kasan: guard release_free_meta() shadow access with kasan_arch_is_ready()
A message that is not in the archive (sent off-list or not yet received): [email protected]
Andrew Morton
maintainerQuoted 27 lines · On Tue, 13 Feb 2024 14:39:58 +1100 Benjamin Gray <[email protected]> wrote:
release_free_meta()accesses the shadow directly through the path
kasan_slab_free
__kasan_slab_free
kasan_release_object_meta
release_free_meta
kasan_mem_to_shadowThere are no
kasan_arch_is_ready()guards here, allowing an oops when the shadow is not initialized. The oops can be seen on a Power8 KVM guest.This patch adds the guard to
release_free_meta(), as it's the first level that specifically requires the shadow.It is safe to put the guard at the start of this function, before the stack put: only
kasan_save_free_info()can initialize the saved stack, which itself is guarded withkasan_arch_is_ready()by its callerpoison_slab_object(). If the arch becomes ready beforerelease_free_meta()then we will not observeKASAN_SLAB_FREE_METAin the object's shadow, so we will not put an uninitialized stack either....
mm/kasan/generic.c static void release_alloc_meta(struct kasan_alloc_meta *meta) · line 530
Quoted 9 lines
mm/kasan/generic.c +3
-0@@ -522,6 +522,9 @@ static void release_alloc_meta(struct kasan_alloc_meta *meta)static void release_free_meta (const void *object,struct kasan_free_meta *meta) { +if (!kasan_arch_is_ready ()) +return ; +/* Check if free meta is valid. */ if (*(u8 *)kasan_mem_to_shadow (object) !=KASAN_SLAB_FREE_META )return ;
I'll add
Andrey Konovalov
Reviewed-byQuoted 23 lines · Benjamin Gray
release_free_meta()accesses the shadow directly through the path
kasan_slab_free
__kasan_slab_free
kasan_release_object_meta
release_free_meta
kasan_mem_to_shadowThere are no
kasan_arch_is_ready()guards here, allowing an oops when the shadow is not initialized. The oops can be seen on a Power8 KVM guest.This patch adds the guard to
release_free_meta(), as it's the first level that specifically requires the shadow.It is safe to put the guard at the start of this function, before the stack put: only
kasan_save_free_info()can initialize the saved stack, which itself is guarded withkasan_arch_is_ready()by its callerpoison_slab_object(). If the arch becomes ready beforerelease_free_meta()then we will not observeKASAN_SLAB_FREE_METAin the object's shadow, so we will not put an uninitialized stack either.
Quoted 6 lines
Signed-off-by Benjamin Gray <[email protected]>---
I am interested in removing the need for
kasan_arch_is_ready()entirely, as it mostly acts like a separate check ofkasan_enabled().
Dropping kasan_arch_is_ready() calls from KASAN internals and instead relying on kasan_enabled() checks in include/linux/kasan.h would be great!
I filed a bug about this a while ago:
https://bugzilla.kernel.org/show_bug.cgi?id=217049
Quoted 4 lines
Currently
both are necessary, but I think adding akasan_enabled()guard tocheck_region_inline()makeskasan_enabled()a superset ofkasan_arch_is_ready().
Sounds good to me. I would also go through the list of other exported KASAN functions to check whether any of them also need a kasan_enabled() check. At least kasan_unpoison_task_stack() seems to be one of them.
Quoted 5 lines
Allowing an arch to override
kasan_enabled()can then let us replace it with a static branch that we enable somewhere in boot (for PowerPC, after we use a bunch of generic code to parse the device tree to determine how we want to configure the MMU). This should generally work OK I think, as HW tags already does this,
We can also add something like CONFIG_ARCH_HAS_KASAN_FLAG_ENABLE and only use a static branch only on those architectures where it's required.
Quoted 2 lines
but I did have to add another
patch for an uninitialised data access it introduces.
What was this data access? Is this something we need to fix in the mainline?
Quoted 2 lines
On the other hand, KASAN does more than shadow based sanitisation, so we'd be disabling that in early boot too.
I think the things that we need to handle before KASAN is enabled is kasan_cache_create() and kasan_metadata_size() (if these can even called before KASAN is enabled). Otherwise, KASAN just collects metadata, which is useless without shadow memory-based reporting anyway.
Quoted 19 lines
--- mm/kasan/generic.c | 3 +++ 1 file changed, 3 insertions(+)mm/kasan/generic.c +3
-1@@ -522,6 +522,9 @@ static void release_alloc_meta(struct kasan_alloc_meta *meta)static void release_free_meta (const void *object,struct kasan_free_meta *meta) { +if (!kasan_arch_is_ready ()) +return ; +/* Check if free meta is valid. */ if (*(u8 *)kasan_mem_to_shadow (object) !=KASAN_SLAB_FREE_META )return ; --
Quoted 2 lines
2.43.0
For the patch itself as a fix:
Thanks!
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 4 lines · Benjamin Gray
Quoted 4 lines
We can also add something like
CONFIG_ARCH_HAS_KASAN_FLAG_ENABLEand only use a static branch only on those architectures where it's required.
Perhaps CONFIG_ARCH_USES_KASAN_ENABLED would be a better name.
Quoted 5 lines
That works too, PowerPC should only need a static branch when
CONFIG_KASANis enabled.Loongarch is also a
kasan_arch_is_ready()user though, so I'm not sure if they'd still need it for something?
And UM as well.
We can start with a change that makes kasan_flag_enabled and kasan_enabled() work for the Generic mode, define a kasan_init_generic() function that switches kasan_flag_enabled (and prints the "KernelAddressSanitizer initialized" message?), and then call kasan_init_generic() from kasan_init() in the arch code (perhaps even for all arches to unify the "initialized" message?).
And then we can ask Loongarch and UM people to test the change.
Both Loongarch and UM define kasan_arch_is_ready() as the value of their global enable flags, which they switch in kasan_init(). So I would think using kasan_enabled() should just work (minding the metadata-related parts).
Quoted 11 lines
Quoted 2 lines
What was this data access? Is this something we need to fix in the mainline?
I don't believe so (though I spent a while debugging it before I realised I had introduced it by changing
kasan_enabled()dynamically).In
kasan_cache_create()we unconditionally allocate a metadata buffer, but thekasan_init_slab_obj()call to initialise it is guarded bykasan_enabled(). But later parts of the code only check the presence of the buffer before using it, so bad things happen ifkasan_enabled()later turns on (I was getting some error about invalid lock state).
Ah, makes sense. Yeah, these metadata init functions should work even before kasan_flag_enabled is switched on then.