Date: Mon, 16 May 2016 15:03:53 +0000 (UTC) From: Andriy Gapon <avg@FreeBSD.org> To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: svn commit: r299945 - head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs Message-ID: <201605161503.u4GF3rRa020856@repo.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: avg Date: Mon May 16 15:03:52 2016 New Revision: 299945 URL: https://svnweb.freebsd.org/changeset/base/299945 Log: avoid deadlock between zfsctl_snapdir_lookup and zfsctl_snapshot_reclaim The former acquired a snap vnode lock while holding sd_lock while the latter does the opposite. The solution is drop sd_lock before acquiring the vnode lock. That should be okay as we are still holding a lock on the 'snapshot' directory in the exclusive mode. That lock ensures that there are no concurrent lookups in the directory and thus no concurrent mount attempts. But now we have to account for the possibility that the snap vnode might get reclaim after we drop sd_lock and before we can get the node lock. So, check for that case and retry. MFC after: 5 weeks Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c Modified: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c ============================================================================== --- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c Mon May 16 14:07:43 2016 (r299944) +++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ctldir.c Mon May 16 15:03:52 2016 (r299945) @@ -1011,6 +1011,7 @@ zfsctl_snapdir_lookup(ap) #endif } +relookup: mutex_enter(&sdp->sd_lock); search.se_name = (char *)nm; if ((sep = avl_find(&sdp->sd_snaps, &search, &where)) != NULL) { @@ -1085,7 +1086,16 @@ domount: (void) snprintf(mountpoint, mountpoint_len, "%s/" ZFS_CTLDIR_NAME "/snapshot/%s", dvp->v_vfsp->mnt_stat.f_mntonname, nm); - VERIFY0(vn_lock(*vpp, LK_EXCLUSIVE)); + mutex_exit(&sdp->sd_lock); + + /* + * The vnode may get reclaimed between dropping sd_lock and + * getting the vnode lock. + * */ + err = vn_lock(*vpp, LK_EXCLUSIVE); + if (err == ENOENT) + goto relookup; + VERIFY0(err); err = mount_snapshot(curthread, vpp, "zfs", mountpoint, snapname, 0); kmem_free(mountpoint, mountpoint_len); if (err == 0) { @@ -1100,7 +1110,6 @@ domount: VTOZ(*vpp)->z_zfsvfs->z_parent = zfsvfs; (*vpp)->v_flag &= ~VROOT; } - mutex_exit(&sdp->sd_lock); ZFS_EXIT(zfsvfs); #ifdef illumos
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201605161503.u4GF3rRa020856>