{"id":"AZL-111194","summary":"CVE-2026-99106 affecting package kernel 6.6.157.1-1","details":"In the Linux kernel, the following vulnerability has been resolved:\n\ndrm/amdgpu: free the userptr ranges when a CS fails after parser_bos\n\namdgpu_cs_parser_bos() allocates an HMM range for every userptr entry of\nthe BO list and looks up the user pages. Those ranges are only freed when\namdgpu_cs_parser_bos() itself fails or when amdgpu_cs_submit() checks\nthem under the notifier lock. If amdgpu_cs_patch_jobs(),\namdgpu_cs_vm_handling(), amdgpu_cs_sync_rings() or the early part of\namdgpu_cs_submit() fails, nothing frees them: neither\namdgpu_cs_parser_fini() nor amdgpu_bo_list_free() touch e-\u003erange.\n\nEach such CS then leaks the range, its hmm_pfns array (8 bytes per page\nof the userptr BO) and the BO reference the range holds. An IB on a ring\nthat patches the CS in place (e.g. VCN) with an unmapped va_start fails\nin amdgpu_cs_patch_ibs() and takes this path.\n\nFree the ranges on those error paths. Ranges amdgpu_cs_submit() already\ndropped are NULL, so freeing them again is a no-op.\n\n(cherry picked from commit c52a9542ed7cfd527feb938bd0accada2ac8d267)","modified":"2026-10-29T14:16:53.411244828Z","published":"2026-10-28T11:17:29Z","upstream":["CVE-2026-99106"],"references":[{"type":"WEB","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-99106"}],"affected":[{"package":{"name":"kernel","ecosystem":"Azure Linux:3","purl":"pkg:rpm/azure-linux/kernel"},"ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"},{"last_affected":"6.6.157.1-1"}]}],"database_specific":{"source":"https://github.com/microsoft/AzureLinuxVulnerabilityData/blob/main/osv/AZL-111194.json"}}],"schema_version":"1.9.2"}