 14 +++++++++++--- mm/vmalloc.c | 1 + 5 files changed, 16 insertions(+), 7 deletions(-) (limited to 'mm') diff --git a/include/linux/kasan.h b/include/linux/kasan.h index 72ba725ddf9c..5fa48a21d73e 100644 --- a/include/linux/kasan.h +++ b/include/linux/kasan.h @@ -5,6 +5,7 @@ struct kmem_cache; struct page; +struct vm_struct; #ifdef CONFIG_KASAN @@ -52,7 +53,7 @@ void kasan_slab_free(struct kmem_cache *s, void *object); #define MODULE_ALIGN (PAGE_SIZE << KASAN_SHADOW_SCALE_SHIFT) int kasan_module_alloc(void *addr, size_t size); -void kasan_module_free(void *addr); +void kasan_free_shadow(const struct vm_struct *vm); #else /* CONFIG_KASAN */ @@ -82,7 +83,7 @@ static inline void kasan_slab_alloc(struct kmem_cache *s, void *object) {} static inline void kasan_slab_free(struct kmem_cache *s, void *object) {} static inline int kasan_module_alloc(void *addr, size_t size) { return 0; } -static inline void kasan_module_free(void *addr) {} +static inline void kasan_free_shadow(const struct vm_struct *vm) {} #endif /* CONFIG_KASAN */ diff --git a/include/linux/vmalloc.h b/include/linux/vmalloc.h index 7d7acb35603d..0ec598381f97 100644 --- a/include/linux/vmalloc.h +++ b/include/linux/vmalloc.h @@ -17,6 +17,7 @@ struct vm_area_struct; /* vma defining user mapping in mm_types.h */ #define VM_VPAGES 0x00000010 /* buffer for pages was vmalloc'ed */ #define VM_UNINITIALIZED 0x00000020 /* vm_struct is not fully initialized */ #define VM_NO_GUARD 0x00000040 /* don't add guard page */ +#define VM_KASAN 0x00000080 /* has allocated kasan shadow memory */ /* bits [20..32] reserved for arch specific ioremap internals */ /* diff --git a/kernel/module.c b/kernel/module.c index cc93cf68653c..b3d634ed06c9 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -56,7 +56,6 @@ #include #include #include -#include #include #include #include @@ -1814,7 +1813,6 @@ static void unset_module_init_ro_nx(struct module *mod) { } void __weak module_memfree(void *module_region) { vfree(module_region); - kasan_module_free(module_region); } void __weak module_arch_cleanup(struct module *mod) diff --git a/mm/kasan/kasan.c b/mm/kasan/kasan.c index 78fee632a7ee..936d81661c47 100644 --- a/mm/kasan/kasan.c +++ b/mm/kasan/kasan.c @@ -29,6 +29,7 @@ #include #include #include +#include #include #include "kasan.h" @@ -414,12 +415,19 @@ int kasan_module_alloc(void *addr, size_t size) GFP_KERNEL | __GFP_HIGHMEM | __GFP_ZERO, PAGE_KERNEL, VM_NO_GUARD, NUMA_NO_NODE, __builtin_return_address(0)); - return ret ? 0 : -ENOMEM; + + if (ret) { + find_vm_area(addr)->flags |= VM_KASAN; + return 0; + } + + return -ENOMEM; } -void kasan_module_free(void *addr) +void kasan_free_shadow(const struct vm_struct *vm) { - vfree(kasan_mem_to_shadow(addr)); + if (vm->flags & VM_KASAN) + vfree(kasan_mem_to_shadow(vm->addr)); } static void register_global(struct kasan_global *global) diff --git a/mm/vmalloc.c b/mm/vmalloc.c index 35b25e1340ca..49abccf29a29 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -1418,6 +1418,7 @@ struct vm_struct *remove_vm_area(const void *addr) spin_unlock(&vmap_area_lock); vmap_debug_free_range(va->va_start, va->va_end); + kasan_free_shadow(vm); free_unmap_vmap_area(va); vm->size -= PAGE_SIZE; -- cgit v1.2.3-58-ga151 From a5a6579db33af91f4f5134e14be758dc71c1b694 Mon Sep 17 00:00:00 2001 From: Jeff Vander Stoep Date: Thu, 12 Mar 2015 16:26:17 -0700 Subject: mm: reorder can_do_mlock to fix audit denial A userspace call to mmap(MAP_LOCKED) may result in the successful locking of memory while also producing a confusing audit log denial. can_do_mlock checks capable and rlimit. If either of these return positive can_do_mlock returns true. The capable check leads to an LSM hook used by apparmour and selinux which produce the audit denial. Reordering so rlimit is checked first eliminates the denial on success, only recording a denial when the lock is unsuccessful as a result of the denial. Signed-off-by: Jeff Vander Stoep Acked-by: Nick Kralevich Cc: Jeff Vander Stoep Cc: Sasha Levin Cc: "Paul E. McKenney" Cc: Rik van Riel Cc: Vlastimil Babka Cc: Paul Cassella Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/mlock.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'mm') diff --git a/mm/mlock.c b/mm/mlock.c index 73cf0987088c..8a54cd214925 100644 --- a/mm/mlock.c +++ b/mm/mlock.c @@ -26,10 +26,10 @@ int can_do_mlock(void) { - if (capable(CAP_IPC_LOCK)) - return 1; if (rlimit(RLIMIT_MEMLOCK) != 0) return 1; + if (capable(CAP_IPC_LOCK)) + return 1; return 0; } EXPORT_SYMBOL(can_do_mlock); -- cgit v1.2.3-58-ga151 From 7feee590bb18ffc42636975f74c2c3120ce1901c Mon Sep 17 00:00:00 2001 From: Vladimir Davydov Date: Thu, 12 Mar 2015 16:26:19 -0700 Subject: memcg: disable hierarchy support if bound to the legacy cgroup hierarchy If the memory cgroup controller is initially mounted in the scope of the default cgroup hierarchy and then remounted to a legacy hierarchy, it will still have hierarchy support enabled, which is incorrect. We should disable hierarchy support if bound to the legacy cgroup hierarchy. Signed-off-by: Vladimir Davydov Signed-off-by: Johannes Weiner Acked-by: Michal Hocko Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/memcontrol.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'mm') diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 9fe07692eaad..b34ef4a32a3b 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -5232,7 +5232,9 @@ static void mem_cgroup_bind(struct cgroup_subsys_state *root_css) * on for the root memcg is enough. */ if (cgroup_on_dfl(root_css->cgroup)) - mem_cgroup_from_css(root_css)->use_hierarchy = true; + root_mem_cgroup->use_hierarchy = true; + else + root_mem_cgroup->use_hierarchy = false; } static u64 memory_current_read(struct cgroup_subsys_state *css, -- cgit v1.2.3-58-ga151