- 16 Mar, 2020 32 commits
-
-
Trond Myklebust authored
Add basic tracing for debugging the sunrpc cache events. Signed-off-by:
Trond Myklebust <trond.myklebust@hammerspace.com> Signed-off-by:
Chuck Lever <chuck.lever@oracle.com>
-
Trond Myklebust authored
If the cache entry never gets initialised, we want the garbage collector to be able to evict it. Otherwise if the upcall daemon fails to initialise the entry, we end up never expiring it. Signed-off-by:
-
Trond Myklebust authored
If the rpc.mountd daemon goes down, then that should not cause all exports to start failing with ESTALE errors. Let's explicitly distinguish between the cache upcall cases that need to time out, and those that do not. Signed-off-by:
-
Trond Myklebust authored
Signed-off-by:
-
Trond Myklebust authored
Add tracepoints for upcalls. Signed-off-by:
-
Trond Myklebust authored
Add tracing to allow us to figure out where any stale filehandle issues may be originating from. Signed-off-by:
-
Trond Myklebust authored
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: 659aefb6 ("nfsd: Ensure we don't recognise lock stateids after freeing them") Signed-off-by:
-
Chuck Lever authored
xprt_sock_sendmsg uses the more efficient iov_iter-enabled kernel socket API, and is a pre-requisite for server send-side support for TLS. Note that svc_process no longer needs to reserve a word for the stream record marker, since the TCP transport now provides the record marker automatically in a separate buffer. The dprintk() in svc_send_common is also removed. It didn't seem crucial for field troubleshooting. If more is needed there, a trace point could be added in xprt_sock_sendmsg(). Signed-off-by: -
Chuck Lever authored
Re-locate xs_sendpages() so that it can be shared with server code. Signed-off-by: -
Chuck Lever authored
On some platforms, DMA mapping part of a page is more costly than copying bytes. Indeed, not involving the I/O MMU can help the RPC/RDMA transport scale better for tiny I/Os across more RDMA devices. This is because interaction with the I/O MMU is eliminated for each of these small I/Os. Without the explicit unmapping, the NIC no longer needs to do a costly internal TLB shoot down for buffers that are just a handful of bytes. Since pull-up is now a more a frequent operation, I've introduced a trace point in the pull-up path. It can be used for debugging or user-space tools that count pull-up frequency. Signed-off-by: -
Chuck Lever authored
Performance optimization: Avoid syncing the transport buffer twice when Reply buffer pull-up is necessary. Signed-off-by: -
Chuck Lever authored
Same idea as the receive-side changes I did a while back: use xdr_stream helpers rather than open-coding the XDR chunk list encoders. This builds the Reply transport header from beginning to end without backtracking. As additional clean-ups, fill in documenting comments for the XDR encoders and sprinkle some trace points in the new encoding functions. Signed-off-by: -
Chuck Lever authored
Clean up. These are taken from the client-side RPC/RDMA transport to a more global header file so they can be used elsewhere. Signed-off-by: -
Chuck Lever authored
These trace points are misnamed: trace_svcrdma_encode_wseg trace_svcrdma_encode_write trace_svcrdma_encode_reply trace_svcrdma_encode_rseg trace_svcrdma_encode_read trace_svcrdma_encode_pzr Because they actually trace posting on the Send Queue. Let's rename them so that I can add trace points in the chunk list encoders that actually do trace chunk list encoding events. Signed-off-by: -
Chuck Lever authored
Preparing for subsequent patches, no behavior change expected. Pass the RPC Call's svc_rdma_recv_ctxt deeper into the sendto() path. This enables passing more information about Requester- provided Write and Reply chunks into those lower-level functions. Signed-off-by: -
Chuck Lever authored
Preparing for subsequent patches, no behavior change expected. Pass the RPC Call's svc_rdma_recv_ctxt deeper into the sendto() path. This enables passing more information about Requester- provided Write and Reply chunks into those lower-level functions. Signed-off-by: -
Chuck Lever authored
Preparing for subsequent patches, no behavior change expected. Pass the RPC Call's svc_rdma_recv_ctxt deeper into the sendto() path. This enables passing more information about Requester- provided Write and Reply chunks into the lower-level send functions. Signed-off-by: -
Chuck Lever authored
Cache the locations of the Requester-provided Write list and Reply chunk so that the Send path doesn't need to parse the Call header again. Signed-off-by: -
Chuck Lever authored
The logic that checks incoming network headers has to be scrupulous. De-duplicate: replace open-coded buffer overflow checks with the use of xdr_stream helpers that are used most everywhere else XDR decoding is done. One minor change to the sanity checks: instead of checking the length of individual segments, cap the length of the whole chunk to be sure it can fit in the set of pages available in rq_pages. This should be a better test of whether the server can handle the chunks in each request. Signed-off-by: -
Chuck Lever authored
Clean up. This trace point is no longer needed because the RDMA/core CMA code has an equivalent trace point that was added by commit ed999f82 ("RDMA/cma: Add trace points in RDMA Connection Manager"). Signed-off-by:
-
Chuck Lever authored
This class can be used to create trace points in either the RPC client or RPC server paths. It simply displays the length of each part of an xdr_buf, which is useful to determine that the transport and XDR codecs are operating correctly. Signed-off-by: -
Chuck Lever authored
Signed-off-by: -
Chuck Lever authored
Introduce a helper function to compute the XDR pad size of a variable-length XDR object. Clean up: Replace open-coded calculation of XDR pad sizes. I'm sure I haven't found every instance of this calculation. Signed-off-by: -
Chuck Lever authored
This error path is almost never executed. Found by code inspection. Fixes: 99722fe4 ("svcrdma: Persistently allocate and DMA-map Send buffers") Signed-off-by:
-
Chuck Lever authored
Address some minor nits I noticed while working on this function. Signed-off-by: -
Chuck Lever authored
svcrdma expects that the payload falls precisely into the xdr_buf page vector. This does not seem to be the case for nfsd4_encode_readv(). This code is called only when fops->splice_read is missing or when RQ_SPLICE_OK is clear, so it's not a noticeable problem in many common cases. Add new transport method: ->xpo_read_payload so that when a READ payload does not fit exactly in rq_res's page vector, the XDR encoder can inform the RPC transport exactly where that payload is, without the payload's XDR pad. That way, when a Write chunk is present, the transport knows what byte range in the Reply message is supposed to be matched with the chunk. Note that the Linux NFS server implementation of NFS/RDMA can currently handle only one Write chunk per RPC-over-RDMA message. This simplifies the implementation of this fix. Fixes: b0420980 ("nfsd4: allow exotic read compounds") Buglink: https://bugzilla.kernel.org/show_bug.cgi?id=198053Signed-off-by:
-
Madhuparna Bhowmik authored
list_for_each_entry_rcu() has built-in RCU and lock checking. Pass cond argument to list_for_each_entry_rcu() to silence false lockdep warning when CONFIG_PROVE_RCU_LIST is enabled by default. Signed-off-by:
Madhuparna Bhowmik <madhuparnabhowmik10@gmail.com> Signed-off-by:
J. Bruce Fields <bfields@redhat.com> Signed-off-by:
-
Madhuparna Bhowmik authored
list_for_each_entry_rcu() has built-in RCU and lock checking. Pass cond argument to list_for_each_entry_rcu() to silence false lockdep warning when CONFIG_PROVE_RCU_LIST is enabled by default. Signed-off-by:
-
Gustavo A. R. Silva authored
The current codebase makes use of the zero-length array language extension to the C90 standard, but the preferred mechanism to declare variable-length types such as these ones is a flexible array member[1][2], introduced in C99: struct foo { int stuff; struct boo array[]; }; By making use of the mechanism above, we will get a compiler warning in case the flexible array does not occur last in the structure, which will help us prevent some kind of undefined behavior bugs from being inadvertently introduced[3] to the codebase from now on. Also, notice that, dynamic memory allocations won't be affected by this change: "Flexible array members have incomplete type, and so the sizeof operator may not be applied. As a quirk of the original implementation of zero-length arrays, sizeof evaluates to zero."[1] This issue was found with the help of Coccinelle. [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html [2] https://github.com/KSPP/linux/issues/21 [3] commit 76497732 ("cxgb3/l2t: Fix undefined behaviour") Signed-off-by:Gustavo A. R. Silva <gustavo@embeddedor.com> Signed-off-by:
-
Amol Grover authored
detail->hash_table[] is traversed using hlist_for_each_entry_rcu outside an RCU read-side critical section but under the protection of detail->hash_lock. Hence, add corresponding lockdep expression to silence false-positive warnings, and harden RCU lists. Signed-off-by:
Amol Grover <frextrite@gmail.com> Signed-off-by:
-
Gustavo A. R. Silva authored
The current codebase makes use of the zero-length array language extension to the C90 standard, but the preferred mechanism to declare variable-length types such as these ones is a flexible array member[1][2], introduced in C99: struct foo { int stuff; struct boo array[]; }; By making use of the mechanism above, we will get a compiler warning in case the flexible array does not occur last in the structure, which will help us prevent some kind of undefined behavior bugs from being inadvertently introduced[3] to the codebase from now on. Also, notice that, dynamic memory allocations won't be affected by this change: "Flexible array members have incomplete type, and so the sizeof operator may not be applied. As a quirk of the original implementation of zero-length arrays, sizeof evaluates to zero."[1] This issue was found with the help of Coccinelle. [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html [2] https://github.com/KSPP/linux/issues/21 [3] commit 76497732 ("cxgb3/l2t: Fix undefined behaviour") Signed-off-by: -
Scott Mayhew authored
Currently, nfsd4_encode_exchange_id() encodes the utsname nodename string in the server_scope field. In a multi-host container environemnt, if an nfsd container is restarted on a different host than it was originally running on, clients will see a server_scope mismatch and will not attempt to reclaim opens. Instead, set the server_scope while we're in a process context during service startup, so we get the utsname nodename of the current process and store that in nfsd_net. Signed-off-by:
Scott Mayhew <smayhew@redhat.com> [bfields: fix up major_id too] Signed-off-by:
-
- 15 Mar, 2020 8 commits
-
-
Linus Torvalds authored
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull irq fix from Thomas Gleixner: "A single commit to handle an erratum in Cavium ThunderX to prevent access to GIC registers which are broken in the implementation" * tag 'irq-urgent-2020-03-15' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: irqchip/gic-v3: Workaround Cavium erratum 38539 when reading GICD_TYPER2
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull futex fix from Thomas Gleixner: "Fix for yet another subtle futex issue. The futex code used ihold() to prevent inodes from vanishing, but ihold() does not guarantee inode persistence. Replace the inode pointer with a per boot, machine wide, unique inode identifier. The second commit fixes the breakage of the hash mechanism which causes a 100% performance regression" * tag 'locking-urgent-2020-03-15' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: futex: Unbreak futex hashing futex: Fix inode life-time issue
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull x86 fixes from Thomas Gleixner: "Two fixes for x86: - Map EFI runtime service data as encrypted when SEV is enabled. Otherwise e.g. SMBIOS data cannot be properly decoded by dmidecode. - Remove the warning in the vector management code which triggered when a managed interrupt affinity changed outside of a CPU hotplug operation. The warning was correct until the recent core code change that introduced a CPU isolation feature which needs to migrate managed interrupts away from online CPUs under certain conditions to achieve the isolation" * tag 'x86-urgent-2020-03-15' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86/vector: Remove warning on managed interrupt migration x86/ioremap: Map EFI runtime services data as encrypted for SEV -
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull perf fixes from Thomas Gleixner: "A pile of perf fixes: Kernel side: - AMD uncore driver: Replace the open coded sanity check with the core variant, which provides the correct error code and also leaves a hint in dmesg Tooling: - Fix the stdio input handling with glibc versions >= 2.28 - Unbreak the futex-wake benchmark which was reduced to 0 test threads due to the conversion to cpumaps - Initialize sigaction structs before invoking sys_sigactio() - Plug the mapfile memory leak in perf jevents - Fix off by one relative directory includes - Fix an undefined string comparison in perf diff" * tag 'perf-urgent-2020-03-15' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: perf/amd/uncore: Replace manual sampling check with CAP_NO_INTERRUPT flag tools: Fix off-by 1 relative directory includes perf jevents: Fix leak of mapfile memory perf bench: Clear struct sigaction before sigaction() syscall perf bench futex-wake: Restore thread count default to online CPU count perf top: Fix stdio interface input handling with glibc 2.28+ perf diff: Fix undefined string comparision spotted by clang's -Wstring-compare perf symbols: Don't try to find a vmlinux file when looking for kernel modules perf bench: Share some global variables to fix build with gcc 10 perf parse-events: Use asprintf() instead of strncpy() to read tracepoint files perf env: Do not return pointers to local variables perf tests bp_account: Make global variable static -
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull timer fix from Thomas Gleixner: "A single fix adding the missing time namespace adjustment in sys/sysinfo which caused sys/sysinfo to be inconsistent with /proc/uptime when read from a task inside a time namespace" * tag 'timers-urgent-2020-03-15' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: sys/sysinfo: Respect boottime inside time namespace
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull RAS fixes from Thomas Gleixner: "Two RAS related fixes: - Shut down the per CPU thermal throttling poll work properly when a CPU goes offline. The missing shutdown caused the poll work to be migrated to a unbound worker which triggered warnings about the usage of smp_processor_id() in preemptible context - Fix the PPIN feature initialization which missed to enable the functionality when PPIN_CTL was enabled but the MSR locked against updates" * tag 'ras-urgent-2020-03-15' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86/mce: Fix logic and comments around MSR_PPIN_CTL x86/mce/therm_throt: Undo thermal polling properly on CPU offline -
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull EFI fixes from Thomas Gleixner: "Two EFI fixes: - Prevent a race and buffer overflow in the sysfs efivars interface which causes kernel memory corruption. - Add the missing NULL pointer checks in efivar_store_raw()" * tag 'efi-urgent-2020-03-15' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: efi: Add a sanity check to efivar_store_raw() efi: Fix a race and a buffer overflow while reading efivars via sysfs
-
