{"api_version":"1","generated_at":"2026-09-13T13:03:48+00:00","cve":"CVE-2026-89589","urls":{"html":"https://cve.report/CVE-2026-89589","api":"https://cve.report/api/cve/CVE-2026-89589.json","docs":"https://cve.report/api","cve_org":"https://www.cve.org/CVERecord?id=CVE-2026-89589","nvd":"https://nvd.nist.gov/vuln/detail/CVE-2026-89589"},"summary":{"title":"acpi/apei/ghes: Use raw_spinlock_t for CXL CPER work locks","description":"In the Linux kernel, the following vulnerability has been resolved:\n\nacpi/apei/ghes: Use raw_spinlock_t for CXL CPER work locks\n\nThe CXL CPER work registration and unregistration helpers acquire\ncxl_cper_work_lock and cxl_cper_prot_err_work_lock with a spinlock\nguard(), which leaves local interrupts enabled. The corresponding post\npaths (cxl_cper_post_event(), cxl_cper_post_prot_err()) execute in hard\nIRQ context (they are called from the GHES error notification path) and\nacquire the same locks with an irqsave guard().\n\nIf a CPU is holding one of these locks via a spinlock guard() when a GHES\ninterrupt arrives on the same CPU, the IRQ handler spins on the held lock\nwaiting for it to release, while the lock holder is preempted by the IRQ.\nThe result is a deadlock.\n\nConvert both locks from spinlock_t to raw_spinlock_t and use guard() at\nall call sites. On PREEMPT_RT kernels spinlock_t is backed by rt_mutex and\nsleeping from hard IRQ context is not permitted; raw_spinlock_t is safe in\nboth contexts.\n\nAdd WARN_ONCE to both register functions to surface double-registration\nbugs at runtime.\n\nRestructure both unregister functions to clear the global work pointer\nunder the lock before calling cancel_work_sync(), closing the window\nwhere a CPER interrupt could schedule work on a pointer about to be\nfreed. Add kfifo_reset() after cancel_work_sync() so stale entries\nare not replayed on next module load.\n\nBoth kfifos are single-consumer: only one work_struct is registered at\na time, enforced by the WARN_ONCE guard in the register functions.\nkfifo_reset() is safe outside the lock because cancel_work_sync() has\nalready quiesced the consumer, and no new consumer can register until\nthe current module exit completes and a fresh module init runs.\n\nRemove the redundant cancel_work_sync() call from cxl_ras_exit() and\ncxl_pci_driver_exit(). The CPER unregister functions now quiesce\nthe work internally.","state":"PUBLISHED","assigner":"Linux","published_at":"2026-09-11 20:19:43","updated_at":"2026-09-11 20:19:43"},"problem_types":[],"metrics":[],"references":[{"url":"https://git.kernel.org/stable/c/c735dbce7ad03df708bf22ec964f67e43cbc823a","name":"https://git.kernel.org/stable/c/c735dbce7ad03df708bf22ec964f67e43cbc823a","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://git.kernel.org/stable/c/0e25382b05de48d09b243d3d447d236a6d735bc2","name":"https://git.kernel.org/stable/c/0e25382b05de48d09b243d3d447d236a6d735bc2","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://git.kernel.org/stable/c/6625ca499c3131ef63be3215f8f942d7a097ea3a","name":"https://git.kernel.org/stable/c/6625ca499c3131ef63be3215f8f942d7a097ea3a","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://www.cve.org/CVERecord?id=CVE-2026-89589","name":"CVE Program record","refsource":"CVE.ORG","tags":["canonical"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-89589","name":"NVD vulnerability detail","refsource":"NVD","tags":["canonical","analysis"]}],"affected":[{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 5e4a264bf8b55d5a1b4c8cd96ac10b99f98d7487 c735dbce7ad03df708bf22ec964f67e43cbc823a git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 5e4a264bf8b55d5a1b4c8cd96ac10b99f98d7487 0e25382b05de48d09b243d3d447d236a6d735bc2 git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 5e4a264bf8b55d5a1b4c8cd96ac10b99f98d7487 6625ca499c3131ef63be3215f8f942d7a097ea3a git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 6.10","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 6.10 semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 6.18.50 6.18.* semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 7.2.4 7.2.* semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 7.3-rc1 * original_commit_for_fix","platforms":[]}],"timeline":[],"solutions":[],"workarounds":[],"exploits":[],"credits":[],"nvd_cpes":[],"vendor_comments":[],"enrichments":{"kev":null,"epss":{"cve_year":"2026","cve_id":"89589","cve":"CVE-2026-89589","epss":"0.001730000","percentile":"0.069430000","score_date":"2026-09-12","updated_at":"2026-09-13 00:08:17"},"legacy_qids":[]},"source_records":{"cve_program":{"containers":{"cna":{"affected":[{"defaultStatus":"unaffected","product":"Linux","programFiles":["drivers/acpi/apei/ghes.c","drivers/cxl/core/ras.c","drivers/cxl/pci.c"],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","vendor":"Linux","versions":[{"lessThan":"c735dbce7ad03df708bf22ec964f67e43cbc823a","status":"affected","version":"5e4a264bf8b55d5a1b4c8cd96ac10b99f98d7487","versionType":"git"},{"lessThan":"0e25382b05de48d09b243d3d447d236a6d735bc2","status":"affected","version":"5e4a264bf8b55d5a1b4c8cd96ac10b99f98d7487","versionType":"git"},{"lessThan":"6625ca499c3131ef63be3215f8f942d7a097ea3a","status":"affected","version":"5e4a264bf8b55d5a1b4c8cd96ac10b99f98d7487","versionType":"git"}]},{"defaultStatus":"affected","product":"Linux","programFiles":["drivers/acpi/apei/ghes.c","drivers/cxl/core/ras.c","drivers/cxl/pci.c"],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","vendor":"Linux","versions":[{"status":"affected","version":"6.10"},{"lessThan":"6.10","status":"unaffected","version":"0","versionType":"semver"},{"lessThanOrEqual":"6.18.*","status":"unaffected","version":"6.18.50","versionType":"semver"},{"lessThanOrEqual":"7.2.*","status":"unaffected","version":"7.2.4","versionType":"semver"},{"lessThanOrEqual":"*","status":"unaffected","version":"7.3-rc1","versionType":"original_commit_for_fix"}]}],"cpeApplicability":[{"nodes":[{"cpeMatch":[{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"6.18.50","versionStartIncluding":"6.10","vulnerable":true},{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"7.2.4","versionStartIncluding":"6.10","vulnerable":true},{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"7.3-rc1","versionStartIncluding":"6.10","vulnerable":true}],"negate":false,"operator":"OR"}]}],"descriptions":[{"lang":"en","value":"In the Linux kernel, the following vulnerability has been resolved:\n\nacpi/apei/ghes: Use raw_spinlock_t for CXL CPER work locks\n\nThe CXL CPER work registration and unregistration helpers acquire\ncxl_cper_work_lock and cxl_cper_prot_err_work_lock with a spinlock\nguard(), which leaves local interrupts enabled. The corresponding post\npaths (cxl_cper_post_event(), cxl_cper_post_prot_err()) execute in hard\nIRQ context (they are called from the GHES error notification path) and\nacquire the same locks with an irqsave guard().\n\nIf a CPU is holding one of these locks via a spinlock guard() when a GHES\ninterrupt arrives on the same CPU, the IRQ handler spins on the held lock\nwaiting for it to release, while the lock holder is preempted by the IRQ.\nThe result is a deadlock.\n\nConvert both locks from spinlock_t to raw_spinlock_t and use guard() at\nall call sites. On PREEMPT_RT kernels spinlock_t is backed by rt_mutex and\nsleeping from hard IRQ context is not permitted; raw_spinlock_t is safe in\nboth contexts.\n\nAdd WARN_ONCE to both register functions to surface double-registration\nbugs at runtime.\n\nRestructure both unregister functions to clear the global work pointer\nunder the lock before calling cancel_work_sync(), closing the window\nwhere a CPER interrupt could schedule work on a pointer about to be\nfreed. Add kfifo_reset() after cancel_work_sync() so stale entries\nare not replayed on next module load.\n\nBoth kfifos are single-consumer: only one work_struct is registered at\na time, enforced by the WARN_ONCE guard in the register functions.\nkfifo_reset() is safe outside the lock because cancel_work_sync() has\nalready quiesced the consumer, and no new consumer can register until\nthe current module exit completes and a fresh module init runs.\n\nRemove the redundant cancel_work_sync() call from cxl_ras_exit() and\ncxl_pci_driver_exit(). The CPER unregister functions now quiesce\nthe work internally."}],"providerMetadata":{"dateUpdated":"2026-09-11T19:44:53.578Z","orgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","shortName":"Linux"},"references":[{"url":"https://git.kernel.org/stable/c/c735dbce7ad03df708bf22ec964f67e43cbc823a"},{"url":"https://git.kernel.org/stable/c/0e25382b05de48d09b243d3d447d236a6d735bc2"},{"url":"https://git.kernel.org/stable/c/6625ca499c3131ef63be3215f8f942d7a097ea3a"}],"title":"acpi/apei/ghes: Use raw_spinlock_t for CXL CPER work locks","x_generator":{"engine":"bippy-1.2.0"}}},"cveMetadata":{"assignerOrgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","assignerShortName":"Linux","cveId":"CVE-2026-89589","datePublished":"2026-09-11T19:44:53.578Z","dateReserved":"2026-09-11T19:38:34.729Z","dateUpdated":"2026-09-11T19:44:53.578Z","state":"PUBLISHED"},"dataType":"CVE_RECORD","dataVersion":"5.2"},"nvd":{"publishedDate":"2026-09-11 20:19:43","lastModifiedDate":"2026-09-11 20:19:43","problem_types":[],"metrics":[],"configurations":[]},"legacy_mitre":{"record":{"CveYear":"2026","CveId":"89589","Ordinal":"1","Title":"acpi/apei/ghes: Use raw_spinlock_t for CXL CPER work locks","CVE":"CVE-2026-89589","Year":"2026"},"notes":[{"CveYear":"2026","CveId":"89589","Ordinal":"1","NoteData":"In the Linux kernel, the following vulnerability has been resolved:\n\nacpi/apei/ghes: Use raw_spinlock_t for CXL CPER work locks\n\nThe CXL CPER work registration and unregistration helpers acquire\ncxl_cper_work_lock and cxl_cper_prot_err_work_lock with a spinlock\nguard(), which leaves local interrupts enabled. The corresponding post\npaths (cxl_cper_post_event(), cxl_cper_post_prot_err()) execute in hard\nIRQ context (they are called from the GHES error notification path) and\nacquire the same locks with an irqsave guard().\n\nIf a CPU is holding one of these locks via a spinlock guard() when a GHES\ninterrupt arrives on the same CPU, the IRQ handler spins on the held lock\nwaiting for it to release, while the lock holder is preempted by the IRQ.\nThe result is a deadlock.\n\nConvert both locks from spinlock_t to raw_spinlock_t and use guard() at\nall call sites. On PREEMPT_RT kernels spinlock_t is backed by rt_mutex and\nsleeping from hard IRQ context is not permitted; raw_spinlock_t is safe in\nboth contexts.\n\nAdd WARN_ONCE to both register functions to surface double-registration\nbugs at runtime.\n\nRestructure both unregister functions to clear the global work pointer\nunder the lock before calling cancel_work_sync(), closing the window\nwhere a CPER interrupt could schedule work on a pointer about to be\nfreed. Add kfifo_reset() after cancel_work_sync() so stale entries\nare not replayed on next module load.\n\nBoth kfifos are single-consumer: only one work_struct is registered at\na time, enforced by the WARN_ONCE guard in the register functions.\nkfifo_reset() is safe outside the lock because cancel_work_sync() has\nalready quiesced the consumer, and no new consumer can register until\nthe current module exit completes and a fresh module init runs.\n\nRemove the redundant cancel_work_sync() call from cxl_ras_exit() and\ncxl_pci_driver_exit(). The CPER unregister functions now quiesce\nthe work internally.","Type":"Description","Title":"acpi/apei/ghes: Use raw_spinlock_t for CXL CPER work locks"}]}}}