CVE-2026-23342 - Vulnerability Details

Vendors	Products
Linux Subscribe	Linux Kernel Subscribe

Link	Providers
https://git.kernel.org/stable/c/7466ae2aeed483de80c5d8dea0913cf74038b652
https://git.kernel.org/stable/c/869c63d5975d55e97f6b168e885452b3da20ea47
https://git.kernel.org/stable/c/e67299e1044349ad0088d52c6bc5764cc1816c06

Type	Values Removed	Values Added
Description		In the Linux kernel, the following vulnerability has been resolved: bpf: Fix race in cpumap on PREEMPT_RT On PREEMPT_RT kernels, the per-CPU xdp_bulk_queue (bq) can be accessed concurrently by multiple preemptible tasks on the same CPU. The original code assumes bq_enqueue() and __cpu_map_flush() run atomically with respect to each other on the same CPU, relying on local_bh_disable() to prevent preemption. However, on PREEMPT_RT, local_bh_disable() only calls migrate_disable() (when PREEMPT_RT_NEEDS_BH_LOCK is not set) and does not disable preemption, which allows CFS scheduling to preempt a task during bq_flush_to_queue(), enabling another task on the same CPU to enter bq_enqueue() and operate on the same per-CPU bq concurrently. This leads to several races: 1. Double __list_del_clearprev(): after bq->count is reset in bq_flush_to_queue(), a preempting task can call bq_enqueue() -> bq_flush_to_queue() on the same bq when bq->count reaches CPU_MAP_BULK_SIZE. Both tasks then call __list_del_clearprev() on the same bq->flush_node, the second call dereferences the prev pointer that was already set to NULL by the first. 2. bq->count and bq->q[] races: concurrent bq_enqueue() can corrupt the packet queue while bq_flush_to_queue() is processing it. The race between task A (__cpu_map_flush -> bq_flush_to_queue) and task B (bq_enqueue -> bq_flush_to_queue) on the same CPU: Task A (xdp_do_flush) Task B (cpu_map_enqueue) ---------------------- ------------------------ bq_flush_to_queue(bq) spin_lock(&q->producer_lock) /* flush bq->q[] to ptr_ring / bq->count = 0 spin_unlock(&q->producer_lock) bq_enqueue(rcpu, xdpf) <-- CFS preempts Task A --> bq->q[bq->count++] = xdpf / ... more enqueues until full ... / bq_flush_to_queue(bq) spin_lock(&q->producer_lock) / flush to ptr_ring / spin_unlock(&q->producer_lock) __list_del_clearprev(flush_node) / sets flush_node.prev = NULL / <-- Task A resumes --> __list_del_clearprev(flush_node) flush_node.prev->next = ... / prev is NULL -> kernel oops */ Fix this by adding a local_lock_t to xdp_bulk_queue and acquiring it in bq_enqueue() and __cpu_map_flush(). These paths already run under local_bh_disable(), so use local_lock_nested_bh() which on non-RT is a pure annotation with no overhead, and on PREEMPT_RT provides a per-CPU sleeping lock that serializes access to the bq. To reproduce, insert an mdelay(100) between bq->count = 0 and __list_del_clearprev() in bq_flush_to_queue(), then run reproducer provided by syzkaller.
Title		bpf: Fix race in cpumap on PREEMPT_RT
First Time appeared		Linux Linux linux Kernel
CPEs		cpe:2.3:o:linux:linux_kernel::::::::
Vendors & Products		Linux Linux linux Kernel
References		https://git.kernel.org/stable/c/7466ae2aeed483de80c5d8dea0913cf74038b652 https://git.kernel.org/stable/c/869c63d5975d55e97f6b168e885452b3da20ea47 https://git.kernel.org/stable/c/e67299e1044349ad0088d52c6bc5764cc1816c06

{"dataType": "CVE_RECORD", "dataVersion": "5.2", "cveMetadata": {"cveId": "CVE-2026-23342", "assignerOrgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "state": "PUBLISHED", "assignerShortName": "Linux", "dateReserved": "2026-01-13T15:37:45.999Z", "datePublished": "2026-03-25T10:27:30.285Z", "dateUpdated": "2026-03-25T10:27:30.285Z"}, "containers": {"cna": {"providerMetadata": {"orgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "shortName": "Linux", "dateUpdated": "2026-03-25T10:27:30.285Z"}, "descriptions": [{"lang": "en", "value": "In the Linux kernel, the following vulnerability has been resolved:\n\nbpf: Fix race in cpumap on PREEMPT_RT\n\nOn PREEMPT_RT kernels, the per-CPU xdp_bulk_queue (bq) can be accessed\nconcurrently by multiple preemptible tasks on the same CPU.\n\nThe original code assumes bq_enqueue() and __cpu_map_flush() run\natomically with respect to each other on the same CPU, relying on\nlocal_bh_disable() to prevent preemption. However, on PREEMPT_RT,\nlocal_bh_disable() only calls migrate_disable() (when\nPREEMPT_RT_NEEDS_BH_LOCK is not set) and does not disable\npreemption, which allows CFS scheduling to preempt a task during\nbq_flush_to_queue(), enabling another task on the same CPU to enter\nbq_enqueue() and operate on the same per-CPU bq concurrently.\n\nThis leads to several races:\n\n1. Double __list_del_clearprev(): after bq->count is reset in\n bq_flush_to_queue(), a preempting task can call bq_enqueue() ->\n bq_flush_to_queue() on the same bq when bq->count reaches\n CPU_MAP_BULK_SIZE. Both tasks then call __list_del_clearprev()\n on the same bq->flush_node, the second call dereferences the\n prev pointer that was already set to NULL by the first.\n\n2. bq->count and bq->q[] races: concurrent bq_enqueue() can corrupt\n the packet queue while bq_flush_to_queue() is processing it.\n\nThe race between task A (__cpu_map_flush -> bq_flush_to_queue) and\ntask B (bq_enqueue -> bq_flush_to_queue) on the same CPU:\n\n Task A (xdp_do_flush) Task B (cpu_map_enqueue)\n ---------------------- ------------------------\n bq_flush_to_queue(bq)\n spin_lock(&q->producer_lock)\n /* flush bq->q[] to ptr_ring */\n bq->count = 0\n spin_unlock(&q->producer_lock)\n bq_enqueue(rcpu, xdpf)\n <-- CFS preempts Task A --> bq->q[bq->count++] = xdpf\n /* ... more enqueues until full ... */\n bq_flush_to_queue(bq)\n spin_lock(&q->producer_lock)\n /* flush to ptr_ring */\n spin_unlock(&q->producer_lock)\n __list_del_clearprev(flush_node)\n /* sets flush_node.prev = NULL */\n <-- Task A resumes -->\n __list_del_clearprev(flush_node)\n flush_node.prev->next = ...\n /* prev is NULL -> kernel oops */\n\nFix this by adding a local_lock_t to xdp_bulk_queue and acquiring it\nin bq_enqueue() and __cpu_map_flush(). These paths already run under\nlocal_bh_disable(), so use local_lock_nested_bh() which on non-RT is\na pure annotation with no overhead, and on PREEMPT_RT provides a\nper-CPU sleeping lock that serializes access to the bq.\n\nTo reproduce, insert an mdelay(100) between bq->count = 0 and\n__list_del_clearprev() in bq_flush_to_queue(), then run reproducer\nprovided by syzkaller."}], "affected": [{"product": "Linux", "vendor": "Linux", "defaultStatus": "unaffected", "repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git", "programFiles": ["kernel/bpf/cpumap.c"], "versions": [{"version": "3253cb49cbad4772389d6ef55be75db1f97da910", "lessThan": "7466ae2aeed483de80c5d8dea0913cf74038b652", "status": "affected", "versionType": "git"}, {"version": "3253cb49cbad4772389d6ef55be75db1f97da910", "lessThan": "e67299e1044349ad0088d52c6bc5764cc1816c06", "status": "affected", "versionType": "git"}, {"version": "3253cb49cbad4772389d6ef55be75db1f97da910", "lessThan": "869c63d5975d55e97f6b168e885452b3da20ea47", "status": "affected", "versionType": "git"}]}, {"product": "Linux", "vendor": "Linux", "defaultStatus": "affected", "repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git", "programFiles": ["kernel/bpf/cpumap.c"], "versions": [{"version": "6.18", "status": "affected"}, {"version": "0", "lessThan": "6.18", "status": "unaffected", "versionType": "semver"}, {"version": "6.18.17", "lessThanOrEqual": "6.18.*", "status": "unaffected", "versionType": "semver"}, {"version": "6.19.7", "lessThanOrEqual": "6.19.*", "status": "unaffected", "versionType": "semver"}, {"version": "7.0-rc2", "lessThanOrEqual": "*", "status": "unaffected", "versionType": "original_commit_for_fix"}]}], "cpeApplicability": [{"nodes": [{"operator": "OR", "negate": false, "cpeMatch": [{"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.18", "versionEndExcluding": "6.18.17"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.18", "versionEndExcluding": "6.19.7"}, {"vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.18", "versionEndExcluding": "7.0-rc2"}]}]}], "references": [{"url": "https://git.kernel.org/stable/c/7466ae2aeed483de80c5d8dea0913cf74038b652"}, {"url": "https://git.kernel.org/stable/c/e67299e1044349ad0088d52c6bc5764cc1816c06"}, {"url": "https://git.kernel.org/stable/c/869c63d5975d55e97f6b168e885452b3da20ea47"}], "title": "bpf: Fix race in cpumap on PREEMPT_RT", "x_generator": {"engine": "bippy-1.2.0"}}}}

{"cveTags": [], "descriptions": [{"lang": "en", "value": "In the Linux kernel, the following vulnerability has been resolved:\n\nbpf: Fix race in cpumap on PREEMPT_RT\n\nOn PREEMPT_RT kernels, the per-CPU xdp_bulk_queue (bq) can be accessed\nconcurrently by multiple preemptible tasks on the same CPU.\n\nThe original code assumes bq_enqueue() and __cpu_map_flush() run\natomically with respect to each other on the same CPU, relying on\nlocal_bh_disable() to prevent preemption. However, on PREEMPT_RT,\nlocal_bh_disable() only calls migrate_disable() (when\nPREEMPT_RT_NEEDS_BH_LOCK is not set) and does not disable\npreemption, which allows CFS scheduling to preempt a task during\nbq_flush_to_queue(), enabling another task on the same CPU to enter\nbq_enqueue() and operate on the same per-CPU bq concurrently.\n\nThis leads to several races:\n\n1. Double __list_del_clearprev(): after bq->count is reset in\n bq_flush_to_queue(), a preempting task can call bq_enqueue() ->\n bq_flush_to_queue() on the same bq when bq->count reaches\n CPU_MAP_BULK_SIZE. Both tasks then call __list_del_clearprev()\n on the same bq->flush_node, the second call dereferences the\n prev pointer that was already set to NULL by the first.\n\n2. bq->count and bq->q[] races: concurrent bq_enqueue() can corrupt\n the packet queue while bq_flush_to_queue() is processing it.\n\nThe race between task A (__cpu_map_flush -> bq_flush_to_queue) and\ntask B (bq_enqueue -> bq_flush_to_queue) on the same CPU:\n\n Task A (xdp_do_flush) Task B (cpu_map_enqueue)\n ---------------------- ------------------------\n bq_flush_to_queue(bq)\n spin_lock(&q->producer_lock)\n /* flush bq->q[] to ptr_ring */\n bq->count = 0\n spin_unlock(&q->producer_lock)\n bq_enqueue(rcpu, xdpf)\n <-- CFS preempts Task A --> bq->q[bq->count++] = xdpf\n /* ... more enqueues until full ... */\n bq_flush_to_queue(bq)\n spin_lock(&q->producer_lock)\n /* flush to ptr_ring */\n spin_unlock(&q->producer_lock)\n __list_del_clearprev(flush_node)\n /* sets flush_node.prev = NULL */\n <-- Task A resumes -->\n __list_del_clearprev(flush_node)\n flush_node.prev->next = ...\n /* prev is NULL -> kernel oops */\n\nFix this by adding a local_lock_t to xdp_bulk_queue and acquiring it\nin bq_enqueue() and __cpu_map_flush(). These paths already run under\nlocal_bh_disable(), so use local_lock_nested_bh() which on non-RT is\na pure annotation with no overhead, and on PREEMPT_RT provides a\nper-CPU sleeping lock that serializes access to the bq.\n\nTo reproduce, insert an mdelay(100) between bq->count = 0 and\n__list_del_clearprev() in bq_flush_to_queue(), then run reproducer\nprovided by syzkaller."}], "id": "CVE-2026-23342", "lastModified": "2026-03-25T11:16:32.147", "metrics": {}, "published": "2026-03-25T11:16:32.147", "references": [{"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "url": "https://git.kernel.org/stable/c/7466ae2aeed483de80c5d8dea0913cf74038b652"}, {"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "url": "https://git.kernel.org/stable/c/869c63d5975d55e97f6b168e885452b3da20ea47"}, {"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "url": "https://git.kernel.org/stable/c/e67299e1044349ad0088d52c6bc5764cc1816c06"}], "sourceIdentifier": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "vulnStatus": "Received"}

Metrics

Affected Vendors & Products

Projects

Metrics

Affected Vendors & Products

Projects

JSON object

JSON object

JSON object

JSON object

JSON object