fkie_cve-2025-38373
Vulnerability from fkie_nvd
Published
2025-07-25 13:15
Modified
2025-07-25 15:29
Severity ?
Summary
In the Linux kernel, the following vulnerability has been resolved: IB/mlx5: Fix potential deadlock in MR deregistration The issue arises when kzalloc() is invoked while holding umem_mutex or any other lock acquired under umem_mutex. This is problematic because kzalloc() can trigger fs_reclaim_aqcuire(), which may, in turn, invoke mmu_notifier_invalidate_range_start(). This function can lead to mlx5_ib_invalidate_range(), which attempts to acquire umem_mutex again, resulting in a deadlock. The problematic flow: CPU0 | CPU1 ---------------------------------------|------------------------------------------------ mlx5_ib_dereg_mr() | → revoke_mr() | → mutex_lock(&umem_odp->umem_mutex) | | mlx5_mkey_cache_init() | → mutex_lock(&dev->cache.rb_lock) | → mlx5r_cache_create_ent_locked() | → kzalloc(GFP_KERNEL) | → fs_reclaim() | → mmu_notifier_invalidate_range_start() | → mlx5_ib_invalidate_range() | → mutex_lock(&umem_odp->umem_mutex) → cache_ent_find_and_store() | → mutex_lock(&dev->cache.rb_lock) | Additionally, when kzalloc() is called from within cache_ent_find_and_store(), we encounter the same deadlock due to re-acquisition of umem_mutex. Solve by releasing umem_mutex in dereg_mr() after umr_revoke_mr() and before acquiring rb_lock. This ensures that we don't hold umem_mutex while performing memory allocations that could trigger the reclaim path. This change prevents the deadlock by ensuring proper lock ordering and avoiding holding locks during memory allocation operations that could trigger the reclaim path. The following lockdep warning demonstrates the deadlock: python3/20557 is trying to acquire lock: ffff888387542128 (&umem_odp->umem_mutex){+.+.}-{4:4}, at: mlx5_ib_invalidate_range+0x5b/0x550 [mlx5_ib] but task is already holding lock: ffffffff82f6b840 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}, at: unmap_vmas+0x7b/0x1a0 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}: fs_reclaim_acquire+0x60/0xd0 mem_cgroup_css_alloc+0x6f/0x9b0 cgroup_init_subsys+0xa4/0x240 cgroup_init+0x1c8/0x510 start_kernel+0x747/0x760 x86_64_start_reservations+0x25/0x30 x86_64_start_kernel+0x73/0x80 common_startup_64+0x129/0x138 -> #2 (fs_reclaim){+.+.}-{0:0}: fs_reclaim_acquire+0x91/0xd0 __kmalloc_cache_noprof+0x4d/0x4c0 mlx5r_cache_create_ent_locked+0x75/0x620 [mlx5_ib] mlx5_mkey_cache_init+0x186/0x360 [mlx5_ib] mlx5_ib_stage_post_ib_reg_umr_init+0x3c/0x60 [mlx5_ib] __mlx5_ib_add+0x4b/0x190 [mlx5_ib] mlx5r_probe+0xd9/0x320 [mlx5_ib] auxiliary_bus_probe+0x42/0x70 really_probe+0xdb/0x360 __driver_probe_device+0x8f/0x130 driver_probe_device+0x1f/0xb0 __driver_attach+0xd4/0x1f0 bus_for_each_dev+0x79/0xd0 bus_add_driver+0xf0/0x200 driver_register+0x6e/0xc0 __auxiliary_driver_register+0x6a/0xc0 do_one_initcall+0x5e/0x390 do_init_module+0x88/0x240 init_module_from_file+0x85/0xc0 idempotent_init_module+0x104/0x300 __x64_sys_finit_module+0x68/0xc0 do_syscall_64+0x6d/0x140 entry_SYSCALL_64_after_hwframe+0x4b/0x53 -> #1 (&dev->cache.rb_lock){+.+.}-{4:4}: __mutex_lock+0x98/0xf10 __mlx5_ib_dereg_mr+0x6f2/0x890 [mlx5_ib] mlx5_ib_dereg_mr+0x21/0x110 [mlx5_ib] ib_dereg_mr_user+0x85/0x1f0 [ib_core] ---truncated---
References
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/2ed25aa7f7711f508b6120e336f05cd9d49943c0
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/727eb1be65a370572edf307558ec3396b8573156
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/beb89ada5715e7bd1518c58863eedce89ec051a7
Impacted products
Vendor Product Version


To clipboard 

{
  "cveTags": [],
  "descriptions": [
    {
      "lang": "en",
      "value": "In the Linux kernel, the following vulnerability has been resolved:\n\nIB/mlx5: Fix potential deadlock in MR deregistration\n\nThe issue arises when kzalloc() is invoked while holding umem_mutex or\nany other lock acquired under umem_mutex. This is problematic because\nkzalloc() can trigger fs_reclaim_aqcuire(), which may, in turn, invoke\nmmu_notifier_invalidate_range_start(). This function can lead to\nmlx5_ib_invalidate_range(), which attempts to acquire umem_mutex again,\nresulting in a deadlock.\n\nThe problematic flow:\n             CPU0                      |              CPU1\n---------------------------------------|------------------------------------------------\nmlx5_ib_dereg_mr()                     |\n \u2192 revoke_mr()                         |\n   \u2192 mutex_lock(\u0026umem_odp-\u003eumem_mutex) |\n                                       | mlx5_mkey_cache_init()\n                                       |  \u2192 mutex_lock(\u0026dev-\u003ecache.rb_lock)\n                                       |  \u2192 mlx5r_cache_create_ent_locked()\n                                       |    \u2192 kzalloc(GFP_KERNEL)\n                                       |      \u2192 fs_reclaim()\n                                       |        \u2192 mmu_notifier_invalidate_range_start()\n                                       |          \u2192 mlx5_ib_invalidate_range()\n                                       |            \u2192 mutex_lock(\u0026umem_odp-\u003eumem_mutex)\n   \u2192 cache_ent_find_and_store()        |\n     \u2192 mutex_lock(\u0026dev-\u003ecache.rb_lock) |\n\nAdditionally, when kzalloc() is called from within\ncache_ent_find_and_store(), we encounter the same deadlock due to\nre-acquisition of umem_mutex.\n\nSolve by releasing umem_mutex in dereg_mr() after umr_revoke_mr()\nand before acquiring rb_lock. This ensures that we don\u0027t hold\numem_mutex while performing memory allocations that could trigger\nthe reclaim path.\n\nThis change prevents the deadlock by ensuring proper lock ordering and\navoiding holding locks during memory allocation operations that could\ntrigger the reclaim path.\n\nThe following lockdep warning demonstrates the deadlock:\n\n python3/20557 is trying to acquire lock:\n ffff888387542128 (\u0026umem_odp-\u003eumem_mutex){+.+.}-{4:4}, at:\n mlx5_ib_invalidate_range+0x5b/0x550 [mlx5_ib]\n\n but task is already holding lock:\n ffffffff82f6b840 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}, at:\n unmap_vmas+0x7b/0x1a0\n\n which lock already depends on the new lock.\n\n the existing dependency chain (in reverse order) is:\n\n -\u003e #3 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}:\n       fs_reclaim_acquire+0x60/0xd0\n       mem_cgroup_css_alloc+0x6f/0x9b0\n       cgroup_init_subsys+0xa4/0x240\n       cgroup_init+0x1c8/0x510\n       start_kernel+0x747/0x760\n       x86_64_start_reservations+0x25/0x30\n       x86_64_start_kernel+0x73/0x80\n       common_startup_64+0x129/0x138\n\n -\u003e #2 (fs_reclaim){+.+.}-{0:0}:\n       fs_reclaim_acquire+0x91/0xd0\n       __kmalloc_cache_noprof+0x4d/0x4c0\n       mlx5r_cache_create_ent_locked+0x75/0x620 [mlx5_ib]\n       mlx5_mkey_cache_init+0x186/0x360 [mlx5_ib]\n       mlx5_ib_stage_post_ib_reg_umr_init+0x3c/0x60 [mlx5_ib]\n       __mlx5_ib_add+0x4b/0x190 [mlx5_ib]\n       mlx5r_probe+0xd9/0x320 [mlx5_ib]\n       auxiliary_bus_probe+0x42/0x70\n       really_probe+0xdb/0x360\n       __driver_probe_device+0x8f/0x130\n       driver_probe_device+0x1f/0xb0\n       __driver_attach+0xd4/0x1f0\n       bus_for_each_dev+0x79/0xd0\n       bus_add_driver+0xf0/0x200\n       driver_register+0x6e/0xc0\n       __auxiliary_driver_register+0x6a/0xc0\n       do_one_initcall+0x5e/0x390\n       do_init_module+0x88/0x240\n       init_module_from_file+0x85/0xc0\n       idempotent_init_module+0x104/0x300\n       __x64_sys_finit_module+0x68/0xc0\n       do_syscall_64+0x6d/0x140\n       entry_SYSCALL_64_after_hwframe+0x4b/0x53\n\n -\u003e #1 (\u0026dev-\u003ecache.rb_lock){+.+.}-{4:4}:\n       __mutex_lock+0x98/0xf10\n       __mlx5_ib_dereg_mr+0x6f2/0x890 [mlx5_ib]\n       mlx5_ib_dereg_mr+0x21/0x110 [mlx5_ib]\n       ib_dereg_mr_user+0x85/0x1f0 [ib_core]\n  \n---truncated---"
    },
    {
      "lang": "es",
      "value": "En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: IB/mlx5: Arreglar un posible bloqueo en la anulaci\u00f3n del registro de MR El problema surge cuando se invoca kzalloc() mientras se mantiene umem_mutex o cualquier otro bloqueo adquirido bajo umem_mutex. Esto es problem\u00e1tico porque kzalloc() puede activar fs_reclaim_aqcuire(), que puede, a su vez, invocar mmu_notifier_invalidate_range_start(). Esta funci\u00f3n puede llevar a mlx5_ib_invalidate_range(), que intenta adquirir umem_mutex de nuevo, lo que resulta en un bloqueo. El flujo problem\u00e1tico: CPU0 | CPU1 ---------------------------------------|------------------------------------------------ mlx5_ib_dereg_mr() | ? revoke_mr() | ? mutex_lock(\u0026amp;umem_odp-\u0026gt;umem_mutex) | | mlx5_mkey_cache_init() | ? mutex_lock(\u0026amp;dev-\u0026gt;cache.rb_lock) | ? mlx5r_cache_create_ent_locked() | ? kzalloc(GFP_KERNEL) | ? fs_reclaim() | ? mmu_notifier_invalidate_range_start() | ? mlx5_ib_invalidate_range() | ? mutex_lock(\u0026amp;umem_odp-\u0026gt;umem_mutex) ? cache_ent_find_and_store() | ? mutex_lock(\u0026amp;dev-\u0026gt;cache.rb_lock) | Adem\u00e1s, cuando se llama a kzalloc() desde dentro de cache_ent_find_and_store(), encontramos el mismo bloqueo debido a la readquisici\u00f3n de umem_mutex. Se resuelve liberando umem_mutex en dereg_mr() despu\u00e9s de umr_revoke_mr() y antes de adquirir rb_lock. Esto garantiza que no se mantenga umem_mutex mientras se realizan asignaciones de memoria que podr\u00edan activar la ruta de recuperaci\u00f3n. Este cambio previene el interbloqueo al asegurar el orden correcto de los bloqueos y evitar mantenerlos durante las operaciones de asignaci\u00f3n de memoria que podr\u00edan activar la ruta de recuperaci\u00f3n. La siguiente advertencia de lockdep demuestra el bloqueo: ython3/20557 is trying to acquire lock: ffff888387542128 (\u0026amp;umem_odp-\u0026gt;umem_mutex){+.+.}-{4:4}, at: mlx5_ib_invalidate_range+0x5b/0x550 [mlx5_ib] but task is already holding lock: ffffffff82f6b840 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}, at: unmap_vmas+0x7b/0x1a0 cuyo bloqueo ya depende del nuevo bloqueo. la cadena de dependencia existente (en orden inverso) es:-\u0026gt; #3 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}: fs_reclaim_acquire+0x60/0xd0 mem_cgroup_css_alloc+0x6f/0x9b0 cgroup_init_subsys+0xa4/0x240 cgroup_init+0x1c8/0x510 start_kernel+0x747/0x760 x86_64_start_reservations+0x25/0x30 x86_64_start_kernel+0x73/0x80 common_startup_64+0x129/0x138 -\u0026gt; #2 (fs_reclaim){+.+.}-{0:0}: fs_reclaim_acquire+0x91/0xd0 __kmalloc_cache_noprof+0x4d/0x4c0 mlx5r_cache_create_ent_locked+0x75/0x620 [mlx5_ib] mlx5_mkey_cache_init+0x186/0x360 [mlx5_ib] mlx5_ib_stage_post_ib_reg_umr_init+0x3c/0x60 [mlx5_ib] __mlx5_ib_add+0x4b/0x190 [mlx5_ib] mlx5r_probe+0xd9/0x320 [mlx5_ib] auxiliary_bus_probe+0x42/0x70 really_probe+0xdb/0x360 __driver_probe_device+0x8f/0x130 driver_probe_device+0x1f/0xb0 __driver_attach+0xd4/0x1f0 bus_for_each_dev+0x79/0xd0 bus_add_driver+0xf0/0x200 driver_register+0x6e/0xc0 __auxiliary_driver_register+0x6a/0xc0 do_one_initcall+0x5e/0x390 do_init_module+0x88/0x240 init_module_from_file+0x85/0xc0 idempotent_init_module+0x104/0x300 __x64_sys_finit_module+0x68/0xc0 do_syscall_64+0x6d/0x140 entry_SYSCALL_64_after_hwframe+0x4b/0x53 -\u0026gt; #1 (\u0026amp;dev-\u0026gt;cache.rb_lock){+.+.}-{4:4}: __mutex_lock+0x98/0xf10 __mlx5_ib_dereg_mr+0x6f2/0x890 [mlx5_ib] mlx5_ib_dereg_mr+0x21/0x110 [mlx5_ib] ib_dereg_mr_user+0x85/0x1f0 [ib_core] ---truncated---"
    }
  ],
  "id": "CVE-2025-38373",
  "lastModified": "2025-07-25T15:29:19.837",
  "metrics": {},
  "published": "2025-07-25T13:15:26.283",
  "references": [
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/2ed25aa7f7711f508b6120e336f05cd9d49943c0"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/727eb1be65a370572edf307558ec3396b8573156"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/beb89ada5715e7bd1518c58863eedce89ec051a7"
    }
  ],
  "sourceIdentifier": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
  "vulnStatus": "Awaiting Analysis"
}


Log in or create an account to share your comment.




Tags
Taxonomy of the tags.




Sightings

Author Source Type Date

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or seen somewhere by the user.
  • Confirmed: The vulnerability is confirmed from an analyst perspective.
  • Exploited: This vulnerability was exploited and seen by the user reporting the sighting.
  • Patched: This vulnerability was successfully patched by the user reporting the sighting.
  • Not exploited: This vulnerability was not exploited or seen by the user reporting the sighting.
  • Not confirmed: The user expresses doubt about the veracity of the vulnerability.
  • Not patched: This vulnerability was not successfully patched by the user reporting the sighting.