Break heap spraying needed for exploiting use-after-free
Subsystems (6)
v2 cover Break heap spraying needed for exploiting use-after-free
Hello everyone! Requesting for your comments.
This is the second version of the heap quarantine prototype for the Linux kernel. I performed a deeper evaluation of its security properties and developed new features like quarantine randomization and integration with init_on_free. That is fun! See below for more details.
Rationale
=========
Use-after-free vulnerabilities in the Linux kernel are very popular for exploitation. There are many examples, some of them:
https://googleprojectzero.blogspot.com/2018/09/a-cache-invalidation-bug-in-linux.html
https://googleprojectzero.blogspot.com/2019/11/bad-binder-android-in-wild-exploit.html?m=1
https://a13xp0p0v.github.io/2020/02/15/CVE-2019-18683.html
Use-after-free exploits usually employ heap spraying technique. Generally it aims to put controlled bytes at a predetermined memory location on the heap.
Heap spraying for exploiting use-after-free in the Linux kernel relies on the fact that on kmalloc(), the slab allocator returns the address of the memory that was recently freed. So allocating a kernel object with the same size and controlled contents allows overwriting the vulnerable freed object.
I've found an easy way to break the heap spraying for use-after-free exploitation. I extracted slab freelist quarantine from KASAN functionality and called it CONFIG_SLAB_QUARANTINE. Please see patch 1/6.
If this feature is enabled, freed allocations are stored in the quarantine queue where they wait for actual freeing. So they can't be instantly reallocated and overwritten by use-after-free exploits.
N.B. Heap spraying for out-of-bounds exploitation is another technique, heap quarantine doesn't break it.
Security properties
===================
For researching security properties of the heap quarantine I developed 2 lkdtm tests (see the patch 5/6).
The first test is called lkdtm_HEAP_SPRAY. It allocates and frees an object from a separate kmem_cache and then allocates 400000 similar objects. I.e. this test performs an original heap spraying technique for use-after-free exploitation.
If CONFIG_SLAB_QUARANTINE is disabled, the freed object is instantly reallocated and overwritten:
# echo HEAP_SPRAY > /sys/kernel/debug/provoke-crash/DIRECT lkdtm: Performing direct entry HEAP_SPRAY lkdtm: Allocated and freed spray_cache object 000000002b5b3ad4 of size 333 lkdtm: Original heap spraying: allocate 400000 objects of size 333... lkdtm: FAIL: attempt 0: freed object is reallocated
If CONFIG_SLAB_QUARANTINE is enabled, 400000 new allocations don't overwrite the freed object:
# echo HEAP_SPRAY > /sys/kernel/debug/provoke-crash/DIRECT lkdtm: Performing direct entry HEAP_SPRAY lkdtm: Allocated and freed spray_cache object 000000009909e777 of size 333 lkdtm: Original heap spraying: allocate 400000 objects of size 333... lkdtm: OK: original heap spraying hasn't succeed
That happens because pushing an object through the quarantine requires _both_ allocating and freeing memory. Objects are released from the quarantine on new memory allocations, but only when the quarantine size is over the limit. And the quarantine size grows on new memory freeing.
That's why I created the second test called lkdtm_PUSH_THROUGH_QUARANTINE. It allocates and frees an object from a separate kmem_cache and then performs kmem_cache_alloc()+kmem_cache_free() for that cache 400000 times. This test effectively pushes the object through the heap quarantine and reallocates it after it returns back to the allocator freelist:
# echo PUSH_THROUGH_QUARANTINE > /sys/kernel/debug/provoke-crash/ lkdtm: Performing direct entry PUSH_THROUGH_QUARANTINE lkdtm: Allocated and freed spray_cache object 000000008fdb15c3 of size 333 lkdtm: Push through quarantine: allocate and free 400000 objects of size 333... lkdtm: Target object is reallocated at attempt 182994 # echo PUSH_THROUGH_QUARANTINE > /sys/kernel/debug/provoke-crash/ lkdtm: Performing direct entry PUSH_THROUGH_QUARANTINE lkdtm: Allocated and freed spray_cache object 000000004e223cbe of size 333 lkdtm: Push through quarantine: allocate and free 400000 objects of size 333... lkdtm: Target object is reallocated at attempt 186830 # echo PUSH_THROUGH_QUARANTINE > /sys/kernel/debug/provoke-crash/ lkdtm: Performing direct entry PUSH_THROUGH_QUARANTINE lkdtm: Allocated and freed spray_cache object 000000007663a058 of size 333 lkdtm: Push through quarantine: allocate and free 400000 objects of size 333... lkdtm: Target object is reallocated at attempt 182010
As you can see, the number of the allocations that are needed for overwriting the vulnerable object is almost the same. That would be good for stable use-after-free exploitation and should not be allowed.
That's why I developed the quarantine randomization (see the patch 4/6).
This randomization required very small hackish changes of the heap quarantine mechanism. At first all quarantine batches are filled by objects. Then during the quarantine reducing I randomly choose and free 1/2 of objects from a randomly chosen batch. Now the randomized quarantine releases the freed object at an unpredictable moment:
lkdtm: Target object is reallocated at attempt 107884
lkdtm: Target object is reallocated at attempt 265641
lkdtm: Target object is reallocated at attempt 100030
lkdtm: Target object is NOT reallocated in 400000 attempts
lkdtm: Target object is reallocated at attempt 204731
lkdtm: Target object is reallocated at attempt 359333
lkdtm: Target object is reallocated at attempt 289349
lkdtm: Target object is reallocated at attempt 119893
lkdtm: Target object is reallocated at attempt 225202
lkdtm: Target object is reallocated at attempt 87343
However, this randomization alone would not disturb the attacker, because the quarantine stores the attacker's data (the payload) in the sprayed objects. I.e. the reallocated and overwritten vulnerable object contains the payload until the next reallocation (very bad).
Hence heap objects should be erased before going to the heap quarantine. Moreover, filling them by zeros gives a chance to detect use-after-free accesses to non-zero data while an object stays in the quarantine (nice!). That functionality already exists in the kernel, it's called init_on_free. I integrated it with CONFIG_SLAB_QUARANTINE in the patch 3/6.
During that work I found a bug: in CONFIG_SLAB init_on_free happens too late, and heap objects go to the KASAN quarantine being dirty. See the fix in the patch 2/6.
For deeper understanding of the heap quarantine inner workings, I attach the patch 6/6, which contains verbose debugging (not for merge).
It's very helpful, see the output example:
quarantine: PUT 508992 to tail batch 123, whole sz 65118872, batch sz 508854
quarantine: whole sz exceed max by 494552, REDUCE head batch 0 by 415392, leave 396304
quarantine: data level in batches:
0 - 77%
1 - 108%
2 - 83%
3 - 21%
...
125 - 75%
126 - 12%
127 - 108% quarantine: whole sz exceed max by 79160, REDUCE head batch 12 by 14160, leave 17608
quarantine: whole sz exceed max by 65000, REDUCE head batch 75 by 218328, leave 195232
quarantine: PUT 508992 to tail batch 124, whole sz 64979984, batch sz 508854
...
Changes in v2
=============
- Added heap quarantine randomization (the patch 4/6).
- Integrated CONFIG_SLAB_QUARANTINE with init_on_free (the patch 3/6).
- Fixed late init_on_free in CONFIG_SLAB (the patch 2/6).
- Added lkdtm_PUSH_THROUGH_QUARANTINE test.
- Added the quarantine verbose debugging (the patch 6/6, not for merge).
- Improved the descriptions according to the feedback from Kees Cook and Matthew Wilcox.
- Made fixes recommended by Kees Cook:
* Avoided BUG_ON() in kasan_cache_create() by handling the error and reporting with WARN_ON().
* Created a separate kmem_cache for new lkdtm tests.
* Fixed kasan_track.pid type to pid_t.
TODO for the next prototypes
============================
1. Performance evaluation and optimization.
I would really appreciate your ideas about performance testing of a kernel with the heap quarantine. The first prototype was tested with hackbench and kernel build timing (which showed very different numbers). Earlier the developers similarly tested init_on_free functionality. However, Brad Spengler says in his twitter that such testing method is poor.
2. Complete separation of CONFIG_SLAB_QUARANTINE from KASAN (feedback from Andrey Konovalov).
3. Adding a kernel boot parameter for enabling/disabling the heap quaranitne (feedback from Kees Cook).
4. Testing the heap quarantine in near-OOM situations (feedback from Pavel Machek).
5. Does this work somehow help or disturb the integration of the Memory Tagging for the Linux kernel?
6. After rebasing the series onto v5.9.0-rc6, CONFIG_SLAB kernel started to show warnings about few slab caches that have no space for additional metadata. It needs more investigation. I believe it affects KASAN bug detection abilities as well. Warning example:
WARNING: CPU: 0 PID: 0 at mm/kasan/slab_quarantine.c:38 kasan_cache_create+0x37/0x50
Modules linked in:
CPU: 0 PID: 0 Comm: swapper Not tainted 5.9.0-rc6+ #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-2.fc32 04/01/2014
RIP: 0010:kasan_cache_create+0x37/0x50
...
Call Trace:
__kmem_cache_create+0x74/0x250
create_boot_cache+0x6d/0x91
create_kmalloc_cache+0x57/0x93
new_kmalloc_cache+0x39/0x47
create_kmalloc_caches+0x33/0xd9
start_kernel+0x25b/0x532
secondary_startup_64+0xb6/0xc0Thanks in advance for your feedback.
Best regards,
Alexander
Alexander Popov (6):
mm: Extract SLAB_QUARANTINE from KASAN
mm/slab: Perform init_on_free earlier
mm: Integrate SLAB_QUARANTINE with init_on_free
mm: Implement slab quarantine randomization
lkdtm: Add heap quarantine tests
mm: Add heap quarantine verbose debugging (not for merge)
drivers/misc/lkdtm/core.c | 2 + drivers/misc/lkdtm/heap.c | 110 +++++++++++++++++++++++++++++++++++++ drivers/misc/lkdtm/lkdtm.h | 2 + include/linux/kasan.h | 107 ++++++++++++++++++++---------------- include/linux/slab_def.h | 2 +- include/linux/slub_def.h | 2 +- init/Kconfig | 14 +++++ mm/Makefile | 3 +- mm/kasan/Makefile | 2 + mm/kasan/kasan.h | 75 +++++++++++++------------ mm/kasan/quarantine.c | 102 ++++++++++++++++++++++++++++++---- mm/kasan/slab_quarantine.c | 106 +++++++++++++++++++++++++++++++++++ mm/page_alloc.c | 22 ++++++++ mm/slab.c | 5 +- mm/slub.c | 2 +- 15 files changed, 455 insertions(+), 101 deletions(-) create mode 100644 mm/kasan/slab_quarantine.c
Signature · 1 line
2.26.2
RFC PATCH v2 1/6 mm: Extract SLAB_QUARANTINE from KASAN
Heap spraying is an exploitation technique that aims to put controlled bytes at a predetermined memory location on the heap. Heap spraying for exploiting use-after-free in the Linux kernel relies on the fact that on kmalloc(), the slab allocator returns the address of the memory that was recently freed. Allocating a kernel object with the same size and controlled contents allows overwriting the vulnerable freed object.
Let's extract slab freelist quarantine from KASAN functionality and call it CONFIG_SLAB_QUARANTINE. This feature breaks widespread heap spraying technique for exploiting use-after-free vulnerabilities in the kernel code.
If this feature is enabled, freed allocations are stored in the quarantine queue where they wait for actual freeing. So they can't be instantly reallocated and overwritten by use-after-free exploits.
N.B. Heap spraying for out-of-bounds exploitation is another technique, heap quarantine doesn't break it.
--- include/linux/kasan.h | 107 ++++++++++++++++++++----------------- include/linux/slab_def.h | 2 +- include/linux/slub_def.h | 2 +- init/Kconfig | 13 +++++ mm/Makefile | 3 +- mm/kasan/Makefile | 2 + mm/kasan/kasan.h | 75 +++++++++++++------------- mm/kasan/quarantine.c | 2 + mm/kasan/slab_quarantine.c | 106 ++++++++++++++++++++++++++++++++++++ mm/slub.c | 2 +- 10 files changed, 225 insertions(+), 89 deletions(-) create mode 100644 mm/kasan/slab_quarantine.c
include/linux/kasan.h +59 -48
@@ -42,32 +42,14 @@ void kasan_unpoison_task_stack(struct task_struct *task);void kasan_alloc_pages (struct page *page,unsigned int order);void kasan_free_pages (struct page *page,unsigned int order); -void kasan_cache_create (struct kmem_cache *cache,unsigned int *size, -slab_flags_t *flags); -void kasan_poison_slab (struct page *page);void kasan_unpoison_object_data (struct kmem_cache *cache,void *object);void kasan_poison_object_data (struct kmem_cache *cache,void *object);void *__must_check kasan_init_slab_obj (struct kmem_cache *cache,const void *object); -void *__must_check kasan_kmalloc_large (const void *ptr,size_t size, -gfp_t flags);void kasan_kfree_large (void *ptr,unsigned long ip);void kasan_poison_kfree (void *ptr,unsigned long ip); -void *__must_check kasan_kmalloc (struct kmem_cache *s,const void *object, -size_t size,gfp_t flags); -void *__must_check kasan_krealloc (const void *object,size_t new_size, -gfp_t flags); - -void *__must_check kasan_slab_alloc (struct kmem_cache *s,void *object, -gfp_t flags); -bool kasan_slab_free (struct kmem_cache *s,void *object,unsigned long ip); - -struct kasan_cache { -int alloc_meta_offset; -int free_meta_offset; -};/* * These functions provide a special case to support backing module
@@ -107,10 +89,6 @@ static inline void kasan_disable_current(void) {}static inline void kasan_alloc_pages (struct page *page,unsigned int order) {}static inline void kasan_free_pages (struct page *page,unsigned int order) {} -static inline void kasan_cache_create (struct kmem_cache *cache, -unsigned int *size, -slab_flags_t *flags) {} -static inline void kasan_poison_slab (struct page *page) {}static inline void kasan_unpoison_object_data (struct kmem_cache *cache,void *object) {}
@@ -122,17 +100,65 @@ static inline void *kasan_init_slab_obj(struct kmem_cache *cache,return (void *)object; } +static inline void kasan_kfree_large (void *ptr,unsigned long ip) {} +static inline void kasan_poison_kfree (void *ptr,unsigned long ip) {} +static inline void kasan_free_shadow (const struct vm_struct *vm) {} +static inline void kasan_remove_zero_shadow (void *start,unsigned long size) {} +static inline void kasan_unpoison_slab (const void *ptr) {} + +static inline int kasan_module_alloc (void *addr,size_t size) +{ +return 0 ; +} + +static inline int kasan_add_zero_shadow (void *start,unsigned long size) +{ +return 0 ; +} + +static inline size_t kasan_metadata_size (struct kmem_cache *cache) +{ +return 0 ; +} + +#endif /* CONFIG_KASAN */ + +struct kasan_cache { +int alloc_meta_offset; +int free_meta_offset; +}; + +#if defined(CONFIG_KASAN ) || defined(CONFIG_SLAB_QUARANTINE ) + +void kasan_cache_create (struct kmem_cache *cache,unsigned int *size, +slab_flags_t *flags); +void *__must_check kasan_kmalloc_large (const void *ptr,size_t size, +gfp_t flags); +void *__must_check kasan_kmalloc (struct kmem_cache *s,const void *object, +size_t size,gfp_t flags); +void *__must_check kasan_krealloc (const void *object,size_t new_size, +gfp_t flags); +void *__must_check kasan_slab_alloc (struct kmem_cache *s,void *object, +gfp_t flags); +bool kasan_slab_free (struct kmem_cache *s,void *object,unsigned long ip); + +#else /* CONFIG_KASAN || CONFIG_SLAB_QUARANTINE */ + +static inline void kasan_cache_create (struct kmem_cache *cache, +unsigned int *size, +slab_flags_t *flags) {} +static inline void *kasan_kmalloc_large (void *ptr,size_t size,gfp_t flags) {return ptr; } -static inline void kasan_kfree_large (void *ptr,unsigned long ip) {} -static inline void kasan_poison_kfree (void *ptr,unsigned long ip) {} +static inline void *kasan_kmalloc (struct kmem_cache *s,const void *object,size_t size,gfp_t flags) {return (void *)object; } +static inline void *kasan_krealloc (const void *object,size_t new_size,gfp_t flags) {
@@ -144,43 +170,28 @@ static inline void *kasan_slab_alloc(struct kmem_cache *s, void *object, {return object; } +static inline bool kasan_slab_free (struct kmem_cache *s,void *object,unsigned long ip) {return false; } - -static inline int kasan_module_alloc (void *addr,size_t size) {return 0 ; } -static inline void kasan_free_shadow (const struct vm_struct *vm) {} - -static inline int kasan_add_zero_shadow (void *start,unsigned long size) -{ -return 0 ; -} -static inline void kasan_remove_zero_shadow (void *start, -unsigned long size) -{} - -static inline void kasan_unpoison_slab (const void *ptr) { } -static inline size_t kasan_metadata_size (struct kmem_cache *cache) {return 0 ; } - -#endif /* CONFIG_KASAN */ +#endif /* CONFIG_KASAN || CONFIG_SLAB_QUARANTINE */ #ifdef CONFIG_KASAN_GENERIC -#define KASAN_SHADOW_INIT 0 - -void kasan_cache_shrink (struct kmem_cache *cache); -void kasan_cache_shutdown (struct kmem_cache *cache);void kasan_record_aux_stack (void *ptr); -#else /* CONFIG_KASAN_GENERIC */ +static inline void kasan_record_aux_stack (void *ptr) {} +#endif /* CONFIG_KASAN_GENERIC */ +#if defined(CONFIG_KASAN_GENERIC ) || defined(CONFIG_SLAB_QUARANTINE ) +void kasan_cache_shrink (struct kmem_cache *cache); +void kasan_cache_shutdown (struct kmem_cache *cache); +#else /* CONFIG_KASAN_GENERIC || CONFIG_SLAB_QUARANTINE */ static inline void kasan_cache_shrink (struct kmem_cache *cache) {}static inline void kasan_cache_shutdown (struct kmem_cache *cache) {} -static inline void kasan_record_aux_stack (void *ptr) {} - -#endif /* CONFIG_KASAN_GENERIC */ +#endif /* CONFIG_KASAN_GENERIC || CONFIG_SLAB_QUARANTINE */ #ifdef CONFIG_KASAN_SW_TAGS
include/linux/slab_def.h +1 -1
@@ -72,7 +72,7 @@ struct kmem_cache {int obj_offset;#endif /* CONFIG_DEBUG_SLAB */ -#ifdef CONFIG_KASAN +#if defined(CONFIG_KASAN ) || defined(CONFIG_SLAB_QUARANTINE )struct kasan_cache kasan_info;#endif
include/linux/slub_def.h +1 -1
@@ -124,7 +124,7 @@ struct kmem_cache {unsigned int *random_seq;#endif -#ifdef CONFIG_KASAN +#if defined(CONFIG_KASAN ) || defined(CONFIG_SLAB_QUARANTINE )struct kasan_cache kasan_info;#endif
init/Kconfig +13 -0
@@ -1931,6 +1931,19 @@ config SLAB_FREELIST_HARDENED sanity-checking than others. This option is most effective with CONFIG_SLUB. +config SLAB_QUARANTINE + bool "Enable slab freelist quarantine" + depends on !KASAN && (SLAB || SLUB) + help + Enable slab freelist quarantine to delay reusing of freed slab + objects. If this feature is enabled, freed objects are stored + in the quarantine queue where they wait for actual freeing. + So they can't be instantly reallocated and overwritten by + use-after-free exploits. In other words, this feature mitigates + heap spraying technique for exploiting use-after-free + vulnerabilities in the kernel code. + KASAN also employs this feature for use-after-free detection. + config SHUFFLE_PAGE_ALLOCATOR bool "Page allocator randomization" default SLAB_FREELIST_RANDOM && ACPI_NUMA
mm/Makefile +1 -2
@@ -52,7 +52,7 @@ obj-y := filemap.o mempool.o oom_kill.o fadvise.o \ mm_init.o percpu.o slab_common.o \ compaction.o vmacache.o \ interval_tree.o list_lru.o workingset.o \ - debug.o gup.o $(mmu-y) + debug.o gup.o kasan/ $(mmu-y) # Give 'page_alloc' its own module-parameter namespace page-alloc-y := page_alloc.o
@@ -80,7 +80,6 @@ obj-$(CONFIG_KSM) += ksm.o obj-$(CONFIG_PAGE_POISONING) += page_poison.o obj-$(CONFIG_SLAB) += slab.o obj-$(CONFIG_SLUB) += slub.o -obj-$(CONFIG_KASAN) += kasan/ obj-$(CONFIG_FAILSLAB) += failslab.o obj-$(CONFIG_MEMORY_HOTPLUG) += memory_hotplug.o obj-$(CONFIG_MEMTEST) += memtest.o
mm/kasan/Makefile +2 -0
@@ -32,3 +32,5 @@ CFLAGS_tags_report.o := $(CC_FLAGS_KASAN_RUNTIME) obj-$(CONFIG_KASAN) := common.o init.o report.o obj-$(CONFIG_KASAN_GENERIC) += generic.o generic_report.o quarantine.o obj-$(CONFIG_KASAN_SW_TAGS) += tags.o tags_report.o + +obj-$(CONFIG_SLAB_QUARANTINE) += slab_quarantine.o quarantine.o
mm/kasan/kasan.h +39 -36
107 lines not on this page · raw message
mm/kasan/quarantine.c +2 -0
@@ -145,7 +145,9 @@ static void qlink_free(struct qlist_node *qlink, struct kmem_cache *cache)if (IS_ENABLED (CONFIG_SLAB ))local_irq_save (flags); +#ifdef CONFIG_KASAN *(u8 *)kasan_mem_to_shadow (object) =KASAN_KMALLOC_FREE ; +#endif ___cache_free (cache, object, _THIS_IP_);if (IS_ENABLED (CONFIG_SLAB ))
mm/kasan/slab_quarantine.c new file +106 -0
106 lines not on this page · raw message
mm/slub.c +1 -1
@@ -3143,7 +3143,7 @@ static __always_inline void slab_free(struct kmem_cache *s, struct page *page,do_slab_free (s, page, head, tail, cnt, addr); } -#ifdef CONFIG_KASAN_GENERIC +#if defined(CONFIG_KASAN_GENERIC ) || defined(CONFIG_SLAB_QUARANTINE )void ___cache_free (struct kmem_cache *cache,void *x,unsigned long addr) {do_slab_free (cache,virt_to_head_page (x), x,NULL ,1 , addr);
Signature · 1 line
2.26.2
RFC PATCH v2 2/6 mm/slab: Perform init_on_free earlier
In reply to Alexander PopovCurrently in CONFIG_SLAB init_on_free happens too late, and heap objects go to the heap quarantine being dirty. Lets move memory clearing before calling kasan_slab_free() to fix that.
--- mm/slab.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
mm/slab.c +3 -2
@@ -3414,6 +3414,9 @@ static void cache_flusharray(struct kmem_cache *cachep, struct array_cache *ac) static __always_inline void __cache_free(struct kmem_cache *cachep, void *objp,unsigned long caller) { +if (unlikely (slab_want_init_on_free (cachep))) +memset (objp,0 , cachep->object_size); +/* Put the object into the quarantine, don't touch it for now. */ if (kasan_slab_free (cachep, objp, _RET_IP_))return ;
@@ -3432,8 +3435,6 @@ void ___cache_free(struct kmem_cache *cachep, void *objp,struct array_cache *ac =cpu_cache_get (cachep);check_irq_off (); -if (unlikely (slab_want_init_on_free (cachep))) -memset (objp,0 , cachep->object_size);kmemleak_free_recursive (objp, cachep->flags); objp =cache_free_debugcheck (cachep, objp, caller);memcg_slab_free_hook (cachep,virt_to_head_page (objp), objp);
Signature · 1 line
2.26.2
Re: [PATCH RFC v2 2/6] mm/slab: Perform init_on_free earlier
Quoted 6 lines · Alexander Popov
Currently in
CONFIG_SLABinit_on_freehappens too late, and heap objects go to the heap quarantine being dirty. Lets move memory clearing before callingkasan_slab_free()to fix that.Signed-off-by Alexander Popov <[email protected]>
Quoted 30 lines
--- mm/slab.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)mm/slab.c +3
-3@@ -3414,6 +3414,9 @@ static void cache_flusharray(struct kmem_cache *cachep, struct array_cache *ac) static __always_inline void __cache_free(struct kmem_cache *cachep, void *objp,unsigned long caller) { +if (unlikely (slab_want_init_on_free (cachep))) +memset (objp,0 , cachep->object_size); +/* Put the object into the quarantine, don't touch it for now. */ if (kasan_slab_free (cachep, objp, _RET_IP_))return ;@@ -3432,8 +3435,6 @@ void ___cache_free(struct kmem_cache *cachep, void *objp,struct array_cache *ac =cpu_cache_get (cachep);check_irq_off (); -if (unlikely (slab_want_init_on_free (cachep))) -memset (objp,0 , cachep->object_size);kmemleak_free_recursive (objp, cachep->flags); objp =cache_free_debugcheck (cachep, objp, caller);memcg_slab_free_hook (cachep,virt_to_head_page (objp), objp); --2.26.2
Signature · 10 lines
Alexander Potapenko Software Engineer Google Germany GmbH Erika-Mann-Straße, 33 80636 München Geschäftsführer: Paul Manicle, Halimah DeLaine Prado Registergericht und -nummer: Hamburg, HRB 86891 Sitz der Gesellschaft: Hamburg
Re: [PATCH RFC v2 2/6] mm/slab: Perform init_on_free earlier
Quoted 8 lines · Alexander Potapenko
Quoted 6 lines · Alexander Popov
Currently in
CONFIG_SLABinit_on_freehappens too late, and heap objects go to the heap quarantine being dirty. Lets move memory clearing before callingkasan_slab_free()to fix that.Signed-off-by Alexander Popov <[email protected]>Reviewed-by Alexander Potapenko <[email protected]>
Thanks for the review, Alexander!
Do you have any idea how this patch series relates to Memory Tagging support that is currently developed?
Best regards,
Alexander
Re: [PATCH RFC v2 2/6] mm/slab: Perform init_on_free earlier
In reply to Alexander PotapenkoQuoted 8 lines · Alexander Potapenko
Quoted 6 lines · Alexander Popov
Currently in
CONFIG_SLABinit_on_freehappens too late, and heap objects go to the heap quarantine being dirty. Lets move memory clearing before callingkasan_slab_free()to fix that.Signed-off-by Alexander Popov <[email protected]>Reviewed-by Alexander Potapenko <[email protected]>
Hello!
Can this particular patch be considered for the mainline kernel?
Note: I summarized the results of the experiment with the Linux kernel heap quarantine in a short article, for future reference:
https://a13xp0p0v.github.io/2020/11/30/slab-quarantine.html
Best regards,
Alexander
Quoted 32 lines
Quoted 30 lines
--- mm/slab.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)mm/slab.c +3
-3@@ -3414,6 +3414,9 @@ static void cache_flusharray(struct kmem_cache *cachep, struct array_cache *ac) static __always_inline void __cache_free(struct kmem_cache *cachep, void *objp,unsigned long caller) { +if (unlikely (slab_want_init_on_free (cachep))) +memset (objp,0 , cachep->object_size); +/* Put the object into the quarantine, don't touch it for now. */ if (kasan_slab_free (cachep, objp, _RET_IP_))return ;@@ -3432,8 +3435,6 @@ void ___cache_free(struct kmem_cache *cachep, void *objp,struct array_cache *ac =cpu_cache_get (cachep);check_irq_off (); -if (unlikely (slab_want_init_on_free (cachep))) -memset (objp,0 , cachep->object_size);kmemleak_free_recursive (objp, cachep->flags); objp =cache_free_debugcheck (cachep, objp, caller);memcg_slab_free_hook (cachep,virt_to_head_page (objp), objp); --2.26.2
Andrew Morton
maintainerRe: [PATCH RFC v2 2/6] mm/slab: Perform init_on_free earlier
Quoted 7 lines · On Thu, 3 Dec 2020 22:50:27 +0300 Alexander Popov <[email protected]> wrote:
Quoted 6 lines · Alexander Potapenko
Quoted 5 lines · Alexander Popov
Currently in
CONFIG_SLABinit_on_freehappens too late, and heap objects go to the heap quarantine being dirty. Lets move memory clearing before callingkasan_slab_free()to fix that.
Quoted 6 lines
Quoted 2 lines
Quoted 1 line
Signed-off-by Alexander Popov <[email protected]>Reviewed-by Alexander Potapenko <[email protected]>Hello!
Can this particular patch be considered for the mainline kernel?
All patches are considered ;) And merged if they're reviewed, tested, judged useful, etc.
If you think this particular patch should be fast-tracked then please send it as a non-RFC, standalone patch. Please also enhance the changelog so that it actually explains what goes wrong. Presumably "objects go to the heap quarantine being dirty" causes some user-visible problem? What is that problem?
Re: [PATCH RFC v2 2/6] mm/slab: Perform init_on_free earlier
Quoted 16 lines · Andrew Morton
Quoted 7 lines · On Thu, 3 Dec 2020 22:50:27 +0300 Alexander Popov <[email protected]> wrote:
Quoted 6 lines · Alexander Potapenko
Quoted 5 lines · Alexander Popov
Currently in
CONFIG_SLABinit_on_freehappens too late, and heap objects go to the heap quarantine being dirty. Lets move memory clearing before callingkasan_slab_free()to fix that.Quoted 6 lines
Quoted 2 lines
Quoted 1 line
Signed-off-by Alexander Popov <[email protected]>Reviewed-by Alexander Potapenko <[email protected]>Hello!
Can this particular patch be considered for the mainline kernel?
Quoted 8 lines
All patches are considered ;) And merged if they're reviewed, tested, judged useful, etc.
If you think this particular patch should be fast-tracked then please send it as a non-RFC, standalone patch. Please also enhance the changelog so that it actually explains what goes wrong. Presumably "objects go to the heap quarantine being dirty" causes some user-visible problem? What is that problem?
Ok, thanks!
I'll improve the commit message and send the patch separately.
Best regards,
Alexander
RFC PATCH v2 3/6 mm: Integrate SLAB_QUARANTINE with init_on_free
In reply to Alexander PopovHaving slab quarantine without memory erasing is harmful. If the quarantined objects are not cleaned and contain data, then:
1. they will be useful for use-after-free exploitation,
2. there is no chance to detect use-after-free access.
So we want the quarantined objects to be erased.
Enable init_on_free that cleans objects before placing them into the quarantine. CONFIG_PAGE_POISONING should be disabled since it cuts off init_on_free.
--- init/Kconfig | 3 ++- mm/page_alloc.c | 22 ++++++++++++++++++++++ 2 files changed, 24 insertions(+), 1 deletion(-)
init/Kconfig +2 -1
@@ -1933,7 +1933,8 @@ config SLAB_FREELIST_HARDENED config SLAB_QUARANTINE bool "Enable slab freelist quarantine" - depends on !KASAN && (SLAB || SLUB) + depends on !KASAN && (SLAB || SLUB) && !PAGE_POISONING + select INIT_ON_FREE_DEFAULT_ON help Enable slab freelist quarantine to delay reusing of freed slab objects. If this feature is enabled, freed objects are stored
mm/page_alloc.c +22 -0
@@ -168,6 +168,27 @@ static int __init early_init_on_alloc(char *buf) }early_param ("init_on_alloc" , early_init_on_alloc); +#ifdef CONFIG_SLAB_QUARANTINE +static int __init early_init_on_free (char *buf) +{ +/* +* Having slab quarantine without memory erasing is harmful. +* If the quarantined objects are not cleaned and contain data, then: +* 1. they will be useful for use-after-free exploitation, +* 2. use-after-free access may not be detected. +* So we want the quarantined objects to be erased. +* +* Enable init_on_free that cleans objects before placing them into +* the quarantine. CONFIG_PAGE_POISONING should be disabled since it +* cuts off init_on_free. +*/ +BUILD_BUG_ON (!IS_ENABLED (CONFIG_INIT_ON_FREE_DEFAULT_ON )); +BUILD_BUG_ON (IS_ENABLED (CONFIG_PAGE_POISONING )); +pr_info ("mem auto-init: init_on_free is on for CONFIG_SLAB_QUARANTINE\n" ); + +return 0 ; +} +#else /* CONFIG_SLAB_QUARANTINE */ static int __init early_init_on_free (char *buf) {int ret;
@@ -184,6 +205,7 @@ static int __init early_init_on_free(char *buf)static_branch_disable (&init_on_free); return ret; } +#endif /* CONFIG_SLAB_QUARANTINE */ early_param ("init_on_free" , early_init_on_free);/*
Signature · 1 line
2.26.2
RFC PATCH v2 4/6 mm: Implement slab quarantine randomization
In reply to Alexander PopovThe randomization is very important for the slab quarantine security properties. Without it the number of kmalloc()+kfree() calls that are needed for overwriting the vulnerable object is almost the same. That would be good for stable use-after-free exploitation, and we should not allow that.
This commit contains very compact and hackish changes that introduce the quarantine randomization. At first all quarantine batches are filled by objects. Then during the quarantine reducing we randomly choose and free 1/2 of objects from a randomly chosen batch. Now the randomized quarantine releases the freed object at an unpredictable moment, which is harmful for the heap spraying technique employed by use-after-free exploits.
--- mm/kasan/quarantine.c | 79 +++++++++++++++++++++++++++++++++++++------ 1 file changed, 69 insertions(+), 10 deletions(-)
mm/kasan/quarantine.c +69 -10
@@ -29,6 +29,7 @@#include <linux/srcu.h> #include <linux/string.h> #include <linux/types.h> +#include <linux/random.h> #include "../slab.h" #include "kasan.h"
@@ -89,8 +90,13 @@ static void qlist_move_all(struct qlist_head *from, struct qlist_head *to) }#define QUARANTINE_PERCPU_SIZE (1 << 20) + +#ifdef CONFIG_KASAN #define QUARANTINE_BATCHES \ (1024 > 4 * CONFIG_NR_CPUS ? 1024 : 4 * CONFIG_NR_CPUS) +#else +#define QUARANTINE_BATCHES 128 +#endif /* * The object quarantine consists of per-cpu queues and a global queue,
@@ -110,10 +116,7 @@ DEFINE_STATIC_SRCU(remove_cache_srcu);/* Maximum size of the global queue. */ static unsigned long quarantine_max_size; -/* -* Target size of a batch in global_quarantine. -* Usually equal to QUARANTINE_PERCPU_SIZE unless we have too much RAM. -*/ +/* Target size of a batch in global_quarantine. */ static unsigned long quarantine_batch_size;/*
@@ -191,7 +194,12 @@ void quarantine_put(struct kasan_free_meta *info, struct kmem_cache *cache) q =this_cpu_ptr (&cpu_quarantine);qlist_put (q, &info->quarantine_link, cache->size); +#ifdef CONFIG_KASAN if (unlikely (q->bytes >QUARANTINE_PERCPU_SIZE )) { +#else +if (unlikely (q->bytes >min_t (size_t,QUARANTINE_PERCPU_SIZE , +READ_ONCE (quarantine_batch_size)))) { +#endif qlist_move_all (q, &temp);raw_spin_lock (&quarantine_lock);
@@ -204,7 +212,7 @@ void quarantine_put(struct kasan_free_meta *info, struct kmem_cache *cache) new_tail = quarantine_tail +1 ;if (new_tail ==QUARANTINE_BATCHES ) new_tail =0 ; -if (new_tail != quarantine_head) +if (new_tail != quarantine_head || !IS_ENABLED (CONFIG_KASAN )) quarantine_tail = new_tail; }raw_spin_unlock (&quarantine_lock);
@@ -213,12 +221,43 @@ void quarantine_put(struct kasan_free_meta *info, struct kmem_cache *cache)local_irq_restore (flags); } +static void qlist_move_random (struct qlist_head *from,struct qlist_head *to) +{ +struct qlist_node *curr; + +if (unlikely (qlist_empty (from))) +return ; + + curr = from->head; +qlist_init (from); +while (curr) { +struct qlist_node *next = curr->next; +struct kmem_cache *obj_cache =qlink_to_cache (curr); +int rnd =get_random_int (); + +/* +* Hackish quarantine randomization, part 2: +* move only 1/2 of objects to the destination list. +* TODO: use random bits sparingly for better performance. +*/ +if (rnd %2 ==0 ) +qlist_put (to, curr, obj_cache->size); +else +qlist_put (from, curr, obj_cache->size); + + curr = next; + } +} +void quarantine_reduce (void ) { -size_t total_size, new_quarantine_size, percpu_quarantines; +size_t total_size;unsigned long flags;int srcu_idx;struct qlist_head to_free =QLIST_INIT ; +#ifdef CONFIG_KASAN +size_t new_quarantine_size, percpu_quarantines; +#endif if (likely (READ_ONCE (quarantine_size) <=READ_ONCE (quarantine_max_size)))
@@ -236,12 +275,12 @@ void quarantine_reduce(void) srcu_idx =srcu_read_lock (&remove_cache_srcu);raw_spin_lock_irqsave (&quarantine_lock, flags); -/* -* Update quarantine size in case of hotplug. Allocate a fraction of -* the installed memory to quarantine minus per-cpu queue limits. -*/ +/* Update quarantine size in case of hotplug */ total_size = (totalram_pages () <<PAGE_SHIFT ) /QUARANTINE_FRACTION ; + +#ifdef CONFIG_KASAN +/* Subtract per-cpu queue limits from total quarantine size */ percpu_quarantines =QUARANTINE_PERCPU_SIZE *num_online_cpus (); new_quarantine_size = (total_size < percpu_quarantines) ?0 : total_size - percpu_quarantines;
@@ -257,6 +296,26 @@ void quarantine_reduce(void)if (quarantine_head ==QUARANTINE_BATCHES ) quarantine_head = 0; } +#else /* CONFIG_KASAN */ +/* +* Don't subtract per-cpu queue limits from total quarantine +* size to consume all quarantine slots. +*/ +WRITE_ONCE (quarantine_max_size, total_size); +WRITE_ONCE (quarantine_batch_size, total_size /QUARANTINE_BATCHES ); + +/* +* Hackish quarantine randomization, part 1: +* pick a random batch for reducing. +*/ +if (likely (quarantine_size > quarantine_max_size)) { +do { + quarantine_head =get_random_int () %QUARANTINE_BATCHES ; + }while (quarantine_head == quarantine_tail); +qlist_move_random (&global_quarantine[quarantine_head], &to_free); +WRITE_ONCE (quarantine_size, quarantine_size - to_free.bytes); + } +#endif raw_spin_unlock_irqrestore (&quarantine_lock, flags);
Signature · 1 line
2.26.2
RFC PATCH v2 5/6 lkdtm: Add heap quarantine tests
In reply to Alexander PopovAdd tests for CONFIG_SLAB_QUARANTINE.
The HEAP_SPRAY test aims to reallocate a recently freed heap object. It allocates and frees an object from a separate kmem_cache and then allocates 400000 similar objects from it. I.e. this test performs an original heap spraying technique for use-after-free exploitation. If CONFIG_SLAB_QUARANTINE is disabled, the freed object is instantly reallocated and overwritten, which is required for a successful attack.
The PUSH_THROUGH_QUARANTINE test allocates and frees an object from a separate kmem_cache and then performs kmem_cache_alloc()+kmem_cache_free() 400000 times. This test pushes the object through the heap quarantine and reallocates it after it returns back to the allocator freelist. If CONFIG_SLAB_QUARANTINE is enabled, this test should show that the randomized quarantine will release the freed object at an unpredictable moment, which makes use-after-free exploitation much harder.
--- drivers/misc/lkdtm/core.c | 2 + drivers/misc/lkdtm/heap.c | 110 +++++++++++++++++++++++++++++++++++++ drivers/misc/lkdtm/lkdtm.h | 2 + 3 files changed, 114 insertions(+)
drivers/misc/lkdtm/core.c +2 -0
@@ -126,6 +126,8 @@ static const struct crashtype crashtypes[] = {CRASHTYPE (SLAB_FREE_DOUBLE ),CRASHTYPE (SLAB_FREE_CROSS ),CRASHTYPE (SLAB_FREE_PAGE ), +CRASHTYPE (HEAP_SPRAY ), +CRASHTYPE (PUSH_THROUGH_QUARANTINE ),CRASHTYPE (SOFTLOCKUP ),CRASHTYPE (HARDLOCKUP ), CRASHTYPE(SPINLOCKUP),
drivers/misc/lkdtm/heap.c +110 -0
132 lines not on this page · raw message
drivers/misc/lkdtm/lkdtm.h +2 -0
@@ -45,6 +45,8 @@ void lkdtm_READ_BUDDY_AFTER_FREE(void);void lkdtm_SLAB_FREE_DOUBLE (void );void lkdtm_SLAB_FREE_CROSS (void );void lkdtm_SLAB_FREE_PAGE (void ); +void lkdtm_HEAP_SPRAY (void ); +void lkdtm_PUSH_THROUGH_QUARANTINE (void );/* lkdtm_perms.c */ void __init lkdtm_perms_init (void );
Signature · 1 line
2.26.2
RFC PATCH v2 6/6 mm: Add heap quarantine verbose debugging (not for merge)
In reply to Alexander PopovAdd verbose debugging for deeper understanding of the heap quarantine inner workings (this patch is not for merge).
--- mm/kasan/quarantine.c | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+)
mm/kasan/quarantine.c +21 -0
@@ -203,6 +203,12 @@ void quarantine_put(struct kasan_free_meta *info, struct kmem_cache *cache)qlist_move_all (q, &temp);raw_spin_lock (&quarantine_lock); + +pr_info ("quarantine: PUT %zu to tail batch %d, whole sz %zu, batch sz %lu\n" , + temp.bytes, quarantine_tail, +READ_ONCE (quarantine_size), +READ_ONCE (quarantine_batch_size)); +WRITE_ONCE (quarantine_size, quarantine_size + temp.bytes);qlist_move_all (&temp, &global_quarantine[quarantine_tail]); if (global_quarantine[quarantine_tail].bytes >=
@@ -313,7 +319,22 @@ void quarantine_reduce(void) quarantine_head = get_random_int() % QUARANTINE_BATCHES; }while (quarantine_head == quarantine_tail);qlist_move_random (&global_quarantine[quarantine_head], &to_free); +pr_info ("quarantine: whole sz exceed max by %lu, REDUCE head batch %d by %zu, leave %zu\n" , + quarantine_size - quarantine_max_size, + quarantine_head, to_free.bytes, + global_quarantine[quarantine_head].bytes);WRITE_ONCE (quarantine_size, quarantine_size - to_free.bytes); + +if (quarantine_head ==0 ) { +unsigned long i; + +pr_info ("quarantine: data level in batches:" ); +for (i =0 ; i <QUARANTINE_BATCHES ; i++) { +pr_info (" %lu - %lu%%\n" , + i, global_quarantine[i].bytes * +100 / quarantine_batch_size); + } + } }#endif
Signature · 1 line
2.26.2
Hello! I have some performance numbers. Please see below.
Quoted 177 lines · Alexander Popov
Hello everyone! Requesting for your comments.
This is the second version of the heap quarantine prototype for the Linux kernel. I performed a deeper evaluation of its security properties and developed new features like quarantine randomization and integration with
init_on_free. That is fun! See below for more details.Rationale
=========Use-after-free vulnerabilities in the Linux kernel are very popular for exploitation. There are many examples, some of them:
https://googleprojectzero.blogspot.com/2018/09/a-cache-invalidation-bug-in-linux.html
https://googleprojectzero.blogspot.com/2019/11/bad-binder-android-in-wild-exploit.html?m=1
https://a13xp0p0v.github.io/2020/02/15/CVE-2019-18683.htmlUse-after-free exploits usually employ heap spraying technique. Generally it aims to put controlled bytes at a predetermined memory location on the heap.
Heap spraying for exploiting use-after-free in the Linux kernel relies on the fact that on
kmalloc(), the slab allocator returns the address of the memory that was recently freed. So allocating a kernel object with the same size and controlled contents allows overwriting the vulnerable freed object.I've found an easy way to break the heap spraying for use-after-free exploitation. I extracted slab freelist quarantine from KASAN functionality and called it
CONFIG_SLAB_QUARANTINE. Please see patch 1/6.If this feature is enabled, freed allocations are stored in the quarantine queue where they wait for actual freeing. So they can't be instantly reallocated and overwritten by use-after-free exploits.
N.B. Heap spraying for out-of-bounds exploitation is another technique, heap quarantine doesn't break it.
Security properties
===================For researching security properties of the heap quarantine I developed 2 lkdtm tests (see the patch 5/6).
The first test is called
lkdtm_HEAP_SPRAY. It allocates and frees an object from a separatekmem_cacheand then allocates 400000 similar objects. I.e. this test performs an original heap spraying technique for use-after-free exploitation.If
CONFIG_SLAB_QUARANTINEis disabled, the freed object is instantly reallocated and overwritten:# echo HEAP_SPRAY > /sys/kernel/debug/provoke-crash/DIRECT lkdtm: Performing direct entry HEAP_SPRAY lkdtm: Allocated and freed spray_cache object 000000002b5b3ad4 of size 333 lkdtm: Original heap spraying: allocate 400000 objects of size 333... lkdtm: FAIL: attempt 0: freed object is reallocatedIf
CONFIG_SLAB_QUARANTINEis enabled, 400000 new allocations don't overwrite the freed object:# echo HEAP_SPRAY > /sys/kernel/debug/provoke-crash/DIRECT lkdtm: Performing direct entry HEAP_SPRAY lkdtm: Allocated and freed spray_cache object 000000009909e777 of size 333 lkdtm: Original heap spraying: allocate 400000 objects of size 333... lkdtm: OK: original heap spraying hasn't succeedThat happens because pushing an object through the quarantine requires _both_ allocating and freeing memory. Objects are released from the quarantine on new memory allocations, but only when the quarantine size is over the limit. And the quarantine size grows on new memory freeing.
That's why I created the second test called
lkdtm_PUSH_THROUGH_QUARANTINE. It allocates and frees an object from a separatekmem_cacheand then performs kmem_cache_alloc()+kmem_cache_free() for that cache 400000 times. This test effectively pushes the object through the heap quarantine and reallocates it after it returns back to the allocator freelist:# echo PUSH_THROUGH_QUARANTINE > /sys/kernel/debug/provoke-crash/ lkdtm: Performing direct entry PUSH_THROUGH_QUARANTINE lkdtm: Allocated and freed spray_cache object 000000008fdb15c3 of size 333 lkdtm: Push through quarantine: allocate and free 400000 objects of size 333... lkdtm: Target object is reallocated at attempt 182994 # echo PUSH_THROUGH_QUARANTINE > /sys/kernel/debug/provoke-crash/ lkdtm: Performing direct entry PUSH_THROUGH_QUARANTINE lkdtm: Allocated and freed spray_cache object 000000004e223cbe of size 333 lkdtm: Push through quarantine: allocate and free 400000 objects of size 333... lkdtm: Target object is reallocated at attempt 186830 # echo PUSH_THROUGH_QUARANTINE > /sys/kernel/debug/provoke-crash/ lkdtm: Performing direct entry PUSH_THROUGH_QUARANTINE lkdtm: Allocated and freed spray_cache object 000000007663a058 of size 333 lkdtm: Push through quarantine: allocate and free 400000 objects of size 333... lkdtm: Target object is reallocated at attempt 182010As you can see, the number of the allocations that are needed for overwriting the vulnerable object is almost the same. That would be good for stable use-after-free exploitation and should not be allowed.
That's why I developed the quarantine randomization (see the patch 4/6).This randomization required very small hackish changes of the heap quarantine mechanism. At first all quarantine batches are filled by objects. Then during the quarantine reducing I randomly choose and free 1/2 of objects from a randomly chosen batch. Now the randomized quarantine releases the freed object at an unpredictable moment:
lkdtm: Target object is reallocated at attempt 107884
lkdtm: Target object is reallocated at attempt 265641
lkdtm: Target object is reallocated at attempt 100030
lkdtm: Target object is NOT reallocated in 400000 attempts
lkdtm: Target object is reallocated at attempt 204731
lkdtm: Target object is reallocated at attempt 359333
lkdtm: Target object is reallocated at attempt 289349
lkdtm: Target object is reallocated at attempt 119893
lkdtm: Target object is reallocated at attempt 225202
lkdtm: Target object is reallocated at attempt 87343However, this randomization alone would not disturb the attacker, because the quarantine stores the attacker's data (the payload) in the sprayed objects. I.e. the reallocated and overwritten vulnerable object contains the payload until the next reallocation (very bad).
Hence heap objects should be erased before going to the heap quarantine. Moreover, filling them by zeros gives a chance to detect use-after-free accesses to non-zero data while an object stays in the quarantine (nice!). That functionality already exists in the kernel, it's called
init_on_free. I integrated it withCONFIG_SLAB_QUARANTINEin the patch 3/6.During that work I found a bug: in
CONFIG_SLABinit_on_freehappens too late, and heap objects go to the KASAN quarantine being dirty. See the fix in the patch 2/6.For deeper understanding of the heap quarantine inner workings, I attach the patch 6/6, which contains verbose debugging (not for merge). It's very helpful, see the output example:
quarantine: PUT 508992 to tail batch 123, whole sz 65118872, batch sz 508854 quarantine: whole sz exceed max by 494552, REDUCE head batch 0 by 415392, leave 396304 quarantine: data level in batches: 0 - 77% 1 - 108% 2 - 83% 3 - 21% ... 125 - 75% 126 - 12% 127 - 108% quarantine: whole sz exceed max by 79160, REDUCE head batch 12 by 14160, leave 17608 quarantine: whole sz exceed max by 65000, REDUCE head batch 75 by 218328, leave 195232 quarantine: PUT 508992 to tail batch 124, whole sz 64979984, batch sz 508854 ...Changes in v2
=============- Added heap quarantine randomization (the patch 4/6).
- Integrated
CONFIG_SLAB_QUARANTINEwithinit_on_free(the patch 3/6).- Fixed late
init_on_freeinCONFIG_SLAB(the patch 2/6).- Added
lkdtm_PUSH_THROUGH_QUARANTINEtest.- Added the quarantine verbose debugging (the patch 6/6, not for merge).
- Improved the descriptions according to the feedback from Kees Cook and Matthew Wilcox.
- Made fixes recommended by Kees Cook:
* Avoided
BUG_ON()inkasan_cache_create()by handling the error and reporting withWARN_ON().* Created a separate
kmem_cachefor new lkdtm tests.* Fixed kasan_track.pid type to
pid_t.TODO for the next prototypes
============================
Quoted 7 lines
1. Performance evaluation and optimization.
I would really appreciate your ideas about performance testing of a kernel with the heap quarantine. The first prototype was tested with hackbench and kernel build timing (which showed very different numbers). Earlier the developers similarly testedinit_on_freefunctionality. However, Brad Spengler says in his twitter that such testing method is poor.
I've made various tests on real hardware and in virtual machines:
1) network throughput test using iperf
server: iperf -s -f K
client: iperf -c 127.0.0.1 -t 60 -f K
2) scheduler stress test
hackbench -s 4000 -l 500 -g 15 -f 25 -P
3) building the defconfig kernel
time make -j2
I compared Linux kernel 5.9.0-rc6 with:
- init_on_free=off,
- init_on_free=on,
- CONFIG_SLAB_QUARANTINE=y (which enables init_on_free).
Each test was performed 5 times. I will show the mean values.
If you are interested, I can share all the results and calculate standard deviation.
Real hardware, Intel Core i7-6500U CPU
1) Network throughput test with iperf
init_on_free=off: 5467152.2 KBytes/sec
init_on_free=on: 3937545 KBytes/sec (-28.0% vs init_on_free=off)
CONFIG_SLAB_QUARANTINE: 3858848.6 KBytes/sec (-2.0% vs init_on_free=on)
2) Scheduler stress test with hackbench
init_on_free=off: 8.5364s
init_on_free=on: 8.9858s (+5.3% vs init_on_free=off)
CONFIG_SLAB_QUARANTINE: 17.2232s (+91.7% vs init_on_free=on)
3) Building the defconfig kernel:
init_on_free=off: 10m54.475s
init_on_free=on: 11m5.745s (+1.7% vs init_on_free=off)
CONFIG_SLAB_QUARANTINE: 11m13.291s (+1.1% vs init_on_free=on)Virtual machine, QEMU/KVM
1) Network throughput test with iperf
init_on_free=off: 3554237.4 KBytes/sec
init_on_free=on: 2828887.4 KBytes/sec (-20.4% vs init_on_free=off)
CONFIG_SLAB_QUARANTINE: 2587308.2 KBytes/sec (-8.5% vs init_on_free=on)
2) Scheduler stress test with hackbench
init_on_free=off: 19.3602s
init_on_free=on: 20.8854s (+7.9% vs init_on_free=off)
CONFIG_SLAB_QUARANTINE: 30.0746s (+44.0% vs init_on_free=on)We can see that the results of these tests are quite diverse. Your interpretation of the results and ideas of other tests are welcome.
N.B. There was NO performance optimization made for this version of the heap quarantine prototype. The main effort was put into researching its security properties (hope for your feedback). Performance optimization will be done in further steps, if we see that my work is worth doing.
Quoted 62 lines
2. Complete separation of
CONFIG_SLAB_QUARANTINEfrom KASAN (feedback from Andrey Konovalov).3. Adding a kernel boot parameter for enabling/disabling the heap quaranitne (feedback from Kees Cook).
4. Testing the heap quarantine in near-OOM situations (feedback from Pavel Machek).
5. Does this work somehow help or disturb the integration of the Memory Tagging for the Linux kernel?
6. After rebasing the series onto v5.9.0-rc6,
CONFIG_SLABkernel started to show warnings about few slab caches that have no space for additional metadata. It needs more investigation. I believe it affects KASAN bug detection abilities as well. Warning example:WARNING: CPU: 0 PID: 0 at mm/kasan/slab_quarantine.c:38 kasan_cache_create+0x37/0x50 Modules linked in: CPU: 0 PID: 0 Comm: swapper Not tainted 5.9.0-rc6+ #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-2.fc32 04/01/2014 RIP: 0010:kasan_cache_create+0x37/0x50 ... Call Trace: __kmem_cache_create+0x74/0x250 create_boot_cache+0x6d/0x91 create_kmalloc_cache+0x57/0x93 new_kmalloc_cache+0x39/0x47 create_kmalloc_caches+0x33/0xd9 start_kernel+0x25b/0x532 secondary_startup_64+0xb6/0xc0Thanks in advance for your feedback.
Best regards,
AlexanderAlexander Popov (6):
mm: ExtractSLAB_QUARANTINEfrom KASAN
mm/slab: Performinit_on_freeearlier
mm: IntegrateSLAB_QUARANTINEwithinit_on_free
mm: Implement slab quarantine randomization
lkdtm: Add heap quarantine tests
mm: Add heap quarantine verbose debugging (not for merge)drivers/misc/lkdtm/core.c | 2 + drivers/misc/lkdtm/heap.c | 110 +++++++++++++++++++++++++++++++++++++ drivers/misc/lkdtm/lkdtm.h | 2 + include/linux/kasan.h | 107 ++++++++++++++++++++---------------- include/linux/slab_def.h | 2 +- include/linux/slub_def.h | 2 +- init/Kconfig | 14 +++++ mm/Makefile | 3 +- mm/kasan/Makefile | 2 + mm/kasan/kasan.h | 75 +++++++++++++------------ mm/kasan/quarantine.c | 102 ++++++++++++++++++++++++++++++---- mm/kasan/slab_quarantine.c | 106 +++++++++++++++++++++++++++++++++++ mm/page_alloc.c | 22 ++++++++ mm/slab.c | 5 +- mm/slub.c | 2 +- 15 files changed, 455 insertions(+), 101 deletions(-) create mode 100644 mm/kasan/slab_quarantine.c
Quoted 7 lines · Alexander Popov
Quoted 6 lines · Alexander Popov
This is the second version of the heap quarantine prototype for the Linux kernel. I performed a deeper evaluation of its security properties and developed new features like quarantine randomization and integration with
init_on_free. That is fun! See below for more details.
Quoted 6 lines
Quoted 6 lines
Rationale
=========Use-after-free vulnerabilities in the Linux kernel are very popular for exploitation. There are many examples, some of them:
https://googleprojectzero.blogspot.com/2018/09/a-cache-invalidation-bug-in-linux.html
I don't think your proposed mitigation would work with much reliability against this bug; the attacker has full control over the timing of the original use and the following use, so an attacker should be able to trigger the kmem_cache_free(), then spam enough new VMAs and delete them to flush out the quarantine, and then do heap spraying as normal, or something like that.
Also, note that here, if the reallocation fails, the kernel still wouldn't crash because the dangling object is not accessed further if the address range stored in it doesn't match the fault address. So an attacker could potentially try multiple times, and if the object happens to be on the quarantine the first time, that wouldn't really be a showstopper, you'd just try again.
Quoted 1 line
I think that here, again, the free() and the dangling pointer use were caused by separate syscalls, meaning the attacker had control over that timing?
Quoted 1 line
Haven't looked at that one in detail.
Quoted 3 lines
Quoted 3 lines
Use-after-free exploits usually employ heap spraying technique. Generally it aims to put controlled bytes at a predetermined memory location on the heap.
Well, not necessarily "predetermined". Depending on the circumstances, you don't necessarily need to know which address you're writing to; and you might not even need to overwrite a specific object, but instead just have to overwrite one out of a bunch of objects, no matter which.
Quoted 3 lines
Quoted 3 lines
Heap spraying for exploiting use-after-free in the Linux kernel relies on the fact that on
kmalloc(), the slab allocator returns the address of the memory that was recently freed.
Yeah; and that behavior is pretty critical for performance. The longer it's been since a newly allocated object was freed, the higher the chance that you'll end up having to go further down the memory cache hierarchy.
Quoted 3 lines
Quoted 3 lines
So allocating a kernel object with
the same size and controlled contents allows overwriting the vulnerable freed object.
The vmacache exploit you linked to doesn't do that, it frees the object all the way back to the page allocator and then sprays 4MiB of memory from the page allocator. (Because VMAs use their own kmem_cache, and the kmem_cache wasn't merged with any interesting ones, and I saw no good way to exploit the bug by reallocating another VMA over the old VMA back then. Although of course that doesn't mean that there is no such way.)
[...]
Quoted 37 lines
Quoted 37 lines
Security properties
===================For researching security properties of the heap quarantine I developed 2 lkdtm tests (see the patch 5/6).
The first test is called
lkdtm_HEAP_SPRAY. It allocates and frees an object from a separatekmem_cacheand then allocates 400000 similar objects. I.e. this test performs an original heap spraying technique for use-after-free exploitation.If
CONFIG_SLAB_QUARANTINEis disabled, the freed object is instantly reallocated and overwritten:# echo HEAP_SPRAY > /sys/kernel/debug/provoke-crash/DIRECT lkdtm: Performing direct entry HEAP_SPRAY lkdtm: Allocated and freed spray_cache object 000000002b5b3ad4 of size 333 lkdtm: Original heap spraying: allocate 400000 objects of size 333... lkdtm: FAIL: attempt 0: freed object is reallocatedIf
CONFIG_SLAB_QUARANTINEis enabled, 400000 new allocations don't overwrite the freed object:# echo HEAP_SPRAY > /sys/kernel/debug/provoke-crash/DIRECT lkdtm: Performing direct entry HEAP_SPRAY lkdtm: Allocated and freed spray_cache object 000000009909e777 of size 333 lkdtm: Original heap spraying: allocate 400000 objects of size 333... lkdtm: OK: original heap spraying hasn't succeedThat happens because pushing an object through the quarantine requires _both_ allocating and freeing memory. Objects are released from the quarantine on new memory allocations, but only when the quarantine size is over the limit. And the quarantine size grows on new memory freeing.
That's why I created the second test called
lkdtm_PUSH_THROUGH_QUARANTINE. It allocates and frees an object from a separatekmem_cacheand then performs kmem_cache_alloc()+kmem_cache_free() for that cache 400000 times. This test effectively pushes the object through the heap quarantine and reallocates it after it returns back to the allocator freelist:
[...]
Quoted 11 lines
Quoted 11 lines
As you can see, the number of the allocations that are needed for overwriting the vulnerable object is almost the same. That would be good for stable use-after-free exploitation and should not be allowed.
That's why I developed the quarantine randomization (see the patch 4/6).This randomization required very small hackish changes of the heap quarantine mechanism. At first all quarantine batches are filled by objects. Then during the quarantine reducing I randomly choose and free 1/2 of objects from a randomly chosen batch. Now the randomized quarantine releases the freed object at an unpredictable moment:
lkdtm: Target object is reallocated at attempt 107884
[...]
Quoted 1 line
Quoted 1 line
lkdtm: Target object is reallocated at attempt 87343
Those numbers are fairly big. At that point you might not even fit into L3 cache anymore, right? You'd often be hitting DRAM for new allocations? And for many slabs, you might end using much more memory for the quarantine than for actual in-use allocations.
It seems to me like, for this to stop attacks with a high probability, you'd have to reserve a huge chunk of kernel memory for the quarantines - even if the attacker doesn't know anything about the status of the quarantine (which isn't necessarily the case, depending on whether the attacker can abuse microarchitectural data leakage, or if the attacker can trigger a pure data read through the dangling pointer), they should still be able to win with a probability around quarantine_size/allocated_memory_size if they have a heap spraying primitive without strict limits.
Quoted 14 lines
Quoted 14 lines
However, this randomization alone would not disturb the attacker, because the quarantine stores the attacker's data (the payload) in the sprayed objects. I.e. the reallocated and overwritten vulnerable object contains the payload until the next reallocation (very bad).
Hence heap objects should be erased before going to the heap quarantine. Moreover, filling them by zeros gives a chance to detect use-after-free accesses to non-zero data while an object stays in the quarantine (nice!). That functionality already exists in the kernel, it's called
init_on_free. I integrated it withCONFIG_SLAB_QUARANTINEin the patch 3/6.During that work I found a bug: in
CONFIG_SLABinit_on_freehappens too late, and heap objects go to the KASAN quarantine being dirty. See the fix in the patch 2/6.
[...]
Quoted 19 lines
I've made various tests on real hardware and in virtual machines:
1) network throughput test using iperf
server: iperf -s -f K
client: iperf -c 127.0.0.1 -t 60 -f K
2) scheduler stress testhackbench -s 4000 -l 500 -g 15 -f 25 -P3) building the defconfig kernel
time make -j2I compared Linux kernel 5.9.0-rc6 with:
- init_on_free=off,
- init_on_free=on,
-CONFIG_SLAB_QUARANTINE=y(which enablesinit_on_free).Each test was performed 5 times. I will show the mean values.
If you are interested, I can share all the results and calculate standard deviation.Real hardware, Intel Core i7-6500U CPU
1) Network throughput test with iperf
Quoted 7 lines
init_on_free=off: 5467152.2 KBytes/sec
init_on_free=on: 3937545 KBytes/sec (-28.0% vs init_on_free=off)
CONFIG_SLAB_QUARANTINE: 3858848.6 KBytes/sec (-2.0% vs init_on_free=on)
2) Scheduler stress test with hackbench
init_on_free=off: 8.5364s
init_on_free=on: 8.9858s (+5.3% vs init_on_free=off)
CONFIG_SLAB_QUARANTINE: 17.2232s (+91.7% vs init_on_free=on)These numbers seem really high for a mitigation, especially if that performance hit does not really buy you deterministic protection against many bugs.
[...]
Quoted 4 lines
N.B. There was NO performance optimization made for this version of the heap quarantine prototype. The main effort was put into researching its security properties (hope for your feedback). Performance optimization will be done in further steps, if we see that my work is worth doing.
But you are pretty much inherently limited in terms of performance by the effect the quarantine has on the data cache, right?
It seems to me like, if you want to make UAF exploitation harder at the heap allocator layer, you could do somewhat more effective things with a probably much smaller performance budget. Things like preventing the reallocation of virtual kernel addresses with different types, such that an attacker can only replace a UAF object with another object of the same type. (That is not an idea I like very much either, but I would like it more than this proposal.) (E.g. some browsers implement things along those lines, I believe.)
Quoted 8 lines · Jann Horn
It seems to me like, if you want to make UAF exploitation harder at the heap allocator layer, you could do somewhat more effective things with a probably much smaller performance budget. Things like preventing the reallocation of virtual kernel addresses with different types, such that an attacker can only replace a UAF object with another object of the same type. (That is not an idea I like very much either, but I would like it more than this proposal.) (E.g. some browsers implement things along those lines, I believe.)
The slab allocator already has that functionality. We call it TYPESAFE_BY_RCU, but if forcing that on by default would enhance security by a measurable amount, it wouldn't be a terribly hard sell ...
Quoted 10 lines · Matthew Wilcox
Quoted 8 lines · Jann Horn
It seems to me like, if you want to make UAF exploitation harder at the heap allocator layer, you could do somewhat more effective things with a probably much smaller performance budget. Things like preventing the reallocation of virtual kernel addresses with different types, such that an attacker can only replace a UAF object with another object of the same type. (That is not an idea I like very much either, but I would like it more than this proposal.) (E.g. some browsers implement things along those lines, I believe.)
Quoted 3 lines
The slab allocator already has that functionality. We call it
TYPESAFE_BY_RCU, but if forcing that on by default would enhance security by a measurable amount, it wouldn't be a terribly hard sell ...
TYPESAFE_BY_RCU just forces an RCU grace period before the reallocation; I'm thinking of something more drastic, like completely refusing to give back the memory, or using vmalloc for slabs where that's safe (reusing physical but not virtual addresses across types). And, to make it more effective, something like a compiler plugin to isolate kmalloc(sizeof(<type>)) allocations by type beyond just size classes.
Quoted 10 lines · Matthew Wilcox
Quoted 8 lines · Jann Horn
It seems to me like, if you want to make UAF exploitation harder at the heap allocator layer, you could do somewhat more effective things with a probably much smaller performance budget. Things like preventing the reallocation of virtual kernel addresses with different types, such that an attacker can only replace a UAF object with another object of the same type. (That is not an idea I like very much either, but I would like it more than this proposal.) (E.g. some browsers implement things along those lines, I believe.)
Quoted 3 lines
The slab allocator already has that functionality. We call it
TYPESAFE_BY_RCU, but if forcing that on by default would enhance security by a measurable amount, it wouldn't be a terribly hard sell ...
Isn't the "easy" version of this already controlled by slab_merge? (i.e. do not share same-sized/flagged kmem_caches between different caches)
The large trouble are the kmalloc caches, which don't have types associated with them. Having implicit kmem caches based on the type being allocated there would need some pretty extensive plumbing, I think?
Signature · 1 line
Kees Cook
Quoted 11 lines · Kees Cook
Quoted 10 lines · Matthew Wilcox
Quoted 8 lines · Jann Horn
It seems to me like, if you want to make UAF exploitation harder at the heap allocator layer, you could do somewhat more effective things with a probably much smaller performance budget. Things like preventing the reallocation of virtual kernel addresses with different types, such that an attacker can only replace a UAF object with another object of the same type. (That is not an idea I like very much either, but I would like it more than this proposal.) (E.g. some browsers implement things along those lines, I believe.)
Quoted 6 lines
Quoted 3 lines
The slab allocator already has that functionality. We call it
TYPESAFE_BY_RCU, but if forcing that on by default would enhance security by a measurable amount, it wouldn't be a terribly hard sell ...Isn't the "easy" version of this already controlled by
slab_merge? (i.e. do not share same-sized/flaggedkmem_cachesbetween different caches)
Yes, but slab_merge still normally frees slab pages to the page allocator.
Quoted 4 lines
The large trouble are the kmalloc caches, which don't have types associated with them. Having implicit kmem caches based on the type being allocated there would need some pretty extensive plumbing, I think?
Well, a bit of plumbing, at least. You'd need to teach the compiler frontend to grab type names from sizeof() and stuff that type information somewhere, e.g. by generating an extra function argument referring to the type, or something like that. Could be as simple as a reference to a bss section variable that encodes the type in the name, and the linker already has the logic to automatically deduplicate those across compilation units - that way, on the compiler side, a pure frontend plugin might do the job?
It will reuse the memory for other things when the whole slab is freed though. Not really realistic to change that without it being backed by virtual memory along with higher-level management of regions to avoid intense fragmentation and metadata waste. It would depend a lot on having much finer-grained slab caches, otherwise it's not going to be much of an alternative to a quarantine feature. Even then, a quarantine feature is still useful, but is less suitable for a mainstream feature due to performance cost. Even a small quarantine has a fairly high performance cost.
Quoted 5 lines · Kees Cook
Quoted 2 lines
TYPESAFE_BY_RCU, but if forcing that on by default would enhance security by a measurable amount, it wouldn't be a terribly hard sell ...Isn't the "easy" version of this already controlled by
slab_merge? (i.e. do not share same-sized/flaggedkmem_cachesbetween different caches)
Right.
Quoted 4 lines
The large trouble are the kmalloc caches, which don't have types associated with them. Having implicit kmem caches based on the type being allocated there would need some pretty extensive plumbing, I think?
Actually typifying those accesses may get rid of a lot of kmalloc allocations and could help to ease the management and control of objects.
It may be a big task though given the ubiquity of kmalloc and the need to create a massive amount of new slab caches. This is going to reduce the cache hit rate significantly.
Quoted 10 lines · Matthew Wilcox
Quoted 8 lines · Jann Horn
It seems to me like, if you want to make UAF exploitation harder at the heap allocator layer, you could do somewhat more effective things with a probably much smaller performance budget. Things like preventing the reallocation of virtual kernel addresses with different types, such that an attacker can only replace a UAF object with another object of the same type. (That is not an idea I like very much either, but I would like it more than this proposal.) (E.g. some browsers implement things along those lines, I believe.)
Quoted 3 lines
The slab allocator already has that functionality. We call it
TYPESAFE_BY_RCU, but if forcing that on by default would enhance security by a measurable amount, it wouldn't be a terribly hard sell ...
TYPESAFE functionality switches a lot of debugging off because that also allows speculative accesses to the object after it was freed (requires for RCU safeness because the object may be freed in an RCU period where it is still accessed). I do not think you would like that.
Quoted 8 lines · Jann Horn
Quoted 7 lines · Alexander Popov
Quoted 6 lines · Alexander Popov
This is the second version of the heap quarantine prototype for the Linux kernel. I performed a deeper evaluation of its security properties and developed new features like quarantine randomization and integration with
init_on_free. That is fun! See below for more details.
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Rationale
=========Use-after-free vulnerabilities in the Linux kernel are very popular for exploitation. There are many examples, some of them:
https://googleprojectzero.blogspot.com/2018/09/a-cache-invalidation-bug-in-linux.html
Hello Jann, thanks for your reply.
Quoted 6 lines
I don't think your proposed mitigation would work with much reliability against this bug; the attacker has full control over the timing of the original use and the following use, so an attacker should be able to trigger the
kmem_cache_free(), then spam enough new VMAs and delete them to flush out the quarantine, and then do heap spraying as normal, or something like that.
The randomized quarantine will release the vulnerable object at an unpredictable moment (patch 4/6).
So I think the control over the time of the use-after-free access doesn't help attackers, if they don't have an "infinite spray" -- unlimited ability to store controlled data in the kernelspace objects of the needed size without freeing them.
"Unlimited", because the quarantine size is 1/32 of whole memory. "Without freeing", because freed objects are erased by init_on_free before going to randomized heap quarantine (patch 3/6).
Would you agree?
Quoted 6 lines
Also, note that here, if the reallocation fails, the kernel still wouldn't crash because the dangling object is not accessed further if the address range stored in it doesn't match the fault address. So an attacker could potentially try multiple times, and if the object happens to be on the quarantine the first time, that wouldn't really be a showstopper, you'd just try again.
Freed objects are filled by zero before going to quarantine (patch 3/6). Would it cause a null pointer dereference on unsuccessful try?
Quoted 5 lines
Quoted 1 line
I think that here, again, the
free()and the dangling pointer use were caused by separate syscalls, meaning the attacker had control over that timing?
As I wrote above, I think attacker's control over this timing is required for a successful attack, but is not enough for bypassing randomized quarantine.
Quoted 8 lines
Quoted 1 line
Haven't looked at that one in detail.
Quoted 3 lines
Quoted 3 lines
Use-after-free exploits usually employ heap spraying technique. Generally it aims to put controlled bytes at a predetermined memory location on the heap.
Quoted 5 lines
Well, not necessarily "predetermined". Depending on the circumstances, you don't necessarily need to know which address you're writing to; and you might not even need to overwrite a specific object, but instead just have to overwrite one out of a bunch of objects, no matter which.
Yes, of course, I didn't mean a "predetermined memory address". Maybe "definite memory location" is a better phrase for that.
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
Heap spraying for exploiting use-after-free in the Linux kernel relies on the fact that on
kmalloc(), the slab allocator returns the address of the memory that was recently freed.Yeah; and that behavior is pretty critical for performance. The longer it's been since a newly allocated object was freed, the higher the chance that you'll end up having to go further down the memory cache hierarchy.
Yes. That behaviour is fast, however very convenient for use-after-free exploitation...
Quoted 11 lines
Quoted 3 lines
Quoted 3 lines
So allocating a kernel object with
the same size and controlled contents allows overwriting the vulnerable freed object.The vmacache exploit you linked to doesn't do that, it frees the object all the way back to the page allocator and then sprays 4MiB of memory from the page allocator. (Because VMAs use their own
kmem_cache, and thekmem_cachewasn't merged with any interesting ones, and I saw no good way to exploit the bug by reallocating another VMA over the old VMA back then. Although of course that doesn't mean that there is no such way.)
Sorry, my mistake.
Exploit examples with heap spraying that fit my description:
- CVE-2017-6074 https://www.openwall.com/lists/oss-security/2017/02/26/2
- CVE-2017-2636 https://a13xp0p0v.github.io/2017/03/24/CVE-2017-2636.html
- CVE-2016-8655 https://seclists.org/oss-sec/2016/q4/607
- CVE-2017-15649
https://ssd-disclosure.com/ssd-advisory-linux-kernel-af_packet-use-after-free/
Quoted 50 lines
[...]
Quoted 37 lines
Quoted 37 lines
Security properties
===================For researching security properties of the heap quarantine I developed 2 lkdtm tests (see the patch 5/6).
The first test is called
lkdtm_HEAP_SPRAY. It allocates and frees an object from a separatekmem_cacheand then allocates 400000 similar objects. I.e. this test performs an original heap spraying technique for use-after-free exploitation.If
CONFIG_SLAB_QUARANTINEis disabled, the freed object is instantly reallocated and overwritten:# echo HEAP_SPRAY > /sys/kernel/debug/provoke-crash/DIRECT lkdtm: Performing direct entry HEAP_SPRAY lkdtm: Allocated and freed spray_cache object 000000002b5b3ad4 of size 333 lkdtm: Original heap spraying: allocate 400000 objects of size 333... lkdtm: FAIL: attempt 0: freed object is reallocatedIf
CONFIG_SLAB_QUARANTINEis enabled, 400000 new allocations don't overwrite the freed object:# echo HEAP_SPRAY > /sys/kernel/debug/provoke-crash/DIRECT lkdtm: Performing direct entry HEAP_SPRAY lkdtm: Allocated and freed spray_cache object 000000009909e777 of size 333 lkdtm: Original heap spraying: allocate 400000 objects of size 333... lkdtm: OK: original heap spraying hasn't succeedThat happens because pushing an object through the quarantine requires _both_ allocating and freeing memory. Objects are released from the quarantine on new memory allocations, but only when the quarantine size is over the limit. And the quarantine size grows on new memory freeing.
That's why I created the second test called
lkdtm_PUSH_THROUGH_QUARANTINE. It allocates and frees an object from a separatekmem_cacheand then performs kmem_cache_alloc()+kmem_cache_free() for that cache 400000 times. This test effectively pushes the object through the heap quarantine and reallocates it after it returns back to the allocator freelist:[...]
Quoted 11 lines
Quoted 11 lines
As you can see, the number of the allocations that are needed for overwriting the vulnerable object is almost the same. That would be good for stable use-after-free exploitation and should not be allowed.
That's why I developed the quarantine randomization (see the patch 4/6).This randomization required very small hackish changes of the heap quarantine mechanism. At first all quarantine batches are filled by objects. Then during the quarantine reducing I randomly choose and free 1/2 of objects from a randomly chosen batch. Now the randomized quarantine releases the freed object at an unpredictable moment:
lkdtm: Target object is reallocated at attempt 107884
Quoted 7 lines
[...]
Quoted 1 line
Quoted 1 line
lkdtm: Target object is reallocated at attempt 87343
Those numbers are fairly big. At that point you might not even fit into L3 cache anymore, right? You'd often be hitting DRAM for new allocations? And for many slabs, you might end using much more memory for the quarantine than for actual in-use allocations.
Yes. The original quarantine size is
(totalram_pages() << PAGE_SHIFT) / QUARANTINE_FRACTION
where
#define QUARANTINE_FRACTION 32
Quoted 3 lines
It seems to me like, for this to stop attacks with a high probability, you'd have to reserve a huge chunk of kernel memory for the quarantines
Yes, that's how it works now.
Quoted 7 lines
- even if the attacker doesn't know anything about the status of the quarantine (which isn't necessarily the case, depending on whether the attacker can abuse microarchitectural data leakage, or if the attacker can trigger a pure data read through the dangling pointer), they should still be able to win with a probability around quarantine_size/allocated_memory_size if they have a heap spraying primitive without strict limits.
Not sure about this probability evaluation.
I will try calculating it taking the quarantine parameters into account.
Quoted 42 lines
Quoted 14 lines
Quoted 14 lines
However, this randomization alone would not disturb the attacker, because the quarantine stores the attacker's data (the payload) in the sprayed objects. I.e. the reallocated and overwritten vulnerable object contains the payload until the next reallocation (very bad).
Hence heap objects should be erased before going to the heap quarantine. Moreover, filling them by zeros gives a chance to detect use-after-free accesses to non-zero data while an object stays in the quarantine (nice!). That functionality already exists in the kernel, it's called
init_on_free. I integrated it withCONFIG_SLAB_QUARANTINEin the patch 3/6.During that work I found a bug: in
CONFIG_SLABinit_on_freehappens too late, and heap objects go to the KASAN quarantine being dirty. See the fix in the patch 2/6.[...]
Quoted 19 lines
I've made various tests on real hardware and in virtual machines:
1) network throughput test using iperf
server: iperf -s -f K
client: iperf -c 127.0.0.1 -t 60 -f K
2) scheduler stress testhackbench -s 4000 -l 500 -g 15 -f 25 -P3) building the defconfig kernel
time make -j2I compared Linux kernel 5.9.0-rc6 with:
- init_on_free=off,
- init_on_free=on,
-CONFIG_SLAB_QUARANTINE=y(which enablesinit_on_free).Each test was performed 5 times. I will show the mean values.
If you are interested, I can share all the results and calculate standard deviation.Real hardware, Intel Core i7-6500U CPU
1) Network throughput test with iperfQuoted 7 lines
init_on_free=off: 5467152.2 KBytes/sec init_on_free=on: 3937545 KBytes/sec (-28.0% vs init_on_free=off) CONFIG_SLAB_QUARANTINE: 3858848.6 KBytes/sec (-2.0% vs init_on_free=on) 2) Scheduler stress test with hackbench init_on_free=off: 8.5364s init_on_free=on: 8.9858s (+5.3% vs init_on_free=off) CONFIG_SLAB_QUARANTINE: 17.2232s (+91.7% vs init_on_free=on)
Quoted 3 lines
These numbers seem really high for a mitigation, especially if that performance hit does not really buy you deterministic protection against many bugs.
Right, I agree.
It's a probabilistic protection, and the probability should be calculated. I'll work on that.
Quoted 8 lines
[...]
Quoted 4 lines
N.B. There was NO performance optimization made for this version of the heap quarantine prototype. The main effort was put into researching its security properties (hope for your feedback). Performance optimization will be done in further steps, if we see that my work is worth doing.
But you are pretty much inherently limited in terms of performance by the effect the quarantine has on the data cache, right?
Yes.
However, the quarantine parameters can be adjusted.
Quoted 8 lines
It seems to me like, if you want to make UAF exploitation harder at the heap allocator layer, you could do somewhat more effective things with a probably much smaller performance budget. Things like preventing the reallocation of virtual kernel addresses with different types, such that an attacker can only replace a UAF object with another object of the same type. (That is not an idea I like very much either, but I would like it more than this proposal.) (E.g. some browsers implement things along those lines, I believe.)
That's interesting, thank you.
Best regards,
Alexander
Quoted 33 lines · Alexander Popov
Quoted 8 lines · Jann Horn
Quoted 7 lines · Alexander Popov
Quoted 6 lines · Alexander Popov
This is the second version of the heap quarantine prototype for the Linux kernel. I performed a deeper evaluation of its security properties and developed new features like quarantine randomization and integration with
init_on_free. That is fun! See below for more details.Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Rationale
=========Use-after-free vulnerabilities in the Linux kernel are very popular for exploitation. There are many examples, some of them:
https://googleprojectzero.blogspot.com/2018/09/a-cache-invalidation-bug-in-linux.htmlHello Jann, thanks for your reply.
Quoted 6 lines
I don't think your proposed mitigation would work with much reliability against this bug; the attacker has full control over the timing of the original use and the following use, so an attacker should be able to trigger the
kmem_cache_free(), then spam enough new VMAs and delete them to flush out the quarantine, and then do heap spraying as normal, or something like that.The randomized quarantine will release the vulnerable object at an unpredictable moment (patch 4/6).
So I think the control over the time of the use-after-free access doesn't help attackers, if they don't have an "infinite spray" -- unlimited ability to store controlled data in the kernelspace objects of the needed size without freeing them.
Quoted 5 lines
"Unlimited", because the quarantine size is 1/32 of whole memory. "Without freeing", because freed objects are erased by
init_on_freebefore going to randomized heap quarantine (patch 3/6).Would you agree?
But you have a single quarantine (per CPU) for all objects, right? So for a UAF on slab A, the attacker can just spam allocations and deallocations on slab B to almost deterministically flush everything in slab A back to the SLUB freelists?
Quoted 9 lines
Quoted 6 lines
Also, note that here, if the reallocation fails, the kernel still wouldn't crash because the dangling object is not accessed further if the address range stored in it doesn't match the fault address. So an attacker could potentially try multiple times, and if the object happens to be on the quarantine the first time, that wouldn't really be a showstopper, you'd just try again.
Freed objects are filled by zero before going to quarantine (patch 3/6). Would it cause a null pointer dereference on unsuccessful try?
Not as far as I can tell.
[...]
Quoted 20 lines
Quoted 7 lines
Quoted 4 lines
N.B. There was NO performance optimization made for this version of the heap quarantine prototype. The main effort was put into researching its security properties (hope for your feedback). Performance optimization will be done in further steps, if we see that my work is worth doing.
But you are pretty much inherently limited in terms of performance by the effect the quarantine has on the data cache, right?
Yes.
However, the quarantine parameters can be adjusted.Quoted 8 lines
It seems to me like, if you want to make UAF exploitation harder at the heap allocator layer, you could do somewhat more effective things with a probably much smaller performance budget. Things like preventing the reallocation of virtual kernel addresses with different types, such that an attacker can only replace a UAF object with another object of the same type. (That is not an idea I like very much either, but I would like it more than this proposal.) (E.g. some browsers implement things along those lines, I believe.)
Quoted 1 line
That's interesting, thank you.
Just as some more context of how I think about this:
Preventing memory corruption, outside of stuff like core memory management code, isn't really all *that* hard. There are schemes out there for hardware that reliably protects the integrity of data pointers, and such things. And if people can do that in hardware, we can also emulate that, and we'll get the same protection in software.
The hard part is making it reasonably fast. And if you are willing to accept the kind of performance impact that comes with gigantic quarantine queues, there might be more effective things to spend that performance on?
Quoted 38 lines · Jann Horn
Quoted 37 lines · Alexander Popov
Quoted 8 lines · Jann Horn
Quoted 7 lines · Alexander Popov
Quoted 6 lines · Alexander Popov
This is the second version of the heap quarantine prototype for the Linux kernel. I performed a deeper evaluation of its security properties and developed new features like quarantine randomization and integration with
init_on_free. That is fun! See below for more details.Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Rationale
=========Use-after-free vulnerabilities in the Linux kernel are very popular for exploitation. There are many examples, some of them:
https://googleprojectzero.blogspot.com/2018/09/a-cache-invalidation-bug-in-linux.htmlHello Jann, thanks for your reply.
Quoted 6 lines
I don't think your proposed mitigation would work with much reliability against this bug; the attacker has full control over the timing of the original use and the following use, so an attacker should be able to trigger the
kmem_cache_free(), then spam enough new VMAs and delete them to flush out the quarantine, and then do heap spraying as normal, or something like that.The randomized quarantine will release the vulnerable object at an unpredictable moment (patch 4/6).
So I think the control over the time of the use-after-free access doesn't help attackers, if they don't have an "infinite spray" -- unlimited ability to store controlled data in the kernelspace objects of the needed size without freeing them.
"Unlimited", because the quarantine size is 1/32 of whole memory. "Without freeing", because freed objects are erased by
init_on_freebefore going to randomized heap quarantine (patch 3/6).
Quoted 6 lines
Quoted 1 line
Would you agree?
But you have a single quarantine (per CPU) for all objects, right? So for a UAF on slab A, the attacker can just spam allocations and deallocations on slab B to almost deterministically flush everything in slab A back to the SLUB freelists?
Aaaahh! Nice shot Jann, I see.
Another slab cache can be used to flush the randomized quarantine, so eventually the vulnerable object returns into the allocator freelist in its cache, and original heap spraying can be used again.
For now I think the idea of a global quarantine for all slab objects is dead.
Thank you.
Best regards,
Alexander