[PATCH AUTOSEL 5.4 061/330] SUNRPC: Capture completion of all RPC tasks

[PATCH AUTOSEL 5.4 061/330] SUNRPC: Capture completion of all RPC tasks

[Sasha Levin]

From: Chuck Lever <chuck.lever@oracle.com>

[ Upstream commit a264abad51d8ecb7954a2f6d9f1885b38daffc74 ]

RPC tasks on the backchannel never invoke xprt_complete_rqst(), so
there is no way to report their tk_status at completion. Also, any
RPC task that exits via rpc_exit_task() before it is replied to will
also disappear without a trace.

Introduce a trace point that is symmetrical with rpc_task_begin that
captures the termination status of each RPC task.

Sample trace output for callback requests initiated on the server:
   kworker/u8:12-448   [003]   127.025240: rpc_task_end:         task:50@3 flags=ASYNC|DYNAMIC|SOFT|SOFTCONN|SENT runstate=RUNNING|ACTIVE status=0 action=rpc_exit_task
   kworker/u8:12-448   [002]   127.567310: rpc_task_end:         task:51@3 flags=ASYNC|DYNAMIC|SOFT|SOFTCONN|SENT runstate=RUNNING|ACTIVE status=0 action=rpc_exit_task
   kworker/u8:12-448   [001]   130.506817: rpc_task_end:         task:52@3 flags=ASYNC|DYNAMIC|SOFT|SOFTCONN|SENT runstate=RUNNING|ACTIVE status=0 action=rpc_exit_task

Odd, though, that I never see trace_rpc_task_complete, either in the
forward or backchannel. Should it be removed?

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 include/trace/events/sunrpc.h | 1 +
 net/sunrpc/sched.c            | 1 +
 2 files changed, 2 insertions(+)

diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h
index ffa3c51dbb1a0..28df77a948e56 100644
--- a/include/trace/events/sunrpc.h
+++ b/include/trace/events/sunrpc.h
@@ -165,6 +165,7 @@ DECLARE_EVENT_CLASS(rpc_task_running,
 DEFINE_RPC_RUNNING_EVENT(begin);
 DEFINE_RPC_RUNNING_EVENT(run_action);
 DEFINE_RPC_RUNNING_EVENT(complete);
+DEFINE_RPC_RUNNING_EVENT(end);
 
 DECLARE_EVENT_CLASS(rpc_task_queued,
 
diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c
index 987c4b1f0b174..9c79548c68474 100644
--- a/net/sunrpc/sched.c
+++ b/net/sunrpc/sched.c
@@ -824,6 +824,7 @@ rpc_reset_task_statistics(struct rpc_task *task)
  */
 void rpc_exit_task(struct rpc_task *task)
 {
+	trace_rpc_task_end(task, task->tk_action);
 	task->tk_action = NULL;
 	if (task->tk_ops->rpc_count_stats)
 		task->tk_ops->rpc_count_stats(task, task->tk_calldata);
-- 
2.25.1

[PATCH AUTOSEL 5.4 100/330] nfsd: Fix a soft lockup race in nfsd_file_mark_find_or_create()

[Sasha Levin]

From: Trond Myklebust <trondmy@gmail.com>

[ Upstream commit 90d2f1da832fd23290ef0c0d964d97501e5e8553 ]

If nfsd_file_mark_find_or_create() keeps winning the race for the
nfsd_file_fsnotify_group->mark_mutex against nfsd_file_mark_put()
then it can soft lock up, since fsnotify_add_inode_mark() ends
up always finding an existing entry.

Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/nfsd/filecache.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/fs/nfsd/filecache.c b/fs/nfsd/filecache.c
index 3007b8945d388..51c08ae79063c 100644
--- a/fs/nfsd/filecache.c
+++ b/fs/nfsd/filecache.c
@@ -133,9 +133,13 @@ nfsd_file_mark_find_or_create(struct nfsd_file *nf)
 						 struct nfsd_file_mark,
 						 nfm_mark));
 			mutex_unlock(&nfsd_file_fsnotify_group->mark_mutex);
-			fsnotify_put_mark(mark);
-			if (likely(nfm))
+			if (nfm) {
+				fsnotify_put_mark(mark);
 				break;
+			}
+			/* Avoid soft lockup race with nfsd_file_mark_put() */
+			fsnotify_destroy_mark(mark, nfsd_file_fsnotify_group);
+			fsnotify_put_mark(mark);
 		} else
 			mutex_unlock(&nfsd_file_fsnotify_group->mark_mutex);
 
-- 
2.25.1

[PATCH AUTOSEL 5.4 118/330] nfsd: Fix a perf warning

[Sasha Levin]

From: Trond Myklebust <trondmy@gmail.com>

[ Upstream commit a9ceb060b3cf37987b6162223575eaf4f4e0fc36 ]

perf does not know how to deal with a __builtin_bswap32() call, and
complains. All other functions just store the xid etc in host endian
form, so let's do that in the tracepoint for nfsd_file_acquire too.

Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/nfsd/trace.h | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/fs/nfsd/trace.h b/fs/nfsd/trace.h
index ffc78a0e28b24..b073bdc2e6e89 100644
--- a/fs/nfsd/trace.h
+++ b/fs/nfsd/trace.h
@@ -228,7 +228,7 @@ TRACE_EVENT(nfsd_file_acquire,
 	TP_ARGS(rqstp, hash, inode, may_flags, nf, status),
 
 	TP_STRUCT__entry(
-		__field(__be32, xid)
+		__field(u32, xid)
 		__field(unsigned int, hash)
 		__field(void *, inode)
 		__field(unsigned int, may_flags)
@@ -236,11 +236,11 @@ TRACE_EVENT(nfsd_file_acquire,
 		__field(unsigned long, nf_flags)
 		__field(unsigned char, nf_may)
 		__field(struct file *, nf_file)
-		__field(__be32, status)
+		__field(u32, status)
 	),
 
 	TP_fast_assign(
-		__entry->xid = rqstp->rq_xid;
+		__entry->xid = be32_to_cpu(rqstp->rq_xid);
 		__entry->hash = hash;
 		__entry->inode = inode;
 		__entry->may_flags = may_flags;
@@ -248,15 +248,15 @@ TRACE_EVENT(nfsd_file_acquire,
 		__entry->nf_flags = nf ? nf->nf_flags : 0;
 		__entry->nf_may = nf ? nf->nf_may : 0;
 		__entry->nf_file = nf ? nf->nf_file : NULL;
-		__entry->status = status;
+		__entry->status = be32_to_cpu(status);
 	),
 
 	TP_printk("xid=0x%x hash=0x%x inode=0x%p may_flags=%s ref=%d nf_flags=%s nf_may=%s nf_file=0x%p status=%u",
-			be32_to_cpu(__entry->xid), __entry->hash, __entry->inode,
+			__entry->xid, __entry->hash, __entry->inode,
 			show_nf_may(__entry->may_flags), __entry->nf_ref,
 			show_nf_flags(__entry->nf_flags),
 			show_nf_may(__entry->nf_may), __entry->nf_file,
-			be32_to_cpu(__entry->status))
+			__entry->status)
 );
 
 DECLARE_EVENT_CLASS(nfsd_file_search_class,
-- 
2.25.1

[PATCH AUTOSEL 5.4 177/330] nfsd: Don't add locks to closed or closing open stateids

[Sasha Levin]

From: Trond Myklebust <trondmy@gmail.com>

[ Upstream commit a451b12311aa8c96c6f6e01c783a86995dc3ec6b ]

In NFSv4, the lock stateids are tied to the lockowner, and the open stateid,
so that the action of closing the file also results in either an automatic
loss of the locks, or an error of the form NFS4ERR_LOCKS_HELD.

In practice this means we must not add new locks to the open stateid
after the close process has been invoked. In fact doing so, can result
in the following panic:

 kernel BUG at lib/list_debug.c:51!
 invalid opcode: 0000 [#1] SMP NOPTI
 CPU: 2 PID: 1085 Comm: nfsd Not tainted 5.6.0-rc3+ #2
 Hardware name: VMware, Inc. VMware7,1/440BX Desktop Reference Platform, BIOS VMW71.00V.14410784.B64.1908150010 08/15/2019
 RIP: 0010:__list_del_entry_valid.cold+0x31/0x55
 Code: 1a 3d 9b e8 74 10 c2 ff 0f 0b 48 c7 c7 f0 1a 3d 9b e8 66 10 c2 ff 0f 0b 48 89 f2 48 89 fe 48 c7 c7 b0 1a 3d 9b e8 52 10 c2 ff <0f> 0b 48 89 fe 4c 89 c2 48 c7 c7 78 1a 3d 9b e8 3e 10 c2 ff 0f 0b
 RSP: 0018:ffffb296c1d47d90 EFLAGS: 00010246
 RAX: 0000000000000054 RBX: ffff8ba032456ec8 RCX: 0000000000000000
 RDX: 0000000000000000 RSI: ffff8ba039e99cc8 RDI: ffff8ba039e99cc8
 RBP: ffff8ba032456e60 R08: 0000000000000781 R09: 0000000000000003
 R10: 0000000000000000 R11: 0000000000000001 R12: ffff8ba009a4abe0
 R13: ffff8ba032456e8c R14: 0000000000000000 R15: ffff8ba00adb01d8
 FS:  0000000000000000(0000) GS:ffff8ba039e80000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
 CR2: 00007fb213f0b008 CR3: 00000001347de006 CR4: 00000000003606e0
 Call Trace:
  release_lock_stateid+0x2b/0x80 [nfsd]
  nfsd4_free_stateid+0x1e9/0x210 [nfsd]
  nfsd4_proc_compound+0x414/0x700 [nfsd]
  ? nfs4svc_decode_compoundargs+0x407/0x4c0 [nfsd]
  nfsd_dispatch+0xc1/0x200 [nfsd]
  svc_process_common+0x476/0x6f0 [sunrpc]
  ? svc_sock_secure_port+0x12/0x30 [sunrpc]
  ? svc_recv+0x313/0x9c0 [sunrpc]
  ? nfsd_svc+0x2d0/0x2d0 [nfsd]
  svc_process+0xd4/0x110 [sunrpc]
  nfsd+0xe3/0x140 [nfsd]
  kthread+0xf9/0x130
  ? nfsd_destroy+0x50/0x50 [nfsd]
  ? kthread_park+0x90/0x90
  ret_from_fork+0x1f/0x40

The fix is to ensure that lock creation tests for whether or not the
open stateid is unhashed, and to fail if that is the case.

Fixes: 659aefb68eca ("nfsd: Ensure we don't recognise lock stateids after freeing them")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/nfsd/nfs4state.c | 73 ++++++++++++++++++++++++++-------------------
 1 file changed, 43 insertions(+), 30 deletions(-)

diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 68cf116607645..8cb2f744dde6b 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -495,6 +495,8 @@ find_any_file(struct nfs4_file *f)
 {
 	struct nfsd_file *ret;
 
+	if (!f)
+		return NULL;
 	spin_lock(&f->fi_lock);
 	ret = __nfs4_get_fd(f, O_RDWR);
 	if (!ret) {
@@ -1273,6 +1275,12 @@ static void nfs4_put_stateowner(struct nfs4_stateowner *sop)
 	nfs4_free_stateowner(sop);
 }
 
+static bool
+nfs4_ol_stateid_unhashed(const struct nfs4_ol_stateid *stp)
+{
+	return list_empty(&stp->st_perfile);
+}
+
 static bool unhash_ol_stateid(struct nfs4_ol_stateid *stp)
 {
 	struct nfs4_file *fp = stp->st_stid.sc_file;
@@ -1343,9 +1351,11 @@ static bool unhash_lock_stateid(struct nfs4_ol_stateid *stp)
 {
 	lockdep_assert_held(&stp->st_stid.sc_client->cl_lock);
 
+	if (!unhash_ol_stateid(stp))
+		return false;
 	list_del_init(&stp->st_locks);
 	nfs4_unhash_stid(&stp->st_stid);
-	return unhash_ol_stateid(stp);
+	return true;
 }
 
 static void release_lock_stateid(struct nfs4_ol_stateid *stp)
@@ -1410,13 +1420,12 @@ static void release_open_stateid_locks(struct nfs4_ol_stateid *open_stp,
 static bool unhash_open_stateid(struct nfs4_ol_stateid *stp,
 				struct list_head *reaplist)
 {
-	bool unhashed;
-
 	lockdep_assert_held(&stp->st_stid.sc_client->cl_lock);
 
-	unhashed = unhash_ol_stateid(stp);
+	if (!unhash_ol_stateid(stp))
+		return false;
 	release_open_stateid_locks(stp, reaplist);
-	return unhashed;
+	return true;
 }
 
 static void release_open_stateid(struct nfs4_ol_stateid *stp)
@@ -6267,21 +6276,21 @@ alloc_init_lock_stateowner(unsigned int strhashval, struct nfs4_client *clp,
 }
 
 static struct nfs4_ol_stateid *
-find_lock_stateid(struct nfs4_lockowner *lo, struct nfs4_file *fp)
+find_lock_stateid(const struct nfs4_lockowner *lo,
+		  const struct nfs4_ol_stateid *ost)
 {
 	struct nfs4_ol_stateid *lst;
-	struct nfs4_client *clp = lo->lo_owner.so_client;
 
-	lockdep_assert_held(&clp->cl_lock);
+	lockdep_assert_held(&ost->st_stid.sc_client->cl_lock);
 
-	list_for_each_entry(lst, &lo->lo_owner.so_stateids, st_perstateowner) {
-		if (lst->st_stid.sc_type != NFS4_LOCK_STID)
-			continue;
-		if (lst->st_stid.sc_file == fp) {
-			refcount_inc(&lst->st_stid.sc_count);
-			return lst;
+	/* If ost is not hashed, ost->st_locks will not be valid */
+	if (!nfs4_ol_stateid_unhashed(ost))
+		list_for_each_entry(lst, &ost->st_locks, st_locks) {
+			if (lst->st_stateowner == &lo->lo_owner) {
+				refcount_inc(&lst->st_stid.sc_count);
+				return lst;
+			}
 		}
-	}
 	return NULL;
 }
 
@@ -6297,11 +6306,11 @@ init_lock_stateid(struct nfs4_ol_stateid *stp, struct nfs4_lockowner *lo,
 	mutex_lock_nested(&stp->st_mutex, OPEN_STATEID_MUTEX);
 retry:
 	spin_lock(&clp->cl_lock);
-	spin_lock(&fp->fi_lock);
-	retstp = find_lock_stateid(lo, fp);
+	if (nfs4_ol_stateid_unhashed(open_stp))
+		goto out_close;
+	retstp = find_lock_stateid(lo, open_stp);
 	if (retstp)
-		goto out_unlock;
-
+		goto out_found;
 	refcount_inc(&stp->st_stid.sc_count);
 	stp->st_stid.sc_type = NFS4_LOCK_STID;
 	stp->st_stateowner = nfs4_get_stateowner(&lo->lo_owner);
@@ -6310,22 +6319,26 @@ retry:
 	stp->st_access_bmap = 0;
 	stp->st_deny_bmap = open_stp->st_deny_bmap;
 	stp->st_openstp = open_stp;
+	spin_lock(&fp->fi_lock);
 	list_add(&stp->st_locks, &open_stp->st_locks);
 	list_add(&stp->st_perstateowner, &lo->lo_owner.so_stateids);
 	list_add(&stp->st_perfile, &fp->fi_stateids);
-out_unlock:
 	spin_unlock(&fp->fi_lock);
 	spin_unlock(&clp->cl_lock);
-	if (retstp) {
-		if (nfsd4_lock_ol_stateid(retstp) != nfs_ok) {
-			nfs4_put_stid(&retstp->st_stid);
-			goto retry;
-		}
-		/* To keep mutex tracking happy */
-		mutex_unlock(&stp->st_mutex);
-		stp = retstp;
-	}
 	return stp;
+out_found:
+	spin_unlock(&clp->cl_lock);
+	if (nfsd4_lock_ol_stateid(retstp) != nfs_ok) {
+		nfs4_put_stid(&retstp->st_stid);
+		goto retry;
+	}
+	/* To keep mutex tracking happy */
+	mutex_unlock(&stp->st_mutex);
+	return retstp;
+out_close:
+	spin_unlock(&clp->cl_lock);
+	mutex_unlock(&stp->st_mutex);
+	return NULL;
 }
 
 static struct nfs4_ol_stateid *
@@ -6340,7 +6353,7 @@ find_or_create_lock_stateid(struct nfs4_lockowner *lo, struct nfs4_file *fi,
 
 	*new = false;
 	spin_lock(&clp->cl_lock);
-	lst = find_lock_stateid(lo, fi);
+	lst = find_lock_stateid(lo, ost);
 	spin_unlock(&clp->cl_lock);
 	if (lst != NULL) {
 		if (nfsd4_lock_ol_stateid(lst) == nfs_ok)
-- 
2.25.1

[PATCH AUTOSEL 5.4 213/330] SUNRPC: Fix a potential buffer overflow in 'svc_print_xprts()'

[Sasha Levin]

From: Christophe JAILLET <christophe.jaillet@wanadoo.fr>

[ Upstream commit b25b60d7bfb02a74bc3c2d998e09aab159df8059 ]

'maxlen' is the total size of the destination buffer. There is only one
caller and this value is 256.

When we compute the size already used and what we would like to add in
the buffer, the trailling NULL character is not taken into account.
However, this trailling character will be added by the 'strcat' once we
have checked that we have enough place.

So, there is a off-by-one issue and 1 byte of the stack could be
erroneously overwridden.

Take into account the trailling NULL, when checking if there is enough
place in the destination buffer.

While at it, also replace a 'sprintf' by a safer 'snprintf', check for
output truncation and avoid a superfluous 'strlen'.

Fixes: dc9a16e49dbba ("svc: Add /proc/sys/sunrpc/transport files")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
[ cel: very minor fix to documenting comment
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/sunrpc/svc_xprt.c | 19 ++++++++++++++-----
 1 file changed, 14 insertions(+), 5 deletions(-)

diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c
index dc74519286be5..fe4cd0b4c4127 100644
--- a/net/sunrpc/svc_xprt.c
+++ b/net/sunrpc/svc_xprt.c
@@ -104,8 +104,17 @@ void svc_unreg_xprt_class(struct svc_xprt_class *xcl)
 }
 EXPORT_SYMBOL_GPL(svc_unreg_xprt_class);
 
-/*
- * Format the transport list for printing
+/**
+ * svc_print_xprts - Format the transport list for printing
+ * @buf: target buffer for formatted address
+ * @maxlen: length of target buffer
+ *
+ * Fills in @buf with a string containing a list of transport names, each name
+ * terminated with '\n'. If the buffer is too small, some entries may be
+ * missing, but it is guaranteed that all lines in the output buffer are
+ * complete.
+ *
+ * Returns positive length of the filled-in string.
  */
 int svc_print_xprts(char *buf, int maxlen)
 {
@@ -118,9 +127,9 @@ int svc_print_xprts(char *buf, int maxlen)
 	list_for_each_entry(xcl, &svc_xprt_class_list, xcl_list) {
 		int slen;
 
-		sprintf(tmpstr, "%s %d\n", xcl->xcl_name, xcl->xcl_max_payload);
-		slen = strlen(tmpstr);
-		if (len + slen > maxlen)
+		slen = snprintf(tmpstr, sizeof(tmpstr), "%s %d\n",
+				xcl->xcl_name, xcl->xcl_max_payload);
+		if (slen >= sizeof(tmpstr) || len + slen >= maxlen)
 			break;
 		len += slen;
 		strcat(buf, tmpstr);
-- 
2.25.1

[PATCH AUTOSEL 5.4 214/330] svcrdma: Fix leak of transport addresses

[Sasha Levin]

From: Chuck Lever <chuck.lever@oracle.com>

[ Upstream commit 1a33d8a284b1e85e03b8c7b1ea8fb985fccd1d71 ]

Kernel memory leak detected:

unreferenced object 0xffff888849cdf480 (size 8):
  comm "kworker/u8:3", pid 2086, jiffies 4297898756 (age 4269.856s)
  hex dump (first 8 bytes):
    30 00 cd 49 88 88 ff ff                          0..I....
  backtrace:
    [<00000000acfc370b>] __kmalloc_track_caller+0x137/0x183
    [<00000000a2724354>] kstrdup+0x2b/0x43
    [<0000000082964f84>] xprt_rdma_format_addresses+0x114/0x17d [rpcrdma]
    [<00000000dfa6ed00>] xprt_setup_rdma_bc+0xc0/0x10c [rpcrdma]
    [<0000000073051a83>] xprt_create_transport+0x3f/0x1a0 [sunrpc]
    [<0000000053531a8e>] rpc_create+0x118/0x1cd [sunrpc]
    [<000000003a51b5f8>] setup_callback_client+0x1a5/0x27d [nfsd]
    [<000000001bd410af>] nfsd4_process_cb_update.isra.7+0x16c/0x1ac [nfsd]
    [<000000007f4bbd56>] nfsd4_run_cb_work+0x4c/0xbd [nfsd]
    [<0000000055c5586b>] process_one_work+0x1b2/0x2fe
    [<00000000b1e3e8ef>] worker_thread+0x1a6/0x25a
    [<000000005205fb78>] kthread+0xf6/0xfb
    [<000000006d2dc057>] ret_from_fork+0x3a/0x50

Introduce a call to xprt_rdma_free_addresses() similar to the way
that the TCP backchannel releases a transport's peer address
strings.

Fixes: 5d252f90a800 ("svcrdma: Add class for RDMA backwards direction transport")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/sunrpc/xprtrdma/svc_rdma_backchannel.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
index cf80394b2db33..325eef1f85824 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
@@ -252,6 +252,7 @@ xprt_rdma_bc_put(struct rpc_xprt *xprt)
 {
 	dprintk("svcrdma: %s: xprt %p\n", __func__, xprt);
 
+	xprt_rdma_free_addresses(xprt);
 	xprt_free(xprt);
 }
 
-- 
2.25.1

[PATCH AUTOSEL 5.4 222/330] NFS: Fix races nfs_page_group_destroy() vs nfs_destroy_unlinked_subrequests()

[Sasha Levin]

From: Trond Myklebust <trond.myklebust@hammerspace.com>

[ Upstream commit 08ca8b21f760c0ed5034a5c122092eec22ccf8f4 ]

When a subrequest is being detached from the subgroup, we want to
ensure that it is not holding the group lock, or in the process
of waiting for the group lock.

Fixes: 5b2b5187fa85 ("NFS: Fix nfs_page_group_destroy() and nfs_lock_and_join_requests() race cases")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/nfs/pagelist.c        | 67 +++++++++++++++++++++++++++-------------
 fs/nfs/write.c           | 10 ++++--
 include/linux/nfs_page.h |  2 ++
 3 files changed, 55 insertions(+), 24 deletions(-)

diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c
index b736912098eee..f4407dd426bf0 100644
--- a/fs/nfs/pagelist.c
+++ b/fs/nfs/pagelist.c
@@ -133,47 +133,70 @@ nfs_async_iocounter_wait(struct rpc_task *task, struct nfs_lock_context *l_ctx)
 EXPORT_SYMBOL_GPL(nfs_async_iocounter_wait);
 
 /*
- * nfs_page_group_lock - lock the head of the page group
- * @req - request in group that is to be locked
+ * nfs_page_set_headlock - set the request PG_HEADLOCK
+ * @req: request that is to be locked
  *
- * this lock must be held when traversing or modifying the page
- * group list
+ * this lock must be held when modifying req->wb_head
  *
  * return 0 on success, < 0 on error
  */
 int
-nfs_page_group_lock(struct nfs_page *req)
+nfs_page_set_headlock(struct nfs_page *req)
 {
-	struct nfs_page *head = req->wb_head;
-
-	WARN_ON_ONCE(head != head->wb_head);
-
-	if (!test_and_set_bit(PG_HEADLOCK, &head->wb_flags))
+	if (!test_and_set_bit(PG_HEADLOCK, &req->wb_flags))
 		return 0;
 
-	set_bit(PG_CONTENDED1, &head->wb_flags);
+	set_bit(PG_CONTENDED1, &req->wb_flags);
 	smp_mb__after_atomic();
-	return wait_on_bit_lock(&head->wb_flags, PG_HEADLOCK,
+	return wait_on_bit_lock(&req->wb_flags, PG_HEADLOCK,
 				TASK_UNINTERRUPTIBLE);
 }
 
 /*
- * nfs_page_group_unlock - unlock the head of the page group
- * @req - request in group that is to be unlocked
+ * nfs_page_clear_headlock - clear the request PG_HEADLOCK
+ * @req: request that is to be locked
  */
 void
-nfs_page_group_unlock(struct nfs_page *req)
+nfs_page_clear_headlock(struct nfs_page *req)
 {
-	struct nfs_page *head = req->wb_head;
-
-	WARN_ON_ONCE(head != head->wb_head);
-
 	smp_mb__before_atomic();
-	clear_bit(PG_HEADLOCK, &head->wb_flags);
+	clear_bit(PG_HEADLOCK, &req->wb_flags);
 	smp_mb__after_atomic();
-	if (!test_bit(PG_CONTENDED1, &head->wb_flags))
+	if (!test_bit(PG_CONTENDED1, &req->wb_flags))
 		return;
-	wake_up_bit(&head->wb_flags, PG_HEADLOCK);
+	wake_up_bit(&req->wb_flags, PG_HEADLOCK);
+}
+
+/*
+ * nfs_page_group_lock - lock the head of the page group
+ * @req: request in group that is to be locked
+ *
+ * this lock must be held when traversing or modifying the page
+ * group list
+ *
+ * return 0 on success, < 0 on error
+ */
+int
+nfs_page_group_lock(struct nfs_page *req)
+{
+	int ret;
+
+	ret = nfs_page_set_headlock(req);
+	if (ret || req->wb_head == req)
+		return ret;
+	return nfs_page_set_headlock(req->wb_head);
+}
+
+/*
+ * nfs_page_group_unlock - unlock the head of the page group
+ * @req: request in group that is to be unlocked
+ */
+void
+nfs_page_group_unlock(struct nfs_page *req)
+{
+	if (req != req->wb_head)
+		nfs_page_clear_headlock(req->wb_head);
+	nfs_page_clear_headlock(req);
 }
 
 /*
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 58c8317dd7d88..613c3ef23e07b 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -425,22 +425,28 @@ nfs_destroy_unlinked_subrequests(struct nfs_page *destroy_list,
 		destroy_list = (subreq->wb_this_page == old_head) ?
 				   NULL : subreq->wb_this_page;
 
+		/* Note: lock subreq in order to change subreq->wb_head */
+		nfs_page_set_headlock(subreq);
 		WARN_ON_ONCE(old_head != subreq->wb_head);
 
 		/* make sure old group is not used */
 		subreq->wb_this_page = subreq;
+		subreq->wb_head = subreq;
 
 		clear_bit(PG_REMOVE, &subreq->wb_flags);
 
 		/* Note: races with nfs_page_group_destroy() */
 		if (!kref_read(&subreq->wb_kref)) {
 			/* Check if we raced with nfs_page_group_destroy() */
-			if (test_and_clear_bit(PG_TEARDOWN, &subreq->wb_flags))
+			if (test_and_clear_bit(PG_TEARDOWN, &subreq->wb_flags)) {
+				nfs_page_clear_headlock(subreq);
 				nfs_free_request(subreq);
+			} else
+				nfs_page_clear_headlock(subreq);
 			continue;
 		}
+		nfs_page_clear_headlock(subreq);
 
-		subreq->wb_head = subreq;
 		nfs_release_request(old_head);
 
 		if (test_and_clear_bit(PG_INODE_REF, &subreq->wb_flags)) {
diff --git a/include/linux/nfs_page.h b/include/linux/nfs_page.h
index 0bbd587fac6a9..7e9419d74b86b 100644
--- a/include/linux/nfs_page.h
+++ b/include/linux/nfs_page.h
@@ -142,6 +142,8 @@ extern	void nfs_unlock_and_release_request(struct nfs_page *);
 extern int nfs_page_group_lock(struct nfs_page *);
 extern void nfs_page_group_unlock(struct nfs_page *);
 extern bool nfs_page_group_sync_on_bit(struct nfs_page *, unsigned int);
+extern	int nfs_page_set_headlock(struct nfs_page *req);
+extern void nfs_page_clear_headlock(struct nfs_page *req);
 extern bool nfs_async_iocounter_wait(struct rpc_task *, struct nfs_lock_context *);
 
 /*
-- 
2.25.1

[PATCH AUTOSEL 5.4 229/330] SUNRPC: Don't start a timer on an already queued rpc task

[Sasha Levin]

From: Trond Myklebust <trond.myklebust@hammerspace.com>

[ Upstream commit 1fab7dc477241c12f977955aa6baea7938b6f08d ]

Move the test for whether a task is already queued to prevent
corruption of the timer list in __rpc_sleep_on_priority_timeout().

Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/sunrpc/sched.c | 19 +++++++++++++------
 1 file changed, 13 insertions(+), 6 deletions(-)

diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c
index 9c79548c68474..53d8b82eda006 100644
--- a/net/sunrpc/sched.c
+++ b/net/sunrpc/sched.c
@@ -204,10 +204,6 @@ static void __rpc_add_wait_queue(struct rpc_wait_queue *queue,
 		struct rpc_task *task,
 		unsigned char queue_priority)
 {
-	WARN_ON_ONCE(RPC_IS_QUEUED(task));
-	if (RPC_IS_QUEUED(task))
-		return;
-
 	INIT_LIST_HEAD(&task->u.tk_wait.timer_list);
 	if (RPC_IS_PRIORITY(queue))
 		__rpc_add_wait_queue_priority(queue, task, queue_priority);
@@ -382,7 +378,7 @@ static void rpc_make_runnable(struct workqueue_struct *wq,
  * NB: An RPC task will only receive interrupt-driven events as long
  * as it's on a wait queue.
  */
-static void __rpc_sleep_on_priority(struct rpc_wait_queue *q,
+static void __rpc_do_sleep_on_priority(struct rpc_wait_queue *q,
 		struct rpc_task *task,
 		unsigned char queue_priority)
 {
@@ -395,12 +391,23 @@ static void __rpc_sleep_on_priority(struct rpc_wait_queue *q,
 
 }
 
+static void __rpc_sleep_on_priority(struct rpc_wait_queue *q,
+		struct rpc_task *task,
+		unsigned char queue_priority)
+{
+	if (WARN_ON_ONCE(RPC_IS_QUEUED(task)))
+		return;
+	__rpc_do_sleep_on_priority(q, task, queue_priority);
+}
+
 static void __rpc_sleep_on_priority_timeout(struct rpc_wait_queue *q,
 		struct rpc_task *task, unsigned long timeout,
 		unsigned char queue_priority)
 {
+	if (WARN_ON_ONCE(RPC_IS_QUEUED(task)))
+		return;
 	if (time_is_after_jiffies(timeout)) {
-		__rpc_sleep_on_priority(q, task, queue_priority);
+		__rpc_do_sleep_on_priority(q, task, queue_priority);
 		__rpc_add_timer(q, task, timeout);
 	} else
 		task->tk_status = -ETIMEDOUT;
-- 
2.25.1

[PATCH AUTOSEL 5.4 271/330] svcrdma: Fix backchannel return code

[Sasha Levin]

From: Chuck Lever <chuck.lever@oracle.com>

[ Upstream commit ea740bd5f58e2912e74f401fd01a9d6aa985ca05 ]

Way back when I was writing the RPC/RDMA server-side backchannel
code, I misread the TCP backchannel reply handler logic. When
svc_tcp_recvfrom() successfully receives a backchannel reply, it
does not return -EAGAIN. It sets XPT_DATA and returns zero.

Update svc_rdma_recvfrom() to return zero. Here, XPT_DATA doesn't
need to be set again: it is set whenever a new message is received,
behind a spin lock in a single threaded context.

Also, if handling the cb reply is not successful, the message is
simply dropped. There's no special message framing to deal with as
there is in the TCP case.

Now that the handle_bc_reply() return value is ignored, I've removed
the dprintk call sites in the error exit of handle_bc_reply() in
favor of trace points in other areas that already report the error
cases.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 include/linux/sunrpc/svc_rdma.h            |  5 ++-
 net/sunrpc/xprtrdma/svc_rdma_backchannel.c | 38 ++++++----------------
 net/sunrpc/xprtrdma/svc_rdma_recvfrom.c    | 11 +++----
 3 files changed, 17 insertions(+), 37 deletions(-)

diff --git a/include/linux/sunrpc/svc_rdma.h b/include/linux/sunrpc/svc_rdma.h
index 26f282e5e0822..77589ed787f5c 100644
--- a/include/linux/sunrpc/svc_rdma.h
+++ b/include/linux/sunrpc/svc_rdma.h
@@ -154,9 +154,8 @@ struct svc_rdma_send_ctxt {
 };
 
 /* svc_rdma_backchannel.c */
-extern int svc_rdma_handle_bc_reply(struct rpc_xprt *xprt,
-				    __be32 *rdma_resp,
-				    struct xdr_buf *rcvbuf);
+extern void svc_rdma_handle_bc_reply(struct svc_rqst *rqstp,
+				     struct svc_rdma_recv_ctxt *rctxt);
 
 /* svc_rdma_recvfrom.c */
 extern void svc_rdma_recv_ctxts_destroy(struct svcxprt_rdma *rdma);
diff --git a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
index 325eef1f85824..68d2dcf0a1be1 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
@@ -15,26 +15,25 @@
 #undef SVCRDMA_BACKCHANNEL_DEBUG
 
 /**
- * svc_rdma_handle_bc_reply - Process incoming backchannel reply
- * @xprt: controlling backchannel transport
- * @rdma_resp: pointer to incoming transport header
- * @rcvbuf: XDR buffer into which to decode the reply
+ * svc_rdma_handle_bc_reply - Process incoming backchannel Reply
+ * @rqstp: resources for handling the Reply
+ * @rctxt: Received message
  *
- * Returns:
- *	%0 if @rcvbuf is filled in, xprt_complete_rqst called,
- *	%-EAGAIN if server should call ->recvfrom again.
  */
-int svc_rdma_handle_bc_reply(struct rpc_xprt *xprt, __be32 *rdma_resp,
-			     struct xdr_buf *rcvbuf)
+void svc_rdma_handle_bc_reply(struct svc_rqst *rqstp,
+			      struct svc_rdma_recv_ctxt *rctxt)
 {
+	struct svc_xprt *sxprt = rqstp->rq_xprt;
+	struct rpc_xprt *xprt = sxprt->xpt_bc_xprt;
 	struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
+	struct xdr_buf *rcvbuf = &rqstp->rq_arg;
 	struct kvec *dst, *src = &rcvbuf->head[0];
+	__be32 *rdma_resp = rctxt->rc_recv_buf;
 	struct rpc_rqst *req;
 	u32 credits;
 	size_t len;
 	__be32 xid;
 	__be32 *p;
-	int ret;
 
 	p = (__be32 *)src->iov_base;
 	len = src->iov_len;
@@ -49,14 +48,10 @@ int svc_rdma_handle_bc_reply(struct rpc_xprt *xprt, __be32 *rdma_resp,
 		__func__, (int)len, p);
 #endif
 
-	ret = -EAGAIN;
-	if (src->iov_len < 24)
-		goto out_shortreply;
-
 	spin_lock(&xprt->queue_lock);
 	req = xprt_lookup_rqst(xprt, xid);
 	if (!req)
-		goto out_notfound;
+		goto out_unlock;
 
 	dst = &req->rq_private_buf.head[0];
 	memcpy(&req->rq_private_buf, &req->rq_rcv_buf, sizeof(struct xdr_buf));
@@ -77,25 +72,12 @@ int svc_rdma_handle_bc_reply(struct rpc_xprt *xprt, __be32 *rdma_resp,
 	spin_unlock(&xprt->transport_lock);
 
 	spin_lock(&xprt->queue_lock);
-	ret = 0;
 	xprt_complete_rqst(req->rq_task, rcvbuf->len);
 	xprt_unpin_rqst(req);
 	rcvbuf->len = 0;
 
 out_unlock:
 	spin_unlock(&xprt->queue_lock);
-out:
-	return ret;
-
-out_shortreply:
-	dprintk("svcrdma: short bc reply: xprt=%p, len=%zu\n",
-		xprt, src->iov_len);
-	goto out;
-
-out_notfound:
-	dprintk("svcrdma: unrecognized bc reply: xprt=%p, xid=%08x\n",
-		xprt, be32_to_cpu(xid));
-	goto out_unlock;
 }
 
 /* Send a backwards direction RPC call.
diff --git a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
index d803d814a03ad..fd5c1f1bb9885 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
@@ -817,12 +817,9 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp)
 		goto out_drop;
 	rqstp->rq_xprt_hlen = ret;
 
-	if (svc_rdma_is_backchannel_reply(xprt, p)) {
-		ret = svc_rdma_handle_bc_reply(xprt->xpt_bc_xprt, p,
-					       &rqstp->rq_arg);
-		svc_rdma_recv_ctxt_put(rdma_xprt, ctxt);
-		return ret;
-	}
+	if (svc_rdma_is_backchannel_reply(xprt, p))
+		goto out_backchannel;
+
 	svc_rdma_get_inv_rkey(rdma_xprt, ctxt);
 
 	p += rpcrdma_fixed_maxsz;
@@ -852,6 +849,8 @@ out_postfail:
 	svc_rdma_recv_ctxt_put(rdma_xprt, ctxt);
 	return ret;
 
+out_backchannel:
+	svc_rdma_handle_bc_reply(rqstp, ctxt);
 out_drop:
 	svc_rdma_recv_ctxt_put(rdma_xprt, ctxt);
 	return 0;
-- 
2.25.1

[PATCH AUTOSEL 5.4 311/330] NFS: nfs_xdr_status should record the procedure name

[Sasha Levin]

From: Chuck Lever <chuck.lever@oracle.com>

[ Upstream commit 5be5945864ea143fda628e8179c8474457af1f43 ]

When sunrpc trace points are not enabled, the recorded task ID
information alone is not helpful.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/nfs/nfstrace.h | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/nfstrace.h b/fs/nfs/nfstrace.h
index 361cc10d6f95d..c8081d2b4166a 100644
--- a/fs/nfs/nfstrace.h
+++ b/fs/nfs/nfstrace.h
@@ -1147,7 +1147,12 @@ TRACE_EVENT(nfs_xdr_status,
 			__field(unsigned int, task_id)
 			__field(unsigned int, client_id)
 			__field(u32, xid)
+			__field(int, version)
 			__field(unsigned long, error)
+			__string(program,
+				 xdr->rqst->rq_task->tk_client->cl_program->name)
+			__string(procedure,
+				 xdr->rqst->rq_task->tk_msg.rpc_proc->p_name)
 		),
 
 		TP_fast_assign(
@@ -1157,13 +1162,19 @@ TRACE_EVENT(nfs_xdr_status,
 			__entry->task_id = task->tk_pid;
 			__entry->client_id = task->tk_client->cl_clid;
 			__entry->xid = be32_to_cpu(rqstp->rq_xid);
+			__entry->version = task->tk_client->cl_vers;
 			__entry->error = error;
+			__assign_str(program,
+				     task->tk_client->cl_program->name)
+			__assign_str(procedure, task->tk_msg.rpc_proc->p_name)
 		),
 
 		TP_printk(
-			"task:%u@%d xid=0x%08x error=%ld (%s)",
+			"task:%u@%d xid=0x%08x %sv%d %s error=%ld (%s)",
 			__entry->task_id, __entry->client_id, __entry->xid,
-			-__entry->error, nfs_show_status(__entry->error)
+			__get_str(program), __entry->version,
+			__get_str(procedure), -__entry->error,
+			nfs_show_status(__entry->error)
 		)
 );
 
-- 
2.25.1