CVE-2026-53399

In the Linux kernel, the following vulnerability has been resolved:

nfsd: release layout stid on setlease failure

nfs4allocstid() publishes the new stid into cl->clstateids via idralloccyclic() under cllock before returning to nfsd4alloclayoutstateid(). When nfsd4layoutsetlease() then fails, the error path frees the layout stateid directly with kmemcachefree() without ever calling idrremove(), leaving the IDR slot pointing at freed slab memory. Any subsequent IDR walker (states_show, client teardown) dereferences the dangling pointer.

The correct teardown for an IDR-published stid is nfs4putstid(), which removes the IDR slot under cllock, dispatches scfree (nfsd4freelayoutstateid) to release ls->lsfile via nfsd4closelayout(), and drops the nfs4_file reference in its tail.

A second issue blocks that switch: nfsd4freelayoutstateid() unconditionally inspects ls->lsfencework via delayedworkpending() under lslock, but INITDELAYEDWORK(&ls->lsfencework, ...) currently runs only after the setlease call. On the setlease-failure path the destructor would touch an uninitialized delayed_work.

nfsd4_alloc_layout_stateid()
  nfs4_alloc_stid()           /* idr_alloc_cyclic under cl_lock */
  nfsd4_layout_setlease()     /* fails */
    nfs4_put_stid()
      nfsd4_free_layout_stateid()
        delayed_work_pending(&ls->ls_fence_work)  /* needs INIT */
        nfsd4_close_layout()  /* nfsd_file_put(ls->ls_file) */
      put_nfs4_file()

Fix by hoisting the lsfenced / lsfencedelay / INITDELAYEDWORK initialization above the nfsd4layoutsetlease() call, and replace the manual nfsdfileput + putnfs4file + kmemcachefree cleanup with a single nfs4put_stid(stp).