CVE-2022-49700

In the Linux kernel, the following vulnerability has been resolved: mm/slub: add missing TID updates on slab deactivation The fastpath in slab_alloc_node() assumes that c->slab is stable as long as the TID stays the same. However, two places in __slab_alloc() currently don't update the TID when deactivating the CPU slab. If multiple operations race the right way, this could lead to an object getting lost; or, in an even more unlikely situation, it could even lead to an object being freed onto the wrong slab's freelist, messing up the `inuse` counter and eventually causing a page to be freed to the page allocator while it still contains slab objects. (I haven't actually tested these cases though, this is just based on looking at the code. Writing testcases for this stuff seems like it'd be a pain...) The race leading to state inconsistency is (all operations on the same CPU and kmem_cache): - task A: begin do_slab_free(): - read TID - read pcpu freelist (==NULL) - check `slab == c->slab` (true) - [PREEMPT A->B] - task B: begin slab_alloc_node(): - fastpath fails (`c->freelist` is NULL) - enter __slab_alloc() - slub_get_cpu_ptr() (disables preemption) - enter ___slab_alloc() - take local_lock_irqsave() - read c->freelist as NULL - get_freelist() returns NULL - write `c->slab = NULL` - drop local_unlock_irqrestore() - goto new_slab - slub_percpu_partial() is NULL - get_partial() returns NULL - slub_put_cpu_ptr() (enables preemption) - [PREEMPT B->A] - task A: finish do_slab_free(): - this_cpu_cmpxchg_double() succeeds() - [CORRUPT STATE: c->slab==NULL, c->freelist!=NULL] From there, the object on c->freelist will get lost if task B is allowed to continue from here: It will proceed to the retry_load_slab label, set c->slab, then jump to load_freelist, which clobbers c->freelist. But if we instead continue as follows, we get worse corruption: - task A: run __slab_free() on object from other struct slab: - CPU_PARTIAL_FREE case (slab was on no list, is now on pcpu partial) - task A: run slab_alloc_node() with NUMA node constraint: - fastpath fails (c->slab is NULL) - call __slab_alloc() - slub_get_cpu_ptr() (disables preemption) - enter ___slab_alloc() - c->slab is NULL: goto new_slab - slub_percpu_partial() is non-NULL - set c->slab to slub_percpu_partial(c) - [CORRUPT STATE: c->slab points to slab-1, c->freelist has objects from slab-2] - goto redo - node_match() fails - goto deactivate_slab - existing c->freelist is passed into deactivate_slab() - inuse count of slab-1 is decremented to account for object from slab-2 At this point, the inuse count of slab-1 is 1 lower than it should be. This means that if we free all allocated objects in slab-1 except for one, SLUB will think that slab-1 is completely unused, and may free its page, leading to use-after-free.

CVSS v3 7.8 HIGH

7.8^/10

CVSS v3 : HIGH

V3 Legend

Vector :

Exploitability : 1.8 / Impact : 5.9

Attack Vector LOCAL

Attack Complexity LOW

Privileges Required LOW

User Interaction NONE

Confidentiality Impact HIGH

Integrity Impact HIGH

Availability Impact HIGH

Scope UNCHANGED

References

Link	Resource
https://git.kernel.org/stable/c/0515cc9b6b24877f59b222ade704bfaa42caa2a6	Patch
https://git.kernel.org/stable/c/197e257da473c725dfe47759c3ee02f2398d8ea5	Patch
https://git.kernel.org/stable/c/308c6d0e1f200fd26c71270c6e6bfcf0fc6ff082	Patch
https://git.kernel.org/stable/c/6c32496964da0dc230cea763a0e934b2e02dabd5	Patch
https://git.kernel.org/stable/c/d6a597450e686d4c6388bd3cdcb17224b4dae7f0	Patch
https://git.kernel.org/stable/c/e2b2f0e2e34d71ae6c2a1114fd3c525930e84bc7	Patch
https://git.kernel.org/stable/c/e7e3e90d671078455a3a08189f89d85b3da2de9e	Patch
https://git.kernel.org/stable/c/eeaa345e128515135ccb864c04482180c08e3259	Patch

Configurations

Configuration 1 (hide)

OR	cpe:2.3:o:linux:linux_kernel::::::::
	cpe:2.3:o:linux:linux_kernel::::::::
	cpe:2.3:o:linux:linux_kernel::::::::
	cpe:2.3:o:linux:linux_kernel::::::::
	cpe:2.3:o:linux:linux_kernel::::::::
	cpe:2.3:o:linux:linux_kernel::::::::
	cpe:2.3:o:linux:linux_kernel::::::::
	cpe:2.3:o:linux:linux_kernel:5.19:rc1::::::
	cpe:2.3:o:linux:linux_kernel:5.19:rc2::::::
	cpe:2.3:o:linux:linux_kernel:5.19:rc3::::::

History

25 Mar 2025, 13:37

Type	Values Removed	Values Added
References	~~() https://git.kernel.org/stable/c/0515cc9b6b24877f59b222ade704bfaa42caa2a6 -~~	() https://git.kernel.org/stable/c/0515cc9b6b24877f59b222ade704bfaa42caa2a6 - Patch
References	~~() https://git.kernel.org/stable/c/197e257da473c725dfe47759c3ee02f2398d8ea5 -~~	() https://git.kernel.org/stable/c/197e257da473c725dfe47759c3ee02f2398d8ea5 - Patch
References	~~() https://git.kernel.org/stable/c/308c6d0e1f200fd26c71270c6e6bfcf0fc6ff082 -~~	() https://git.kernel.org/stable/c/308c6d0e1f200fd26c71270c6e6bfcf0fc6ff082 - Patch
References	~~() https://git.kernel.org/stable/c/6c32496964da0dc230cea763a0e934b2e02dabd5 -~~	() https://git.kernel.org/stable/c/6c32496964da0dc230cea763a0e934b2e02dabd5 - Patch
References	~~() https://git.kernel.org/stable/c/d6a597450e686d4c6388bd3cdcb17224b4dae7f0 -~~	() https://git.kernel.org/stable/c/d6a597450e686d4c6388bd3cdcb17224b4dae7f0 - Patch
References	~~() https://git.kernel.org/stable/c/e2b2f0e2e34d71ae6c2a1114fd3c525930e84bc7 -~~	() https://git.kernel.org/stable/c/e2b2f0e2e34d71ae6c2a1114fd3c525930e84bc7 - Patch
References	~~() https://git.kernel.org/stable/c/e7e3e90d671078455a3a08189f89d85b3da2de9e -~~	() https://git.kernel.org/stable/c/e7e3e90d671078455a3a08189f89d85b3da2de9e - Patch
References	~~() https://git.kernel.org/stable/c/eeaa345e128515135ccb864c04482180c08e3259 -~~	() https://git.kernel.org/stable/c/eeaa345e128515135ccb864c04482180c08e3259 - Patch
CPE		cpe:2.3:o:linux:linux_kernel:::::::: cpe:2.3:o:linux:linux_kernel:5.19:rc1:::::: cpe:2.3:o:linux:linux_kernel:5.19:rc2:::::: cpe:2.3:o:linux:linux_kernel:5.19:rc3::::::
First Time		Linux Linux linux Kernel

27 Feb 2025, 19:15

Type	Values Removed	Values Added
CWE		CWE-416
CVSS	v2 : ~~unknown~~ v3 : ~~unknown~~	v2 : unknown v3 : 7.8
Summary		(es) En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: mm/slub: añadir actualizaciones de TID faltantes en la desactivación de slab La ruta rápida en slab_alloc_node() asume que c->slab es estable siempre que el TID permanezca igual. Sin embargo, dos lugares en __slab_alloc() actualmente no actualizan el TID al desactivar el slab de la CPU. Si varias operaciones se ejecutan en la dirección correcta, esto podría provocar que se pierda un objeto; o, en una situación aún más improbable, incluso podría provocar que se libere un objeto en la lista libre del slab incorrecto, lo que arruina el contador `inuse` y eventualmente provoca que se libere una página al asignador de páginas mientras aún contiene objetos slab. (En realidad no he probado estos casos, esto se basa simplemente en mirar el código. Escribir casos de prueba para estas cosas parece que sería una molestia...) La ejecución que lleva a la inconsistencia de estado es (todas las operaciones en la misma CPU y kmem_cache): - task A: begin do_slab_free(): - read TID - read pcpu freelist (==NULL) - check `slab == c->slab` (true) - [PREEMPT A->B] - task B: begin slab_alloc_node(): - fastpath fails (`c->freelist` is NULL) - enter __slab_alloc() - slub_get_cpu_ptr() (disables preemption) - enter ___slab_alloc() - take local_lock_irqsave() - read c->freelist as NULL - get_freelist() returns NULL - write `c->slab = NULL` - drop local_unlock_irqrestore() - goto new_slab - slub_percpu_partial() is NULL - get_partial() returns NULL - slub_put_cpu_ptr() (enables preemption) - [PREEMPT B->A] - task A: finish do_slab_free(): - this_cpu_cmpxchg_double() succeeds() - [CORRUPT STATE: c->slab==NULL, c->freelist!=NULL] From there, the object on c->freelist will get lost if task B is allowed to continue from here: It will proceed to the retry_load_slab label, set c->slab, then jump to load_freelist, which clobbers c->freelist. But if we instead continue as follows, we get worse corruption: - task A: run __slab_free() on object from other struct slab: - CPU_PARTIAL_FREE case (slab was on no list, is now on pcpu partial) - task A: run slab_alloc_node() with NUMA node constraint: - fastpath fails (c->slab is NULL) - call __slab_alloc() - slub_get_cpu_ptr() (disables preemption) - enter ___slab_alloc() - c->slab is NULL: goto new_slab - slub_percpu_partial() is non-NULL - set c->slab to slub_percpu_partial(c) - [CORRUPT STATE: c->slab points to slab-1, c->freelist has objects from slab-2] - goto redo - node_match() fails - goto deactivate_slab - existing c->freelist is passed into deactivate_slab() - inuse count of slab-1 se decrementa para tener en cuenta el objeto de slab-2 En este punto, el conteo de objetos en uso de slab-1 es 1 menos de lo que debería ser. Esto significa que si liberamos todos los objetos asignados en slab-1 excepto uno, SLUB pensará que slab-1 está completamente sin usar y puede liberar su página, lo que lleva a un use-after-free.

26 Feb 2025, 07:01

Type	Values Removed	Values Added
New CVE

Information

Published : 2025-02-26 07:01

Updated : 2025-03-25 13:37

NVD link : CVE-2022-49700

Mitre link : CVE-2022-49700

CVE.ORG link : CVE-2022-49700

JSON object : View

Products Affected

linux

linux_kernel

CWE

CWE-416

Use After Free

{"id": "CVE-2022-49700", "cveTags": [], "metrics": {"cvssMetricV31": [{"type": "Secondary", "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0", "cvssData": {"scope": "UNCHANGED", "version": "3.1", "baseScore": 7.8, "attackVector": "LOCAL", "baseSeverity": "HIGH", "vectorString": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H", "integrityImpact": "HIGH", "userInteraction": "NONE", "attackComplexity": "LOW", "availabilityImpact": "HIGH", "privilegesRequired": "LOW", "confidentialityImpact": "HIGH"}, "impactScore": 5.9, "exploitabilityScore": 1.8}]}, "published": "2025-02-26T07:01:44.740", "references": [{"url": "https://git.kernel.org/stable/c/0515cc9b6b24877f59b222ade704bfaa42caa2a6", "tags": ["Patch"], "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"}, {"url": "https://git.kernel.org/stable/c/197e257da473c725dfe47759c3ee02f2398d8ea5", "tags": ["Patch"], "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"}, {"url": "https://git.kernel.org/stable/c/308c6d0e1f200fd26c71270c6e6bfcf0fc6ff082", "tags": ["Patch"], "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"}, {"url": "https://git.kernel.org/stable/c/6c32496964da0dc230cea763a0e934b2e02dabd5", "tags": ["Patch"], "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"}, {"url": "https://git.kernel.org/stable/c/d6a597450e686d4c6388bd3cdcb17224b4dae7f0", "tags": ["Patch"], "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"}, {"url": "https://git.kernel.org/stable/c/e2b2f0e2e34d71ae6c2a1114fd3c525930e84bc7", "tags": ["Patch"], "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"}, {"url": "https://git.kernel.org/stable/c/e7e3e90d671078455a3a08189f89d85b3da2de9e", "tags": ["Patch"], "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"}, {"url": "https://git.kernel.org/stable/c/eeaa345e128515135ccb864c04482180c08e3259", "tags": ["Patch"], "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"}], "vulnStatus": "Analyzed", "weaknesses": [{"type": "Secondary", "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0", "description": [{"lang": "en", "value": "CWE-416"}]}], "descriptions": [{"lang": "en", "value": "In the Linux kernel, the following vulnerability has been resolved:\n\nmm/slub: add missing TID updates on slab deactivation\n\nThe fastpath in slab_alloc_node() assumes that c->slab is stable as long as\nthe TID stays the same. However, two places in __slab_alloc() currently\ndon't update the TID when deactivating the CPU slab.\n\nIf multiple operations race the right way, this could lead to an object\ngetting lost; or, in an even more unlikely situation, it could even lead to\nan object being freed onto the wrong slab's freelist, messing up the\n`inuse` counter and eventually causing a page to be freed to the page\nallocator while it still contains slab objects.\n\n(I haven't actually tested these cases though, this is just based on\nlooking at the code. Writing testcases for this stuff seems like it'd be\na pain...)\n\nThe race leading to state inconsistency is (all operations on the same CPU\nand kmem_cache):\n\n - task A: begin do_slab_free():\n - read TID\n - read pcpu freelist (==NULL)\n - check `slab == c->slab` (true)\n - [PREEMPT A->B]\n - task B: begin slab_alloc_node():\n - fastpath fails (`c->freelist` is NULL)\n - enter __slab_alloc()\n - slub_get_cpu_ptr() (disables preemption)\n - enter ___slab_alloc()\n - take local_lock_irqsave()\n - read c->freelist as NULL\n - get_freelist() returns NULL\n - write `c->slab = NULL`\n - drop local_unlock_irqrestore()\n - goto new_slab\n - slub_percpu_partial() is NULL\n - get_partial() returns NULL\n - slub_put_cpu_ptr() (enables preemption)\n - [PREEMPT B->A]\n - task A: finish do_slab_free():\n - this_cpu_cmpxchg_double() succeeds()\n - [CORRUPT STATE: c->slab==NULL, c->freelist!=NULL]\n\nFrom there, the object on c->freelist will get lost if task B is allowed to\ncontinue from here: It will proceed to the retry_load_slab label,\nset c->slab, then jump to load_freelist, which clobbers c->freelist.\n\nBut if we instead continue as follows, we get worse corruption:\n\n - task A: run __slab_free() on object from other struct slab:\n - CPU_PARTIAL_FREE case (slab was on no list, is now on pcpu partial)\n - task A: run slab_alloc_node() with NUMA node constraint:\n - fastpath fails (c->slab is NULL)\n - call __slab_alloc()\n - slub_get_cpu_ptr() (disables preemption)\n - enter ___slab_alloc()\n - c->slab is NULL: goto new_slab\n - slub_percpu_partial() is non-NULL\n - set c->slab to slub_percpu_partial(c)\n - [CORRUPT STATE: c->slab points to slab-1, c->freelist has objects\n from slab-2]\n - goto redo\n - node_match() fails\n - goto deactivate_slab\n - existing c->freelist is passed into deactivate_slab()\n - inuse count of slab-1 is decremented to account for object from\n slab-2\n\nAt this point, the inuse count of slab-1 is 1 lower than it should be.\nThis means that if we free all allocated objects in slab-1 except for one,\nSLUB will think that slab-1 is completely unused, and may free its page,\nleading to use-after-free."}, {"lang": "es", "value": "En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: mm/slub: a\u00f1adir actualizaciones de TID faltantes en la desactivaci\u00f3n de slab La ruta r\u00e1pida en slab_alloc_node() asume que c->slab es estable siempre que el TID permanezca igual. Sin embargo, dos lugares en __slab_alloc() actualmente no actualizan el TID al desactivar el slab de la CPU. Si varias operaciones se ejecutan en la direcci\u00f3n correcta, esto podr\u00eda provocar que se pierda un objeto; o, en una situaci\u00f3n a\u00fan m\u00e1s improbable, incluso podr\u00eda provocar que se libere un objeto en la lista libre del slab incorrecto, lo que arruina el contador `inuse` y eventualmente provoca que se libere una p\u00e1gina al asignador de p\u00e1ginas mientras a\u00fan contiene objetos slab. (En realidad no he probado estos casos, esto se basa simplemente en mirar el c\u00f3digo. Escribir casos de prueba para estas cosas parece que ser\u00eda una molestia...) La ejecuci\u00f3n que lleva a la inconsistencia de estado es (todas las operaciones en la misma CPU y kmem_cache): - task A: begin do_slab_free(): - read TID - read pcpu freelist (==NULL) - check `slab == c->slab` (true) - [PREEMPT A->B] - task B: begin slab_alloc_node(): - fastpath fails (`c->freelist` is NULL) - enter __slab_alloc() - slub_get_cpu_ptr() (disables preemption) - enter ___slab_alloc() - take local_lock_irqsave() - read c->freelist as NULL - get_freelist() returns NULL - write `c->slab = NULL` - drop local_unlock_irqrestore() - goto new_slab - slub_percpu_partial() is NULL - get_partial() returns NULL - slub_put_cpu_ptr() (enables preemption) - [PREEMPT B->A] - task A: finish do_slab_free(): - this_cpu_cmpxchg_double() succeeds() - [CORRUPT STATE: c->slab==NULL, c->freelist!=NULL] From there, the object on c->freelist will get lost if task B is allowed to continue from here: It will proceed to the retry_load_slab label, set c->slab, then jump to load_freelist, which clobbers c->freelist. But if we instead continue as follows, we get worse corruption: - task A: run __slab_free() on object from other struct slab: - CPU_PARTIAL_FREE case (slab was on no list, is now on pcpu partial) - task A: run slab_alloc_node() with NUMA node constraint: - fastpath fails (c->slab is NULL) - call __slab_alloc() - slub_get_cpu_ptr() (disables preemption) - enter ___slab_alloc() - c->slab is NULL: goto new_slab - slub_percpu_partial() is non-NULL - set c->slab to slub_percpu_partial(c) - [CORRUPT STATE: c->slab points to slab-1, c->freelist has objects from slab-2] - goto redo - node_match() fails - goto deactivate_slab - existing c->freelist is passed into deactivate_slab() - inuse count of slab-1 se decrementa para tener en cuenta el objeto de slab-2 En este punto, el conteo de objetos en uso de slab-1 es 1 menos de lo que deber\u00eda ser. Esto significa que si liberamos todos los objetos asignados en slab-1 excepto uno, SLUB pensar\u00e1 que slab-1 est\u00e1 completamente sin usar y puede liberar su p\u00e1gina, lo que lleva a un use-after-free."}], "lastModified": "2025-03-25T13:37:51.227", "configurations": [{"nodes": [{"negate": false, "cpeMatch": [{"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "2B9BA632-0605-4759-8828-79C0E7F757FD", "versionEndExcluding": "4.9.323", "versionStartIncluding": "3.1"}, {"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "11E86C4E-715B-4E25-81E9-2FD98431E3FA", "versionEndExcluding": "4.14.288", "versionStartIncluding": "4.10"}, {"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "2CC8E927-649A-40AD-AB62-F7EE444BEF0F", "versionEndExcluding": "4.19.252", "versionStartIncluding": "4.15"}, {"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "844199C4-DEBE-4DA1-AB77-5A7984F9393B", "versionEndExcluding": "5.4.205", "versionStartIncluding": "4.20"}, {"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "8BF24898-7C80-443F-93F3-F82029BBFF72", "versionEndExcluding": "5.10.130", "versionStartIncluding": "5.5"}, {"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "B2197EF1-3D9C-4EBA-9F94-6C8668E719B6", "versionEndExcluding": "5.15.54", "versionStartIncluding": "5.11"}, {"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "0172D3FA-DDEB-482A-A270-4A1495A8798C", "versionEndExcluding": "5.18.8", "versionStartIncluding": "5.16"}, {"criteria": "cpe:2.3:o:linux:linux_kernel:5.19:rc1:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "A8C30C2D-F82D-4D37-AB48-D76ABFBD5377"}, {"criteria": "cpe:2.3:o:linux:linux_kernel:5.19:rc2:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "BF8547FC-C849-4F1B-804B-A93AE2F04A92"}, {"criteria": "cpe:2.3:o:linux:linux_kernel:5.19:rc3:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "F3068028-F453-4A1C-B80F-3F5609ACEF60"}], "operator": "OR"}]}], "sourceIdentifier": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"}

7.8 /10