[PATCH AUTOSEL 5.8 020/132] NFS4: Fix oops when copy_file_range is attempted with NFS4.0 source

[PATCH AUTOSEL 5.8 020/132] NFS4: Fix oops when copy_file_range is attempted with NFS4.0 source

[Sasha Levin]

From: Dave Wysochanski <dwysocha@redhat.com>

[ Upstream commit d8a6ad913c286d4763ae20b14c02fe6f39d7cd9f ]

The following oops is seen during xfstest/565 when the 'test'
(source of the copy) is NFS4.0 and 'scratch' (destination) is NFS4.2
[   59.692458] run fstests generic/565 at 2020-08-01 05:50:35
[   60.613588] BUG: kernel NULL pointer dereference, address: 0000000000000008
[   60.624970] #PF: supervisor read access in kernel mode
[   60.627671] #PF: error_code(0x0000) - not-present page
[   60.630347] PGD 0 P4D 0
[   60.631853] Oops: 0000 [#1] SMP PTI
[   60.634086] CPU: 6 PID: 2828 Comm: xfs_io Kdump: loaded Not tainted 5.8.0-rc3 #1
[   60.637676] Hardware name: Red Hat KVM, BIOS 0.5.1 01/01/2011
[   60.639901] RIP: 0010:nfs4_check_serverowner_major_id+0x5/0x30 [nfsv4]
[   60.642719] Code: 89 ff e8 3e b3 b8 e1 e9 71 fe ff ff 41 bc da d8 ff ff e9 c3 fe ff ff e8 e9 9d 08 e2 66 0f 1f 84 00 00 00 00 00 66 66 66 66 90 <8b> 57 08 31 c0 3b 56 08 75 12 48 83 c6 0c 48 83 c7 0c e8 c4 97 bb
[   60.652629] RSP: 0018:ffffc265417f7e10 EFLAGS: 00010287
[   60.655379] RAX: ffffa0664b066400 RBX: 0000000000000000 RCX: 0000000000000001
[   60.658754] RDX: ffffa066725fb000 RSI: ffffa066725fd000 RDI: 0000000000000000
[   60.662292] RBP: 0000000000020000 R08: 0000000000020000 R09: 0000000000000000
[   60.666189] R10: 0000000000000003 R11: 0000000000000000 R12: ffffa06648258d00
[   60.669914] R13: 0000000000000000 R14: 0000000000000000 R15: ffffa06648258100
[   60.673645] FS:  00007faa9fb35800(0000) GS:ffffa06677d80000(0000) knlGS:0000000000000000
[   60.677698] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   60.680773] CR2: 0000000000000008 CR3: 0000000203f14000 CR4: 00000000000406e0
[   60.684476] Call Trace:
[   60.685809]  nfs4_copy_file_range+0xfc/0x230 [nfsv4]
[   60.688704]  vfs_copy_file_range+0x2ee/0x310
[   60.691104]  __x64_sys_copy_file_range+0xd6/0x210
[   60.693527]  do_syscall_64+0x4d/0x90
[   60.695512]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
[   60.698006] RIP: 0033:0x7faa9febc1bd

Signed-off-by: Dave Wysochanski <dwysocha@redhat.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/nfs/nfs4file.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/nfs/nfs4file.c b/fs/nfs/nfs4file.c
index a339707654673..af84787aa0631 100644
--- a/fs/nfs/nfs4file.c
+++ b/fs/nfs/nfs4file.c
@@ -145,7 +145,8 @@ static ssize_t __nfs4_copy_file_range(struct file *file_in, loff_t pos_in,
 	/* Only offload copy if superblock is the same */
 	if (file_in->f_op != &nfs4_file_operations)
 		return -EXDEV;
-	if (!nfs_server_capable(file_inode(file_out), NFS_CAP_COPY))
+	if (!nfs_server_capable(file_inode(file_out), NFS_CAP_COPY) ||
+	    !nfs_server_capable(file_inode(file_in), NFS_CAP_COPY))
 		return -EOPNOTSUPP;
 	if (file_inode(file_in) == file_inode(file_out))
 		return -EOPNOTSUPP;
-- 
2.25.1

[PATCH AUTOSEL 5.8 071/132] SUNRPC: Mitigate cond_resched() in xprt_transmit()

[Sasha Levin]

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

[ Upstream commit 6f9f17287e78e5049931af2037b15b26d134a32a ]

The original purpose of this expensive call is to prevent a long
queue of requests from blocking other work.

The cond_resched() call is unnecessary after just a single send
operation.

For longer queues, instead of invoking the kernel scheduler, simply
release the transport send lock and return to the RPC scheduler.

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>
---
 net/sunrpc/xprt.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index d5cc5db9dbf39..b44099958c8bb 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -1511,10 +1511,13 @@ xprt_transmit(struct rpc_task *task)
 {
 	struct rpc_rqst *next, *req = task->tk_rqstp;
 	struct rpc_xprt	*xprt = req->rq_xprt;
-	int status;
+	int counter, status;
 
 	spin_lock(&xprt->queue_lock);
+	counter = 0;
 	while (!list_empty(&xprt->xmit_queue)) {
+		if (++counter == 20)
+			break;
 		next = list_first_entry(&xprt->xmit_queue,
 				struct rpc_rqst, rq_xmit);
 		xprt_pin_rqst(next);
@@ -1522,7 +1525,6 @@ xprt_transmit(struct rpc_task *task)
 		status = xprt_request_transmit(next, task);
 		if (status == -EBADMSG && next != req)
 			status = 0;
-		cond_resched();
 		spin_lock(&xprt->queue_lock);
 		xprt_unpin_rqst(next);
 		if (status == 0) {
-- 
2.25.1

[PATCH AUTOSEL 5.8 104/132] nfsd: rename delegation related tracepoints to make them less confusing

[Sasha Levin]

From: Hou Tao <houtao1@huawei.com>

[ Upstream commit 3caf91757ced158e6c4a44d8b105bd7b3e1767d8 ]

Now when a read delegation is given, two delegation related traces
will be printed:

    nfsd_deleg_open: client 5f45b854:e6058001 stateid 00000030:00000001
    nfsd_deleg_none: client 5f45b854:e6058001 stateid 0000002f:00000001

Although the intention is to let developers know two stateid are
returned, the traces are confusing about whether or not a read delegation
is handled out. So renaming trace_nfsd_deleg_none() to trace_nfsd_open()
and trace_nfsd_deleg_open() to trace_nfsd_deleg_read() to make
the intension clearer.

The patched traces will be:

    nfsd_deleg_read: client 5f48a967:b55b21cd stateid 00000003:00000001
    nfsd_open: client 5f48a967:b55b21cd stateid 00000002:00000001

Suggested-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Hou Tao <houtao1@huawei.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/nfsd/nfs4state.c | 4 ++--
 fs/nfsd/trace.h     | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index cea682ce8aa12..2d568d0e55f78 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -5100,7 +5100,7 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open,
 
 	memcpy(&open->op_delegate_stateid, &dp->dl_stid.sc_stateid, sizeof(dp->dl_stid.sc_stateid));
 
-	trace_nfsd_deleg_open(&dp->dl_stid.sc_stateid);
+	trace_nfsd_deleg_read(&dp->dl_stid.sc_stateid);
 	open->op_delegate_type = NFS4_OPEN_DELEGATE_READ;
 	nfs4_put_stid(&dp->dl_stid);
 	return;
@@ -5217,7 +5217,7 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf
 	nfs4_open_delegation(current_fh, open, stp);
 nodeleg:
 	status = nfs_ok;
-	trace_nfsd_deleg_none(&stp->st_stid.sc_stateid);
+	trace_nfsd_open(&stp->st_stid.sc_stateid);
 out:
 	/* 4.1 client trying to upgrade/downgrade delegation? */
 	if (open->op_delegate_type == NFS4_OPEN_DELEGATE_NONE && dp &&
diff --git a/fs/nfsd/trace.h b/fs/nfsd/trace.h
index 1861db1bdc670..99bf07800cd09 100644
--- a/fs/nfsd/trace.h
+++ b/fs/nfsd/trace.h
@@ -289,8 +289,8 @@ DEFINE_STATEID_EVENT(layout_recall_done);
 DEFINE_STATEID_EVENT(layout_recall_fail);
 DEFINE_STATEID_EVENT(layout_recall_release);
 
-DEFINE_STATEID_EVENT(deleg_open);
-DEFINE_STATEID_EVENT(deleg_none);
+DEFINE_STATEID_EVENT(open);
+DEFINE_STATEID_EVENT(deleg_read);
 DEFINE_STATEID_EVENT(deleg_break);
 DEFINE_STATEID_EVENT(deleg_recall);
 
-- 
2.25.1