| Message ID | 20190111185842.13978-1-aryabinin@virtuozzo.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | kasan: Remove use after scope bugs detection. | expand |
On Fri, Jan 11, 2019 at 7:58 PM Andrey Ryabinin <aryabinin@virtuozzo.com> wrote: > > Use after scope bugs detector seems to be almost entirely useless > for the linux kernel. It exists over two years, but I've seen only > one valid bug so far [1]. And the bug was fixed before it has been > reported. There were some other use-after-scope reports, but they > were false-positives due to different reasons like incompatibility > with structleak plugin. > > This feature significantly increases stack usage, especially with > GCC < 9 version, and causes a 32K stack overflow. It probably > adds performance penalty too. > > Given all that, let's remove use-after-scope detector entirely. > > While preparing this patch I've noticed that we mistakenly enable > use-after-scope detection for clang compiler regardless of > CONFIG_KASAN_EXTRA setting. This is also fixed now. Hi Andrey, I am on a fence. On one hand removing bug detection sucks and each case of a missed memory corruption leads to a splash of assorted bug reports by syzbot. On the other hand everything you said is true. Maybe support for CONFIG_VMAP_STACK will enable stacks larger then PAGE_ALLOC_COSTLY_ORDER? > [1] http://lkml.kernel.org/r/<20171129052106.rhgbjhhis53hkgfn@wfg-t540p.sh.intel.com> > > Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com> > Cc: Qian Cai <cai@lca.pw> > Cc: Alexander Potapenko <glider@google.com> > Cc: Dmitry Vyukov <dvyukov@google.com> > Cc: Catalin Marinas <catalin.marinas@arm.com> > Cc: Will Deacon <will.deacon@arm.com> > --- > arch/arm64/include/asm/memory.h | 4 ---- > lib/Kconfig.debug | 1 - > lib/Kconfig.kasan | 10 ---------- > lib/test_kasan.c | 24 ------------------------ > mm/kasan/generic.c | 19 ------------------- > mm/kasan/generic_report.c | 3 --- > mm/kasan/kasan.h | 3 --- > scripts/Makefile.kasan | 5 ----- > scripts/gcc-plugins/Kconfig | 4 ---- > 9 files changed, 73 deletions(-) > > diff --git a/arch/arm64/include/asm/memory.h b/arch/arm64/include/asm/memory.h > index e1ec947e7c0c..0e236a99b3ef 100644 > --- a/arch/arm64/include/asm/memory.h > +++ b/arch/arm64/include/asm/memory.h > @@ -80,11 +80,7 @@ > */ > #ifdef CONFIG_KASAN > #define KASAN_SHADOW_SIZE (UL(1) << (VA_BITS - KASAN_SHADOW_SCALE_SHIFT)) > -#ifdef CONFIG_KASAN_EXTRA > -#define KASAN_THREAD_SHIFT 2 > -#else > #define KASAN_THREAD_SHIFT 1 > -#endif /* CONFIG_KASAN_EXTRA */ > #else > #define KASAN_SHADOW_SIZE (0) > #define KASAN_THREAD_SHIFT 0 > diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug > index d4df5b24d75e..a219f3488ad7 100644 > --- a/lib/Kconfig.debug > +++ b/lib/Kconfig.debug > @@ -222,7 +222,6 @@ config ENABLE_MUST_CHECK > config FRAME_WARN > int "Warn for stack frames larger than (needs gcc 4.4)" > range 0 8192 > - default 3072 if KASAN_EXTRA > default 2048 if GCC_PLUGIN_LATENT_ENTROPY > default 1280 if (!64BIT && PARISC) > default 1024 if (!64BIT && !PARISC) > diff --git a/lib/Kconfig.kasan b/lib/Kconfig.kasan > index d8c474b6691e..67d7d1309c52 100644 > --- a/lib/Kconfig.kasan > +++ b/lib/Kconfig.kasan > @@ -78,16 +78,6 @@ config KASAN_SW_TAGS > > endchoice > > -config KASAN_EXTRA > - bool "KASAN: extra checks" > - depends on KASAN_GENERIC && DEBUG_KERNEL && !COMPILE_TEST > - help > - This enables further checks in generic KASAN, for now it only > - includes the address-use-after-scope check that can lead to > - excessive kernel stack usage, frame size warnings and longer > - compile time. > - See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81715 > - > choice > prompt "Instrumentation type" > depends on KASAN > diff --git a/lib/test_kasan.c b/lib/test_kasan.c > index 51b78405bf24..7de2702621dc 100644 > --- a/lib/test_kasan.c > +++ b/lib/test_kasan.c > @@ -480,29 +480,6 @@ static noinline void __init copy_user_test(void) > kfree(kmem); > } > > -static noinline void __init use_after_scope_test(void) > -{ > - volatile char *volatile p; > - > - pr_info("use-after-scope on int\n"); > - { > - int local = 0; > - > - p = (char *)&local; > - } > - p[0] = 1; > - p[3] = 1; > - > - pr_info("use-after-scope on array\n"); > - { > - char local[1024] = {0}; > - > - p = local; > - } > - p[0] = 1; > - p[1023] = 1; > -} > - > static noinline void __init kasan_alloca_oob_left(void) > { > volatile int i = 10; > @@ -682,7 +659,6 @@ static int __init kmalloc_tests_init(void) > kasan_alloca_oob_right(); > ksize_unpoisons_memory(); > copy_user_test(); > - use_after_scope_test(); > kmem_cache_double_free(); > kmem_cache_invalid_free(); > kasan_memchr(); > diff --git a/mm/kasan/generic.c b/mm/kasan/generic.c > index ccb6207276e3..504c79363a34 100644 > --- a/mm/kasan/generic.c > +++ b/mm/kasan/generic.c > @@ -275,25 +275,6 @@ EXPORT_SYMBOL(__asan_storeN_noabort); > void __asan_handle_no_return(void) {} > EXPORT_SYMBOL(__asan_handle_no_return); > > -/* Emitted by compiler to poison large objects when they go out of scope. */ > -void __asan_poison_stack_memory(const void *addr, size_t size) > -{ > - /* > - * Addr is KASAN_SHADOW_SCALE_SIZE-aligned and the object is surrounded > - * by redzones, so we simply round up size to simplify logic. > - */ > - kasan_poison_shadow(addr, round_up(size, KASAN_SHADOW_SCALE_SIZE), > - KASAN_USE_AFTER_SCOPE); > -} > -EXPORT_SYMBOL(__asan_poison_stack_memory); > - > -/* Emitted by compiler to unpoison large objects when they go into scope. */ > -void __asan_unpoison_stack_memory(const void *addr, size_t size) > -{ > - kasan_unpoison_shadow(addr, size); > -} > -EXPORT_SYMBOL(__asan_unpoison_stack_memory); > - > /* Emitted by compiler to poison alloca()ed objects. */ > void __asan_alloca_poison(unsigned long addr, size_t size) > { > diff --git a/mm/kasan/generic_report.c b/mm/kasan/generic_report.c > index 5e12035888f2..36c645939bc9 100644 > --- a/mm/kasan/generic_report.c > +++ b/mm/kasan/generic_report.c > @@ -82,9 +82,6 @@ static const char *get_shadow_bug_type(struct kasan_access_info *info) > case KASAN_KMALLOC_FREE: > bug_type = "use-after-free"; > break; > - case KASAN_USE_AFTER_SCOPE: > - bug_type = "use-after-scope"; > - break; > case KASAN_ALLOCA_LEFT: > case KASAN_ALLOCA_RIGHT: > bug_type = "alloca-out-of-bounds"; > diff --git a/mm/kasan/kasan.h b/mm/kasan/kasan.h > index ea51b2d898ec..3e0c11f7d7a1 100644 > --- a/mm/kasan/kasan.h > +++ b/mm/kasan/kasan.h > @@ -34,7 +34,6 @@ > #define KASAN_STACK_MID 0xF2 > #define KASAN_STACK_RIGHT 0xF3 > #define KASAN_STACK_PARTIAL 0xF4 > -#define KASAN_USE_AFTER_SCOPE 0xF8 > > /* > * alloca redzone shadow values > @@ -187,8 +186,6 @@ void __asan_unregister_globals(struct kasan_global *globals, size_t size); > void __asan_loadN(unsigned long addr, size_t size); > void __asan_storeN(unsigned long addr, size_t size); > void __asan_handle_no_return(void); > -void __asan_poison_stack_memory(const void *addr, size_t size); > -void __asan_unpoison_stack_memory(const void *addr, size_t size); > void __asan_alloca_poison(unsigned long addr, size_t size); > void __asan_allocas_unpoison(const void *stack_top, const void *stack_bottom); > > diff --git a/scripts/Makefile.kasan b/scripts/Makefile.kasan > index 25c259df8ffa..f1fb8e502657 100644 > --- a/scripts/Makefile.kasan > +++ b/scripts/Makefile.kasan > @@ -27,14 +27,9 @@ else > $(call cc-param,asan-globals=1) \ > $(call cc-param,asan-instrumentation-with-call-threshold=$(call_threshold)) \ > $(call cc-param,asan-stack=1) \ > - $(call cc-param,asan-use-after-scope=1) \ > $(call cc-param,asan-instrument-allocas=1) > endif > > -ifdef CONFIG_KASAN_EXTRA > -CFLAGS_KASAN += $(call cc-option, -fsanitize-address-use-after-scope) > -endif > - > endif # CONFIG_KASAN_GENERIC > > ifdef CONFIG_KASAN_SW_TAGS > diff --git a/scripts/gcc-plugins/Kconfig b/scripts/gcc-plugins/Kconfig > index d45f7f36b859..d9fd9988ef27 100644 > --- a/scripts/gcc-plugins/Kconfig > +++ b/scripts/gcc-plugins/Kconfig > @@ -68,10 +68,6 @@ config GCC_PLUGIN_LATENT_ENTROPY > > config GCC_PLUGIN_STRUCTLEAK > bool "Force initialization of variables containing userspace addresses" > - # Currently STRUCTLEAK inserts initialization out of live scope of > - # variables from KASAN point of view. This leads to KASAN false > - # positive reports. Prohibit this combination for now. > - depends on !KASAN_EXTRA > help > This plugin zero-initializes any structures containing a > __user attribute. This can prevent some classes of information > -- > 2.19.2 > > -- > You received this message because you are subscribed to the Google Groups "kasan-dev" group. > To unsubscribe from this group and stop receiving emails from it, send an email to kasan-dev+unsubscribe@googlegroups.com. > To post to this group, send email to kasan-dev@googlegroups.com. > To view this discussion on the web visit https://groups.google.com/d/msgid/kasan-dev/20190111185842.13978-1-aryabinin%40virtuozzo.com. > For more options, visit https://groups.google.com/d/optout.
On 1/14/19 4:24 PM, Dmitry Vyukov wrote: > On Fri, Jan 11, 2019 at 7:58 PM Andrey Ryabinin <aryabinin@virtuozzo.com> wrote: >> >> Use after scope bugs detector seems to be almost entirely useless >> for the linux kernel. It exists over two years, but I've seen only >> one valid bug so far [1]. And the bug was fixed before it has been >> reported. There were some other use-after-scope reports, but they >> were false-positives due to different reasons like incompatibility >> with structleak plugin. >> >> This feature significantly increases stack usage, especially with >> GCC < 9 version, and causes a 32K stack overflow. It probably >> adds performance penalty too. >> >> Given all that, let's remove use-after-scope detector entirely. >> >> While preparing this patch I've noticed that we mistakenly enable >> use-after-scope detection for clang compiler regardless of >> CONFIG_KASAN_EXTRA setting. This is also fixed now. > > Hi Andrey, > > I am on a fence. On one hand removing bug detection sucks and each > case of a missed memory corruption leads to a splash of assorted bug > reports by syzbot. On the other hand everything you said is true. > Maybe support for CONFIG_VMAP_STACK will enable stacks larger then > PAGE_ALLOC_COSTLY_ORDER? > Yes, with vmap stacks higher order won't be a problem, since vmalloc() does only 0-order allocations. But even with vmap stacks use-after-scope won't become useful, thus I don't see the point of re-enabling it with vmap stacks. If feature doesn't detect bugs, but waste resources, than it's bad for detecting bugs. We wasting our limited resources for useless checks, instead of using these resources for doing more useful checks, running tests faster hence detecting more bugs per-time.
On Fri, Jan 11, 2019 at 09:58:42PM +0300, Andrey Ryabinin wrote: > Use after scope bugs detector seems to be almost entirely useless > for the linux kernel. It exists over two years, but I've seen only > one valid bug so far [1]. And the bug was fixed before it has been > reported. There were some other use-after-scope reports, but they > were false-positives due to different reasons like incompatibility > with structleak plugin. > > This feature significantly increases stack usage, especially with > GCC < 9 version, and causes a 32K stack overflow. It probably > adds performance penalty too. > > Given all that, let's remove use-after-scope detector entirely. > > While preparing this patch I've noticed that we mistakenly enable > use-after-scope detection for clang compiler regardless of > CONFIG_KASAN_EXTRA setting. This is also fixed now. > > [1] http://lkml.kernel.org/r/<20171129052106.rhgbjhhis53hkgfn@wfg-t540p.sh.intel.com> > > Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com> > Cc: Qian Cai <cai@lca.pw> > Cc: Alexander Potapenko <glider@google.com> > Cc: Dmitry Vyukov <dvyukov@google.com> > Cc: Catalin Marinas <catalin.marinas@arm.com> > Cc: Will Deacon <will.deacon@arm.com> > --- > arch/arm64/include/asm/memory.h | 4 ---- > lib/Kconfig.debug | 1 - > lib/Kconfig.kasan | 10 ---------- > lib/test_kasan.c | 24 ------------------------ > mm/kasan/generic.c | 19 ------------------- > mm/kasan/generic_report.c | 3 --- > mm/kasan/kasan.h | 3 --- > scripts/Makefile.kasan | 5 ----- > scripts/gcc-plugins/Kconfig | 4 ---- > 9 files changed, 73 deletions(-) For the arm64 part: Acked-by: Will Deacon <will.deacon@arm.com> but I defer to you and Dmitry as to whether or not you go ahead with this. Will
diff --git a/arch/arm64/include/asm/memory.h b/arch/arm64/include/asm/memory.h index e1ec947e7c0c..0e236a99b3ef 100644 --- a/arch/arm64/include/asm/memory.h +++ b/arch/arm64/include/asm/memory.h @@ -80,11 +80,7 @@ */ #ifdef CONFIG_KASAN #define KASAN_SHADOW_SIZE (UL(1) << (VA_BITS - KASAN_SHADOW_SCALE_SHIFT)) -#ifdef CONFIG_KASAN_EXTRA -#define KASAN_THREAD_SHIFT 2 -#else #define KASAN_THREAD_SHIFT 1 -#endif /* CONFIG_KASAN_EXTRA */ #else #define KASAN_SHADOW_SIZE (0) #define KASAN_THREAD_SHIFT 0 diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index d4df5b24d75e..a219f3488ad7 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -222,7 +222,6 @@ config ENABLE_MUST_CHECK config FRAME_WARN int "Warn for stack frames larger than (needs gcc 4.4)" range 0 8192 - default 3072 if KASAN_EXTRA default 2048 if GCC_PLUGIN_LATENT_ENTROPY default 1280 if (!64BIT && PARISC) default 1024 if (!64BIT && !PARISC) diff --git a/lib/Kconfig.kasan b/lib/Kconfig.kasan index d8c474b6691e..67d7d1309c52 100644 --- a/lib/Kconfig.kasan +++ b/lib/Kconfig.kasan @@ -78,16 +78,6 @@ config KASAN_SW_TAGS endchoice -config KASAN_EXTRA - bool "KASAN: extra checks" - depends on KASAN_GENERIC && DEBUG_KERNEL && !COMPILE_TEST - help - This enables further checks in generic KASAN, for now it only - includes the address-use-after-scope check that can lead to - excessive kernel stack usage, frame size warnings and longer - compile time. - See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81715 - choice prompt "Instrumentation type" depends on KASAN diff --git a/lib/test_kasan.c b/lib/test_kasan.c index 51b78405bf24..7de2702621dc 100644 --- a/lib/test_kasan.c +++ b/lib/test_kasan.c @@ -480,29 +480,6 @@ static noinline void __init copy_user_test(void) kfree(kmem); } -static noinline void __init use_after_scope_test(void) -{ - volatile char *volatile p; - - pr_info("use-after-scope on int\n"); - { - int local = 0; - - p = (char *)&local; - } - p[0] = 1; - p[3] = 1; - - pr_info("use-after-scope on array\n"); - { - char local[1024] = {0}; - - p = local; - } - p[0] = 1; - p[1023] = 1; -} - static noinline void __init kasan_alloca_oob_left(void) { volatile int i = 10; @@ -682,7 +659,6 @@ static int __init kmalloc_tests_init(void) kasan_alloca_oob_right(); ksize_unpoisons_memory(); copy_user_test(); - use_after_scope_test(); kmem_cache_double_free(); kmem_cache_invalid_free(); kasan_memchr(); diff --git a/mm/kasan/generic.c b/mm/kasan/generic.c index ccb6207276e3..504c79363a34 100644 --- a/mm/kasan/generic.c +++ b/mm/kasan/generic.c @@ -275,25 +275,6 @@ EXPORT_SYMBOL(__asan_storeN_noabort); void __asan_handle_no_return(void) {} EXPORT_SYMBOL(__asan_handle_no_return); -/* Emitted by compiler to poison large objects when they go out of scope. */ -void __asan_poison_stack_memory(const void *addr, size_t size) -{ - /* - * Addr is KASAN_SHADOW_SCALE_SIZE-aligned and the object is surrounded - * by redzones, so we simply round up size to simplify logic. - */ - kasan_poison_shadow(addr, round_up(size, KASAN_SHADOW_SCALE_SIZE), - KASAN_USE_AFTER_SCOPE); -} -EXPORT_SYMBOL(__asan_poison_stack_memory); - -/* Emitted by compiler to unpoison large objects when they go into scope. */ -void __asan_unpoison_stack_memory(const void *addr, size_t size) -{ - kasan_unpoison_shadow(addr, size); -} -EXPORT_SYMBOL(__asan_unpoison_stack_memory); - /* Emitted by compiler to poison alloca()ed objects. */ void __asan_alloca_poison(unsigned long addr, size_t size) { diff --git a/mm/kasan/generic_report.c b/mm/kasan/generic_report.c index 5e12035888f2..36c645939bc9 100644 --- a/mm/kasan/generic_report.c +++ b/mm/kasan/generic_report.c @@ -82,9 +82,6 @@ static const char *get_shadow_bug_type(struct kasan_access_info *info) case KASAN_KMALLOC_FREE: bug_type = "use-after-free"; break; - case KASAN_USE_AFTER_SCOPE: - bug_type = "use-after-scope"; - break; case KASAN_ALLOCA_LEFT: case KASAN_ALLOCA_RIGHT: bug_type = "alloca-out-of-bounds"; diff --git a/mm/kasan/kasan.h b/mm/kasan/kasan.h index ea51b2d898ec..3e0c11f7d7a1 100644 --- a/mm/kasan/kasan.h +++ b/mm/kasan/kasan.h @@ -34,7 +34,6 @@ #define KASAN_STACK_MID 0xF2 #define KASAN_STACK_RIGHT 0xF3 #define KASAN_STACK_PARTIAL 0xF4 -#define KASAN_USE_AFTER_SCOPE 0xF8 /* * alloca redzone shadow values @@ -187,8 +186,6 @@ void __asan_unregister_globals(struct kasan_global *globals, size_t size); void __asan_loadN(unsigned long addr, size_t size); void __asan_storeN(unsigned long addr, size_t size); void __asan_handle_no_return(void); -void __asan_poison_stack_memory(const void *addr, size_t size); -void __asan_unpoison_stack_memory(const void *addr, size_t size); void __asan_alloca_poison(unsigned long addr, size_t size); void __asan_allocas_unpoison(const void *stack_top, const void *stack_bottom); diff --git a/scripts/Makefile.kasan b/scripts/Makefile.kasan index 25c259df8ffa..f1fb8e502657 100644 --- a/scripts/Makefile.kasan +++ b/scripts/Makefile.kasan @@ -27,14 +27,9 @@ else $(call cc-param,asan-globals=1) \ $(call cc-param,asan-instrumentation-with-call-threshold=$(call_threshold)) \ $(call cc-param,asan-stack=1) \ - $(call cc-param,asan-use-after-scope=1) \ $(call cc-param,asan-instrument-allocas=1) endif -ifdef CONFIG_KASAN_EXTRA -CFLAGS_KASAN += $(call cc-option, -fsanitize-address-use-after-scope) -endif - endif # CONFIG_KASAN_GENERIC ifdef CONFIG_KASAN_SW_TAGS diff --git a/scripts/gcc-plugins/Kconfig b/scripts/gcc-plugins/Kconfig index d45f7f36b859..d9fd9988ef27 100644 --- a/scripts/gcc-plugins/Kconfig +++ b/scripts/gcc-plugins/Kconfig @@ -68,10 +68,6 @@ config GCC_PLUGIN_LATENT_ENTROPY config GCC_PLUGIN_STRUCTLEAK bool "Force initialization of variables containing userspace addresses" - # Currently STRUCTLEAK inserts initialization out of live scope of - # variables from KASAN point of view. This leads to KASAN false - # positive reports. Prohibit this combination for now. - depends on !KASAN_EXTRA help This plugin zero-initializes any structures containing a __user attribute. This can prevent some classes of information
Use after scope bugs detector seems to be almost entirely useless for the linux kernel. It exists over two years, but I've seen only one valid bug so far [1]. And the bug was fixed before it has been reported. There were some other use-after-scope reports, but they were false-positives due to different reasons like incompatibility with structleak plugin. This feature significantly increases stack usage, especially with GCC < 9 version, and causes a 32K stack overflow. It probably adds performance penalty too. Given all that, let's remove use-after-scope detector entirely. While preparing this patch I've noticed that we mistakenly enable use-after-scope detection for clang compiler regardless of CONFIG_KASAN_EXTRA setting. This is also fixed now. [1] http://lkml.kernel.org/r/<20171129052106.rhgbjhhis53hkgfn@wfg-t540p.sh.intel.com> Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com> Cc: Qian Cai <cai@lca.pw> Cc: Alexander Potapenko <glider@google.com> Cc: Dmitry Vyukov <dvyukov@google.com> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Will Deacon <will.deacon@arm.com> --- arch/arm64/include/asm/memory.h | 4 ---- lib/Kconfig.debug | 1 - lib/Kconfig.kasan | 10 ---------- lib/test_kasan.c | 24 ------------------------ mm/kasan/generic.c | 19 ------------------- mm/kasan/generic_report.c | 3 --- mm/kasan/kasan.h | 3 --- scripts/Makefile.kasan | 5 ----- scripts/gcc-plugins/Kconfig | 4 ---- 9 files changed, 73 deletions(-)