From 1cbdf60bd1b74e397d48aa877367cfc621f45ffe Mon Sep 17 00:00:00 2001 From: Peter Collingbourne Date: Wed, 26 May 2021 10:49:27 -0700 Subject: kasan: arm64: support specialized outlined tag mismatch checks By using outlined checks we can achieve a significant code size improvement by moving the tag-based ASAN checks into separate functions. Unlike the existing CONFIG_KASAN_OUTLINE mode these functions have a custom calling convention that preserves most registers and is specialized to the register containing the address and the type of access, and as a result we can eliminate the code size and performance overhead of a standard calling convention such as AAPCS for these functions. This change depends on a separate series of changes to Clang [1] to support outlined checks in the kernel, although the change works fine without them (we just don't get outlined checks). This is because the flag -mllvm -hwasan-inline-all-checks=0 has no effect until the Clang changes land. The flag was introduced in the Clang 9.0 timeframe as part of the support for outlined checks in userspace and because our minimum Clang version is 10.0 we can pass it unconditionally. Outlined checks require a new runtime function with a custom calling convention. Add this function to arch/arm64/lib. I measured the code size of defconfig + tag-based KASAN, as well as boot time (i.e. time to init launch) on a DragonBoard 845c with an Android arm64 GKI kernel. The results are below: code size boot time CONFIG_KASAN_INLINE=y before 92824064 6.18s CONFIG_KASAN_INLINE=y after 38822400 6.65s CONFIG_KASAN_OUTLINE=y 39215616 11.48s We can see straight away that specialized outlined checks beat the existing CONFIG_KASAN_OUTLINE=y on both code size and boot time for tag-based ASAN. As for the comparison between CONFIG_KASAN_INLINE=y before and after we saw similar performance numbers in userspace [2] and decided that since the performance overhead is minimal compared to the overhead of tag-based ASAN itself as well as compared to the code size improvements we would just replace the inlined checks with the specialized outlined checks without the option to select between them, and that is what I have implemented in this patch. Signed-off-by: Peter Collingbourne Acked-by: Andrey Konovalov Reviewed-by: Mark Rutland Tested-by: Mark Rutland Link: https://linux-review.googlesource.com/id/I1a30036c70ab3c3ee78d75ed9b87ef7cdc3fdb76 Link: [1] https://reviews.llvm.org/D90426 Link: [2] https://reviews.llvm.org/D56954 Link: https://lore.kernel.org/r/20210526174927.2477847-3-pcc@google.com Signed-off-by: Will Deacon --- mm/kasan/sw_tags.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'mm/kasan') diff --git a/mm/kasan/sw_tags.c b/mm/kasan/sw_tags.c index 9df8e7f69e87..9362938abbfa 100644 --- a/mm/kasan/sw_tags.c +++ b/mm/kasan/sw_tags.c @@ -207,3 +207,10 @@ struct kasan_track *kasan_get_free_track(struct kmem_cache *cache, return &alloc_meta->free_track[i]; } + +void kasan_tag_mismatch(unsigned long addr, unsigned long access_info, + unsigned long ret_ip) +{ + kasan_report(addr, 1 << (access_info & 0xf), access_info & 0x10, + ret_ip); +} -- cgit v1.2.3 From 7a3b835371883558eb63e069d891bd87f562380d Mon Sep 17 00:00:00 2001 From: Peter Collingbourne Date: Wed, 2 Jun 2021 16:52:28 -0700 Subject: kasan: use separate (un)poison implementation for integrated init Currently with integrated init page_alloc.c needs to know whether kasan_alloc_pages() will zero initialize memory, but this will start becoming more complicated once we start adding tag initialization support for user pages. To avoid page_alloc.c needing to know more details of what integrated init will do, move the unpoisoning logic for integrated init into the HW tags implementation. Currently the logic is identical but it will diverge in subsequent patches. For symmetry do the same for poisoning although this logic will be unaffected by subsequent patches. Signed-off-by: Peter Collingbourne Reviewed-by: Andrey Konovalov Link: https://linux-review.googlesource.com/id/I2c550234c6c4a893c48c18ff0c6ce658c7c67056 Link: https://lore.kernel.org/r/20210602235230.3928842-3-pcc@google.com Signed-off-by: Will Deacon --- include/linux/kasan.h | 64 ++++++++++++++++++++++++++++++--------------------- mm/kasan/common.c | 4 ++-- mm/kasan/hw_tags.c | 22 ++++++++++++++++++ mm/mempool.c | 6 +++-- mm/page_alloc.c | 55 ++++++++++++++++++++++--------------------- 5 files changed, 95 insertions(+), 56 deletions(-) (limited to 'mm/kasan') diff --git a/include/linux/kasan.h b/include/linux/kasan.h index b1678a61e6a7..a1c7ce5f3e4f 100644 --- a/include/linux/kasan.h +++ b/include/linux/kasan.h @@ -2,6 +2,7 @@ #ifndef _LINUX_KASAN_H #define _LINUX_KASAN_H +#include #include #include @@ -79,14 +80,6 @@ static inline void kasan_disable_current(void) {} #endif /* CONFIG_KASAN_GENERIC || CONFIG_KASAN_SW_TAGS */ -#ifdef CONFIG_KASAN - -struct kasan_cache { - int alloc_meta_offset; - int free_meta_offset; - bool is_kmalloc; -}; - #ifdef CONFIG_KASAN_HW_TAGS DECLARE_STATIC_KEY_FALSE(kasan_flag_enabled); @@ -101,11 +94,14 @@ static inline bool kasan_has_integrated_init(void) return kasan_enabled(); } +void kasan_alloc_pages(struct page *page, unsigned int order, gfp_t flags); +void kasan_free_pages(struct page *page, unsigned int order); + #else /* CONFIG_KASAN_HW_TAGS */ static inline bool kasan_enabled(void) { - return true; + return IS_ENABLED(CONFIG_KASAN); } static inline bool kasan_has_integrated_init(void) @@ -113,8 +109,30 @@ static inline bool kasan_has_integrated_init(void) return false; } +static __always_inline void kasan_alloc_pages(struct page *page, + unsigned int order, gfp_t flags) +{ + /* Only available for integrated init. */ + BUILD_BUG(); +} + +static __always_inline void kasan_free_pages(struct page *page, + unsigned int order) +{ + /* Only available for integrated init. */ + BUILD_BUG(); +} + #endif /* CONFIG_KASAN_HW_TAGS */ +#ifdef CONFIG_KASAN + +struct kasan_cache { + int alloc_meta_offset; + int free_meta_offset; + bool is_kmalloc; +}; + slab_flags_t __kasan_never_merge(void); static __always_inline slab_flags_t kasan_never_merge(void) { @@ -130,20 +148,20 @@ static __always_inline void kasan_unpoison_range(const void *addr, size_t size) __kasan_unpoison_range(addr, size); } -void __kasan_alloc_pages(struct page *page, unsigned int order, bool init); -static __always_inline void kasan_alloc_pages(struct page *page, +void __kasan_poison_pages(struct page *page, unsigned int order, bool init); +static __always_inline void kasan_poison_pages(struct page *page, unsigned int order, bool init) { if (kasan_enabled()) - __kasan_alloc_pages(page, order, init); + __kasan_poison_pages(page, order, init); } -void __kasan_free_pages(struct page *page, unsigned int order, bool init); -static __always_inline void kasan_free_pages(struct page *page, - unsigned int order, bool init) +void __kasan_unpoison_pages(struct page *page, unsigned int order, bool init); +static __always_inline void kasan_unpoison_pages(struct page *page, + unsigned int order, bool init) { if (kasan_enabled()) - __kasan_free_pages(page, order, init); + __kasan_unpoison_pages(page, order, init); } void __kasan_cache_create(struct kmem_cache *cache, unsigned int *size, @@ -285,21 +303,15 @@ void kasan_restore_multi_shot(bool enabled); #else /* CONFIG_KASAN */ -static inline bool kasan_enabled(void) -{ - return false; -} -static inline bool kasan_has_integrated_init(void) -{ - return false; -} static inline slab_flags_t kasan_never_merge(void) { return 0; } static inline void kasan_unpoison_range(const void *address, size_t size) {} -static inline void kasan_alloc_pages(struct page *page, unsigned int order, bool init) {} -static inline void kasan_free_pages(struct page *page, unsigned int order, bool init) {} +static inline void kasan_poison_pages(struct page *page, unsigned int order, + bool init) {} +static inline void kasan_unpoison_pages(struct page *page, unsigned int order, + bool init) {} static inline void kasan_cache_create(struct kmem_cache *cache, unsigned int *size, slab_flags_t *flags) {} diff --git a/mm/kasan/common.c b/mm/kasan/common.c index 6bb87f2acd4e..0ecd293af344 100644 --- a/mm/kasan/common.c +++ b/mm/kasan/common.c @@ -97,7 +97,7 @@ slab_flags_t __kasan_never_merge(void) return 0; } -void __kasan_alloc_pages(struct page *page, unsigned int order, bool init) +void __kasan_unpoison_pages(struct page *page, unsigned int order, bool init) { u8 tag; unsigned long i; @@ -111,7 +111,7 @@ void __kasan_alloc_pages(struct page *page, unsigned int order, bool init) kasan_unpoison(page_address(page), PAGE_SIZE << order, init); } -void __kasan_free_pages(struct page *page, unsigned int order, bool init) +void __kasan_poison_pages(struct page *page, unsigned int order, bool init) { if (likely(!PageHighMem(page))) kasan_poison(page_address(page), PAGE_SIZE << order, diff --git a/mm/kasan/hw_tags.c b/mm/kasan/hw_tags.c index 4004388b4e4b..9d0f6f934016 100644 --- a/mm/kasan/hw_tags.c +++ b/mm/kasan/hw_tags.c @@ -238,6 +238,28 @@ struct kasan_track *kasan_get_free_track(struct kmem_cache *cache, return &alloc_meta->free_track[0]; } +void kasan_alloc_pages(struct page *page, unsigned int order, gfp_t flags) +{ + /* + * This condition should match the one in post_alloc_hook() in + * page_alloc.c. + */ + bool init = !want_init_on_free() && want_init_on_alloc(flags); + + kasan_unpoison_pages(page, order, init); +} + +void kasan_free_pages(struct page *page, unsigned int order) +{ + /* + * This condition should match the one in free_pages_prepare() in + * page_alloc.c. + */ + bool init = want_init_on_free(); + + kasan_poison_pages(page, order, init); +} + #if IS_ENABLED(CONFIG_KASAN_KUNIT_TEST) void kasan_set_tagging_report_once(bool state) diff --git a/mm/mempool.c b/mm/mempool.c index a258cf4de575..0b8afbec3e35 100644 --- a/mm/mempool.c +++ b/mm/mempool.c @@ -106,7 +106,8 @@ static __always_inline void kasan_poison_element(mempool_t *pool, void *element) if (pool->alloc == mempool_alloc_slab || pool->alloc == mempool_kmalloc) kasan_slab_free_mempool(element); else if (pool->alloc == mempool_alloc_pages) - kasan_free_pages(element, (unsigned long)pool->pool_data, false); + kasan_poison_pages(element, (unsigned long)pool->pool_data, + false); } static void kasan_unpoison_element(mempool_t *pool, void *element) @@ -114,7 +115,8 @@ static void kasan_unpoison_element(mempool_t *pool, void *element) if (pool->alloc == mempool_alloc_slab || pool->alloc == mempool_kmalloc) kasan_unpoison_range(element, __ksize(element)); else if (pool->alloc == mempool_alloc_pages) - kasan_alloc_pages(element, (unsigned long)pool->pool_data, false); + kasan_unpoison_pages(element, (unsigned long)pool->pool_data, + false); } static __always_inline void add_element(mempool_t *pool, void *element) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index aaa1655cf682..4fddb7cac3c6 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -382,7 +382,7 @@ int page_group_by_mobility_disabled __read_mostly; static DEFINE_STATIC_KEY_TRUE(deferred_pages); /* - * Calling kasan_free_pages() only after deferred memory initialization + * Calling kasan_poison_pages() only after deferred memory initialization * has completed. Poisoning pages during deferred memory init will greatly * lengthen the process and cause problem in large memory systems as the * deferred pages initialization is done with interrupt disabled. @@ -394,15 +394,11 @@ static DEFINE_STATIC_KEY_TRUE(deferred_pages); * on-demand allocation and then freed again before the deferred pages * initialization is done, but this is not likely to happen. */ -static inline void kasan_free_nondeferred_pages(struct page *page, int order, - bool init, fpi_t fpi_flags) +static inline bool should_skip_kasan_poison(fpi_t fpi_flags) { - if (static_branch_unlikely(&deferred_pages)) - return; - if (!IS_ENABLED(CONFIG_KASAN_GENERIC) && - (fpi_flags & FPI_SKIP_KASAN_POISON)) - return; - kasan_free_pages(page, order, init); + return static_branch_unlikely(&deferred_pages) || + (!IS_ENABLED(CONFIG_KASAN_GENERIC) && + (fpi_flags & FPI_SKIP_KASAN_POISON)); } /* Returns true if the struct page for the pfn is uninitialised */ @@ -453,13 +449,10 @@ defer_init(int nid, unsigned long pfn, unsigned long end_pfn) return false; } #else -static inline void kasan_free_nondeferred_pages(struct page *page, int order, - bool init, fpi_t fpi_flags) +static inline bool should_skip_kasan_poison(fpi_t fpi_flags) { - if (!IS_ENABLED(CONFIG_KASAN_GENERIC) && - (fpi_flags & FPI_SKIP_KASAN_POISON)) - return; - kasan_free_pages(page, order, init); + return (!IS_ENABLED(CONFIG_KASAN_GENERIC) && + (fpi_flags & FPI_SKIP_KASAN_POISON)); } static inline bool early_page_uninitialised(unsigned long pfn) @@ -1245,7 +1238,7 @@ static __always_inline bool free_pages_prepare(struct page *page, unsigned int order, bool check_free, fpi_t fpi_flags) { int bad = 0; - bool init; + bool skip_kasan_poison = should_skip_kasan_poison(fpi_flags); VM_BUG_ON_PAGE(PageTail(page), page); @@ -1314,10 +1307,17 @@ static __always_inline bool free_pages_prepare(struct page *page, * With hardware tag-based KASAN, memory tags must be set before the * page becomes unavailable via debug_pagealloc or arch_free_page. */ - init = want_init_on_free(); - if (init && !kasan_has_integrated_init()) - kernel_init_free_pages(page, 1 << order); - kasan_free_nondeferred_pages(page, order, init, fpi_flags); + if (kasan_has_integrated_init()) { + if (!skip_kasan_poison) + kasan_free_pages(page, order); + } else { + bool init = want_init_on_free(); + + if (init) + kernel_init_free_pages(page, 1 << order); + if (!skip_kasan_poison) + kasan_poison_pages(page, order, init); + } /* * arch_free_page() can make the page's contents inaccessible. s390 @@ -2324,8 +2324,6 @@ static bool check_new_pages(struct page *page, unsigned int order) inline void post_alloc_hook(struct page *page, unsigned int order, gfp_t gfp_flags) { - bool init; - set_page_private(page, 0); set_page_refcounted(page); @@ -2344,10 +2342,15 @@ inline void post_alloc_hook(struct page *page, unsigned int order, * kasan_alloc_pages and kernel_init_free_pages must be * kept together to avoid discrepancies in behavior. */ - init = !want_init_on_free() && want_init_on_alloc(gfp_flags); - kasan_alloc_pages(page, order, init); - if (init && !kasan_has_integrated_init()) - kernel_init_free_pages(page, 1 << order); + if (kasan_has_integrated_init()) { + kasan_alloc_pages(page, order, gfp_flags); + } else { + bool init = !want_init_on_free() && want_init_on_alloc(gfp_flags); + + kasan_unpoison_pages(page, order, init); + if (init) + kernel_init_free_pages(page, 1 << order); + } set_page_owner(page, order, gfp_flags); } -- cgit v1.2.3 From 013bb59dbb7cf876449df860946458a595a96d51 Mon Sep 17 00:00:00 2001 From: Peter Collingbourne Date: Wed, 2 Jun 2021 16:52:29 -0700 Subject: arm64: mte: handle tags zeroing at page allocation time Currently, on an anonymous page fault, the kernel allocates a zeroed page and maps it in user space. If the mapping is tagged (PROT_MTE), set_pte_at() additionally clears the tags. It is, however, more efficient to clear the tags at the same time as zeroing the data on allocation. To avoid clearing the tags on any page (which may not be mapped as tagged), only do this if the vma flags contain VM_MTE. This requires introducing a new GFP flag that is used to determine whether to clear the tags. The DC GZVA instruction with a 0 top byte (and 0 tag) requires top-byte-ignore. Set the TCR_EL1.{TBI1,TBID1} bits irrespective of whether KASAN_HW is enabled. Signed-off-by: Peter Collingbourne Co-developed-by: Catalin Marinas Signed-off-by: Catalin Marinas Link: https://linux-review.googlesource.com/id/Id46dc94e30fe11474f7e54f5d65e7658dbdddb26 Reviewed-by: Catalin Marinas Reviewed-by: Andrey Konovalov Link: https://lore.kernel.org/r/20210602235230.3928842-4-pcc@google.com Signed-off-by: Will Deacon --- arch/arm64/include/asm/mte.h | 4 ++++ arch/arm64/include/asm/page.h | 8 ++++++-- arch/arm64/lib/mte.S | 20 ++++++++++++++++++++ arch/arm64/mm/fault.c | 26 ++++++++++++++++++++++++++ arch/arm64/mm/proc.S | 10 +++++++--- include/linux/gfp.h | 9 +++++++-- include/linux/highmem.h | 8 ++++++++ mm/kasan/hw_tags.c | 9 ++++++++- mm/page_alloc.c | 13 ++++++++++--- 9 files changed, 96 insertions(+), 11 deletions(-) (limited to 'mm/kasan') diff --git a/arch/arm64/include/asm/mte.h b/arch/arm64/include/asm/mte.h index bc88a1ced0d7..67bf259ae768 100644 --- a/arch/arm64/include/asm/mte.h +++ b/arch/arm64/include/asm/mte.h @@ -37,6 +37,7 @@ void mte_free_tag_storage(char *storage); /* track which pages have valid allocation tags */ #define PG_mte_tagged PG_arch_2 +void mte_zero_clear_page_tags(void *addr); void mte_sync_tags(pte_t *ptep, pte_t pte); void mte_copy_page_tags(void *kto, const void *kfrom); void mte_thread_init_user(void); @@ -53,6 +54,9 @@ int mte_ptrace_copy_tags(struct task_struct *child, long request, /* unused if !CONFIG_ARM64_MTE, silence the compiler */ #define PG_mte_tagged 0 +static inline void mte_zero_clear_page_tags(void *addr) +{ +} static inline void mte_sync_tags(pte_t *ptep, pte_t pte) { } diff --git a/arch/arm64/include/asm/page.h b/arch/arm64/include/asm/page.h index e1fc0f60e79f..ed1b9dcf12b2 100644 --- a/arch/arm64/include/asm/page.h +++ b/arch/arm64/include/asm/page.h @@ -13,6 +13,7 @@ #ifndef __ASSEMBLY__ #include /* for READ_IMPLIES_EXEC */ +#include /* for gfp_t */ #include struct page; @@ -28,10 +29,13 @@ void copy_user_highpage(struct page *to, struct page *from, void copy_highpage(struct page *to, struct page *from); #define __HAVE_ARCH_COPY_HIGHPAGE -#define alloc_zeroed_user_highpage_movable(vma, vaddr) \ - alloc_page_vma(GFP_HIGHUSER_MOVABLE | __GFP_ZERO, vma, vaddr) +struct page *alloc_zeroed_user_highpage_movable(struct vm_area_struct *vma, + unsigned long vaddr); #define __HAVE_ARCH_ALLOC_ZEROED_USER_HIGHPAGE_MOVABLE +void tag_clear_highpage(struct page *to); +#define __HAVE_ARCH_TAG_CLEAR_HIGHPAGE + #define clear_user_page(page, vaddr, pg) clear_page(page) #define copy_user_page(to, from, vaddr, pg) copy_page(to, from) diff --git a/arch/arm64/lib/mte.S b/arch/arm64/lib/mte.S index 351537c12f36..e83643b3995f 100644 --- a/arch/arm64/lib/mte.S +++ b/arch/arm64/lib/mte.S @@ -36,6 +36,26 @@ SYM_FUNC_START(mte_clear_page_tags) ret SYM_FUNC_END(mte_clear_page_tags) +/* + * Zero the page and tags at the same time + * + * Parameters: + * x0 - address to the beginning of the page + */ +SYM_FUNC_START(mte_zero_clear_page_tags) + mrs x1, dczid_el0 + and w1, w1, #0xf + mov x2, #4 + lsl x1, x2, x1 + and x0, x0, #(1 << MTE_TAG_SHIFT) - 1 // clear the tag + +1: dc gzva, x0 + add x0, x0, x1 + tst x0, #(PAGE_SIZE - 1) + b.ne 1b + ret +SYM_FUNC_END(mte_zero_clear_page_tags) + /* * Copy the tags from the source page to the destination one * x0 - address of the destination page diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c index 871c82ab0a30..180c0343d82a 100644 --- a/arch/arm64/mm/fault.c +++ b/arch/arm64/mm/fault.c @@ -921,3 +921,29 @@ void do_debug_exception(unsigned long addr_if_watchpoint, unsigned int esr, debug_exception_exit(regs); } NOKPROBE_SYMBOL(do_debug_exception); + +/* + * Used during anonymous page fault handling. + */ +struct page *alloc_zeroed_user_highpage_movable(struct vm_area_struct *vma, + unsigned long vaddr) +{ + gfp_t flags = GFP_HIGHUSER_MOVABLE | __GFP_ZERO; + + /* + * If the page is mapped with PROT_MTE, initialise the tags at the + * point of allocation and page zeroing as this is usually faster than + * separate DC ZVA and STGM. + */ + if (vma->vm_flags & VM_MTE) + flags |= __GFP_ZEROTAGS; + + return alloc_page_vma(flags, vma, vaddr); +} + +void tag_clear_highpage(struct page *page) +{ + mte_zero_clear_page_tags(page_address(page)); + page_kasan_tag_reset(page); + set_bit(PG_mte_tagged, &page->flags); +} diff --git a/arch/arm64/mm/proc.S b/arch/arm64/mm/proc.S index 97d7bcd8d4f2..48fd1df3d05a 100644 --- a/arch/arm64/mm/proc.S +++ b/arch/arm64/mm/proc.S @@ -46,9 +46,13 @@ #endif #ifdef CONFIG_KASAN_HW_TAGS -#define TCR_KASAN_HW_FLAGS SYS_TCR_EL1_TCMA1 | TCR_TBI1 | TCR_TBID1 +#define TCR_MTE_FLAGS SYS_TCR_EL1_TCMA1 | TCR_TBI1 | TCR_TBID1 #else -#define TCR_KASAN_HW_FLAGS 0 +/* + * The mte_zero_clear_page_tags() implementation uses DC GZVA, which relies on + * TBI being enabled at EL1. + */ +#define TCR_MTE_FLAGS TCR_TBI1 | TCR_TBID1 #endif /* @@ -464,7 +468,7 @@ SYM_FUNC_START(__cpu_setup) msr_s SYS_TFSRE0_EL1, xzr /* set the TCR_EL1 bits */ - mov_q x10, TCR_KASAN_HW_FLAGS + mov_q x10, TCR_MTE_FLAGS orr tcr, tcr, x10 1: #endif diff --git a/include/linux/gfp.h b/include/linux/gfp.h index 11da8af06704..68ba237365dc 100644 --- a/include/linux/gfp.h +++ b/include/linux/gfp.h @@ -53,8 +53,9 @@ struct vm_area_struct; #define ___GFP_HARDWALL 0x100000u #define ___GFP_THISNODE 0x200000u #define ___GFP_ACCOUNT 0x400000u +#define ___GFP_ZEROTAGS 0x800000u #ifdef CONFIG_LOCKDEP -#define ___GFP_NOLOCKDEP 0x800000u +#define ___GFP_NOLOCKDEP 0x1000000u #else #define ___GFP_NOLOCKDEP 0 #endif @@ -229,16 +230,20 @@ struct vm_area_struct; * %__GFP_COMP address compound page metadata. * * %__GFP_ZERO returns a zeroed page on success. + * + * %__GFP_ZEROTAGS returns a page with zeroed memory tags on success, if + * __GFP_ZERO is set. */ #define __GFP_NOWARN ((__force gfp_t)___GFP_NOWARN) #define __GFP_COMP ((__force gfp_t)___GFP_COMP) #define __GFP_ZERO ((__force gfp_t)___GFP_ZERO) +#define __GFP_ZEROTAGS ((__force gfp_t)___GFP_ZEROTAGS) /* Disable lockdep for GFP context tracking */ #define __GFP_NOLOCKDEP ((__force gfp_t)___GFP_NOLOCKDEP) /* Room for N __GFP_FOO bits */ -#define __GFP_BITS_SHIFT (23 + IS_ENABLED(CONFIG_LOCKDEP)) +#define __GFP_BITS_SHIFT (24 + IS_ENABLED(CONFIG_LOCKDEP)) #define __GFP_BITS_MASK ((__force gfp_t)((1 << __GFP_BITS_SHIFT) - 1)) /** diff --git a/include/linux/highmem.h b/include/linux/highmem.h index 54d0643b8fcf..8c6e8e996c87 100644 --- a/include/linux/highmem.h +++ b/include/linux/highmem.h @@ -185,6 +185,14 @@ static inline void clear_highpage(struct page *page) kunmap_atomic(kaddr); } +#ifndef __HAVE_ARCH_TAG_CLEAR_HIGHPAGE + +static inline void tag_clear_highpage(struct page *page) +{ +} + +#endif + /* * If we pass in a base or tail page, we can zero up to PAGE_SIZE. * If we pass in a head page, we can zero up to the size of the compound page. diff --git a/mm/kasan/hw_tags.c b/mm/kasan/hw_tags.c index 9d0f6f934016..41fd5326ee0a 100644 --- a/mm/kasan/hw_tags.c +++ b/mm/kasan/hw_tags.c @@ -246,7 +246,14 @@ void kasan_alloc_pages(struct page *page, unsigned int order, gfp_t flags) */ bool init = !want_init_on_free() && want_init_on_alloc(flags); - kasan_unpoison_pages(page, order, init); + if (flags & __GFP_ZEROTAGS) { + int i; + + for (i = 0; i != 1 << order; ++i) + tag_clear_highpage(page + i); + } else { + kasan_unpoison_pages(page, order, init); + } } void kasan_free_pages(struct page *page, unsigned int order) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 4fddb7cac3c6..13937e793fda 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1219,10 +1219,16 @@ out: return ret; } -static void kernel_init_free_pages(struct page *page, int numpages) +static void kernel_init_free_pages(struct page *page, int numpages, bool zero_tags) { int i; + if (zero_tags) { + for (i = 0; i < numpages; i++) + tag_clear_highpage(page + i); + return; + } + /* s390's use of memset() could override KASAN redzones. */ kasan_disable_current(); for (i = 0; i < numpages; i++) { @@ -1314,7 +1320,7 @@ static __always_inline bool free_pages_prepare(struct page *page, bool init = want_init_on_free(); if (init) - kernel_init_free_pages(page, 1 << order); + kernel_init_free_pages(page, 1 << order, false); if (!skip_kasan_poison) kasan_poison_pages(page, order, init); } @@ -2349,7 +2355,8 @@ inline void post_alloc_hook(struct page *page, unsigned int order, kasan_unpoison_pages(page, order, init); if (init) - kernel_init_free_pages(page, 1 << order); + kernel_init_free_pages(page, 1 << order, + gfp_flags & __GFP_ZEROTAGS); } set_page_owner(page, order, gfp_flags); -- cgit v1.2.3 From c275c5c6d50a0518cdb0584e85905d10e7cefc6e Mon Sep 17 00:00:00 2001 From: Peter Collingbourne Date: Wed, 2 Jun 2021 16:52:30 -0700 Subject: kasan: disable freed user page poisoning with HW tags Poisoning freed pages protects against kernel use-after-free. The likelihood of such a bug involving kernel pages is significantly higher than that for user pages. At the same time, poisoning freed pages can impose a significant performance cost, which cannot always be justified for user pages given the lower probability of finding a bug. Therefore, disable freed user page poisoning when using HW tags. We identify "user" pages via the flag set GFP_HIGHUSER_MOVABLE, which indicates a strong likelihood of not being directly accessible to the kernel. Signed-off-by: Peter Collingbourne Reviewed-by: Andrey Konovalov Link: https://linux-review.googlesource.com/id/I716846e2de8ef179f44e835770df7e6307be96c9 Link: https://lore.kernel.org/r/20210602235230.3928842-5-pcc@google.com Signed-off-by: Will Deacon --- include/linux/gfp.h | 13 ++++++++++--- include/linux/page-flags.h | 9 +++++++++ include/trace/events/mmflags.h | 9 ++++++++- mm/kasan/hw_tags.c | 3 +++ mm/page_alloc.c | 12 +++++++----- 5 files changed, 37 insertions(+), 9 deletions(-) (limited to 'mm/kasan') diff --git a/include/linux/gfp.h b/include/linux/gfp.h index 68ba237365dc..e6102dfa4faa 100644 --- a/include/linux/gfp.h +++ b/include/linux/gfp.h @@ -54,8 +54,9 @@ struct vm_area_struct; #define ___GFP_THISNODE 0x200000u #define ___GFP_ACCOUNT 0x400000u #define ___GFP_ZEROTAGS 0x800000u +#define ___GFP_SKIP_KASAN_POISON 0x1000000u #ifdef CONFIG_LOCKDEP -#define ___GFP_NOLOCKDEP 0x1000000u +#define ___GFP_NOLOCKDEP 0x2000000u #else #define ___GFP_NOLOCKDEP 0 #endif @@ -233,17 +234,22 @@ struct vm_area_struct; * * %__GFP_ZEROTAGS returns a page with zeroed memory tags on success, if * __GFP_ZERO is set. + * + * %__GFP_SKIP_KASAN_POISON returns a page which does not need to be poisoned + * on deallocation. Typically used for userspace pages. Currently only has an + * effect in HW tags mode. */ #define __GFP_NOWARN ((__force gfp_t)___GFP_NOWARN) #define __GFP_COMP ((__force gfp_t)___GFP_COMP) #define __GFP_ZERO ((__force gfp_t)___GFP_ZERO) #define __GFP_ZEROTAGS ((__force gfp_t)___GFP_ZEROTAGS) +#define __GFP_SKIP_KASAN_POISON ((__force gfp_t)___GFP_SKIP_KASAN_POISON) /* Disable lockdep for GFP context tracking */ #define __GFP_NOLOCKDEP ((__force gfp_t)___GFP_NOLOCKDEP) /* Room for N __GFP_FOO bits */ -#define __GFP_BITS_SHIFT (24 + IS_ENABLED(CONFIG_LOCKDEP)) +#define __GFP_BITS_SHIFT (25 + IS_ENABLED(CONFIG_LOCKDEP)) #define __GFP_BITS_MASK ((__force gfp_t)((1 << __GFP_BITS_SHIFT) - 1)) /** @@ -324,7 +330,8 @@ struct vm_area_struct; #define GFP_DMA __GFP_DMA #define GFP_DMA32 __GFP_DMA32 #define GFP_HIGHUSER (GFP_USER | __GFP_HIGHMEM) -#define GFP_HIGHUSER_MOVABLE (GFP_HIGHUSER | __GFP_MOVABLE) +#define GFP_HIGHUSER_MOVABLE (GFP_HIGHUSER | __GFP_MOVABLE | \ + __GFP_SKIP_KASAN_POISON) #define GFP_TRANSHUGE_LIGHT ((GFP_HIGHUSER_MOVABLE | __GFP_COMP | \ __GFP_NOMEMALLOC | __GFP_NOWARN) & ~__GFP_RECLAIM) #define GFP_TRANSHUGE (GFP_TRANSHUGE_LIGHT | __GFP_DIRECT_RECLAIM) diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h index 04a34c08e0a6..40e2c5000585 100644 --- a/include/linux/page-flags.h +++ b/include/linux/page-flags.h @@ -137,6 +137,9 @@ enum pageflags { #endif #ifdef CONFIG_64BIT PG_arch_2, +#endif +#ifdef CONFIG_KASAN_HW_TAGS + PG_skip_kasan_poison, #endif __NR_PAGEFLAGS, @@ -443,6 +446,12 @@ TESTCLEARFLAG(Young, young, PF_ANY) PAGEFLAG(Idle, idle, PF_ANY) #endif +#ifdef CONFIG_KASAN_HW_TAGS +PAGEFLAG(SkipKASanPoison, skip_kasan_poison, PF_HEAD) +#else +PAGEFLAG_FALSE(SkipKASanPoison) +#endif + /* * PageReported() is used to track reported free pages within the Buddy * allocator. We can use the non-atomic version of the test and set diff --git a/include/trace/events/mmflags.h b/include/trace/events/mmflags.h index 629c7a0eaff2..390270e00a1d 100644 --- a/include/trace/events/mmflags.h +++ b/include/trace/events/mmflags.h @@ -85,6 +85,12 @@ #define IF_HAVE_PG_ARCH_2(flag,string) #endif +#ifdef CONFIG_KASAN_HW_TAGS +#define IF_HAVE_PG_SKIP_KASAN_POISON(flag,string) ,{1UL << flag, string} +#else +#define IF_HAVE_PG_SKIP_KASAN_POISON(flag,string) +#endif + #define __def_pageflag_names \ {1UL << PG_locked, "locked" }, \ {1UL << PG_waiters, "waiters" }, \ @@ -112,7 +118,8 @@ IF_HAVE_PG_UNCACHED(PG_uncached, "uncached" ) \ IF_HAVE_PG_HWPOISON(PG_hwpoison, "hwpoison" ) \ IF_HAVE_PG_IDLE(PG_young, "young" ) \ IF_HAVE_PG_IDLE(PG_idle, "idle" ) \ -IF_HAVE_PG_ARCH_2(PG_arch_2, "arch_2" ) +IF_HAVE_PG_ARCH_2(PG_arch_2, "arch_2" ) \ +IF_HAVE_PG_SKIP_KASAN_POISON(PG_skip_kasan_poison, "skip_kasan_poison") #define show_page_flags(flags) \ (flags) ? __print_flags(flags, "|", \ diff --git a/mm/kasan/hw_tags.c b/mm/kasan/hw_tags.c index 41fd5326ee0a..ed5e5b833d61 100644 --- a/mm/kasan/hw_tags.c +++ b/mm/kasan/hw_tags.c @@ -246,6 +246,9 @@ void kasan_alloc_pages(struct page *page, unsigned int order, gfp_t flags) */ bool init = !want_init_on_free() && want_init_on_alloc(flags); + if (flags & __GFP_SKIP_KASAN_POISON) + SetPageSkipKASanPoison(page); + if (flags & __GFP_ZEROTAGS) { int i; diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 13937e793fda..5ad76e540a22 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -394,11 +394,12 @@ static DEFINE_STATIC_KEY_TRUE(deferred_pages); * on-demand allocation and then freed again before the deferred pages * initialization is done, but this is not likely to happen. */ -static inline bool should_skip_kasan_poison(fpi_t fpi_flags) +static inline bool should_skip_kasan_poison(struct page *page, fpi_t fpi_flags) { return static_branch_unlikely(&deferred_pages) || (!IS_ENABLED(CONFIG_KASAN_GENERIC) && - (fpi_flags & FPI_SKIP_KASAN_POISON)); + (fpi_flags & FPI_SKIP_KASAN_POISON)) || + PageSkipKASanPoison(page); } /* Returns true if the struct page for the pfn is uninitialised */ @@ -449,10 +450,11 @@ defer_init(int nid, unsigned long pfn, unsigned long end_pfn) return false; } #else -static inline bool should_skip_kasan_poison(fpi_t fpi_flags) +static inline bool should_skip_kasan_poison(struct page *page, fpi_t fpi_flags) { return (!IS_ENABLED(CONFIG_KASAN_GENERIC) && - (fpi_flags & FPI_SKIP_KASAN_POISON)); + (fpi_flags & FPI_SKIP_KASAN_POISON)) || + PageSkipKASanPoison(page); } static inline bool early_page_uninitialised(unsigned long pfn) @@ -1244,7 +1246,7 @@ static __always_inline bool free_pages_prepare(struct page *page, unsigned int order, bool check_free, fpi_t fpi_flags) { int bad = 0; - bool skip_kasan_poison = should_skip_kasan_poison(fpi_flags); + bool skip_kasan_poison = should_skip_kasan_poison(page, fpi_flags); VM_BUG_ON_PAGE(PageTail(page), page); -- cgit v1.2.3