blk-wbt: call rq_qos_add() after wb_normal is initialized
Our test found a problem that wbt inflight counter is negative, which
will cause io hang(noted that this problem doesn't exist in mainline):
t1: device create t2: issue io
add_disk
blk_register_queue
wbt_enable_default
wbt_init
rq_qos_add
// wb_normal is still 0
/*
* in mainline, disk can't be opened before
* bdev_add(), however, in old kernels, disk
* can be opened before blk_register_queue().
*/
blkdev_issue_flush
// disk size is 0, however, it's not checked
submit_bio_wait
submit_bio
blk_mq_submit_bio
rq_qos_throttle
wbt_wait
bio_to_wbt_flags
rwb_enabled
// wb_normal is 0, inflight is not increased
wbt_queue_depth_changed(&rwb->rqos);
wbt_update_limits
// wb_normal is initialized
rq_qos_track
wbt_track
rq->wbt_flags |= bio_to_wbt_flags(rwb, bio);
// wb_normal is not 0，wbt_flags will be set
t3: io completion
blk_mq_free_request
rq_qos_done
wbt_done
wbt_is_tracked
// return true
__wbt_done
wbt_rqw_done
atomic_dec_return(&rqw->inflight);
// inflight is decreased
commit 8235b5c1e8c1 ("block: call bdev_add later in device_add_disk") can
avoid this problem, however it's better to fix this problem in wbt:
1) Lower kernel can't backport this patch due to lots of refactor.
2) Root cause is that wbt call rq_qos_add() before wb_normal is
initialized.
Given that rnbd_srv_sess_dev already has an open_flags member, there
is no need for the rnbd_dev indirection as a simple block_device pointer
works just as well.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com> Acked-by: Jack Wang <jinpu.wang@ionos.com> Link: https://lore.kernel.org/r/20220909131509.3263924-5-hch@lst.de Signed-off-by: Jens Axboe <axboe@kernel.dk>
block: Fix the enum blk_eh_timer_return documentation
The documentation of the blk_eh_timer_return enumeration values does not
reflect correctly how e.g. the SCSI core uses these values. Fix the
documentation.
Cc: Christoph Hellwig <hch@lst.de> Cc: Ming Lei <ming.lei@redhat.com> Cc: Hannes Reinecke <hare@suse.de> Cc: Damien Le Moal <damien.lemoal@wdc.com> Cc: Johannes Thumshirn <johannes.thumshirn@wdc.com> Fixes: 88b0cfad2888 ("block: document the blk_eh_timer_return values") Signed-off-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com> Link: https://lore.kernel.org/r/20220920200626.3422296-1-bvanassche@acm.org Signed-off-by: Jens Axboe <axboe@kernel.dk>
Stefan Haberland [Tue, 20 Sep 2022 19:26:16 +0000 (21:26 +0200)]
s390/dasd: add device ping attribute
Add a function to check if a device is accessible.
This makes mostly sense for copy pair secondary devices but it will work
for all devices.
The sysfs attribute ping is a write only attribute and will issue a NOP
CCW to the device.
In case of success it will return zero. If the device is not accessible
it will return an error code.
Stefan Haberland [Tue, 20 Sep 2022 19:26:15 +0000 (21:26 +0200)]
s390/dasd: suppress generic error messages for PPRC secondary devices
Suppress generic command reject messages and dump of sense data for
Peer-To-Peer-Remote-Copy (PPRC) secondary errors.
If IO is issued on a PPRC secondary device, a specific
error message is printed instead.
Stefan Haberland [Tue, 20 Sep 2022 19:26:14 +0000 (21:26 +0200)]
s390/dasd: add ioctl to perform a swap of the drivers copy pair
The newly defined ioctl BIODASDCOPYPAIRSWAP takes a structure that
specifies a copy pair that should be swapped. It will call the device
discipline function to perform the swap operation.
Stefan Haberland [Tue, 20 Sep 2022 19:26:13 +0000 (21:26 +0200)]
s390/dasd: add copy pair swap capability
In case of errors or misbehaviour of the primary device a controlled
failover to one of the configured secondary devices needs to be
performed.
The swap processing stops I/O on the primary device, all requests are
re-queued to the blocklayer queue, the entries in the copy relation are
swapped and finally the link to the blockdevice is moved from primary to
secondary dasd device.
After this, the secondary becomes the new primary device and I/O is
restarted on that device.
Stefan Haberland [Tue, 20 Sep 2022 19:26:12 +0000 (21:26 +0200)]
s390/dasd: add copy pair setup
A copy relation that is configured on the storage server side needs to be
enabled separately in the device driver. A sysfs interface is created
that allows userspace tooling to control such setup.
The following sysfs entries are added to store and read copy relation
information:
copy_pair
- Add/Delete a copy pair relation to the DASD device driver
- Query all previously added copy pair relations
copy_role
- Query the copy pair role of the device
To add a copy pair to the DASD device driver it has to be specified
through the sysfs attribute copy_pair. Only one secondary device can be
specified at a time together with the primary device. Both, secondary
and primary can be used equally to define the copy pair.
The secondary devices have to be offline when adding the copy relation.
The primary device needs to be specified first followed by the comma
separated secondary device.
Read from the copy_pair attribute to get the current setup and write
"clear" to the attribute to delete any existing setup.
During device online processing the required data will be read from the
storage server and the information will be compared to the setup
requested through the copy_pair attribute. The registration of the
primary and secondary device will be handled accordingly.
A blockdevice is only allocated for copy relation primary devices.
To query the copy role of a device read from the copy_role sysfs
attribute. Possible values are primary, secondary, and none.
erofs: add manual PSI accounting for the compressed address space
erofs uses an additional address space for compressed data read from disk
in addition to the one directly associated with the inode. Reading into
the lower address space is open coded using add_to_page_cache_lru instead
of using the filemap.c helper for page allocation micro-optimizations,
which means it is not covered by the MM PSI annotations for ->read_folio
and ->readahead, so add manual ones instead.
Signed-off-by: Christoph Hellwig <hch@lst.de> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Gao Xiang <hsiangkao@linux.alibaba.com> Link: https://lore.kernel.org/r/20220915094200.139713-5-hch@lst.de Signed-off-by: Jens Axboe <axboe@kernel.dk>
btrfs: add manual PSI accounting for compressed reads
btrfs compressed reads try to always read the entire compressed chunk,
even if only a subset is requested. Currently this is covered by the
magic PSI accounting underneath submit_bio, but that is about to go
away. Instead add manual psi_memstall_{enter,leave} annotations.
Note that for readahead this really should be using readahead_expand,
but the additionals reads are also done for plain ->read_folio where
readahead_expand can't work, so this overall logic is left as-is for
now.
Signed-off-by: Christoph Hellwig <hch@lst.de> Acked-by: David Sterba <dsterba@suse.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Link: https://lore.kernel.org/r/20220915094200.139713-4-hch@lst.de Signed-off-by: Jens Axboe <axboe@kernel.dk>
mm: add PSI accounting around ->read_folio and ->readahead calls
PSI tries to account for the cost of bringing back in pages discarded by
the MM LRU management. Currently the prime place for that is hooked into
the bio submission path, which is a rather bad place:
- it does not actually account I/O for non-block file systems, of which
we have many
- it adds overhead and a layering violation to the block layer
Add the accounting into the two places in the core MM code that read
pages into an address space by calling into ->read_folio and ->readahead
so that the entire file system operations are covered, to broaden
the coverage and allow removing the accounting in the block layer going
forward.
As psi_memstall_enter can deal with nested calls this will not lead to
double accounting even while the bio annotations are still present.
Merge tag 'nvme-6.1-2022-09-20' of git://git.infradead.org/nvme into for-6.1/block
Pull NVMe updates from Christoph:
"nvme updates for Linux 6.1
- handle number of queue changes in the TCP and RDMA drivers
(Daniel Wagner)
- allow changing the number of queues in nvmet (Daniel Wagner)
- also consider host_iface when checking ip options (Daniel Wagner)
- don't map pages which can't come from HIGHMEM (Fabio M. De Francesco)
- avoid unnecessary flush bios in nvmet (Guixin Liu)
- shrink and better pack the nvme_iod structure (Keith Busch)
- add comment for unaligned "fake" nqn (Linjun Bao)
- print actual source IP address through sysfs "address" attr
(Martin Belanger)
- various cleanups (Jackie Liu, Wolfram Sang, Genjian Zhang)"
* tag 'nvme-6.1-2022-09-20' of git://git.infradead.org/nvme:
nvme-tcp: print actual source IP address through sysfs "address" attr
nvmet-tcp: don't map pages which can't come from HIGHMEM
nvme-pci: move iod dma_len fill gaps
nvme-pci: iod npages fits in s8
nvme-pci: iod's 'aborted' is a bool
nvme-pci: remove nvme_queue from nvme_iod
nvme: consider also host_iface when checking ip options
nvme-rdma: handle number of queue changes
nvme-tcp: handle number of queue changes
nvmet: expose max queues to configfs
nvmet: avoid unnecessary flush bio
nvmet-auth: remove redundant parameters req
nvmet-auth: clean up with done_kfree
nvme-auth: remove the redundant req->cqe->result.u16 assignment operation
nvme: move from strlcpy with unused retval to strscpy
nvme: add comment for unaligned "fake" nqn
Coly Li [Mon, 19 Sep 2022 16:16:47 +0000 (00:16 +0800)]
bcache: fix set_at_max_writeback_rate() for multiple attached devices
Inside set_at_max_writeback_rate() the calculation in following if()
check is wrong,
if (atomic_inc_return(&c->idle_counter) <
atomic_read(&c->attached_dev_nr) * 6)
Because each attached backing device has its own writeback thread
running and increasing c->idle_counter, the counter increates much
faster than expected. The correct calculation should be,
(counter / dev_nr) < dev_nr * 6
which equals to,
counter < dev_nr * dev_nr * 6
This patch fixes the above mistake with correct calculation, and helper
routine idle_counter_exceeded() is added to make code be more clear.
Reported-by: Mingzhe Zou <mingzhe.zou@easystack.cn> Signed-off-by: Coly Li <colyli@suse.de> Acked-by: Mingzhe Zou <mingzhe.zou@easystack.cn> Link: https://lore.kernel.org/r/20220919161647.81238-6-colyli@suse.de Signed-off-by: Jens Axboe <axboe@kernel.dk>
Lin Feng [Mon, 19 Sep 2022 16:16:44 +0000 (00:16 +0800)]
bcache: remove unused bch_mark_cache_readahead function def in stats.h
This is a cleanup for commit 1616a4c2ab1a ("bcache: remove bcache device
self-defined readahead")', currently no user for
bch_mark_cache_readahead() since that commit.
Martin Belanger [Wed, 7 Sep 2022 12:27:37 +0000 (08:27 -0400)]
nvme-tcp: print actual source IP address through sysfs "address" attr
TCP transport relies on the routing table to determine which source
address and interface to use when making a connection. Currently, there
is no way to tell from userspace where a connection was made. This
patch exposes the actual source address using a new field named
"src_addr=" in the "address" attribute.
This is needed to diagnose and identify connectivity issues. With the
source address we can infer the interface associated with each
connection.
This was tested with nvme-cli 2.0 to verify it does not have any
adverse effect. The new "src_addr=" field will simply be displayed in
the output of the "list-subsys" or "list -v" commands as shown here.
nvmet-tcp: don't map pages which can't come from HIGHMEM
kmap() is being deprecated in favor of kmap_local_page().[1]
There are two main problems with kmap(): (1) It comes with an overhead as
mapping space is restricted and protected by a global lock for
synchronization and (2) it also requires global TLB invalidation when the
kmap’s pool wraps and it might block when the mapping space is fully
utilized until a slot becomes available.
The pages which will be mapped are allocated in nvmet_tcp_map_data(),
using the GFP_KERNEL flag. This assures that they cannot come from
HIGHMEM. This imply that a straight page_address() can replace the kmap()
of sg_page(sg) in nvmet_tcp_map_pdu_iovec(). As a side effect, we might
also delete the field "nr_mapped" from struct "nvmet_tcp_cmd" because,
after removing the kmap() calls, there would be no longer any need of it.
In addition, there is no reason to use a kvec for the command receive
data buffers iovec, use a bio_vec instead and let iov_iter handle the
buffer mapping and data copy.
Test with blktests on a QEMU/KVM x86_32 VM, 6GB RAM, booting a kernel with
HIGHMEM64GB enabled.
[1] "[PATCH] checkpatch: Add kmap and kmap_atomic to the deprecated
list" https://lore.kernel.org/all/20220813220034.806698-1-ira.weiny@intel.com/
Cc: Chaitanya Kulkarni <chaitanyak@nvidia.com> Cc: Keith Busch <kbusch@kernel.org> Suggested-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Fabio M. De Francesco <fmdefrancesco@gmail.com> Suggested-by: Christoph Hellwig <hch@lst.de> Suggested-by: Al Viro <viro@zeniv.linux.org.uk>
[sagi: added bio_vec plus minor naming changes] Signed-off-by: Sagi Grimberg <sagi@grimberg.me> Signed-off-by: Christoph Hellwig <hch@lst.de>
Keith Busch [Tue, 6 Sep 2022 16:07:37 +0000 (09:07 -0700)]
nvme-pci: iod npages fits in s8
The largest allowed transfer is 4MB, which can use at most 1025 PRPs.
Each PRP is 8 bytes, so the maximum number of 4k nvme pages needed for
the iod_list is 3, which fits in an 's8' type.
While modifying this field, change the name to "nr_allocations" to
better represent that this is referring to the number of units allocated
from a dma_pool.
Also introduce a BUILD_BUG_ON to ensure we never accidently increase the
largest transfer limit beyond 127 chained prp lists.
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com> Signed-off-by: Keith Busch <kbusch@kernel.org> Signed-off-by: Christoph Hellwig <hch@lst.de>
Daniel Wagner [Fri, 29 Jul 2022 14:26:30 +0000 (16:26 +0200)]
nvme: consider also host_iface when checking ip options
It's perfectly fine to use the same traddr and trsvcid more than once
as long we use different host interface. This is used in setups where
the host has more than one interface but the target exposes only one
traddr/trsvcid combination.
Use the same acceptance rules for host_iface as we have for
host_traddr.
Signed-off-by: Daniel Wagner <dwagner@suse.de> Reviewed-by: Chao Leng <lengchao@huawei.com> Signed-off-by: Christoph Hellwig <hch@lst.de>
Daniel Wagner [Mon, 29 Aug 2022 09:28:41 +0000 (11:28 +0200)]
nvme-rdma: handle number of queue changes
On reconnect, the number of queues might have changed.
In the case where we have more queues available than previously we try
to access queues which are not initialized yet.
The other case where we have less queues than previously, the
connection attempt will fail because the target doesn't support the
old number of queues and we end up in a reconnect loop.
Thus, only start queues which are currently present in the tagset
limited by the number of available queues. Then we update the tagset
and we can start any new queue.
Signed-off-by: Daniel Wagner <dwagner@suse.de> Reviewed-by: Sagi Grimberg <sagi@grimberg.me> Reviewed-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Christoph Hellwig <hch@lst.de>
Daniel Wagner [Mon, 29 Aug 2022 09:28:40 +0000 (11:28 +0200)]
nvme-tcp: handle number of queue changes
On reconnect, the number of queues might have changed.
In the case where we have more queues available than previously we try
to access queues which are not initialized yet.
The other case where we have less queues than previously, the
connection attempt will fail because the target doesn't support the
old number of queues and we end up in a reconnect loop.
Thus, only start queues which are currently present in the tagset
limited by the number of available queues. Then we update the tagset
and we can start any new queue.
Signed-off-by: Daniel Wagner <dwagner@suse.de> Reviewed-by: Sagi Grimberg <sagi@grimberg.me> Reviewed-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Christoph Hellwig <hch@lst.de>
Daniel Wagner [Mon, 29 Aug 2022 09:28:39 +0000 (11:28 +0200)]
nvmet: expose max queues to configfs
Allow to set the max queues the target supports. This is useful for
testing the reconnect attempt of the host with changing numbers of
supported queues.
Signed-off-by: Daniel Wagner <dwagner@suse.de> Reviewed-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Christoph Hellwig <hch@lst.de>
Wolfram Sang [Thu, 18 Aug 2022 21:00:52 +0000 (23:00 +0200)]
nvme: move from strlcpy with unused retval to strscpy
Follow the advice of the below link and prefer 'strscpy' in this
subsystem. Conversion is 1:1 because the return value is not used.
Generated by a coccinelle script.
Yu Kuai [Sat, 27 Aug 2022 10:16:37 +0000 (18:16 +0800)]
blk-throttle: cleanup tg_update_disptime()
tg_update_disptime() only need to adjust postion for 'tg' in
'parent_sq', there is no need to call throtl_enqueue/dequeue_tg(),
since they will set/clear flag THROTL_TG_PENDING and increase/decrease
nr_pending, which is useless. By the way, clear the flag/decrease
nr_pending while there are still throttled bios is not good for debugging.
Yu Kuai [Sat, 27 Aug 2022 10:16:36 +0000 (18:16 +0800)]
blk-throttle: calling throtl_dequeue/enqueue_tg in pairs
It's a little weird to call throtl_dequeue_tg() unconditionally in
throtl_select_dispatch(), since it will be called in tg_update_disptime()
again if some bio is still throttled. Thus call it later if there are no
throttled bio. There are no functional changes.
Yu Kuai [Mon, 29 Aug 2022 02:22:40 +0000 (10:22 +0800)]
blk-throttle: fix io hung due to configuration updates
If new configuration is submitted while a bio is throttled, then new
waiting time is recalculated regardless that the bio might already wait
for some time:
Then io hung can be triggered by always submmiting new configuration
before the throttled bio is dispatched.
Fix the problem by respecting the time that throttled bio already waited.
In order to do that, add new fields to record how many bytes/io are
waited, and use it to calculate wait time for throttled bio under new
configuration.
Yu Kuai [Mon, 29 Aug 2022 02:22:38 +0000 (10:22 +0800)]
blk-throttle: prevent overflow while calculating wait time
There is a problem found by code review in tg_with_in_bps_limit() that
'bps_limit * jiffy_elapsed_rnd' might overflow. Fix the problem by
calling mul_u64_u64_div_u64() instead.
The problem is that the second bio is finished after 10s instead of 20s.
Root cause:
1) second bio will be flagged:
__blk_throtl_bio
while (true) {
...
if (sq->nr_queued[rw]) -> some bio is throttled already
break
};
bio_set_flag(bio, BIO_THROTTLED); -> flag the bio
2) flagged bio will be dispatched without waiting:
throtl_dispatch_tg
tg_may_dispatch
tg_with_in_bps_limit
if (bps_limit == U64_MAX || bio_flagged(bio, BIO_THROTTLED))
*wait = 0; -> wait time is zero
return true;
commit 9f5ede3c01f9 ("block: throttle split bio in case of iops limit")
support to count split bios for iops limit, thus it adds flagged bio
checking in tg_with_in_bps_limit() so that split bios will only count
once for bps limit, however, it introduce a new problem that io throttle
won't work if multiple bios are throttled.
In order to fix the problem, handle iops/bps limit in different ways:
1) for iops limit, there is no flag to record if the bio is throttled,
and iops is always applied.
2) for bps limit, original bio will be flagged with BIO_BPS_THROTTLED,
and io throttle will ignore bio with the flag.
Noted this patch also remove the code to set flag in __bio_clone(), it's
introduced in commit 111be8839817 ("block-throttle: avoid double
charge"), and author thinks split bio can be resubmited and throttled
again, which is wrong because split bio will continue to dispatch from
caller.
Fixes: 9f5ede3c01f9 ("block: throttle split bio in case of iops limit") Cc: <stable@vger.kernel.org> Signed-off-by: Yu Kuai <yukuai3@huawei.com> Acked-by: Tejun Heo <tj@kernel.org> Link: https://lore.kernel.org/r/20220829022240.3348319-2-yukuai1@huaweicloud.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
Keith Busch [Fri, 9 Sep 2022 18:40:22 +0000 (11:40 -0700)]
sbitmap: fix batched wait_cnt accounting
Batched completions can clear multiple bits, but we're only decrementing
the wait_cnt by one each time. This can cause waiters to never be woken,
stalling IO. Use the batched count instead.
sbitmap: Use atomic_long_try_cmpxchg in __sbitmap_queue_get_batch
Use atomic_long_try_cmpxchg instead of
atomic_long_cmpxchg (*ptr, old, new) == old in __sbitmap_queue_get_batch.
x86 CMPXCHG instruction returns success in ZF flag, so this change
saves a compare after cmpxchg (and related move instruction in front
of cmpxchg).
Also, atomic_long_cmpxchg implicitly assigns old *ptr value to "old"
when cmpxchg fails, enabling further code simplifications, e.g.
an extra memory read can be avoided in the loop.
Shigeru Yoshida [Wed, 7 Sep 2022 16:35:02 +0000 (01:35 +0900)]
nbd: Fix hung when signal interrupts nbd_start_device_ioctl()
syzbot reported hung task [1]. The following program is a simplified
version of the reproducer:
int main(void)
{
int sv[2], fd;
if (socketpair(AF_UNIX, SOCK_STREAM, 0, sv) < 0)
return 1;
if ((fd = open("/dev/nbd0", 0)) < 0)
return 1;
if (ioctl(fd, NBD_SET_SIZE_BLOCKS, 0x81) < 0)
return 1;
if (ioctl(fd, NBD_SET_SOCK, sv[0]) < 0)
return 1;
if (ioctl(fd, NBD_DO_IT) < 0)
return 1;
return 0;
}
When signal interrupt nbd_start_device_ioctl() waiting the condition
atomic_read(&config->recv_threads) == 0, the task can hung because it
waits the completion of the inflight IOs.
This patch fixes the issue by clearing queue, not just shutdown, when
signal interrupt nbd_start_device_ioctl().
Jan Kara [Thu, 8 Sep 2022 13:09:37 +0000 (15:09 +0200)]
sbitmap: Avoid leaving waitqueue in invalid state in __sbq_wake_up()
When __sbq_wake_up() decrements wait_cnt to 0 but races with someone
else waking the waiter on the waitqueue (so the waitqueue becomes
empty), it exits without reseting wait_cnt to wake_batch number. Once
wait_cnt is 0, nobody will ever reset the wait_cnt or wake the new
waiters resulting in possible deadlocks or busyloops. Fix the problem by
making sure we reset wait_cnt even if we didn't wake up anybody in the
end.
Fixes: 040b83fcecfb ("sbitmap: fix possible io hung due to lost wakeup") Reported-by: Keith Busch <kbusch@kernel.org> Signed-off-by: Jan Kara <jack@suse.cz> Link: https://lore.kernel.org/r/20220908130937.2795-1-jack@suse.cz Signed-off-by: Jens Axboe <axboe@kernel.dk>
This is causing issues with CPU stalls on my test box, revert it for
now until we understand what is going on. It looks like infinite
looping off sbitmap_queue_wake_up(), but hard to tell with a lot of
CPUs hitting this issue and the console scrolling infinitely.
block: enable per-cpu bio caching for the fs bio set
This is useful for polled IO on a file, or for polled IO with the
io_uring passthrough mechanism. If bio allocations are done with
REQ_POLLED for those cases, then initializing the bio set with
BIOSET_PERCPU_CACHE enables the local per-cpu cache which eliminates
allocations (and frees) of bio structs when possible.
Keith Busch [Thu, 25 Aug 2022 14:53:12 +0000 (07:53 -0700)]
sbitmap: fix batched wait_cnt accounting
Batched completions can clear multiple bits, but we're only decrementing
the wait_cnt by one each time. This can cause waiters to never be woken,
stalling IO. Use the batched count instead.
Yu Kuai [Wed, 3 Aug 2022 12:15:04 +0000 (20:15 +0800)]
sbitmap: fix possible io hung due to lost wakeup
There are two problems can lead to lost wakeup:
1) invalid wakeup on the wrong waitqueue:
For example, 2 * wake_batch tags are put, while only wake_batch threads
are woken:
__sbq_wake_up
atomic_cmpxchg -> reset wait_cnt
__sbq_wake_up -> decrease wait_cnt
...
__sbq_wake_up -> wait_cnt is decreased to 0 again
atomic_cmpxchg
sbq_index_atomic_inc -> increase wake_index
wake_up_nr -> wake up and waitqueue might be empty
sbq_index_atomic_inc -> increase again, one waitqueue is skipped
wake_up_nr -> invalid wake up because old wakequeue might be empty
To fix the problem, increasing 'wake_index' before resetting 'wait_cnt'.
2) 'wait_cnt' can be decreased while waitqueue is empty
As pointed out by Jan Kara, following race is possible:
CPU1 CPU2
__sbq_wake_up __sbq_wake_up
sbq_wake_ptr() sbq_wake_ptr() -> the same
wait_cnt = atomic_dec_return()
/* decreased to 0 */
sbq_index_atomic_inc()
/* move to next waitqueue */
atomic_set()
/* reset wait_cnt */
wake_up_nr()
/* wake up on the old waitqueue */
wait_cnt = atomic_dec_return()
/*
* decrease wait_cnt in the old
* waitqueue, while it can be
* empty.
*/
Fix the problem by waking up before updating 'wake_index' and
'wait_cnt'.
With this patch, noted that 'wait_cnt' is still decreased in the old
empty waitqueue, however, the wakeup is redirected to a active waitqueue,
and the extra decrement on the old empty waitqueue is not handled.
Jens Axboe [Fri, 5 Aug 2022 22:44:34 +0000 (16:44 -0600)]
block: use on-stack page vec for <= UIO_FASTIOV
Avoid a kmalloc+kfree for each page array, if we only have a few pages
that are mapped. An alloc+free for each IO is quite expensive, and
it's pretty pointless if we're only dealing with 1 or a few vecs.
Use UIO_FASTIOV like we do in other spots to set a sane limit for how
big of an IO we want to avoid allocations for.
Jens Axboe [Fri, 5 Aug 2022 22:43:09 +0000 (16:43 -0600)]
block: enable bio caching use for passthru IO
bdev based polled O_DIRECT is currently quite a bit faster than
passthru on the same device, and one of the reaons is that we're not
able to use the bio caching for passthru IO.
If REQ_POLLED is set on the request, use the fs bio set for grabbing a
bio from the caches, if available. This saves 5-6% of CPU over head
for polled passthru IO.
Jens Axboe [Fri, 5 Aug 2022 22:39:04 +0000 (16:39 -0600)]
block: shrink rq_map_data a bit
We don't need full ints for several of these members. Change the
page_order and nr_entries to unsigned shorts, and the true/false from_user
and null_mapped to booleans.
This shrinks the struct from 32 to 24 bytes on 64-bit archs.
Bart Van Assche [Mon, 15 Aug 2022 17:00:43 +0000 (10:00 -0700)]
block: Change the return type of blk_mq_map_queues() into void
Since blk_mq_map_queues() and the .map_queues() callbacks always return 0,
change their return type into void. Most callers ignore the returned value
anyway.
Cc: Christoph Hellwig <hch@lst.de> Cc: Jason Wang <jasowang@redhat.com> Cc: Keith Busch <kbusch@kernel.org> Cc: Martin K. Petersen <martin.petersen@oracle.com> Cc: Doug Gilbert <dgilbert@interlog.com> Cc: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: John Garry <john.garry@huawei.com> Acked-by: Md Haris Iqbal <haris.iqbal@ionos.com> Reviewed-by: Sagi Grimberg <sagi@grimberg.me> Link: https://lore.kernel.org/r/20220815170043.19489-3-bvanassche@acm.org
[axboe: fold in fix from Bart] Signed-off-by: Jens Axboe <axboe@kernel.dk>
Bart Van Assche [Mon, 15 Aug 2022 17:00:42 +0000 (10:00 -0700)]
null_blk: Modify the behavior of null_map_queues()
Instead of returning -EINVAL if an internal inconsistency is detected,
fall back to a single submission queue. This patch prepares for changing
the return value of the .map_queues() callbacks into void.
Cc: Christoph Hellwig <hch@lst.de> Cc: Keith Busch <kbusch@kernel.org> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com> Link: https://lore.kernel.org/r/20220815170043.19489-2-bvanassche@acm.org Signed-off-by: Jens Axboe <axboe@kernel.dk>
Santosh Pradhan [Thu, 18 Aug 2022 10:55:51 +0000 (12:55 +0200)]
block/rnbd-srv: Add event tracing support
Add event tracing mechanism for following routines:
- create_sess()
- destroy_sess()
- process_rdma()
- process_msg_sess_info()
- process_msg_open()
- process_msg_close()
How to use:
1. Load the rnbd_server module
2. cd /sys/kernel/debug/tracing
3. If all the events need to be enabled:
echo 1 > events/rnbd_srv/enable
4. OR only speific routine/event needs to be enabled e.g.
echo 1 > events/rnbd_srv/create_sess/enable
5. cat trace
5. Run some workload which can trigger create_sess() routine/event
block: sed-opal: Add ioctl to return device status
Provide a mechanism to retrieve basic status information about
the device, including the "supported" flag indicating whether
SED-OPAL is supported. The information returned is from the various
feature descriptors received during the discovery0 step, and so
this ioctl does nothing more than perform the discovery0 step
and then save the information received. See "struct opal_status"
and OPAL_FL_* bits for the status information currently returned.
This is necessary to be able to check whether a device is OPAL
enabled, set up, locked or unlocked from userspace programs
like systemd-cryptsetup and libcryptsetup. Right now we just
have to assume the user 'knows' or blindly attempt setup/lock/unlock
operations.
Signed-off-by: Douglas Miller <dougmill@linux.vnet.ibm.com> Tested-by: Luca Boccassi <bluca@debian.org> Reviewed-by: Scott Bauer <sbauer@plzdonthack.me> Acked-by: Christian Brauner (Microsoft) <brauner@kernel.org> Link: https://lore.kernel.org/r/20220816140713.84893-1-luca.boccassi@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
Linus Torvalds [Sun, 21 Aug 2022 22:09:55 +0000 (15:09 -0700)]
Merge tag 'irq-urgent-2022-08-21' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull irq fixes from Ingo Molnar:
"Misc irqchip fixes: LoongArch driver fixes and a Hyper-V IOMMU fix"
* tag 'irq-urgent-2022-08-21' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
irqchip/loongson-liointc: Fix an error handling path in liointc_init()
irqchip/loongarch: Fix irq_domain_alloc_fwnode() abuse
irqchip/loongson-pch-pic: Move find_pch_pic() into CONFIG_ACPI
irqchip/loongson-eiointc: Fix a build warning
irqchip/loongson-eiointc: Fix irq affinity setting
iommu/hyper-v: Use helper instead of directly accessing affinity
Linus Torvalds [Sun, 21 Aug 2022 22:01:51 +0000 (15:01 -0700)]
Merge tag 'perf-urgent-2022-08-21' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 kprobes fix from Ingo Molnar:
"Fix a kprobes bug in JNG/JNLE emulation when a kprobe is installed at
such instructions, possibly resulting in incorrect execution (the
wrong branch taken)"
* tag 'perf-urgent-2022-08-21' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/kprobes: Fix JNG/JNLE emulation
Linus Torvalds [Sun, 21 Aug 2022 21:49:42 +0000 (14:49 -0700)]
Merge tag 'trace-v6.0-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull tracing fixes from Steven Rostedt:
"Various fixes for tracing:
- Fix a return value of traceprobe_parse_event_name()
- Fix NULL pointer dereference from failed ftrace enabling
- Fix NULL pointer dereference when asking for registers from eprobes
- Make eprobes consistent with kprobes/uprobes, filters and
histograms"
* tag 'trace-v6.0-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
tracing: Have filter accept "common_cpu" to be consistent
tracing/probes: Have kprobes and uprobes use $COMM too
tracing/eprobes: Have event probes be consistent with kprobes and uprobes
tracing/eprobes: Fix reading of string fields
tracing/eprobes: Do not hardcode $comm as a string
tracing/eprobes: Do not allow eprobes to use $stack, or % for regs
ftrace: Fix NULL pointer dereference in is_ftrace_trampoline when ftrace is dead
tracing/perf: Fix double put of trace event when init fails
tracing: React to error return from traceprobe_parse_event_name()
tracing: Have filter accept "common_cpu" to be consistent
Make filtering consistent with histograms. As "cpu" can be a field of an
event, allow for "common_cpu" to keep it from being confused with the
"cpu" field of the event.
tracing/probes: Have kprobes and uprobes use $COMM too
Both $comm and $COMM can be used to get current->comm in eprobes and the
filtering and histogram logic. Make kprobes and uprobes consistent in this
regard and allow both $comm and $COMM as well. Currently kprobes and
uprobes only handle $comm, which is inconsistent with the other utilities,
and can be confusing to users.
tracing/eprobes: Have event probes be consistent with kprobes and uprobes
Currently, if a symbol "@" is attempted to be used with an event probe
(eprobes), it will cause a NULL pointer dereference crash.
Both kprobes and uprobes can reference data other than the main registers.
Such as immediate address, symbols and the current task name. Have eprobes
do the same thing.
For "comm", if "comm" is used and the event being attached to does not
have the "comm" field, then make it the "$comm" that kprobes has. This is
consistent to the way histograms and filters work.
Link: https://lkml.kernel.org/r/20220820134401.136924220@goodmis.org Cc: stable@vger.kernel.org Cc: Ingo Molnar <mingo@kernel.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Tzvetomir Stoyanov <tz.stoyanov@gmail.com> Cc: Tom Zanussi <zanussi@kernel.org> Fixes: 7491e2c44278 ("tracing: Add a probe that attaches to trace events") Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Currently when an event probe (eprobe) hooks to a string field, it does
not display it as a string, but instead as a number. This makes the field
rather useless. Handle the different kinds of strings, dynamic, static,
relational/dynamic etc.
Now when a string field is used, the ":string" type can be used to display
it:
tracing/eprobes: Do not hardcode $comm as a string
The variable $comm is hard coded as a string, which is true for both
kprobes and uprobes, but for event probes (eprobes) it is a field name. In
most cases the "comm" field would be a string, but there's no guarantee of
that fact.
Do not assume that comm is a string. Not to mention, it currently forces
comm fields to fault, as string processing for event probes is currently
broken.
Link: https://lkml.kernel.org/r/20220820134400.756152112@goodmis.org Cc: stable@vger.kernel.org Cc: Ingo Molnar <mingo@kernel.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Tzvetomir Stoyanov <tz.stoyanov@gmail.com> Cc: Tom Zanussi <zanussi@kernel.org> Fixes: 7491e2c44278 ("tracing: Add a probe that attaches to trace events") Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
tracing/eprobes: Do not allow eprobes to use $stack, or % for regs
While playing with event probes (eprobes), I tried to see what would
happen if I attempted to retrieve the instruction pointer (%rip) knowing
that event probes do not use pt_regs. The result was:
Move the testing for TPARG_FL_TPOINT which is only used for event probes
to the top of the "$" variable check, as all the other variables are not
used for event probes. Also add a check in the register parsing "%" to
fail if an event probe is used.
Link: https://lkml.kernel.org/r/20220820134400.564426983@goodmis.org Cc: stable@vger.kernel.org Cc: Ingo Molnar <mingo@kernel.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Tzvetomir Stoyanov <tz.stoyanov@gmail.com> Cc: Tom Zanussi <zanussi@kernel.org> Fixes: 7491e2c44278 ("tracing: Add a probe that attaches to trace events") Acked-by: Masami Hiramatsu (Google) <mhiramat@kernel.org> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Yang Jihong [Thu, 18 Aug 2022 03:26:59 +0000 (11:26 +0800)]
ftrace: Fix NULL pointer dereference in is_ftrace_trampoline when ftrace is dead
ftrace_startup does not remove ops from ftrace_ops_list when
ftrace_startup_enable fails:
register_ftrace_function
ftrace_startup
__register_ftrace_function
...
add_ftrace_ops(&ftrace_ops_list, ops)
...
...
ftrace_startup_enable // if ftrace failed to modify, ftrace_disabled is set to 1
...
return 0 // ops is in the ftrace_ops_list.
When ftrace_disabled = 1, unregister_ftrace_function simply returns without doing anything:
unregister_ftrace_function
ftrace_shutdown
if (unlikely(ftrace_disabled))
return -ENODEV; // return here, __unregister_ftrace_function is not executed,
// as a result, ops is still in the ftrace_ops_list
__unregister_ftrace_function
...
If ops is dynamically allocated, it will be free later, in this case,
is_ftrace_trampoline accesses NULL pointer:
is_ftrace_trampoline
ftrace_ops_trampoline
do_for_each_ftrace_op(op, ftrace_ops_list) // OOPS! op may be NULL!
tracing/perf: Fix double put of trace event when init fails
If in perf_trace_event_init(), the perf_trace_event_open() fails, then it
will call perf_trace_event_unreg() which will not only unregister the perf
trace event, but will also call the put() function of the tp_event.
The problem here is that the trace_event_try_get_ref() is called by the
caller of perf_trace_event_init() and if perf_trace_event_init() returns a
failure, it will then call trace_event_put(). But since the
perf_trace_event_unreg() already called the trace_event_put() function, it
triggers a WARN_ON().
WARNING: CPU: 1 PID: 30309 at kernel/trace/trace_dynevent.c:46 trace_event_dyn_put_ref+0x15/0x20
If perf_trace_event_reg() does not call the trace_event_try_get_ref() then
the perf_trace_event_unreg() should not be calling trace_event_put(). This
breaks symmetry and causes bugs like these.
Pull out the trace_event_put() from perf_trace_event_unreg() and call it
in the locations that perf_trace_event_unreg() is called. This not only
fixes this bug, but also brings back the proper symmetry of the reg/unreg
vs get/put logic.
Lukas Bulwahn [Thu, 11 Aug 2022 07:17:34 +0000 (09:17 +0200)]
tracing: React to error return from traceprobe_parse_event_name()
The function traceprobe_parse_event_name() may set the first two function
arguments to a non-null value and still return -EINVAL to indicate an
unsuccessful completion of the function. Hence, it is not sufficient to
just check the result of the two function arguments for being not null,
but the return value also needs to be checked.
Commit 95c104c378dc ("tracing: Auto generate event name when creating a
group of events") changed the error-return-value checking of the second
traceprobe_parse_event_name() invocation in __trace_eprobe_create() and
removed checking the return value to jump to the error handling case.
Reinstate using the return value in the error-return-value checking.
Link: https://lkml.kernel.org/r/20220811071734.20700-1-lukas.bulwahn@gmail.com Fixes: 95c104c378dc ("tracing: Auto generate event name when creating a group of events") Acked-by: Linyu Yuan <quic_linyyuan@quicinc.com> Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Linus Torvalds [Sun, 21 Aug 2022 18:18:33 +0000 (11:18 -0700)]
Merge tag 'i2c-for-6.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux
Pull i2c fixes from Wolfram Sang:
"A revert to fix a regression introduced this merge window and a fix
for proper error handling in the remove path of the iMX driver"
* tag 'i2c-for-6.0-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
i2c: imx: Make sure to unregister adapter on remove()
Revert "i2c: scmi: Replace open coded device_get_match_data()"
Linus Torvalds [Sun, 21 Aug 2022 17:21:16 +0000 (10:21 -0700)]
Merge tag '6.0-rc1-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6
Pull cifs client fixes from Steve French:
- memory leak fix
- two small cleanups
- trivial strlcpy removal
- update missing entry for cifs headers in MAINTAINERS file
* tag '6.0-rc1-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
cifs: move from strlcpy with unused retval to strscpy
cifs: Fix memory leak on the deferred close
cifs: remove useless parameter 'is_fsctl' from SMB2_ioctl()
cifs: remove unused server parameter from calc_smb_size()
cifs: missing directory in MAINTAINERS file
Nick Desaulniers [Fri, 19 Aug 2022 19:06:40 +0000 (12:06 -0700)]
asm goto: eradicate CC_HAS_ASM_GOTO
GCC has supported asm goto since 4.5, and Clang has since version 9.0.0.
The minimum supported versions of these tools for the build according to
Documentation/process/changes.rst are 5.1 and 11.0.0 respectively.
Remove the feature detection script, Kconfig option, and clean up some
fallback code that is no longer supported.
The removed script was also testing for a GCC specific bug that was
fixed in the 4.7 release.
Also remove workarounds for bpftrace using clang older than 9.0.0, since
other BPF backend fixes are required at this point.
i2c: imx: Make sure to unregister adapter on remove()
If for whatever reasons pm_runtime_resume_and_get() fails and .remove() is
exited early, the i2c adapter stays around and the irq still calls its
handler, while the driver data and the register mapping go away. So if
later the i2c adapter is accessed or the irq triggers this results in
havoc accessing freed memory and unmapped registers.
So unregister the software resources even if resume failed, and only skip
the hardware access in that case.
Fixes: 588eb93ea49f ("i2c: imx: add runtime pm support to improve the performance") Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Acked-by: Oleksij Rempel <o.rempel@pengutronix.de> Signed-off-by: Wolfram Sang <wsa@kernel.org>
Wolfram Sang [Thu, 18 Aug 2022 20:31:13 +0000 (22:31 +0200)]
Revert "i2c: scmi: Replace open coded device_get_match_data()"
This reverts commit 9ae551ded5ba55f96a83cd0811f7ef8c2f329d0c. We got a
regression report, so ensure this machine boots again. We will come back
with a better version hopefully.
Linus Torvalds [Sat, 20 Aug 2022 21:55:38 +0000 (14:55 -0700)]
Merge tag 'kbuild-fixes-v6.0' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild
Pull Kbuild fixes from Masahiro Yamada:
- Fix module versioning broken on some architectures
- Make dummy-tools enable CONFIG_PPC_LONG_DOUBLE_128
- Remove -Wformat-zero-length, which has no warning instance
- Fix the order between drivers and libs in modules.order
- Fix false-positive warnings in clang-analyzer
* tag 'kbuild-fixes-v6.0' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
scripts/clang-tools: Remove DeprecatedOrUnsafeBufferHandling check
kbuild: fix the modules order between drivers and libs
scripts/Makefile.extrawarn: Do not disable clang's -Wformat-zero-length
kbuild: dummy-tools: pretend we understand __LONG_DOUBLE_128__
modpost: fix module versioning when a symbol lacks valid CRC
* tag 'perf-tools-fixes-for-v6.0-2022-08-19' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
perf tools: Support reading PERF_FORMAT_LOST
libperf: Add a test case for read formats
libperf: Handle read format in perf_evsel__read()
tools headers UAPI: Sync linux/perf_event.h with the kernel sources
tools headers UAPI: Sync x86's asm/kvm.h with the kernel sources
tools headers UAPI: Sync KVM's vmx.h header with the kernel sources
tools include UAPI: Sync linux/vhost.h with the kernel sources
tools headers kvm s390: Sync headers with the kernel sources
tools headers UAPI: Sync linux/kvm.h with the kernel sources
tools headers UAPI: Sync drm/i915_drm.h with the kernel sources
tools headers cpufeatures: Sync with the kernel sources
tools headers UAPI: Sync linux/fscrypt.h with the kernel sources
tools arch x86: Sync the msr-index.h copy with the kernel sources
perf beauty: Update copy of linux/socket.h with the kernel sources
perf cpumap: Fix alignment for masks in event encoding
perf cpumap: Compute mask size in constant time
perf cpumap: Synthetic events and const/static
perf cpumap: Const map for max()
Linus Torvalds [Sat, 20 Aug 2022 18:29:01 +0000 (11:29 -0700)]
Merge tag 's390-6.0-1' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux
Pull s390 updates from Alexander Gordeev:
- Fix a KVM crash on z12 and older machines caused by a wrong
assumption that Query AP Configuration Information is always
available.
- Lower severity of excessive Hypervisor filesystem error messages
when booting under KVM.
* tag 's390-6.0-1' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
s390/ap: fix crash on older machines based on QCI info missing
s390/hypfs: avoid error message under KVM
Linus Torvalds [Sat, 20 Aug 2022 17:49:02 +0000 (10:49 -0700)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Pull rdma fixes from Jason Gunthorpe:
"A few minor fixes:
- Fix buffer management in SRP to correct a regression with the login
authentication feature from v5.17
- Don't iterate over non-present ports in mlx5
- Fix an error introduced by the foritify work in cxgb4
- Two bug fixes for the recently merged ERDMA driver
- Unbreak RDMA dmabuf support, a regresion from v5.19"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
RDMA: Handle the return code from dma_resv_wait_timeout() properly
RDMA/erdma: Correct the max_qp and max_cq capacities of the device
RDMA/erdma: Using the key in FMR WR instead of MR structure
RDMA/cxgb4: fix accept failure due to increased cpl_t5_pass_accept_rpl size
RDMA/mlx5: Use the proper number of ports
IB/iser: Fix login with authentication
projects / kernel.git / log