CVE-2026-74584

Sin historial EPSS suficiente todavía.

In the Linux kernel, the following vulnerability has been resolved: RDMA/bnxt_re: zero shared page before exposing to userspace bnxt_re_alloc_ucontext() allocates uctx->shpg via __get_free_page(GFP_KERNEL). The buddy allocator does not zero pages without __GFP_ZERO, so the page contains stale kernel data from whatever object most recently freed it. The page is then mapped into userspace via vm_insert_page() under BNXT_RE_MMAP_SH_PAGE in bnxt_re_mmap(). The driver only ever writes 4 bytes (a u32 AVID) at offset BNXT_RE_AVID_OFFT (0x10) inside bnxt_re_create_ah(); the remaining 4092 bytes of the page are exposed to userspace unsanitised, leaking kernel memory contents. Any user with access to /dev/infiniband/uverbsX on a host with a bnxt_re device (typically rdma group membership) can read this data via a single mmap() at pgoff 0 after IB_USER_VERBS_CMD_GET_CONTEXT. Other shared pages in the same file already use get_zeroed_page() correctly: drivers/infiniband/hw/bnxt_re/ib_verbs.c srq->uctx_srq_page = (void *)get_zeroed_page(GFP_KERNEL); cq->uctx_cq_page = (void *)get_zeroed_page(GFP_KERNEL); uctx->shpg is the only outlier. Bring it in line with the existing convention by switching to get_zeroed_page().

• git.kernel.orghttps://git.kernel.org/stable/c/53c97e9882f4e747b4ac31b211317c2eba541af9
• git.kernel.orghttps://git.kernel.org/stable/c/9128c2411b83a64c0a69d2ff059c741bde25a9cc
• git.kernel.orghttps://git.kernel.org/stable/c/9896bdfd21d918e9f26a52bc6109cc77970ee0b1
• git.kernel.orghttps://git.kernel.org/stable/c/a3ed2daab02b2a706e882ad31b5c3c4f33cb5bb1
• git.kernel.orghttps://git.kernel.org/stable/c/c19b360fa10c521c0b681875cdaa51545d45a491
• git.kernel.orghttps://git.kernel.org/stable/c/c75f8ce4baa29ae57fe615c6a2c5101f59b8b89a
• git.kernel.orghttps://git.kernel.org/stable/c/e2b143df29003d2704b51f62e9297006953dbacb
• git.kernel.orghttps://git.kernel.org/stable/c/f6b079629becfa977f9c51fe53ad2e6dcc55ef44

Sin CVEs relacionados por CWE o producto.