CVE-2026-72473

published 15 Aug 2026 · updated 14 Nov 2026

xprtrdma: Decouple req recycling from RPC completion

9.8CriticalCVSS 9.8EPSS 1%

No exploitation in the wild has been reported. EPSS puts the chance of exploitation in the next 30 days at 1%. Fixed in Kernel e786233d2e0bbff9a82e43f02ae3a46ab4b08ec3, Kernel 53442c7d0c888e51b8bc3da196970a669cc6b294, Kernel 740975054a1970c0cf15f70ac39724a064f45847 and 8 more.

In the Linux kernel, the following vulnerability has been resolved: xprtrdma: Decouple req recycling from RPC completion rl_kref formerly served two distinct lifetimes through a single refcount: it gated when a Reply could wake its RPC task, and it gated when an rpcrdma_req could return to its free pool. The marshal path took the Send-side reference only when SGEs needed DMA-unmap (sc_unmap_count > 0), which made a Send carrying only pre-registered buffers an exception: the Reply handler dropped rl_kref from 1 to 0 and freed the req while the HCA might still be DMA-reading from its send buffer. Give rl_kref a narrower job. The RPC layer takes one reference when slot allocation hands a req out. rpcrdma_prepare_send_sges() takes a Send-side reference unconditionally after WR preparation succeeds. xprt_rdma_free_slot() and xprt_rdma_bc_free_rqst() drop the RPC-layer reference; rpcrdma_sendctx_unmap() drops the Send-side reference. The req returns to its free pool only after both owners have signed off. The existing kref_init(&req->rl_kref) call in rpcrdma_prepare_send_sges() is removed. Initialization moves to the slot-allocation paths (xprt_rdma_alloc_slot and rpcrdma_bc_rqst_get), and the release callback re-arms rl_kref before the req returns to a free pool. A re-init in the marshal path would discard the RPC-layer reference that already exists on entry. Three invariants follow: - Any rpcrdma_req held by an rpc_rqst has rl_kref >= 1. xprt_rdma_alloc_slot(), rpcrdma_bc_rqst_get(), and the backlog-wake branch in xprt_rdma_alloc_slot() each kref_init rl_kref before publishing the req. Without this invariant, an RPC task that aborts between slot allocation and marshal (gss_refresh failure or signal during call_connect, for example) would drive xprt_release() -> xprt_rdma_free_slot() -> kref_put against a refcount of zero, saturating refcount_t and stranding the slot. - The Send-side reference is taken only after WR prep succeeds. A mapping failure in rpcrdma_prepare_send_sges() runs rpcrdma_sendctx_cancel(), which DMA-unmaps the sendctx and clears sc_req without touching rl_kref. The sendctx ring walks in rpcrdma_sendctx_put_locked() and rpcrdma_sendctxs_destroy() skip entries with sc_req == NULL, so a burst of -EIO marshal failures cannot hold reqs off rb_send_bufs. - The release callback re-arms rl_kref so the next consumer enters with the invariant satisfied. Replies now complete the RPC directly. rpcrdma_reply_handler() calls rpcrdma_complete_rqst() in place of kref_put on the non-LocalInv branch. The LocalInv branch already completes the RPC from frwr_unmap_async() and is unaffected. Because Send-side references can now outlive RPC completion, connection teardown drains sendctx entries whose unsignaled Sends never had a later signaled completion to walk the ring. rpcrdma_sendctxs_destroy() walks the active range and runs rpcrdma_sendctx_unmap() on each entry with a non-NULL sc_req before the request buffers are reset, and is moved ahead of rpcrdma_reqs_reset() in rpcrdma_xprt_disconnect() so the reqs are still in their pre-reset state when the Send-side refs are released. The drain creates a teardown-ordering hazard on the backchannel path. With the new lifetime, releasing a bc_prealloc req from rpcrdma_req_release() re-adds it to bc_pa_list. The disconnect in xprt_rdma_destroy() runs after xprt_destroy_backchannel() has already emptied bc_pa_list, so the drained reqs would otherwise leak. xprt_rdma_destroy() now runs xprt_rdma_bc_destroy(xprt, 0) a second time after the disconnect to reclaim them.

Affected

Vendors, products and packages with version ranges

VendorProductVersionsFixed
LinuxKernel>=0ab115237025f5e379620bbcd56a02697d07b002 <e786233d2e0bbff9a82e43f02ae3a46ab4b08ec3, >=0ab115237025f5e379620bbcd56a02697d07b002 <53442c7d0c888e51b8bc3da196970a669cc6b294, >=0ab115237025f5e379620bbcd56a02697d07b002 <740975054a1970c0cf15f70ac39724a064f45847, >=0ab115237025f5e379620bbcd56a02697d07b002 <8203f760a72bd39a3b66bc4eff0aa272a99fe22b, >=0ab115237025f5e379620bbcd56a02697d07b002 <9f3d9b68c1c6c51746e5ecdb52b2e6a2901de37e, >=0ab115237025f5e379620bbcd56a02697d07b002 <e7632089523acddcdd8f090ad19e96fb3107b04d, =5.3e786233d2e0bbff9a82e43f02ae3a46ab4b08ec3, 53442c7d0c888e51b8bc3da196970a669cc6b294, 740975054a1970c0cf15f70ac39724a064f45847, 8203f760a72bd39a3b66bc4eff0aa272a99fe22b, 9f3d9b68c1c6c51746e5ecdb52b2e6a2901de37e, e7632089523acddcdd8f090ad19e96fb3107b04d
linux-kernel: Kernel>=6.13.0 <6.18.40, >=6.19.0 <7.1.5, >=6.2.0 <6.6.145, >=6.7.0 <6.12.97, >=5.3.0 <6.1.1786.18.40, 7.1.5, 6.6.145, 6.12.97, 6.1.178

Timeline

Every event on this vulnerability

Related

Other vulnerabilities in Kernel, strongest first

Embed

A live severity badge for an advisory, README or ticket

CVE-2026-72473 severity on Zeptric[![CVE-2026-72473](https://zeptric.io/badge/cve/CVE-2026-72473.svg)](https://zeptric.io/cve/CVE-2026-72473)