- 05 Nov, 2021 8 commits
-
-
Anna Schumaker authored
For creating fattrs with the label field already allocated for us. I also update nfs_free_fattr() to free the label in the end. Signed-off-by:
Anna Schumaker <Anna.Schumaker@Netapp.com> Signed-off-by:
Trond Myklebust <trond.myklebust@hammerspace.com>
-
Trond Myklebust authored
We're about to add a check in nfs_free_fattr() for whether or not the label is non-zero. Signed-off-by: -
Trond Myklebust authored
Signed-off-by: -
Trond Myklebust authored
The preferred behaviour is always to allocate struct nfs_fattr from the slab. Signed-off-by: -
Trond Myklebust authored
It is completely redundant to the server capability check. Signed-off-by: -
Trond Myklebust authored
The return value of xdr_inline_decode() is not being checked, leading to a potential Oops. Just replace the open coded array decode with the generic XDR version. Reported-by: <rtm@csail.mit.edu> Signed-off-by: -
Trond Myklebust authored
The server is supposed to return the same tag that the client sends in the outgoing RPC call, but we should still sanity check the length just in case. Reported-by: <rtm@csail.mit.edu> Signed-off-by: -
Trond Myklebust authored
If fhandle is NULL or fattr is NULL, then 'error' is uninitialised. Reported-by:
Nathan Chancellor <nathan@kernel.org> Signed-off-by:
-
- 04 Nov, 2021 10 commits
-
-
Trond Myklebust authored
If we have already set up the socket and are waiting for it to connect, then don't immediately close and retry. Signed-off-by: -
Trond Myklebust authored
Ensure that we bump the xprt->connect_cookie when we set the XPRT_CLOSE_WAIT flag so that another call to xprt_conditional_disconnect() won't race with the reconnection. Signed-off-by: -
Olga Kornievskaia authored
Add tracepoint to OFFLOAD_CANCEL operation. Signed-off-by:
Olga Kornievskaia <kolga@netapp.com> Signed-off-by:
-
Olga Kornievskaia authored
Add a tracepoint to COPY_NOTIFY operation. Signed-off-by:
-
Olga Kornievskaia authored
Add a tracepoint to the CB_OFFLOAD operation. Signed-off-by:
-
Olga Kornievskaia authored
Add a tracepoint to the CLONE operation. Signed-off-by:
-
Olga Kornievskaia authored
Add a tracepoint to the COPY operation. Signed-off-by:
-
Olga Kornievskaia authored
Add a tracepoint to the FALLOCATE/DEALLOCATE operations. Signed-off-by:
-
Olga Kornievskaia authored
Add a tracepoint to the SEEK operation. Signed-off-by:
-
Anna Schumaker authored
xprts don't immediately reconnect when changing the "dstaddr" property, instead this gets handled the next time an operation uses the transport. This could lead to NULL pointer dereferences when trying to read sysfs files between the disconnect and reconnect operations. Fix this by returning an error if the xprt is not connected. Signed-off-by:
-
- 03 Nov, 2021 2 commits
-
-
Jonathan Corbet authored
Commit 19fcae3d ("scsi: remove the SCSI OSD library") deleted the last file that included <linux/pnfs_osd_xdr.h> but left that file behind. It's unused, get rid of it now. Cc: Christoph Hellwig <hch@lst.de> Cc: Trond Myklebust <trond.myklebust@hammerspace.com> Cc: Anna Schumaker <anna.schumaker@netapp.com> Cc: linux-nfs@vger.kernel.org Signed-off-by:
Jonathan Corbet <corbet@lwn.net> Signed-off-by:
-
Scott Mayhew authored
During umount, the session slot tables are freed. If there are outstanding FREE_STATEID tasks, a use-after-free and slab corruption can occur when rpc_exit_task calls rpc_call_done -> nfs41_sequence_done -> nfs4_sequence_process/nfs41_sequence_free_slot. Prevent that from happening by taking a reference on the nfs_client in nfs41_free_stateid and putting it in nfs41_free_stateid_release. Signed-off-by:
Scott Mayhew <smayhew@redhat.com> Signed-off-by:
-
- 02 Nov, 2021 4 commits
-
-
Dave Wysochanski authored
Add the byte offset of the readahead request to the tracepoint output so we know where the read starts. Before this patch: cat-8104 [002] ..... 813.168775: nfs_aop_readahead: fileid=00:31:141 fhandle=0xe55807f6 version=1756509392533525500 nr_pages=256 cat-8104 [002] ..... 813.174973: nfs_aop_readahead_done: fileid=00:31:141 fhandle=0xe55807f6 version=1756509392533525500 nr_pages=256 ret=0 cat-8104 [002] ..... 813.175963: nfs_aop_readahead: fileid=00:31:141 fhandle=0xe55807f6 version=1756509392533525500 nr_pages=256 cat-8104 [002] ..... 813.183742: nfs_aop_readahead_done: fileid=00:31:141 fhandle=0xe55807f6 version=1756509392533525500 nr_pages=1 ret=0 After this patch: cat-6392 [001] ..... 73.107782: nfs_aop_readahead: fileid=00:31:141 fhandle=0xed22403f version=1756511950029502774 offset=5242880 nr_pages=256 cat-6392 [001] ..... 73.112466: nfs_aop_readahead_done: fileid=00:31:141 fhandle=0xed22403f version=1756511950029502774 nr_pages=256 ret=0 cat-6392 [001] ..... 73.115692: nfs_aop_readahead: fileid=00:31:141 fhandle=0xed22403f version=1756511950029502774 offset=6291456 nr_pages=256 cat-6392 [001] ..... 73.123283: nfs_aop_readahead_done: fileid=00:31:141 fhandle=0xed22403f version=1756511950029502774 nr_pages=256 ret=0 Signed-off-by:
Dave Wysochanski <dwysocha@redhat.com> Signed-off-by:
-
Benjamin Coddington authored
This minor fix-up keeps GCC from complaining that "last' may be used uninitialized", which breaks some build workflows that have been running with all warnings treated as errors. Signed-off-by:
Benjamin Coddington <bcodding@redhat.com> Signed-off-by:
-
Chuck Lever authored
Refactor: surface useful show_ macros so they can be shared between the client and server trace code. Additional clean up: - Housekeeping: ensure the correct #include files are pulled in and add proper TRACE_DEFINE_ENUM where they are missing - Use a consistent naming scheme for the helpers - Store values to be displayed symbolically as unsigned long, as that is the type that the __print_yada() functions take Signed-off-by:
Chuck Lever <chuck.lever@oracle.com> Signed-off-by:
-
Chuck Lever authored
Refactor: Surface useful show_ macros for use by other trace subsystems. Signed-off-by:
-
- 01 Nov, 2021 2 commits
-
-
Trond Myklebust authored
Move the error handling into a single switch statement for clarity. Signed-off-by: -
Trond Myklebust authored
Since we do things like setting flags, etc it really is more appropriate to use sock_lock(). Signed-off-by:
-
- 27 Oct, 2021 1 commit
-
-
Trond Myklebust authored
If we already hold open state on the client, yet the server gives us a completely different stateid to the one we already hold, then we currently treat it as if it were an out-of-sequence update, and wait for 5 seconds for other updates to come in. This commit fixes the behaviour so that we immediately start processing of the new stateid, and then leave it to the call to nfs4_test_and_free_stateid() to decide what to do with the old stateid. Fixes: b4868b44 ("NFSv4: Wait for stateid updates after CLOSE/OPEN_DOWNGRADE") Signed-off-by:
-
- 21 Oct, 2021 1 commit
-
-
Trond Myklebust authored
Remove a redundant call in nfs_updatepage(). nfs_writepage_setup() will have already called nfs_mark_request_dirty() on success. Signed-off-by:
-
- 20 Oct, 2021 12 commits
-
-
Thiago Rafael Becker authored
In rpc_task_set_client(), testing for a NULL clnt is not necessary, as clnt should always be a valid pointer to a rpc_client. Signed-off-by:
Thiago Rafael Becker <trbecker@gmail.com> Signed-off-by:
-
Anna Schumaker authored
All the callers are now in client.c so we can remove the EXPORT_SYMBOL_GPL() and make it static. Signed-off-by:
-
Anna Schumaker authored
This lets us update the server's attributes when the user does a "mount -o remount" on the filesystem. Signed-off-by:
-
Anna Schumaker authored
Clean up. There are a few places where we want to probe the server, but don't actually care about the fsinfo result. Change these to use nfs_probe_server(), which handles the fattr allocation for us. Signed-off-by:
-
Anna Schumaker authored
And rename it to nfs_probe_server(). I also change it to take the nfs_fh as an argument so callers can choose what filehandle to probe. Signed-off-by:
-
Anna Schumaker authored
And call it before doing an FSINFO probe to reset to the baseline capabilities before probing. Signed-off-by:
-
Chuck Lever authored
dprintk call sites that display no other information than the function name can be replaced with use of the trace "function" or "function_graph" plug-ins. Signed-off-by:
-
Chuck Lever authored
Introduce a single tracepoint that can replace simple dprintk call sites in upper layer "rpc_call_done" callbacks. Example: kworker/u24:2-1254 [001] 771.026677: rpc_stats_latency: task:00000001@00000002 xid=0x16a6f3c0 rpcbindv2 GETPORT backlog=446 rtt=101 execute=555 kworker/u24:2-1254 [001] 771.026677: rpc_task_call_done: task:00000001@00000002 flags=ASYNC|DYNAMIC|SOFT|SOFTCONN|SENT runstate=RUNNING|ACTIVE status=0 action=rpcb_getport_done kworker/u24:2-1254 [001] 771.026678: rpcb_setport: task:00000001@00000002 status=0 port=20048 Signed-off-by:
-
Chuck Lever authored
These new events report slightly different information for readpage and readpages/readahead. For readpage: fsx-1387 [006] 380.761896: nfs_aop_readpage: fileid=00:28:2 fhandle=0x36fbbe51 version=1752899355910932437 offset=131072 fsx-1387 [006] 380.761900: nfs_aop_readpage_done: fileid=00:28:2 fhandle=0x36fbbe51 version=1752899355910932437 offset=131072 ret=0 The index of a synchronous single-page read is reported. For readpages: fsx-1387 [006] 380.760847: nfs_aop_readahead: fileid=00:28:2 fhandle=0x36fbbe51 version=1752899355909932456 nr_pages=3 fsx-1387 [006] 380.760853: nfs_aop_readahead_done: fileid=00:28:2 fhandle=0x36fbbe51 version=1752899355909932456 nr_pages=3 ret=0 The count of pages requested is reported. nfs_readpages does not wait for the READ requests to complete. Signed-off-by: -
Chuck Lever authored
Clean up: BIT() is preferred over open-coding the shift. Signed-off-by:
-
Chuck Lever authored
For certain special cases, RPC-related tracepoints record a -1 as the task ID or the client ID. It's ugly for a trace event to display 4 billion in these cases. To help keep SUNRPC tracepoints consistent, create a macro that defines the print format specifiers for tk_pid and cl_clid. At some point in the future we might try tk_pid with a wider range of values than 0..64K so this makes it easier to make that change. RPC tracepoints now look like this: <...>-1276 [009] 149.720358: rpc_clnt_new: client=00000005 peer=[192.168.2.55]:20049 program=nfs server=klimt.ib <...>-1342 [004] 149.921234: rpc_xdr_recvfrom: task:0000001a@00000005 head=[0xff1242d9ab6dc01c,144] page=0 tail=[(nil),0] len=144 <...>-1342 [004] 149.921235: xprt_release_cong: task:0000001a@00000005 snd_task:ffffffff cong=256 cwnd=16384 <...>-1342 [004] 149.921235: xprt_put_cong: task:0000001a@00000005 snd_task:ffffffff cong=0 cwnd=16384 Signed-off-by:
-
Chuck Lever authored
Clean up: this field is no longer used. xprt_rdma_pad_optimize is also no longer used, but is left in place because it is part of the kernel/userspace API. Signed-off-by:
-
