| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
In commit edb854a3680b ("scsi: core: Reallocate device's budget map on
queue depth change"), the sbitmap for the device budget map may be
reallocated after the slave device depth is configured.
When the sbitmap is reallocated we use the result from
scsi_device_max_queue_depth() for the sbitmap size, but don't resize to
match the actual device queue depth.
Fix by resizing the sbitmap after reallocating the budget sbitmap. We do
this instead of init'ing the sbitmap to the device queue depth as the user
may want to change the queue depth later via sysfs or other.
Link: https://lore.kernel.org/r/[email protected]
Fixes: edb854a3680b ("scsi: core: Reallocate device's budget map on queue depth change")
Tested-by: Damien Le Moal <[email protected]>
Reviewed-by: Ming Lei <[email protected]>
Reviewed-by: Bart Van Assche <[email protected]>
Signed-off-by: John Garry <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
Remove aha152x_cmd_priv.scsi_pointer by moving the necessary members into
aha152x_cmd_priv proper.
Tested with an Adaptec SlimSCSI APA-1460A card.
Link: https://lore.kernel.org/r/bdc1264b6dd331150bffb737958cab8c9c068fa1.1648070977.git.fthain@linux-m68k.org
Cc: Christoph Hellwig <[email protected]>
Suggested-by: Christoph Hellwig <[email protected]>
Reviewed-by: Christoph Hellwig <[email protected]>
Reviewed-by: Hannes Reinecke <[email protected]>
Signed-off-by: Finn Thain <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
The adapter request_limit is hardcoded to be INITIAL_SRP_LIMIT which is
currently an arbitrary value of 800. Increase this value to 1024 which
better matches the characteristics of the typical IBMi Initiator that
supports 32 LUNs and a queue depth of 32.
This change also has the secondary benefit of being a power of two as
required by the kfifo API. Since, Commit ab9bb6318b09 ("Partially revert
"kfifo: fix kfifo_alloc() and kfifo_init()"") the size of IU pool for each
target has been rounded down to 512 when attempting to kfifo_init() those
pools with the current request_limit size of 800.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Tyrel Datwyler <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
Commit 2e27f576abc6 ("scsi: scsi_ioctl: Call scsi_cmd_ioctl() from
scsi_ioctl()") seems to have a typo as it is checking ret instead of cmd in
the if statement checking for CDROMCLOSETRAY and CDROMEJECT. This changes
the behaviour of these ioctls as the cdrom_ioctl handling of these is more
restrictive than the scsi_ioctl version.
Link: https://lore.kernel.org/r/[email protected]
Fixes: 2e27f576abc6 ("scsi: scsi_ioctl: Call scsi_cmd_ioctl() from scsi_ioctl()")
Reviewed-by: Christoph Hellwig <[email protected]>
Signed-off-by: Kevin Groeneveld <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
The whole passthrough ioctl path looks completely broken. For example it
DMA maps the scatterlist and after that copies data to it, which is
prohibited by the DMA API contract.
Moreover, in pmcraid_alloc_sglist(), the pointer returned by a
sgl_alloc_order() call is not recorded anywhere which is pointless.
So remove the PMCRAID_PASSTHROUGH_IOCTL ioctl implementation entirely.
Should it be needed, we should reimplement it using the proper block layer
request mapping helpers.
Link: https://lore.kernel.org/r/7f27a70bec3f3dcaf46a29b1c630edd4792e71c0.1648298857.git.christophe.jaillet@wanadoo.fr
Suggested-by: Christoph Hellwig <[email protected]>
Reviewed-by: Christoph Hellwig <[email protected]>
Signed-off-by: Christophe JAILLET <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
The request_queue may be NULL in a request, for example when it comes from
scsi_ioctl_reset(). Check it before use.
Fixes: f3fa33acca9f ("block: remove the ->rq_disk field in struct request")
Link: https://lore.kernel.org/r/[email protected]
Reported-by: Changhui Zhong <[email protected]>
Reviewed-by: Christoph Hellwig <[email protected]>
Signed-off-by: Tomas Henzl <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
Remove unused fields 'rpm_lvl_attr' and 'spm_lvl_attr' in struct ufs_hba.
Commit cbb6813ee771 ("scsi: ufs: sysfs: attribute group for existing sysfs
entries.") removed all code using that field.
Link: https://lore.kernel.org/r/413601558.101648105683746.JavaMail.epsvc@epcpadp4
Reviewed-by: Bart Van Assche <[email protected]>
Acked-by: Avri Altman <[email protected]>
Signed-off-by: Keoseong Park <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
The rules changed for lpfc_sli_iocbq_lookup() vs locking. Prior, the
routine properly took out the lock. In newly refactored code, the locks
must be held when calling the routine.
Fix lpfc_sli_process_sol_iocb() to take the locks before calling the
routine.
Fix lpfc_sli_handle_fast_ring_event() to not release the locks to call the
routine.
Link: https://lore.kernel.org/r/[email protected]
Fixes: 1b64aa9eae28 ("scsi: lpfc: SLI path split: Refactor fast and slow paths to native SLI4")
Co-developed-by: Dick Kennedy <[email protected]>
Signed-off-by: Dick Kennedy <[email protected]>
Signed-off-by: James Smart <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
There was a merge error in ther 14.2.0.0 patches that resulted in the SLI4
path using the SLI3 issue_abort_iotag() routine. This resulted in txcmplq
corruption.
Fix to use the SLI4 routine when SLI4.
Link: https://lore.kernel.org/r/[email protected]
Fixes: 31a59f75702f ("scsi: lpfc: SLI path split: Refactor Abort paths")
Cc: <[email protected]> # v5.2+
Co-developed-by: Dick Kennedy <[email protected]>
Signed-off-by: Dick Kennedy <[email protected]>
Signed-off-by: James Smart <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
Update lpfc version to 14.2.0.1
Link: https://lore.kernel.org/r/[email protected]
Co-developed-by: Justin Tee <[email protected]>
Signed-off-by: Justin Tee <[email protected]>
Signed-off-by: James Smart <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
When recovering from a pci-parity error the driver is failing to re-create
queues, causing recovery to fail. Looking deeper, it was found that the
interrupt vector count allocated on the recovery was fewer than the vectors
originally allocated. This disparity resulted in CPU map entries with stale
information. When the driver tries to re-create the queues, it attempts to
use the stale information which indicates an eq/interrupt vector that was
no longer created.
Fix by clearng the cpup map array before enabling and requesting the IRQs
in the lpfc_sli_reset_slot_s4 routine().
Link: https://lore.kernel.org/r/[email protected]
Co-developed-by: Justin Tee <[email protected]>
Signed-off-by: Justin Tee <[email protected]>
Signed-off-by: James Smart <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
When injecting EEH errors the port is getting hung up waiting on the node
list to empty, message number 0233. The driver is stuck at this point and
also can't unload. The driver makes transport remoteport delete calls which
try to abort I/O's, but the EEH daemon has already called the driver to
detach and the detachment has set the global FC_UNLOADING flag. There are
several code paths that will avoid I/O cleanup if the FC_UNLOADING flag is
set, resulting in transports waiting for I/O while the driver is waiting on
transports to clean up.
Additionally, during study of the list, a locking issue was found in
lpfc_sli_abort_iocb_ring that could corrupt the list.
A special case was added to the lpfc_cleanup() routine to call
lpfc_sli_flush_rings() if the driver is FC_UNLOADING and if the pci-slot
is offline (e.g. EEH).
The SLI4 part of lpfc_sli_abort_iocb_ring() is changed to use the
ring_lock. Also added code to cancel the I/Os if the pci-slot is offline
and added checks and returns for the FC_UNLOADING and HBA_IOQ_FLUSH flags
to prevent trying to send an I/O that we cannot handle.
Link: https://lore.kernel.org/r/[email protected]
Co-developed-by: Justin Tee <[email protected]>
Signed-off-by: Justin Tee <[email protected]>
Signed-off-by: James Smart <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
Following EEH errors, the driver can crash or hang when deleting the
localport or when attempting to unload.
The EEH handlers in the driver did not notify the NVMe-FC transport before
tearing the driver down. This was delayed until the resume steps. This
worked for SCSI because lpfc_block_scsi() would notify the
scsi_fc_transport that the target was not available but it would not clean
up all the references to the ndlp.
The SLI3 prep for dev reset handler did the lpfc_offline_prep() and
lpfc_offline() calls to get the port stopped before restarting. The SLI4
version of the prep for dev reset just destroyed the queues and did not
stop NVMe from continuing. Also because the port was not really stopped
the localport destroy would hang because the transport was still waiting
for I/O. Additionally, a devloss tmo can fire and post events to a stopped
worker thread creating another hang condition.
lpfc_sli4_prep_dev_for_reset() is modified to call lpfc_offline_prep() and
lpfc_offline() rather than just lpfc_scsi_dev_block() to ensure both SCSI
and NVMe transports are notified to block I/O to the driver.
Logic is added to devloss handler and worker thread to clean up ndlp
references and quiesce appropriately.
Link: https://lore.kernel.org/r/[email protected]
Co-developed-by: Justin Tee <[email protected]>
Signed-off-by: Justin Tee <[email protected]>
Signed-off-by: James Smart <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
tcmu_try_get_data_page() looks up pages under cmdr_lock, but it does not
take refcount properly and just returns page pointer. When
tcmu_try_get_data_page() returns, the returned page may have been freed by
tcmu_blocks_release().
We need to get_page() under cmdr_lock to avoid concurrent
tcmu_blocks_release().
Link: https://lore.kernel.org/r/[email protected]
Reviewed-by: Bodo Stroesser <[email protected]>
Signed-off-by: Xiaoguang Wang <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
iscsi_cls_conn is alloced by kzalloc(), the whole iscsi_cls_conn is zero
filled already including the dd_data. So it is unnecessary to call memset
again.
Link: https://lore.kernel.org/r/[email protected]
Reviewed-by: Wu Bo <[email protected]>
Reviewed-by: Lee Duncan <[email protected]>
Reviewed-by: Mike Christie <[email protected]>
Signed-off-by: Wenchao Hao <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
The function mpt3sas_transport_port_remove() called in
_scsih_expander_node_remove() frees the port field of the sas_expander
structure, leading to the following use-after-free splat from KASAN when
the ioc_info() call following that function is executed (e.g. when doing
rmmod of the driver module):
[ 3479.371167] ==================================================================
[ 3479.378496] BUG: KASAN: use-after-free in _scsih_expander_node_remove+0x710/0x750 [mpt3sas]
[ 3479.386936] Read of size 1 at addr ffff8881c037691c by task rmmod/1531
[ 3479.393524]
[ 3479.395035] CPU: 18 PID: 1531 Comm: rmmod Not tainted 5.17.0-rc8+ #1436
[ 3479.401712] Hardware name: Supermicro Super Server/H12SSL-NT, BIOS 2.1 06/02/2021
[ 3479.409263] Call Trace:
[ 3479.411743] <TASK>
[ 3479.413875] dump_stack_lvl+0x45/0x59
[ 3479.417582] print_address_description.constprop.0+0x1f/0x120
[ 3479.423389] ? _scsih_expander_node_remove+0x710/0x750 [mpt3sas]
[ 3479.429469] kasan_report.cold+0x83/0xdf
[ 3479.433438] ? _scsih_expander_node_remove+0x710/0x750 [mpt3sas]
[ 3479.439514] _scsih_expander_node_remove+0x710/0x750 [mpt3sas]
[ 3479.445411] ? _raw_spin_unlock_irqrestore+0x2d/0x40
[ 3479.452032] scsih_remove+0x525/0xc90 [mpt3sas]
[ 3479.458212] ? mpt3sas_expander_remove+0x1d0/0x1d0 [mpt3sas]
[ 3479.465529] ? down_write+0xde/0x150
[ 3479.470746] ? up_write+0x14d/0x460
[ 3479.475840] ? kernfs_find_ns+0x137/0x310
[ 3479.481438] pci_device_remove+0x65/0x110
[ 3479.487013] __device_release_driver+0x316/0x680
[ 3479.493180] driver_detach+0x1ec/0x2d0
[ 3479.498499] bus_remove_driver+0xe7/0x2d0
[ 3479.504081] pci_unregister_driver+0x26/0x250
[ 3479.510033] _mpt3sas_exit+0x2b/0x6cf [mpt3sas]
[ 3479.516144] __x64_sys_delete_module+0x2fd/0x510
[ 3479.522315] ? free_module+0xaa0/0xaa0
[ 3479.527593] ? __cond_resched+0x1c/0x90
[ 3479.532951] ? lockdep_hardirqs_on_prepare+0x273/0x3e0
[ 3479.539607] ? syscall_enter_from_user_mode+0x21/0x70
[ 3479.546161] ? trace_hardirqs_on+0x1c/0x110
[ 3479.551828] do_syscall_64+0x35/0x80
[ 3479.556884] entry_SYSCALL_64_after_hwframe+0x44/0xae
[ 3479.563402] RIP: 0033:0x7f1fc482483b
...
[ 3479.943087] ==================================================================
Fix this by introducing the local variable port_id to store the port ID
value before executing mpt3sas_transport_port_remove(). This local variable
is then used in the call to ioc_info() instead of dereferencing the freed
port structure.
Link: https://lore.kernel.org/r/[email protected]
Fixes: 7d310f241001 ("scsi: mpt3sas: Get device objects using sas_address & portID")
Cc: [email protected]
Acked-by: Sreekanth Reddy <[email protected]>
Signed-off-by: Damien Le Moal <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
Alexei Starovoitov says:
====================
pull-request: bpf 2022-03-29
We've added 16 non-merge commits during the last 1 day(s) which contain
a total of 24 files changed, 354 insertions(+), 187 deletions(-).
The main changes are:
1) x86 specific bits of fprobe/rethook, from Masami and Peter.
2) ice/xsk fixes, from Maciej and Magnus.
3) Various small fixes, from Andrii, Yonghong, Geliang and others.
* https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
selftests/bpf: Fix clang compilation errors
ice: xsk: Fix indexing in ice_tx_xsk_pool()
ice: xsk: Stop Rx processing when ntc catches ntu
ice: xsk: Eliminate unnecessary loop iteration
xsk: Do not write NULL in SW ring at allocation failure
x86,kprobes: Fix optprobe trampoline to generate complete pt_regs
x86,rethook: Fix arch_rethook_trampoline() to generate a complete pt_regs
x86,rethook,kprobes: Replace kretprobe with rethook on x86
kprobes: Use rethook for kretprobe if possible
bpftool: Fix generated code in codegen_asserts
selftests/bpf: fix selftest after random: Urandom_read tracepoint removal
bpf: Fix maximum permitted number of arguments check
bpf: Sync comments for bpf_get_stack
fprobe: Fix sparse warning for acccessing __rcu ftrace_hash
fprobe: Fix smatch type mismatch warning
bpf/bpftool: Add unprivileged_bpf_disabled check against value of 2
====================
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
Pull NFS client updates from Trond Myklebust:
"Highlights include:
Features:
- Switch NFS to use readahead instead of the obsolete readpages.
- Readdir fixes to improve cacheability of large directories when
there are multiple readers and writers.
- Readdir performance improvements when doing a seekdir() immediately
after opening the directory (common when re-exporting NFS).
- NFS swap improvements from Neil Brown.
- Loosen up memory allocation to permit direct reclaim and write back
in cases where there is no danger of deadlocking the writeback code
or NFS swap.
- Avoid sillyrename when the NFSv4 server claims to support the
necessary features to recover the unlinked but open file after
reboot.
Bugfixes:
- Patch from Olga to add a mount option to control NFSv4.1 session
trunking discovery, and default it to being off.
- Fix a lockup in nfs_do_recoalesce().
- Two fixes for list iterator variables being used when pointing to
the list head.
- Fix a kernel memory scribble when reading from a non-socket
transport in /sys/kernel/sunrpc.
- Fix a race where reconnecting to a server could leave the TCP
socket stuck forever in the connecting state.
- Patch from Neil to fix a shutdown race which can leave the SUNRPC
transport timer primed after we free the struct xprt itself.
- Patch from Xin Xiong to fix reference count leaks in the NFSv4.2
copy offload.
- Sunrpc patch from Olga to avoid resending a task on an offlined
transport.
Cleanups:
- Patches from Dave Wysochanski to clean up the fscache code"
* tag 'nfs-for-5.18-1' of git://git.linux-nfs.org/projects/trondmy/linux-nfs: (91 commits)
NFSv4/pNFS: Fix another issue with a list iterator pointing to the head
NFS: Don't loop forever in nfs_do_recoalesce()
SUNRPC: Don't return error values in sysfs read of closed files
SUNRPC: Do not dereference non-socket transports in sysfs
NFSv4.1: don't retry BIND_CONN_TO_SESSION on session error
SUNRPC don't resend a task on an offlined transport
NFS: replace usage of found with dedicated list iterator variable
SUNRPC: avoid race between mod_timer() and del_timer_sync()
pNFS/files: Ensure pNFS allocation modes are consistent with nfsiod
pNFS/flexfiles: Ensure pNFS allocation modes are consistent with nfsiod
NFSv4/pnfs: Ensure pNFS allocation modes are consistent with nfsiod
NFS: Avoid writeback threads getting stuck in mempool_alloc()
NFS: nfsiod should not block forever in mempool_alloc()
SUNRPC: Make the rpciod and xprtiod slab allocation modes consistent
SUNRPC: Fix unx_lookup_cred() allocation
NFS: Fix memory allocation in rpc_alloc_task()
NFS: Fix memory allocation in rpc_malloc()
SUNRPC: Improve accuracy of socket ENOBUFS determination
SUNRPC: Replace internal use of SOCKWQ_ASYNC_NOSPACE
SUNRPC: Fix socket waits for write buffer space
...
|
|
Jan Kara reported a performance regression in dbench that he
bisected down to commit bad77c375e8d ("xfs: CIL checkpoint
flushes caches unconditionally").
Whilst developing the journal flush/fua optimisations this cache was
part of, it appeared to made a significant difference to
performance. However, now that this patchset has settled and all the
correctness issues fixed, there does not appear to be any
significant performance benefit to asynchronous cache flushes.
In fact, the opposite is true on some storage types and workloads,
where additional cache flushes that can occur from fsync heavy
workloads have measurable and significant impact on overall
throughput.
Local dbench testing shows little difference on dbench runs with
sync vs async cache flushes on either fast or slow SSD storage, and
no difference in streaming concurrent async transaction workloads
like fs-mark.
Fast NVME storage.
From `dbench -t 30`, CIL scale:
clients async sync
BW Latency BW Latency
1 935.18 0.855 915.64 0.903
8 2404.51 6.873 2341.77 6.511
16 3003.42 6.460 2931.57 6.529
32 3697.23 7.939 3596.28 7.894
128 7237.43 15.495 7217.74 11.588
512 5079.24 90.587 5167.08 95.822
fsmark, 32 threads, create w/ 64 byte xattr w/32k logbsize
create chown unlink
async 1m41s 1m16s 2m03s
sync 1m40s 1m19s 1m54s
Slower SATA SSD storage:
From `dbench -t 30`, CIL scale:
clients async sync
BW Latency BW Latency
1 78.59 15.792 83.78 10.729
8 367.88 92.067 404.63 59.943
16 564.51 72.524 602.71 76.089
32 831.66 105.984 870.26 110.482
128 1659.76 102.969 1624.73 91.356
512 2135.91 223.054 2603.07 161.160
fsmark, 16 threads, create w/32k logbsize
create unlink
async 5m06s 4m15s
sync 5m00s 4m22s
And on Jan's test machine:
5.18-rc8-vanilla 5.18-rc8-patched
Amean 1 71.22 ( 0.00%) 64.94 * 8.81%*
Amean 2 93.03 ( 0.00%) 84.80 * 8.85%*
Amean 4 150.54 ( 0.00%) 137.51 * 8.66%*
Amean 8 252.53 ( 0.00%) 242.24 * 4.08%*
Amean 16 454.13 ( 0.00%) 439.08 * 3.31%*
Amean 32 835.24 ( 0.00%) 829.74 * 0.66%*
Amean 64 1740.59 ( 0.00%) 1686.73 * 3.09%*
Performance and cache flush behaviour is restored to pre-regression
levels.
As such, we can now consider the async cache flush mechanism an
unnecessary exercise in premature optimisation and hence we can
now remove it and the infrastructure it requires completely.
Fixes: bad77c375e8d ("xfs: CIL checkpoint flushes caches unconditionally")
Reported-and-tested-by: Jan Kara <[email protected]>
Signed-off-by: Dave Chinner <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
|
|
When a checkpoint writeback is run by log recovery, corruption
propagated from the log can result in writeback verifiers failing
and calling xfs_force_shutdown() from
xfs_buf_delwri_submit_buffers().
This results in the mount being marked as shutdown, but the log does
not get marked as shut down because:
/*
* If this happens during log recovery then we aren't using the runtime
* log mechanisms yet so there's nothing to shut down.
*/
if (!log || xlog_in_recovery(log))
return false;
If there are other buffers that then fail (say due to detecting the
mount shutdown), they will now hang in xfs_do_force_shutdown()
waiting for the log to shut down like this:
__schedule+0x30d/0x9e0
schedule+0x55/0xd0
xfs_do_force_shutdown+0x1cd/0x200
? init_wait_var_entry+0x50/0x50
xfs_buf_ioend+0x47e/0x530
__xfs_buf_submit+0xb0/0x240
xfs_buf_delwri_submit_buffers+0xfe/0x270
xfs_buf_delwri_submit+0x3a/0xc0
xlog_do_recovery_pass+0x474/0x7b0
? do_raw_spin_unlock+0x30/0xb0
xlog_do_log_recovery+0x91/0x140
xlog_do_recover+0x38/0x1e0
xlog_recover+0xdd/0x170
xfs_log_mount+0x17e/0x2e0
xfs_mountfs+0x457/0x930
xfs_fs_fill_super+0x476/0x830
xlog_force_shutdown() always needs to mark the log as shut down,
regardless of whether recovery is in progress or not, so that
multiple calls to xfs_force_shutdown() during recovery don't end
up waiting for the log to be shut down like this.
Signed-off-by: Dave Chinner <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
|
|
If a shut races with xfs_trans_commit() and we have shut down the
filesystem but not the log, we will still cancel the transaction.
This can result in aborting dirty log items instead of committing and
pinning them whilst the log is still running. Hence we can end up
with dirty, unlogged metadata that isn't in the AIL in memory that
can be flushed to disk via writeback clustering.
This was discovered from a g/388 trace where an inode log item was
having IO completed on it and it wasn't in the AIL, hence tripping
asserts xfs_ail_check(). Inode cluster writeback started long after
the filesystem shutdown started, and long after the transaction
containing the dirty inode was aborted and the log item marked
XFS_LI_ABORTED. The inode was seen as dirty and unpinned, so it
was flushed. IO completion tried to remove the inode from the AIL,
at which point stuff went bad:
XFS (pmem1): Log I/O Error (0x6) detected at xfs_fs_goingdown+0xa3/0xf0 (fs/xfs/xfs_fsops.c:500). Shutting down filesystem.
XFS: Assertion failed: in_ail, file: fs/xfs/xfs_trans_ail.c, line: 67
XFS (pmem1): Please unmount the filesystem and rectify the problem(s)
Workqueue: xfs-buf/pmem1 xfs_buf_ioend_work
RIP: 0010:assfail+0x27/0x2d
Call Trace:
<TASK>
xfs_ail_check+0xa8/0x180
xfs_ail_delete_one+0x3b/0xf0
xfs_buf_inode_iodone+0x329/0x3f0
xfs_buf_ioend+0x1f8/0x530
xfs_buf_ioend_work+0x15/0x20
process_one_work+0x1ac/0x390
worker_thread+0x56/0x3c0
kthread+0xf6/0x120
ret_from_fork+0x1f/0x30
</TASK>
xfs_trans_commit() needs to check log state for shutdown, not mount
state. It cannot abort dirty log items while the log is still
running as dirty items must remained pinned in memory until they are
either committed to the journal or the log has shut down and they
can be safely tossed away. Hence if the log has not shut down, the
xfs_trans_commit() path must allow completed transactions to commit
to the CIL and pin the dirty items even if a mount shutdown has
started.
Signed-off-by: Dave Chinner <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
|
|
When we call xfs_forced_shutdown(), the caller often expects the
filesystem to be completely shut down when it returns. However,
if we have racing xfs_forced_shutdown() calls, the first caller sets
the mount shutdown flag then goes to shutdown the log. The second
caller sees the mount shutdown flag and returns immediately - it
does not wait for the log to be shut down.
Unfortunately, xfs_forced_shutdown() is used in some places that
expect it to completely shut down the filesystem before it returns
(e.g. xfs_trans_log_inode()). As such, returning before the log has
been shut down leaves us in a place where the transaction failed to
complete correctly but we still call xfs_trans_commit(). This
situation arises because xfs_trans_log_inode() does not return an
error and instead calls xfs_force_shutdown() to ensure that the
transaction being committed is aborted.
Unfortunately, we have a race condition where xfs_trans_commit()
needs to check xlog_is_shutdown() because it can't abort log items
before the log is shut down, but it needs to use xfs_is_shutdown()
because xfs_forced_shutdown() does not block waiting for the log to
shut down.
To fix this conundrum, first we make all calls to
xfs_forced_shutdown() block until the log is also shut down. This
means we can then safely use xfs_forced_shutdown() as a mechanism
that ensures the currently running transaction will be aborted by
xfs_trans_commit() regardless of the shutdown check it uses.
Signed-off-by: Dave Chinner <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
|
|
We've got a mess on our hands.
1. xfs_trans_commit() cannot cancel transactions because the mount is
shut down - that causes dirty, aborted, unlogged log items to sit
unpinned in memory and potentially get written to disk before the
log is shut down. Hence xfs_trans_commit() can only abort
transactions when xlog_is_shutdown() is true.
2. xfs_force_shutdown() is used in places to cause the current
modification to be aborted via xfs_trans_commit() because it may be
impractical or impossible to cancel the transaction directly, and
hence xfs_trans_commit() must cancel transactions when
xfs_is_shutdown() is true in this situation. But we can't do that
because of #1.
3. Log IO errors cause log shutdowns by calling xfs_force_shutdown()
to shut down the mount and then the log from log IO completion.
4. xfs_force_shutdown() can result in a log force being issued,
which has to wait for log IO completion before it will mark the log
as shut down. If #3 races with some other shutdown trigger that runs
a log force, we rely on xfs_force_shutdown() silently ignoring #3
and avoiding shutting down the log until the failed log force
completes.
5. To ensure #2 always works, we have to ensure that
xfs_force_shutdown() does not return until the the log is shut down.
But in the case of #4, this will result in a deadlock because the
log Io completion will block waiting for a log force to complete
which is blocked waiting for log IO to complete....
So the very first thing we have to do here to untangle this mess is
dissociate log shutdown triggers from mount shutdowns. We already
have xlog_forced_shutdown, which will atomically transistion to the
log a shutdown state. Due to internal asserts it cannot be called
multiple times, but was done simply because the only place that
could call it was xfs_do_force_shutdown() (i.e. the mount shutdown!)
and that could only call it once and once only. So the first thing
we do is remove the asserts.
We then convert all the internal log shutdown triggers to call
xlog_force_shutdown() directly instead of xfs_force_shutdown(). This
allows the log shutdown triggers to shut down the log without
needing to care about mount based shutdown constraints. This means
we shut down the log independently of the mount and the mount may
not notice this until it's next attempt to read or modify metadata.
At that point (e.g. xfs_trans_commit()) it will see that the log is
shutdown, error out and shutdown the mount.
To ensure that all the unmount behaviours and asserts track
correctly as a result of a log shutdown, propagate the shutdown up
to the mount if it is not already set. This keeps the mount and log
state in sync, and saves a huge amount of hassle where code fails
because of a log shutdown but only checks for mount shutdowns and
hence ends up doing the wrong thing. Cleaning up that mess is
an exercise for another day.
This enables us to address the other problems noted above in
followup patches.
Signed-off-by: Dave Chinner <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
|
|
Brian reported a null pointer dereference failure during unmount in
xfs/006. He tracked the problem down to the AIL being torn down
before a log shutdown had completed and removed all the items from
the AIL. The failure occurred in this path while unmount was
proceeding in another task:
xfs_trans_ail_delete+0x102/0x130 [xfs]
xfs_buf_item_done+0x22/0x30 [xfs]
xfs_buf_ioend+0x73/0x4d0 [xfs]
xfs_trans_committed_bulk+0x17e/0x2f0 [xfs]
xlog_cil_committed+0x2a9/0x300 [xfs]
xlog_cil_process_committed+0x69/0x80 [xfs]
xlog_state_shutdown_callbacks+0xce/0xf0 [xfs]
xlog_force_shutdown+0xdf/0x150 [xfs]
xfs_do_force_shutdown+0x5f/0x150 [xfs]
xlog_ioend_work+0x71/0x80 [xfs]
process_one_work+0x1c5/0x390
worker_thread+0x30/0x350
kthread+0xd7/0x100
ret_from_fork+0x1f/0x30
This is processing an EIO error to a log write, and it's
triggering a force shutdown. This causes the log to be shut down,
and then it is running attached iclog callbacks from the shutdown
context. That means the fs and log has already been marked as
xfs_is_shutdown/xlog_is_shutdown and so high level code will abort
(e.g. xfs_trans_commit(), xfs_log_force(), etc) with an error
because of shutdown.
The umount would have been blocked waiting for a log force
completion inside xfs_log_cover() -> xfs_sync_sb(). The first thing
for this situation to occur is for xfs_sync_sb() to exit without
waiting for the iclog buffer to be comitted to disk. The
above trace is the completion routine for the iclog buffer, and
it is shutting down the filesystem.
xlog_state_shutdown_callbacks() does this:
{
struct xlog_in_core *iclog;
LIST_HEAD(cb_list);
spin_lock(&log->l_icloglock);
iclog = log->l_iclog;
do {
if (atomic_read(&iclog->ic_refcnt)) {
/* Reference holder will re-run iclog callbacks. */
continue;
}
list_splice_init(&iclog->ic_callbacks, &cb_list);
>>>>>> wake_up_all(&iclog->ic_write_wait);
>>>>>> wake_up_all(&iclog->ic_force_wait);
} while ((iclog = iclog->ic_next) != log->l_iclog);
wake_up_all(&log->l_flush_wait);
spin_unlock(&log->l_icloglock);
>>>>>> xlog_cil_process_committed(&cb_list);
}
This wakes any thread waiting on IO completion of the iclog (in this
case the umount log force) before shutdown processes all the pending
callbacks. That means the xfs_sync_sb() waiting on a sync
transaction in xfs_log_force() on iclog->ic_force_wait will get
woken before the callbacks attached to that iclog are run. This
results in xfs_sync_sb() returning an error, and so unmount unblocks
and continues to run whilst the log shutdown is still in progress.
Normally this is just fine because the force waiter has nothing to
do with AIL operations. But in the case of this unmount path, the
log force waiter goes on to tear down the AIL because the log is now
shut down and so nothing ever blocks it again from the wait point in
xfs_log_cover().
Hence it's a race to see who gets to the AIL first - the unmount
code or xlog_cil_process_committed() killing the superblock buffer.
To fix this, we just have to change the order of processing in
xlog_state_shutdown_callbacks() to run the callbacks before it wakes
any task waiting on completion of the iclog.
Reported-by: Brian Foster <[email protected]>
Fixes: aad7272a9208 ("xfs: separate out log shutdown callback processing")
Signed-off-by: Dave Chinner <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
|
|
generic/388 triggered a failure in RUI recovery due to a corrupted
btree record and the system then locked up hard due to a subsequent
assert failure while holding a spinlock cancelling intents:
XFS (pmem1): Corruption of in-memory data (0x8) detected at xfs_do_force_shutdown+0x1a/0x20 (fs/xfs/xfs_trans.c:964). Shutting down filesystem.
XFS (pmem1): Please unmount the filesystem and rectify the problem(s)
XFS: Assertion failed: !xlog_item_is_intent(lip), file: fs/xfs/xfs_log_recover.c, line: 2632
Call Trace:
<TASK>
xlog_recover_cancel_intents.isra.0+0xd1/0x120
xlog_recover_finish+0xb9/0x110
xfs_log_mount_finish+0x15a/0x1e0
xfs_mountfs+0x540/0x910
xfs_fs_fill_super+0x476/0x830
get_tree_bdev+0x171/0x270
? xfs_init_fs_context+0x1e0/0x1e0
xfs_fs_get_tree+0x15/0x20
vfs_get_tree+0x24/0xc0
path_mount+0x304/0xba0
? putname+0x55/0x60
__x64_sys_mount+0x108/0x140
do_syscall_64+0x35/0x80
entry_SYSCALL_64_after_hwframe+0x44/0xae
Essentially, there's dirty metadata in the AIL from intent recovery
transactions, so when we go to cancel the remaining intents we assume
that all objects after the first non-intent log item in the AIL are
not intents.
This is not true. Intent recovery can log new intents to continue
the operations the original intent could not complete in a single
transaction. The new intents are committed before they are deferred,
which means if the CIL commits in the background they will get
inserted into the AIL at the head.
Hence if we shut down the filesystem while processing intent
recovery, the AIL may have new intents active at the current head.
Hence this check:
/*
* We're done when we see something other than an intent.
* There should be no intents left in the AIL now.
*/
if (!xlog_item_is_intent(lip)) {
#ifdef DEBUG
for (; lip; lip = xfs_trans_ail_cursor_next(ailp, &cur))
ASSERT(!xlog_item_is_intent(lip));
#endif
break;
}
in both xlog_recover_process_intents() and
log_recover_cancel_intents() is simply not valid. It was valid back
when we only had EFI/EFD intents and didn't chain intents, but it
hasn't been valid ever since intent recovery could create and commit
new intents.
Given that crashing the mount task like this pretty much prevents
diagnosing what went wrong that lead to the initial failure that
triggered intent cancellation, just remove the checks altogether.
Signed-off-by: Dave Chinner <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
|
|
Most buffer io list operations are run with the bp->b_lock held, but
xfs_iflush_abort() can be called without the buffer lock being held
resulting in inodes being removed from the buffer list while other
list operations are occurring. This causes problems with corrupted
bp->b_io_list inode lists during filesystem shutdown, leading to
traversals that never end, double removals from the AIL, etc.
Fix this by passing the buffer to xfs_iflush_abort() if we have
it locked. If the inode is attached to the buffer, we're going to
have to remove it from the buffer list and we'd have to get the
buffer off the inode log item to do that anyway.
If we don't have a buffer passed in (e.g. from xfs_reclaim_inode())
then we can determine if the inode has a log item and if it is
attached to a buffer before we do anything else. If it does have an
attached buffer, we can lock it safely (because the inode has a
reference to it) and then perform the inode abort.
Signed-off-by: Dave Chinner <[email protected]>
Reviewed-by: Darrick J. Wong <[email protected]>
Signed-off-by: Darrick J. Wong <[email protected]>
|
|
Pull jfs updates from Dave Kleikamp:
"A couple bug fixes"
* tag 'jfs-5.18' of https://github.com/kleikamp/linux-shaggy:
jfs: prevent NULL deref in diFree
jfs: fix divide error in dbNextAG
|
|
SM8150 has an ethernet controller and it needs a different
configuration, so add a new compatible for this.
Acked-by: Rob Herring <[email protected]>
Signed-off-by: Vinod Koul <[email protected]>
[bhsharma: Massage the commit log]
Signed-off-by: Bhupesh Sharma <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
The "test_dev" pointer is freed but then returned to the caller.
Fixes: d9c6a72d6fa2 ("kmod: add test driver to stress test the module loader")
Signed-off-by: Dan Carpenter <[email protected]>
Signed-off-by: Luis Chamberlain <[email protected]>
|
|
This has always been the rule: fdtables have several bitmaps in them,
and as a result they have to be sized properly for bitmaps. We walk
those bitmaps in chunks of 'unsigned long' in serveral cases, but even
when we don't, we use the regular kernel bitops that are defined to work
on arrays of 'unsigned long', not on some byte array.
Now, the distinction between arrays of bytes and 'unsigned long'
normally only really ends up being noticeable on big-endian systems, but
Fedor Pchelkin and Alexey Khoroshilov reported that copy_fd_bitmaps()
could be called with an argument that wasn't even a multiple of
BITS_PER_BYTE. And then it fails to do the proper copy even on
little-endian machines.
The bug wasn't in copy_fd_bitmap(), but in sane_fdtable_size(), which
didn't actually sanitize the fdtable size sufficiently, and never made
sure it had the proper BITS_PER_LONG alignment.
That's partly because the alignment historically came not from having to
explicitly align things, but simply from previous fdtable sizes, and
from count_open_files(), which counts the file descriptors by walking
them one 'unsigned long' word at a time and thus naturally ends up doing
sizing in the proper 'chunks of unsigned long'.
But with the introduction of close_range(), we now have an external
source of "this is how many files we want to have", and so
sane_fdtable_size() needs to do a better job.
This also adds that explicit alignment to alloc_fdtable(), although
there it is mainly just for documentation at a source code level. The
arithmetic we do there to pick a reasonable fdtable size already aligns
the result sufficiently.
In fact,clang notices that the added ALIGN() in that function doesn't
actually do anything, and does not generate any extra code for it.
It turns out that gcc ends up confusing itself by combining a previous
constant-sized shift operation with the variable-sized shift operations
in roundup_pow_of_two(). And probably due to that doesn't notice that
the ALIGN() is a no-op. But that's a (tiny) gcc misfeature that doesn't
matter. Having the explicit alignment makes sense, and would actually
matter on a 128-bit architecture if we ever go there.
This also adds big comments above both functions about how fdtable sizes
have to have that BITS_PER_LONG alignment.
Fixes: 60997c3d45d9 ("close_range: add CLOSE_RANGE_UNSHARE")
Reported-by: Fedor Pchelkin <[email protected]>
Reported-by: Alexey Khoroshilov <[email protected]>
Link: https://lore.kernel.org/all/[email protected]/
Tested-and-acked-by: Christian Brauner <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
On ELF, (NOLOAD) sets the section type to SHT_NOBITS[1]. It is conceptually
inappropriate for .plt, .got, and .got.plt sections which are always
SHT_PROGBITS.
In GNU ld, if PLT entries are needed, .plt will be SHT_PROGBITS anyway
and (NOLOAD) will be essentially ignored. In ld.lld, since
https://reviews.llvm.org/D118840 ("[ELF] Support (TYPE=<value>) to
customize the output section type"), ld.lld will report a `section type
mismatch` error (later changed to a warning). Just remove (NOLOAD) to
fix the warning.
[1] https://lld.llvm.org/ELF/linker_script.html As of today, "The
section should be marked as not loadable" on
https://sourceware.org/binutils/docs/ld/Output-Section-Type.html is
outdated for ELF.
Link: https://github.com/ClangBuiltLinux/linux/issues/1597
Fixes: ab1ef68e5401 ("RISC-V: Add sections of PLT and GOT for kernel module")
Reported-by: Nathan Chancellor <[email protected]>
Signed-off-by: Fangrui Song <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
|
|
Clang static analysis reports this issue
interface.c:810:8: warning: Passed-by-value struct
argument contains uninitialized data
now = rtc_tm_to_ktime(tm);
^~~~~~~~~~~~~~~~~~~
tm is set by a successful call to __rtc_read_time()
but its return status is not checked. Check if
it was successful before setting the enabled flag.
Move the decl of err to function scope.
Fixes: 2b2f5ff00f63 ("rtc: interface: ignore expired timers when enqueuing new timers")
Signed-off-by: Tom Rix <[email protected]>
Signed-off-by: Alexandre Belloni <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
|
|
The of_find_compatible_node() function returns a node pointer with
refcount incremented, We should use of_node_put() on it when done
Add the missing of_node_put() to release the refcount.
Fixes: 86559400b3ef ("rtc: gamecube: Add a RTC driver for the GameCube, Wii and Wii U")
Signed-off-by: Miaoqian Lin <[email protected]>
Signed-off-by: Alexandre Belloni <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
|
|
Setting the century forward has been failing on AMD platforms.
There was a previous attempt at fixing this for family 0x17 as part of
commit 7ad295d5196a ("rtc: Fix the AltCentury value on AMD/Hygon
platform") but this was later reverted due to some problems reported
that appeared to stem from an FW bug on a family 0x17 desktop system.
The same comments mentioned in the previous commit continue to apply
to the newer platforms as well.
```
MC146818 driver use function mc146818_set_time() to set register
RTC_FREQ_SELECT(RTC_REG_A)'s bit4-bit6 field which means divider stage
reset value on Intel platform to 0x7.
While AMD/Hygon RTC_REG_A(0Ah)'s bit4 is defined as DV0 [Reference]:
DV0 = 0 selects Bank 0, DV0 = 1 selects Bank 1. Bit5-bit6 is defined
as reserved.
DV0 is set to 1, it will select Bank 1, which will disable AltCentury
register(0x32) access. As UEFI pass acpi_gbl_FADT.century 0x32
(AltCentury), the CMOS write will be failed on code:
CMOS_WRITE(century, acpi_gbl_FADT.century).
Correct RTC_REG_A bank select bit(DV0) to 0 on AMD/Hygon CPUs, it will
enable AltCentury(0x32) register writing and finally setup century as
expected.
```
However in closer examination the change previously submitted was also
modifying bits 5 & 6 which are declared reserved in the AMD documentation.
So instead modify just the DV0 bank selection bit.
Being cognizant that there was a failure reported before, split the code
change out to a static function that can also be used for exclusions if
any regressions such as Mikhail's pop up again.
Cc: Jinke Fan <[email protected]>
Cc: Mikhail Gavrilov <[email protected]>
Link: https://lore.kernel.org/all/CABXGCsMLob0DC25JS8wwAYydnDoHBSoMh2_YLPfqm3TTvDE-Zw@mail.gmail.com/
Link: https://www.amd.com/system/files/TechDocs/51192_Bolton_FCH_RRG.pdf
Signed-off-by: Raul E Rangel <[email protected]>
Signed-off-by: Mario Limonciello <[email protected]>
Signed-off-by: Alexandre Belloni <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
|
|
In preparation for not using the file at prep time, defer checking if this
file refers to a valid io_uring instance until issue time.
Signed-off-by: Jens Axboe <[email protected]>
|
|
This change fixes the following:
1) The flags variable is not initialized. Always use raw_spin_lock_irqsave
and raw_spin_unlock_irqrestore to serialize patching.
2) flush_kernel_vmap_range is primarily intended for DMA flushes. Since
__patch_text_multiple is often called with interrupts disabled, it is
better to directly call flush_kernel_dcache_range_asm and
flush_kernel_icache_range_asm. This avoids an extra call.
3) The final call to flush_icache_range is unnecessary.
Signed-off-by: John David Anglin <[email protected]>
Signed-off-by: Helge Deller <[email protected]>
|
|
When CPU hotplugging is enabled, the user may want to remove the
current CPU which is providing the timer ticks. If this happens
we need to find a new timesync master.
Signed-off-by: Helge Deller <[email protected]>
|
|
Move the common_stext function into the non-init text section if hotplug
is enabled. This function is called from the firmware when hotplugged
CPUs are brought up.
Signed-off-by: Helge Deller <[email protected]>
|
|
Let the PDC firmware put the CPU into firmware idle loop with the
pdc_cpu_rendezvous() function.
Signed-off-by: Helge Deller <[email protected]>
|
|
Add relevant code to __cpu_die() and __cpu_disable() to finally enable
the CPU hotplugging features. Reset the irq count values in smp_callin()
to zero before bringing up the CPU.
It seems that the firmware may need up to 8 seconds to fully stop a CPU
in which no other PDC calls are allowed to be made. Use a timeout
__cpu_die() to accommodate for this.
Use "chcpu -d 1" to bring CPU1 down, and "chcpu -e 1" to bring it up.
Signed-off-by: Helge Deller <[email protected]>
|
|
Add pdc_cpu_rendezvous_lock() and pdc_cpu_rendezvous_unlock()
to lock PDC while CPU is transitioning into rendezvous state.
This is needed, because the transition phase may take up to 8 seconds.
Add pdc_pat_get_PDC_entrypoint() to get PDC entry point for current CPU.
Signed-off-by: Helge Deller <[email protected]>
|
|
Signed-off-by: Helge Deller <[email protected]>
|
|
If CONFIG_HOTPLUG_CPU is enabled, those functions will be run again
after bootup. So they need to reside in the .text section.
Signed-off-by: Helge Deller <[email protected]>
|
|
Signed-off-by: Helge Deller <[email protected]>
|
|
Switch away from the own cpu topology code to common code which is used
by ARM64 and RISCV. That will allow us to enable CPU hotplug later on.
Signed-off-by: Helge Deller <[email protected]>
|
|
Call set_firmware_width() only once at runtime.
This prevents that hotplugged CPUs will get stuck in spinlocks later on.
Signed-off-by: Helge Deller <[email protected]>
|
|
Clean up the code for the mfctl() and mtctl() functions and add often
used constants.
Signed-off-by: Helge Deller <[email protected]>
|
|
Allow the system to find the SUSE hppa compiler and linker to set
CROSS32_COMPILE and CROSS_COMPILE.
Suggested-by: Jiri Slaby <[email protected]>
Signed-off-by: Helge Deller <[email protected]>
|
|
The cpu_set_affinity_irq() isn't needed. Not the CPU irqs need to
change, but the slave irq chips simply need to be reprogrammed to
a new CPU irq with the txn_* functions.
Signed-off-by: Helge Deller <[email protected]>
|
|
Add the missing logic to allow Lasi, WAX and Dino to set the
CPU affinity. This fixes IRQ migration to other CPUs when a
CPU is shutdown which currently holds the IRQs for one of those
chips.
Signed-off-by: Helge Deller <[email protected]>
|