From ac3a2585f018f10039b4a856dcb122da88c1c1c9 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Sun, 11 Dec 2022 06:19:33 -0500 Subject: [PATCH 1/3] nfsd: rework refcounting in filecache The filecache refcounting is a bit non-standard for something searchable by RCU, in that we maintain a sentinel reference while it's hashed. This in turn requires that we have to do things differently in the "put" depending on whether its hashed, which we believe to have led to races. There are other problems in here too. nfsd_file_close_inode_sync can end up freeing an nfsd_file while there are still outstanding references to it, and there are a number of subtle ToC/ToU races. Rework the code so that the refcount is what drives the lifecycle. When the refcount goes to zero, then unhash and rcu free the object. A task searching for a nfsd_file is allowed to bump its refcount, but only if it's not already 0. Ensure that we don't make any other changes to it until a reference is held. With this change, the LRU carries a reference. Take special care to deal with it when removing an entry from the list, and ensure that we only repurpose the nf_lru list_head when the refcount is 0 to ensure exclusive access to it. Signed-off-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/filecache.c | 328 +++++++++++++++++++++++--------------------- fs/nfsd/trace.h | 51 +++---- 2 files changed, 194 insertions(+), 185 deletions(-) diff --git a/fs/nfsd/filecache.c b/fs/nfsd/filecache.c index 1998b4d5f692..45b2c9e3f636 100644 --- a/fs/nfsd/filecache.c +++ b/fs/nfsd/filecache.c @@ -324,8 +324,7 @@ nfsd_file_alloc(struct nfsd_file_lookup_key *key, unsigned int may) if (key->gc) __set_bit(NFSD_FILE_GC, &nf->nf_flags); nf->nf_inode = key->inode; - /* nf_ref is pre-incremented for hash table */ - refcount_set(&nf->nf_ref, 2); + refcount_set(&nf->nf_ref, 1); nf->nf_may = key->need; nf->nf_mark = NULL; } @@ -377,24 +376,35 @@ nfsd_file_unhash(struct nfsd_file *nf) return false; } -static bool +static void nfsd_file_free(struct nfsd_file *nf) { s64 age = ktime_to_ms(ktime_sub(ktime_get(), nf->nf_birthtime)); - bool flush = false; trace_nfsd_file_free(nf); this_cpu_inc(nfsd_file_releases); this_cpu_add(nfsd_file_total_age, age); + nfsd_file_unhash(nf); + + /* + * We call fsync here in order to catch writeback errors. It's not + * strictly required by the protocol, but an nfsd_file could get + * evicted from the cache before a COMMIT comes in. If another + * task were to open that file in the interim and scrape the error, + * then the client may never see it. By calling fsync here, we ensure + * that writeback happens before the entry is freed, and that any + * errors reported result in the write verifier changing. + */ + nfsd_file_fsync(nf); + if (nf->nf_mark) nfsd_file_mark_put(nf->nf_mark); if (nf->nf_file) { get_file(nf->nf_file); filp_close(nf->nf_file, NULL); fput(nf->nf_file); - flush = true; } /* @@ -402,10 +412,9 @@ nfsd_file_free(struct nfsd_file *nf) * WARN and leak it to preserve system stability. */ if (WARN_ON_ONCE(!list_empty(&nf->nf_lru))) - return flush; + return; call_rcu(&nf->nf_rcu, nfsd_file_slab_free); - return flush; } static bool @@ -421,17 +430,23 @@ nfsd_file_check_writeback(struct nfsd_file *nf) mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK); } -static void nfsd_file_lru_add(struct nfsd_file *nf) +static bool nfsd_file_lru_add(struct nfsd_file *nf) { set_bit(NFSD_FILE_REFERENCED, &nf->nf_flags); - if (list_lru_add(&nfsd_file_lru, &nf->nf_lru)) + if (list_lru_add(&nfsd_file_lru, &nf->nf_lru)) { trace_nfsd_file_lru_add(nf); + return true; + } + return false; } -static void nfsd_file_lru_remove(struct nfsd_file *nf) +static bool nfsd_file_lru_remove(struct nfsd_file *nf) { - if (list_lru_del(&nfsd_file_lru, &nf->nf_lru)) + if (list_lru_del(&nfsd_file_lru, &nf->nf_lru)) { trace_nfsd_file_lru_del(nf); + return true; + } + return false; } struct nfsd_file * @@ -442,54 +457,48 @@ nfsd_file_get(struct nfsd_file *nf) return NULL; } -static void -nfsd_file_unhash_and_queue(struct nfsd_file *nf, struct list_head *dispose) -{ - trace_nfsd_file_unhash_and_queue(nf); - if (nfsd_file_unhash(nf)) { - /* caller must call nfsd_file_dispose_list() later */ - nfsd_file_lru_remove(nf); - list_add(&nf->nf_lru, dispose); - } -} - -static void -nfsd_file_put_noref(struct nfsd_file *nf) -{ - trace_nfsd_file_put(nf); - - if (refcount_dec_and_test(&nf->nf_ref)) { - WARN_ON(test_bit(NFSD_FILE_HASHED, &nf->nf_flags)); - nfsd_file_lru_remove(nf); - nfsd_file_free(nf); - } -} - -static void -nfsd_file_unhash_and_put(struct nfsd_file *nf) -{ - if (nfsd_file_unhash(nf)) - nfsd_file_put_noref(nf); -} - +/** + * nfsd_file_put - put the reference to a nfsd_file + * @nf: nfsd_file of which to put the reference + * + * Put a reference to a nfsd_file. In the non-GC case, we just put the + * reference immediately. In the GC case, if the reference would be + * the last one, the put it on the LRU instead to be cleaned up later. + */ void nfsd_file_put(struct nfsd_file *nf) { might_sleep(); + trace_nfsd_file_put(nf); - if (test_bit(NFSD_FILE_GC, &nf->nf_flags)) - nfsd_file_lru_add(nf); - else if (refcount_read(&nf->nf_ref) == 2) - nfsd_file_unhash_and_put(nf); + if (test_bit(NFSD_FILE_GC, &nf->nf_flags) && + test_bit(NFSD_FILE_HASHED, &nf->nf_flags)) { + /* + * If this is the last reference (nf_ref == 1), then try to + * transfer it to the LRU. + */ + if (refcount_dec_not_one(&nf->nf_ref)) + return; - if (!test_bit(NFSD_FILE_HASHED, &nf->nf_flags)) { - nfsd_file_fsync(nf); - nfsd_file_put_noref(nf); - } else if (nf->nf_file && test_bit(NFSD_FILE_GC, &nf->nf_flags)) { - nfsd_file_put_noref(nf); - nfsd_file_schedule_laundrette(); - } else - nfsd_file_put_noref(nf); + /* Try to add it to the LRU. If that fails, decrement. */ + if (nfsd_file_lru_add(nf)) { + /* If it's still hashed, we're done */ + if (test_bit(NFSD_FILE_HASHED, &nf->nf_flags)) { + nfsd_file_schedule_laundrette(); + return; + } + + /* + * We're racing with unhashing, so try to remove it from + * the LRU. If removal fails, then someone else already + * has our reference. + */ + if (!nfsd_file_lru_remove(nf)) + return; + } + } + if (refcount_dec_and_test(&nf->nf_ref)) + nfsd_file_free(nf); } static void @@ -497,33 +506,13 @@ nfsd_file_dispose_list(struct list_head *dispose) { struct nfsd_file *nf; - while(!list_empty(dispose)) { + while (!list_empty(dispose)) { nf = list_first_entry(dispose, struct nfsd_file, nf_lru); list_del_init(&nf->nf_lru); - nfsd_file_fsync(nf); - nfsd_file_put_noref(nf); + nfsd_file_free(nf); } } -static void -nfsd_file_dispose_list_sync(struct list_head *dispose) -{ - bool flush = false; - struct nfsd_file *nf; - - while(!list_empty(dispose)) { - nf = list_first_entry(dispose, struct nfsd_file, nf_lru); - list_del_init(&nf->nf_lru); - nfsd_file_fsync(nf); - if (!refcount_dec_and_test(&nf->nf_ref)) - continue; - if (nfsd_file_free(nf)) - flush = true; - } - if (flush) - flush_delayed_fput(); -} - static void nfsd_file_list_remove_disposal(struct list_head *dst, struct nfsd_fcache_disposal *l) @@ -591,21 +580,8 @@ nfsd_file_lru_cb(struct list_head *item, struct list_lru_one *lru, struct list_head *head = arg; struct nfsd_file *nf = list_entry(item, struct nfsd_file, nf_lru); - /* - * Do a lockless refcount check. The hashtable holds one reference, so - * we look to see if anything else has a reference, or if any have - * been put since the shrinker last ran. Those don't get unhashed and - * released. - * - * Note that in the put path, we set the flag and then decrement the - * counter. Here we check the counter and then test and clear the flag. - * That order is deliberate to ensure that we can do this locklessly. - */ - if (refcount_read(&nf->nf_ref) > 1) { - list_lru_isolate(lru, &nf->nf_lru); - trace_nfsd_file_gc_in_use(nf); - return LRU_REMOVED; - } + /* We should only be dealing with GC entries here */ + WARN_ON_ONCE(!test_bit(NFSD_FILE_GC, &nf->nf_flags)); /* * Don't throw out files that are still undergoing I/O or @@ -616,40 +592,30 @@ nfsd_file_lru_cb(struct list_head *item, struct list_lru_one *lru, return LRU_SKIP; } + /* If it was recently added to the list, skip it */ if (test_and_clear_bit(NFSD_FILE_REFERENCED, &nf->nf_flags)) { trace_nfsd_file_gc_referenced(nf); return LRU_ROTATE; } - if (!test_and_clear_bit(NFSD_FILE_HASHED, &nf->nf_flags)) { - trace_nfsd_file_gc_hashed(nf); - return LRU_SKIP; + /* + * Put the reference held on behalf of the LRU. If it wasn't the last + * one, then just remove it from the LRU and ignore it. + */ + if (!refcount_dec_and_test(&nf->nf_ref)) { + trace_nfsd_file_gc_in_use(nf); + list_lru_isolate(lru, &nf->nf_lru); + return LRU_REMOVED; } + /* Refcount went to zero. Unhash it and queue it to the dispose list */ + nfsd_file_unhash(nf); list_lru_isolate_move(lru, &nf->nf_lru, head); this_cpu_inc(nfsd_file_evictions); trace_nfsd_file_gc_disposed(nf); return LRU_REMOVED; } -/* - * Unhash items on @dispose immediately, then queue them on the - * disposal workqueue to finish releasing them in the background. - * - * cel: Note that between the time list_lru_shrink_walk runs and - * now, these items are in the hash table but marked unhashed. - * Why release these outside of lru_cb ? There's no lock ordering - * problem since lru_cb currently takes no lock. - */ -static void nfsd_file_gc_dispose_list(struct list_head *dispose) -{ - struct nfsd_file *nf; - - list_for_each_entry(nf, dispose, nf_lru) - nfsd_file_hash_remove(nf); - nfsd_file_dispose_list_delayed(dispose); -} - static void nfsd_file_gc(void) { @@ -659,7 +625,7 @@ nfsd_file_gc(void) ret = list_lru_walk(&nfsd_file_lru, nfsd_file_lru_cb, &dispose, list_lru_count(&nfsd_file_lru)); trace_nfsd_file_gc_removed(ret, list_lru_count(&nfsd_file_lru)); - nfsd_file_gc_dispose_list(&dispose); + nfsd_file_dispose_list_delayed(&dispose); } static void @@ -685,7 +651,7 @@ nfsd_file_lru_scan(struct shrinker *s, struct shrink_control *sc) ret = list_lru_shrink_walk(&nfsd_file_lru, sc, nfsd_file_lru_cb, &dispose); trace_nfsd_file_shrinker_removed(ret, list_lru_count(&nfsd_file_lru)); - nfsd_file_gc_dispose_list(&dispose); + nfsd_file_dispose_list_delayed(&dispose); return ret; } @@ -695,72 +661,111 @@ static struct shrinker nfsd_file_shrinker = { .seeks = 1, }; -/* - * Find all cache items across all net namespaces that match @inode and - * move them to @dispose. The lookup is atomic wrt nfsd_file_acquire(). +/** + * nfsd_file_queue_for_close: try to close out any open nfsd_files for an inode + * @inode: inode on which to close out nfsd_files + * @dispose: list on which to gather nfsd_files to close out + * + * An nfsd_file represents a struct file being held open on behalf of nfsd. An + * open file however can block other activity (such as leases), or cause + * undesirable behavior (e.g. spurious silly-renames when reexporting NFS). + * + * This function is intended to find open nfsd_files when this sort of + * conflicting access occurs and then attempt to close those files out. + * + * Populates the dispose list with entries that have already had their + * refcounts go to zero. The actual free of an nfsd_file can be expensive, + * so we leave it up to the caller whether it wants to wait or not. */ -static unsigned int -__nfsd_file_close_inode(struct inode *inode, struct list_head *dispose) +static void +nfsd_file_queue_for_close(struct inode *inode, struct list_head *dispose) { struct nfsd_file_lookup_key key = { .type = NFSD_FILE_KEY_INODE, .inode = inode, }; - unsigned int count = 0; struct nfsd_file *nf; rcu_read_lock(); do { + int decrement = 1; + nf = rhashtable_lookup(&nfsd_file_rhash_tbl, &key, nfsd_file_rhash_params); if (!nf) break; - nfsd_file_unhash_and_queue(nf, dispose); - count++; + + /* If we raced with someone else unhashing, ignore it */ + if (!nfsd_file_unhash(nf)) + continue; + + /* If we can't get a reference, ignore it */ + if (!nfsd_file_get(nf)) + continue; + + /* Extra decrement if we remove from the LRU */ + if (nfsd_file_lru_remove(nf)) + ++decrement; + + /* If refcount goes to 0, then put on the dispose list */ + if (refcount_sub_and_test(decrement, &nf->nf_ref)) { + list_add(&nf->nf_lru, dispose); + trace_nfsd_file_closing(nf); + } } while (1); rcu_read_unlock(); - return count; -} - -/** - * nfsd_file_close_inode_sync - attempt to forcibly close a nfsd_file - * @inode: inode of the file to attempt to remove - * - * Unhash and put, then flush and fput all cache items associated with @inode. - */ -void -nfsd_file_close_inode_sync(struct inode *inode) -{ - LIST_HEAD(dispose); - unsigned int count; - - count = __nfsd_file_close_inode(inode, &dispose); - trace_nfsd_file_close_inode_sync(inode, count); - nfsd_file_dispose_list_sync(&dispose); } /** * nfsd_file_close_inode - attempt a delayed close of a nfsd_file * @inode: inode of the file to attempt to remove * - * Unhash and put all cache item associated with @inode. + * Close out any open nfsd_files that can be reaped for @inode. The + * actual freeing is deferred to the dispose_list_delayed infrastructure. + * + * This is used by the fsnotify callbacks and setlease notifier. */ static void nfsd_file_close_inode(struct inode *inode) { LIST_HEAD(dispose); - unsigned int count; - count = __nfsd_file_close_inode(inode, &dispose); - trace_nfsd_file_close_inode(inode, count); + nfsd_file_queue_for_close(inode, &dispose); nfsd_file_dispose_list_delayed(&dispose); } +/** + * nfsd_file_close_inode_sync - attempt to forcibly close a nfsd_file + * @inode: inode of the file to attempt to remove + * + * Close out any open nfsd_files that can be reaped for @inode. The + * nfsd_files are closed out synchronously. + * + * This is called from nfsd_rename and nfsd_unlink to avoid silly-renames + * when reexporting NFS. + */ +void +nfsd_file_close_inode_sync(struct inode *inode) +{ + struct nfsd_file *nf; + LIST_HEAD(dispose); + + trace_nfsd_file_close(inode); + + nfsd_file_queue_for_close(inode, &dispose); + while (!list_empty(&dispose)) { + nf = list_first_entry(&dispose, struct nfsd_file, nf_lru); + list_del_init(&nf->nf_lru); + nfsd_file_free(nf); + } + flush_delayed_fput(); +} + /** * nfsd_file_delayed_close - close unused nfsd_files * @work: dummy * - * Walk the LRU list and close any entries that have not been used since + * Walk the LRU list and destroy any entries that have not been used since * the last scan. */ static void @@ -782,7 +787,7 @@ nfsd_file_lease_notifier_call(struct notifier_block *nb, unsigned long arg, /* Only close files for F_SETLEASE leases */ if (fl->fl_flags & FL_LEASE) - nfsd_file_close_inode_sync(file_inode(fl->fl_file)); + nfsd_file_close_inode(file_inode(fl->fl_file)); return 0; } @@ -903,6 +908,13 @@ out_err: goto out; } +/** + * __nfsd_file_cache_purge: clean out the cache for shutdown + * @net: net-namespace to shut down the cache (may be NULL) + * + * Walk the nfsd_file cache and close out any that match @net. If @net is NULL, + * then close out everything. Called when an nfsd instance is being shut down. + */ static void __nfsd_file_cache_purge(struct net *net) { @@ -916,8 +928,11 @@ __nfsd_file_cache_purge(struct net *net) nf = rhashtable_walk_next(&iter); while (!IS_ERR_OR_NULL(nf)) { - if (!net || nf->nf_net == net) - nfsd_file_unhash_and_queue(nf, &dispose); + if (!net || nf->nf_net == net) { + nfsd_file_unhash(nf); + nfsd_file_lru_remove(nf); + list_add(&nf->nf_lru, &dispose); + } nf = rhashtable_walk_next(&iter); } @@ -1084,8 +1099,12 @@ retry: if (nf) nf = nfsd_file_get(nf); rcu_read_unlock(); - if (nf) + + if (nf) { + if (nfsd_file_lru_remove(nf)) + WARN_ON_ONCE(refcount_dec_and_test(&nf->nf_ref)); goto wait_for_construction; + } nf = nfsd_file_alloc(&key, may_flags); if (!nf) { @@ -1118,11 +1137,11 @@ wait_for_construction: goto out; } open_retry = false; - nfsd_file_put_noref(nf); + if (refcount_dec_and_test(&nf->nf_ref)) + nfsd_file_free(nf); goto retry; } - nfsd_file_lru_remove(nf); this_cpu_inc(nfsd_file_cache_hits); status = nfserrno(nfsd_open_break_lease(file_inode(nf->nf_file), may_flags)); @@ -1132,7 +1151,8 @@ out: this_cpu_inc(nfsd_file_acquisitions); *pnf = nf; } else { - nfsd_file_put(nf); + if (refcount_dec_and_test(&nf->nf_ref)) + nfsd_file_free(nf); nf = NULL; } @@ -1158,8 +1178,10 @@ open_file: * If construction failed, or we raced with a call to unlink() * then unhash. */ - if (status != nfs_ok || key.inode->i_nlink == 0) - nfsd_file_unhash_and_put(nf); + if (status == nfs_ok && key.inode->i_nlink == 0) + status = nfserr_jukebox; + if (status != nfs_ok) + nfsd_file_unhash(nf); clear_bit_unlock(NFSD_FILE_PENDING, &nf->nf_flags); smp_mb__after_atomic(); wake_up_bit(&nf->nf_flags, NFSD_FILE_PENDING); diff --git a/fs/nfsd/trace.h b/fs/nfsd/trace.h index 46b8f68a2497..c852ae8eaf37 100644 --- a/fs/nfsd/trace.h +++ b/fs/nfsd/trace.h @@ -876,8 +876,8 @@ DEFINE_CLID_EVENT(confirmed_r); __print_flags(val, "|", \ { 1 << NFSD_FILE_HASHED, "HASHED" }, \ { 1 << NFSD_FILE_PENDING, "PENDING" }, \ - { 1 << NFSD_FILE_REFERENCED, "REFERENCED"}, \ - { 1 << NFSD_FILE_GC, "GC"}) + { 1 << NFSD_FILE_REFERENCED, "REFERENCED" }, \ + { 1 << NFSD_FILE_GC, "GC" }) DECLARE_EVENT_CLASS(nfsd_file_class, TP_PROTO(struct nfsd_file *nf), @@ -912,6 +912,7 @@ DEFINE_EVENT(nfsd_file_class, name, \ DEFINE_NFSD_FILE_EVENT(nfsd_file_free); DEFINE_NFSD_FILE_EVENT(nfsd_file_unhash); DEFINE_NFSD_FILE_EVENT(nfsd_file_put); +DEFINE_NFSD_FILE_EVENT(nfsd_file_closing); DEFINE_NFSD_FILE_EVENT(nfsd_file_unhash_and_queue); TRACE_EVENT(nfsd_file_alloc, @@ -1103,35 +1104,6 @@ TRACE_EVENT(nfsd_file_open, __entry->nf_file) ) -DECLARE_EVENT_CLASS(nfsd_file_search_class, - TP_PROTO( - const struct inode *inode, - unsigned int count - ), - TP_ARGS(inode, count), - TP_STRUCT__entry( - __field(const struct inode *, inode) - __field(unsigned int, count) - ), - TP_fast_assign( - __entry->inode = inode; - __entry->count = count; - ), - TP_printk("inode=%p count=%u", - __entry->inode, __entry->count) -); - -#define DEFINE_NFSD_FILE_SEARCH_EVENT(name) \ -DEFINE_EVENT(nfsd_file_search_class, name, \ - TP_PROTO( \ - const struct inode *inode, \ - unsigned int count \ - ), \ - TP_ARGS(inode, count)) - -DEFINE_NFSD_FILE_SEARCH_EVENT(nfsd_file_close_inode_sync); -DEFINE_NFSD_FILE_SEARCH_EVENT(nfsd_file_close_inode); - TRACE_EVENT(nfsd_file_is_cached, TP_PROTO( const struct inode *inode, @@ -1209,7 +1181,6 @@ DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_lru_del_disposed); DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_in_use); DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_writeback); DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_referenced); -DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_hashed); DEFINE_NFSD_FILE_GC_EVENT(nfsd_file_gc_disposed); DECLARE_EVENT_CLASS(nfsd_file_lruwalk_class, @@ -1241,6 +1212,22 @@ DEFINE_EVENT(nfsd_file_lruwalk_class, name, \ DEFINE_NFSD_FILE_LRUWALK_EVENT(nfsd_file_gc_removed); DEFINE_NFSD_FILE_LRUWALK_EVENT(nfsd_file_shrinker_removed); +TRACE_EVENT(nfsd_file_close, + TP_PROTO( + const struct inode *inode + ), + TP_ARGS(inode), + TP_STRUCT__entry( + __field(const void *, inode) + ), + TP_fast_assign( + __entry->inode = inode; + ), + TP_printk("inode=%p", + __entry->inode + ) +); + TRACE_EVENT(nfsd_file_fsync, TP_PROTO( const struct nfsd_file *nf, From 3bc8edc98bd43540dbe648e4ef91f443d6d20a24 Mon Sep 17 00:00:00 2001 From: Dan Aloni Date: Mon, 12 Dec 2022 13:11:06 +0200 Subject: [PATCH 2/3] nfsd: under NFSv4.1, fix double svc_xprt_put on rpc_create failure On error situation `clp->cl_cb_conn.cb_xprt` should not be given a reference to the xprt otherwise both client cleanup and the error handling path of the caller call to put it. Better to delay handing over the reference to a later branch. [ 72.530665] refcount_t: underflow; use-after-free. [ 72.531933] WARNING: CPU: 0 PID: 173 at lib/refcount.c:28 refcount_warn_saturate+0xcf/0x120 [ 72.533075] Modules linked in: nfsd(OE) nfsv4(OE) nfsv3(OE) nfs(OE) lockd(OE) compat_nfs_ssc(OE) nfs_acl(OE) rpcsec_gss_krb5(OE) auth_rpcgss(OE) rpcrdma(OE) dns_resolver fscache netfs grace rdma_cm iw_cm ib_cm sunrpc(OE) mlx5_ib mlx5_core mlxfw pci_hyperv_intf ib_uverbs ib_core xt_MASQUERADE nf_conntrack_netlink nft_counter xt_addrtype nft_compat br_netfilter bridge stp llc nft_reject_inet nf_reject_ipv4 nf_reject_ipv6 nft_reject nft_ct nft_chain_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 ip_set overlay nf_tables nfnetlink crct10dif_pclmul crc32_pclmul ghash_clmulni_intel xfs serio_raw virtio_net virtio_blk net_failover failover fuse [last unloaded: sunrpc] [ 72.540389] CPU: 0 PID: 173 Comm: kworker/u16:5 Tainted: G OE 5.15.82-dan #1 [ 72.541511] Hardware name: Red Hat KVM/RHEL-AV, BIOS 1.16.0-3.module+el8.7.0+1084+97b81f61 04/01/2014 [ 72.542717] Workqueue: nfsd4_callbacks nfsd4_run_cb_work [nfsd] [ 72.543575] RIP: 0010:refcount_warn_saturate+0xcf/0x120 [ 72.544299] Code: 55 00 0f 0b 5d e9 01 50 98 00 80 3d 75 9e 39 08 00 0f 85 74 ff ff ff 48 c7 c7 e8 d1 60 8e c6 05 61 9e 39 08 01 e8 f6 51 55 00 <0f> 0b 5d e9 d9 4f 98 00 80 3d 4b 9e 39 08 00 0f 85 4c ff ff ff 48 [ 72.546666] RSP: 0018:ffffb3f841157cf0 EFLAGS: 00010286 [ 72.547393] RAX: 0000000000000026 RBX: ffff89ac6231d478 RCX: 0000000000000000 [ 72.548324] RDX: ffff89adb7c2c2c0 RSI: ffff89adb7c205c0 RDI: ffff89adb7c205c0 [ 72.549271] RBP: ffffb3f841157cf0 R08: 0000000000000000 R09: c0000000ffefffff [ 72.550209] R10: 0000000000000001 R11: ffffb3f841157ad0 R12: ffff89ac6231d180 [ 72.551142] R13: ffff89ac6231d478 R14: ffff89ac40c06180 R15: ffff89ac6231d4b0 [ 72.552089] FS: 0000000000000000(0000) GS:ffff89adb7c00000(0000) knlGS:0000000000000000 [ 72.553175] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 72.553934] CR2: 0000563a310506a8 CR3: 0000000109a66000 CR4: 0000000000350ef0 [ 72.554874] Call Trace: [ 72.555278] [ 72.555614] svc_xprt_put+0xaf/0xe0 [sunrpc] [ 72.556276] nfsd4_process_cb_update.isra.11+0xb7/0x410 [nfsd] [ 72.557087] ? update_load_avg+0x82/0x610 [ 72.557652] ? cpuacct_charge+0x60/0x70 [ 72.558212] ? dequeue_entity+0xdb/0x3e0 [ 72.558765] ? queued_spin_unlock+0x9/0x20 [ 72.559358] nfsd4_run_cb_work+0xfc/0x270 [nfsd] [ 72.560031] process_one_work+0x1df/0x390 [ 72.560600] worker_thread+0x37/0x3b0 [ 72.561644] ? process_one_work+0x390/0x390 [ 72.562247] kthread+0x12f/0x150 [ 72.562710] ? set_kthread_struct+0x50/0x50 [ 72.563309] ret_from_fork+0x22/0x30 [ 72.563818] [ 72.564189] ---[ end trace 031117b1c72ec616 ]--- [ 72.566019] list_add corruption. next->prev should be prev (ffff89ac4977e538), but was ffff89ac4763e018. (next=ffff89ac4763e018). [ 72.567647] ------------[ cut here ]------------ Fixes: a4abc6b12eb1 ("nfsd: Fix svc_xprt refcnt leak when setup callback client failed") Cc: Xiyu Yang Cc: J. Bruce Fields Signed-off-by: Dan Aloni Reviewed-by: Jeff Layton Signed-off-by: Chuck Lever --- fs/nfsd/nfs4callback.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 1b57f2c2f0bb..905d66acf6ab 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -988,7 +988,6 @@ static int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *c } else { if (!conn->cb_xprt) return -EINVAL; - clp->cl_cb_conn.cb_xprt = conn->cb_xprt; clp->cl_cb_session = ses; args.bc_xprt = conn->cb_xprt; args.prognumber = clp->cl_cb_session->se_cb_prog; @@ -1008,6 +1007,9 @@ static int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *c rpc_shutdown_client(client); return -ENOMEM; } + + if (clp->cl_minorversion != 0) + clp->cl_cb_conn.cb_xprt = conn->cb_xprt; clp->cl_cb_client = client; clp->cl_cb_cred = cred; rcu_read_lock(); From 75333d48f92256a0dec91dbf07835e804fc411c0 Mon Sep 17 00:00:00 2001 From: Dai Ngo Date: Mon, 12 Dec 2022 14:50:11 -0800 Subject: [PATCH 3/3] NFSD: fix use-after-free in __nfs42_ssc_open() Problem caused by source's vfsmount being unmounted but remains on the delayed unmount list. This happens when nfs42_ssc_open() return errors. Fixed by removing nfsd4_interssc_connect(), leave the vfsmount for the laundromat to unmount when idle time expires. We don't need to call nfs_do_sb_deactive when nfs42_ssc_open return errors since the file was not opened so nfs_server->active was not incremented. Same as in nfsd4_copy, if we fail to launch nfsd4_do_async_copy thread then there's no need to call nfs_do_sb_deactive Reported-by: Xingyuan Mo Signed-off-by: Dai Ngo Tested-by: Xingyuan Mo Signed-off-by: Chuck Lever --- fs/nfsd/nfs4proc.c | 20 +++++--------------- 1 file changed, 5 insertions(+), 15 deletions(-) diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 73ed32ad23a2..bd880d55f565 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -1461,13 +1461,6 @@ out_err: return status; } -static void -nfsd4_interssc_disconnect(struct vfsmount *ss_mnt) -{ - nfs_do_sb_deactive(ss_mnt->mnt_sb); - mntput(ss_mnt); -} - /* * Verify COPY destination stateid. * @@ -1570,11 +1563,6 @@ nfsd4_cleanup_inter_ssc(struct vfsmount *ss_mnt, struct file *filp, { } -static void -nfsd4_interssc_disconnect(struct vfsmount *ss_mnt) -{ -} - static struct file *nfs42_ssc_open(struct vfsmount *ss_mnt, struct nfs_fh *src_fh, nfs4_stateid *stateid) @@ -1770,7 +1758,7 @@ static int nfsd4_do_async_copy(void *data) default: nfserr = nfserr_offload_denied; } - nfsd4_interssc_disconnect(copy->ss_mnt); + /* ss_mnt will be unmounted by the laundromat */ goto do_callback; } nfserr = nfsd4_do_copy(copy, filp, copy->nf_dst->nf_file, @@ -1851,8 +1839,10 @@ out_err: if (async_copy) cleanup_async_copy(async_copy); status = nfserrno(-ENOMEM); - if (nfsd4_ssc_is_inter(copy)) - nfsd4_interssc_disconnect(copy->ss_mnt); + /* + * source's vfsmount of inter-copy will be unmounted + * by the laundromat + */ goto out; }