Recent commit 52fde2c07d ("nvme: set dma alignment to dword") has
caused a regression on our platform.
It turned out that the nvme_get_log() method invocation caused the
nvme_hwmon_data structure instance corruption. In particular the
nvme_hwmon_data.ctrl pointer was overwritten either with zeros or with
garbage. After some research we discovered that the problem happened
even before the actual NVME DMA execution, but during the buffer mapping.
Since our platform is DMA-noncoherent, the mapping implied the cache-line
invalidations or write-backs depending on the DMA-direction parameter.
In case of the NVME SMART log getting the DMA was performed
from-device-to-memory, thus the cache-invalidation was activated during
the buffer mapping. Since the log-buffer isn't cache-line aligned, the
cache-invalidation caused the neighbour data to be discarded. The
neighbouring data turned to be the data surrounding the buffer in the
framework of the nvme_hwmon_data structure.
In order to fix that we need to make sure that the whole log-buffer is
defined within the cache-line-aligned memory region so the
cache-invalidation procedure wouldn't involve the adjacent data. One of
the option to guarantee that is to kmalloc the DMA-buffer [1]. Seeing the
rest of the NVME core driver prefer that method it has been chosen to fix
this problem too.
Note after a deeper researches we found out that the denoted commit wasn't
a root cause of the problem. It just revealed the invalidity by activating
the DMA-based NVME SMART log getting performed in the framework of the
NVME hwmon driver. The problem was here since the initial commit of the
driver.
[1] Documentation/core-api/dma-api-howto.rst
Fixes: 400b6a7b13 ("nvme: Add hardware monitoring support")
Signed-off-by: Serge Semin <Sergey.Semin@baikalelectronics.ru>
Signed-off-by: Christoph Hellwig <hch@lst.de>
An NVMe controller works perfectly fine even when the hwmon
initialization fails. Stop returning errors that do not come from a
controller reset from nvme_hwmon_init to handle this case consistently.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Serge Semin <fancer.lancer@gmail.com>
NVMe uses PRPs for data transfers and has no specific limit for a single
DMA segement. Limiting the size will cause problems because the block
layer assumes PRP-ish devices using a virt boundary mask don't have a
segment limit. And while this is true, we also really need to tell the
DMA mapping layer about it, otherwise dma-debug will trip over it.
Fixes: 5bd2927ace ("nvme-apple: Add initial Apple SoC NVMe driver")
Suggested-by: Sven Peter <sven@svenpeter.dev>
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
[hch: rewrote the commit message based on the PCIe commit]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Eric Curtin <ecurtin@redhat.com>
Reviewed-by: Sven Peter <sven@svenpeter.dev>
Kingston SSDs do support NVMe Write_Zeroes cmd but take long time to
process. The firmware version is locked by these SSDs, we can not expect
firmware improvement, so disable Write_Zeroes cmd.
Signed-off-by: Xander Li <xander_li@kingston.com.tw>
Signed-off-by: Christoph Hellwig <hch@lst.de>
There is typo here so it releases the wrong variable. "ctrl->admin_q"
was intended instead of "ctrl->fabrics_q".
Fixes: fe60e8c534 ("nvme: add common helpers to allocate and free tagsets")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
When we revalidate paths as part of ns size change (as of commit
e7d65803e2), it is possible that during the path revalidation, the
only paths that is IO capable (i.e. optimized/non-optimized) are the
ones that ns resize was not yet informed to the host, which will cause
inflight requests to be requeued (as we have available paths but none
are IO capable). These requests on the requeue list are waiting for
someone to resubmit them at some point.
The IO capable paths will eventually notify the ns resize change to the
host, but there is nothing that will kick the requeue list to resubmit
the queued requests.
Fix this by always kicking the requeue list, and if no IO capable path
exists, these requests will be queued again.
A typical log that indicates that IOs are requeued:
--
nvme nvme1: creating 4 I/O queues.
nvme nvme1: new ctrl: "testnqn1"
nvme nvme2: creating 4 I/O queues.
nvme nvme2: mapped 4/0/0 default/read/poll queues.
nvme nvme2: new ctrl: NQN "testnqn1", addr 127.0.0.1:8009
nvme nvme1: rescanning namespaces.
nvme1n1: detected capacity change from 2097152 to 4194304
block nvme1n1: no usable path - requeuing I/O
block nvme1n1: no usable path - requeuing I/O
block nvme1n1: no usable path - requeuing I/O
block nvme1n1: no usable path - requeuing I/O
block nvme1n1: no usable path - requeuing I/O
block nvme1n1: no usable path - requeuing I/O
block nvme1n1: no usable path - requeuing I/O
block nvme1n1: no usable path - requeuing I/O
block nvme1n1: no usable path - requeuing I/O
block nvme1n1: no usable path - requeuing I/O
nvme nvme2: rescanning namespaces.
--
Reported-by: Yogev Cohen <yogev@lightbitslabs.com>
Fixes: e7d65803e2 ("nvme-multipath: revalidate paths during rescan")
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Cc: <stable@vger.kernel.org> # v5.15+
Signed-off-by: Christoph Hellwig <hch@lst.de>
When we delete a controller, we execute the following:
1. nvme_stop_ctrl() - stop some work elements that may be
inflight or scheduled (specifically also .stop_ctrl
which cancels ctrl error recovery work)
2. nvme_remove_namespaces() - which first flushes scan_work
to avoid competing ns addition/removal
3. continue to teardown the controller
However, if err_work was scheduled to run in (1), it is designed to
cancel any inflight I/O, particularly I/O that is originating from ns
scan_work in (2), but because it is cancelled in .stop_ctrl(), we can
prevent forward progress of (2) as ns scanning is blocking on I/O
(that will never be cancelled).
The race is:
1. transport layer error observed -> err_work is scheduled
2. scan_work executes, discovers ns, generate I/O to it
3. nvme_ctop_ctrl() -> .stop_ctrl() -> cancel_work_sync(err_work)
- err_work never executed
4. nvme_remove_namespaces() -> flush_work(scan_work)
--> deadlock, because scan_work is blocked on I/O that was supposed
to be cancelled by err_work, but was cancelled before executing (see
stack trace [1]).
Fix this by flushing err_work instead of cancelling it, to force it
to execute and cancel all inflight I/O.
[1]:
--
Call Trace:
<TASK>
__schedule+0x390/0x910
? scan_shadow_nodes+0x40/0x40
schedule+0x55/0xe0
io_schedule+0x16/0x40
do_read_cache_page+0x55d/0x850
? __page_cache_alloc+0x90/0x90
read_cache_page+0x12/0x20
read_part_sector+0x3f/0x110
amiga_partition+0x3d/0x3e0
? osf_partition+0x33/0x220
? put_partition+0x90/0x90
bdev_disk_changed+0x1fe/0x4d0
blkdev_get_whole+0x7b/0x90
blkdev_get_by_dev+0xda/0x2d0
device_add_disk+0x356/0x3b0
nvme_mpath_set_live+0x13c/0x1a0 [nvme_core]
? nvme_parse_ana_log+0xae/0x1a0 [nvme_core]
nvme_update_ns_ana_state+0x3a/0x40 [nvme_core]
nvme_mpath_add_disk+0x120/0x160 [nvme_core]
nvme_alloc_ns+0x594/0xa00 [nvme_core]
nvme_validate_or_alloc_ns+0xb9/0x1a0 [nvme_core]
? __nvme_submit_sync_cmd+0x1d2/0x210 [nvme_core]
nvme_scan_work+0x281/0x410 [nvme_core]
process_one_work+0x1be/0x380
worker_thread+0x37/0x3b0
? process_one_work+0x380/0x380
kthread+0x12d/0x150
? set_kthread_struct+0x50/0x50
ret_from_fork+0x1f/0x30
</TASK>
INFO: task nvme:6725 blocked for more than 491 seconds.
Not tainted 5.15.65-f0.el7.x86_64 #1
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:nvme state:D
stack: 0 pid: 6725 ppid: 1761 flags:0x00004000
Call Trace:
<TASK>
__schedule+0x390/0x910
? sched_clock+0x9/0x10
schedule+0x55/0xe0
schedule_timeout+0x24b/0x2e0
? try_to_wake_up+0x358/0x510
? finish_task_switch+0x88/0x2c0
wait_for_completion+0xa5/0x110
__flush_work+0x144/0x210
? worker_attach_to_pool+0xc0/0xc0
flush_work+0x10/0x20
nvme_remove_namespaces+0x41/0xf0 [nvme_core]
nvme_do_delete_ctrl+0x47/0x66 [nvme_core]
nvme_sysfs_delete.cold.96+0x8/0xd [nvme_core]
dev_attr_store+0x14/0x30
sysfs_kf_write+0x38/0x50
kernfs_fop_write_iter+0x146/0x1d0
new_sync_write+0x114/0x1b0
? intel_pmu_handle_irq+0xe0/0x420
vfs_write+0x18d/0x270
ksys_write+0x61/0xe0
__x64_sys_write+0x1a/0x20
do_syscall_64+0x37/0x90
entry_SYSCALL_64_after_hwframe+0x61/0xcb
--
Fixes: 3f2304f8c6 ("nvme-tcp: add NVMe over TCP host driver")
Reported-by: Jonathan Nicklin <jnicklin@blockbridge.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Tested-by: Jonathan Nicklin <jnicklin@blockbridge.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
When we delete a controller, we execute the following:
1. nvme_stop_ctrl() - stop some work elements that may be
inflight or scheduled (specifically also .stop_ctrl
which cancels ctrl error recovery work)
2. nvme_remove_namespaces() - which first flushes scan_work
to avoid competing ns addition/removal
3. continue to teardown the controller
However, if err_work was scheduled to run in (1), it is designed to
cancel any inflight I/O, particularly I/O that is originating from ns
scan_work in (2), but because it is cancelled in .stop_ctrl(), we can
prevent forward progress of (2) as ns scanning is blocking on I/O
(that will never be cancelled).
The race is:
1. transport layer error observed -> err_work is scheduled
2. scan_work executes, discovers ns, generate I/O to it
3. nvme_ctop_ctrl() -> .stop_ctrl() -> cancel_work_sync(err_work)
- err_work never executed
4. nvme_remove_namespaces() -> flush_work(scan_work)
--> deadlock, because scan_work is blocked on I/O that was supposed
to be cancelled by err_work, but was cancelled before executing.
Fix this by flushing err_work instead of cancelling it, to force it
to execute and cancel all inflight I/O.
Fixes: b435ecea2a ("nvme: Add .stop_ctrl to nvme ctrl ops")
Fixes: f6c8e432cb ("nvme: flush namespace scanning work just before removing namespaces")
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
-----BEGIN PGP SIGNATURE-----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=nwLk
-----END PGP SIGNATURE-----
Merge tag 'for-6.1/passthrough-2022-10-04' of git://git.kernel.dk/linux
Pull passthrough updates from Jens Axboe:
"With these changes, passthrough NVMe support over io_uring now
performs at the same level as block device O_DIRECT, and in many cases
6-8% better.
This contains:
- Add support for fixed buffers for passthrough (Anuj, Kanchan)
- Enable batched allocations and freeing on passthrough, similarly to
what we support on the normal storage path (me)
- Fix from Geert fixing an issue with !CONFIG_IO_URING"
* tag 'for-6.1/passthrough-2022-10-04' of git://git.kernel.dk/linux:
io_uring: Add missing inline to io_uring_cmd_import_fixed() dummy
nvme: wire up fixed buffer support for nvme passthrough
nvme: pass ubuffer as an integer
block: extend functionality to map bvec iterator
block: factor out blk_rq_map_bio_alloc helper
block: rename bio_map_put to blk_mq_map_bio_put
nvme: refactor nvme_alloc_request
nvme: refactor nvme_add_user_metadata
nvme: Use blk_rq_map_user_io helper
scsi: Use blk_rq_map_user_io helper
block: add blk_rq_map_user_io
io_uring: introduce fixed buffer support for io_uring_cmd
io_uring: add io_uring_cmd_import_fixed
nvme: enable batched completions of passthrough IO
nvme: split out metadata vs non metadata end_io uring_cmd completions
block: allow end_io based requests in the completion batch handling
block: change request end_io handler to pass back a return value
block: enable batched allocation for blk_mq_alloc_request()
block: kill deprecated BUG_ON() in the flush handling
-----BEGIN PGP SIGNATURE-----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=R05e
-----END PGP SIGNATURE-----
Merge tag 'for-6.1/block-2022-10-03' of git://git.kernel.dk/linux
Pull block updates from Jens Axboe:
- NVMe pull requests via Christoph:
- 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)
- handle effects after freeing the request (Keith Busch)
- copy firmware_rev on each init (Keith Busch)
- restrict management ioctls to admin (Keith Busch)
- ensure subsystem reset is single threaded (Keith Busch)
- report the actual number of tagset maps in nvme-pci (Keith
Busch)
- small fabrics authentication fixups (Christoph Hellwig)
- add common code for tagset allocation and freeing (Christoph
Hellwig)
- stop using the request_queue in nvmet (Christoph Hellwig)
- set min_align_mask before calculating max_hw_sectors (Rishabh
Bhatnagar)
- send a rediscover uevent when a persistent discovery controller
reconnects (Sagi Grimberg)
- misc nvmet-tcp fixes (Varun Prakash, zhenwei pi)
- MD pull request via Song:
- Various raid5 fix and clean up, by Logan Gunthorpe and David
Sloan.
- Raid10 performance optimization, by Yu Kuai.
- sbitmap wakeup hang fixes (Hugh, Keith, Jan, Yu)
- IO scheduler switching quisce fix (Keith)
- s390/dasd block driver updates (Stefan)
- support for recovery for the ublk driver (ZiyangZhang)
- rnbd drivers fixes and updates (Guoqing, Santosh, ye, Christoph)
- blk-mq and null_blk map fixes (Bart)
- various bcache fixes (Coly, Jilin, Jules)
- nbd signal hang fix (Shigeru)
- block writeback throttling fix (Yu)
- optimize the passthrough mapping handling (me)
- prepare block cgroups to being gendisk based (Christoph)
- get rid of an old PSI hack in the block layer, moving it to the
callers instead where it belongs (Christoph)
- blk-throttle fixes and cleanups (Yu)
- misc fixes and cleanups (Liu Shixin, Liu Song, Miaohe, Pankaj,
Ping-Xiang, Wolfram, Saurabh, Li Jinlin, Li Lei, Lin, Li zeming,
Miaohe, Bart, Coly, Gaosheng
* tag 'for-6.1/block-2022-10-03' of git://git.kernel.dk/linux: (162 commits)
sbitmap: fix lockup while swapping
block: add rationale for not using blk_mq_plug() when applicable
block: adapt blk_mq_plug() to not plug for writes that require a zone lock
s390/dasd: use blk_mq_alloc_disk
blk-cgroup: don't update the blkg lookup hint in blkg_conf_prep
nvmet: don't look at the request_queue in nvmet_bdev_set_limits
nvmet: don't look at the request_queue in nvmet_bdev_zone_mgmt_emulate_all
blk-mq: use quiesced elevator switch when reinitializing queues
block: replace blk_queue_nowait with bdev_nowait
nvme: remove nvme_ctrl_init_connect_q
nvme-loop: use the tagset alloc/free helpers
nvme-loop: store the generic nvme_ctrl in set->driver_data
nvme-loop: initialize sqsize later
nvme-fc: use the tagset alloc/free helpers
nvme-fc: store the generic nvme_ctrl in set->driver_data
nvme-fc: keep ctrl->sqsize in sync with opts->queue_size
nvme-rdma: use the tagset alloc/free helpers
nvme-rdma: store the generic nvme_ctrl in set->driver_data
nvme-tcp: use the tagset alloc/free helpers
nvme-tcp: store the generic nvme_ctrl in set->driver_data
...
-----BEGIN PGP SIGNATURE-----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=m9kX
-----END PGP SIGNATURE-----
Merge tag 'for-6.1/io_uring-2022-10-03' of git://git.kernel.dk/linux
Pull io_uring updates from Jens Axboe:
- Add supported for more directly managed task_work running.
This is beneficial for real world applications that end up issuing
lots of system calls as part of handling work. Normal task_work will
always execute as we transition in and out of the kernel, even for
"unrelated" system calls. It's more efficient to defer the handling
of io_uring's deferred work until the application wants it to be run,
generally in batches.
As part of ongoing work to write an io_uring network backend for
Thrift, this has been shown to greatly improve performance. (Dylan)
- Add IOPOLL support for passthrough (Kanchan)
- Improvements and fixes to the send zero-copy support (Pavel)
- Partial IO handling fixes (Pavel)
- CQE ordering fixes around CQ ring overflow (Pavel)
- Support sendto() for non-zc as well (Pavel)
- Support sendmsg for zerocopy (Pavel)
- Networking iov_iter fix (Stefan)
- Misc fixes and cleanups (Pavel, me)
* tag 'for-6.1/io_uring-2022-10-03' of git://git.kernel.dk/linux: (56 commits)
io_uring/net: fix notif cqe reordering
io_uring/net: don't update msg_name if not provided
io_uring: don't gate task_work run on TIF_NOTIFY_SIGNAL
io_uring/rw: defer fsnotify calls to task context
io_uring/net: fix fast_iov assignment in io_setup_async_msg()
io_uring/net: fix non-zc send with address
io_uring/net: don't skip notifs for failed requests
io_uring/rw: don't lose short results on io_setup_async_rw()
io_uring/rw: fix unexpected link breakage
io_uring/net: fix cleanup double free free_iov init
io_uring: fix CQE reordering
io_uring/net: fix UAF in io_sendrecv_fail()
selftest/net: adjust io_uring sendzc notif handling
io_uring: ensure local task_work marks task as running
io_uring/net: zerocopy sendmsg
io_uring/net: combine fail handlers
io_uring/net: rename io_sendzc()
io_uring/net: support non-zerocopy sendto
io_uring/net: refactor io_setup_async_addr
io_uring/net: don't lose partial send_zc on fail
...
-----BEGIN PGP SIGNATURE-----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=SZZ1
-----END PGP SIGNATURE-----
Merge tag 'block-6.0-2022-09-29' of git://git.kernel.dk/linux
Pull block fixes from Jens Axboe:
"A single NVMe pull request via Christoph with a few fixes that should
go into the 6.0 release:
- Fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices
(Michael Kelley)
- Disable Write Zeroes on Phison E3C/E4C (Tina Hsu)"
* tag 'block-6.0-2022-09-29' of git://git.kernel.dk/linux:
nvme-pci: disable Write Zeroes on Phison E3C/E4C
nvme: Fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices
if io_uring sends passthrough command with IORING_URING_CMD_FIXED flag,
use the pre-registered buffer for IO (non-vectored variant). Pass the
buffer/length to io_uring and get the bvec iterator for the range. Next,
pass this bvec to block-layer and obtain a bio/request for subsequent
processing.
Signed-off-by: Kanchan Joshi <joshi.k@samsung.com>
Link: https://lore.kernel.org/r/20220930062749.152261-13-anuj20.g@samsung.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This is a prep patch. Modify nvme_submit_user_cmd and
nvme_map_user_request to take ubuffer as plain integer
argument, and do away with nvme_to_user_ptr conversion in callers.
Signed-off-by: Anuj Gupta <anuj20.g@samsung.com>
Signed-off-by: Kanchan Joshi <joshi.k@samsung.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220930062749.152261-12-anuj20.g@samsung.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
nvme_alloc_request expects a large number of parameters.
Split this out into two functions to reduce number of parameters.
First one retains the name nvme_alloc_request, while second one is
named nvme_map_user_request.
Signed-off-by: Kanchan Joshi <joshi.k@samsung.com>
Signed-off-by: Anuj Gupta <anuj20.g@samsung.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220930062749.152261-8-anuj20.g@samsung.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pass struct request rather than bio. It helps to kill a parameter, and
some processing clean-up too.
Signed-off-by: Kanchan Joshi <joshi.k@samsung.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220930062749.152261-7-anuj20.g@samsung.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now that the normal passthrough end_io path doesn't need the request
anymore, we can kill the explicit blk_mq_free_request() and just pass
back RQ_END_IO_FREE instead. This enables the batched completion from
freeing batches of requests at the time.
This brings passthrough IO performance at least on par with bdev based
O_DIRECT with io_uring. With this and batche allocations, peak performance
goes from 110M IOPS to 122M IOPS. For IRQ based, passthrough is now also
about 10% faster than previously, going from ~61M to ~67M IOPS.
Reviewed-by: Anuj Gupta <anuj20.g@samsung.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Co-developed-by: Stefan Roesch <shr@fb.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
By splitting up the metadata and non-metadata end_io handling, we can
remove any request dependencies on the normal non-metadata IO path. This
is in preparation for enabling the normal IO passthrough path to pass
the ownership of the request back to the block layer.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Anuj Gupta <anuj20.g@samsung.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Co-developed-by: Stefan Roesch <shr@fb.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Everything is just converted to returning RQ_END_IO_NONE, and there
should be no functional changes with this patch.
In preparation for allowing the end_io handler to pass ownership back
to the block layer, rather than retain ownership of the request.
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
* for-6.1/block: (162 commits)
sbitmap: fix lockup while swapping
block: add rationale for not using blk_mq_plug() when applicable
block: adapt blk_mq_plug() to not plug for writes that require a zone lock
s390/dasd: use blk_mq_alloc_disk
blk-cgroup: don't update the blkg lookup hint in blkg_conf_prep
nvmet: don't look at the request_queue in nvmet_bdev_set_limits
nvmet: don't look at the request_queue in nvmet_bdev_zone_mgmt_emulate_all
blk-mq: use quiesced elevator switch when reinitializing queues
block: replace blk_queue_nowait with bdev_nowait
nvme: remove nvme_ctrl_init_connect_q
nvme-loop: use the tagset alloc/free helpers
nvme-loop: store the generic nvme_ctrl in set->driver_data
nvme-loop: initialize sqsize later
nvme-fc: use the tagset alloc/free helpers
nvme-fc: store the generic nvme_ctrl in set->driver_data
nvme-fc: keep ctrl->sqsize in sync with opts->queue_size
nvme-rdma: use the tagset alloc/free helpers
nvme-rdma: store the generic nvme_ctrl in set->driver_data
nvme-tcp: use the tagset alloc/free helpers
nvme-tcp: store the generic nvme_ctrl in set->driver_data
...
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Use the common helpers to allocate and free the tagsets. To make this
work the generic nvme_ctrl now needs to be stored in the hctx private
data instead of the nvme_fc_ctrl.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: James Smart <jsmart2021@gmail.com>
Point the private data to the generic controller structure in preparation
of using the common tagset init/exit code and use the chance the cleanup
the init_hctx methods a bit.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: James Smart <jsmart2021@gmail.com>
Also update the sqsize field when capping the queue size, and remove the
check a queue size that is larger than sqsize given that sqsize is only
initialized from opts->queue_size.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: James Smart <jsmart2021@gmail.com>
Use the common helpers to allocate and free the tagsets. To make this
work the generic nvme_ctrl now needs to be stored in the hctx private
data instead of the nvme_rdma_ctrl.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Point the private data to the generic controller structure in preparation
of using the common tagset init/exit code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Use the common helpers to allocate and free the tagsets. To make this
work the generic nvme_ctrl now needs to be stored in the hctx private
data instead of the nvme_tcp_ctrl.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Point the private data to the generic controller structure in preparation
of using the common tagset init/exit code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
->nvme_tcp_queue is not used anywhere, so remove it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Add common helpers to allocate and tear down the admin and I/O tag sets,
including the special queues allocated with them.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
We've been reporting 2 maps regardless of whether the module parameter
asked for anything beyond the default queues. A consequence of this
means that blk-mq will reinitialize the all the hardware contexts and io
schedulers on every controller reset when the mapping is exactly the
same as before. This unnecessary overhead is adding several milliseconds
on a reset for environments that don't need it. Report the actual number
of mappings in use.
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
If swiotlb is force enabled dma_max_mapping_size ends up calling
swiotlb_max_mapping_size which takes into account the min align mask for
the device. Set the min align mask for nvme driver before calling
dma_max_mapping_size while calculating max hw sectors.
Signed-off-by: Rishabh Bhatnagar <risbhat@amazon.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
When a discovery controller is disconnected, no AENs will arrive to
notify the host about discovery log change events.
In order to solve this, send a uevent notification when a
persistent discovery controller reconnects. We add a new ctrl
flag NVME_CTRL_STARTED_ONCE that will be set on the first
start, and consecutive calls will find it set, and send the
event to userspace if the controller is a discovery controller.
Upon the event reception, userspace will re-read the discovery
log page and will act upon changes as it sees fit.
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Daniel Wagner <dwagner@suse.de>
Reviewed-by: James Smart <jsmart2021@gmail.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
We expect to grow a few of these flags for various purposes
so make them a proper enumeration.
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: James Smart <jsmart2021@gmail.com>
Reviewed-by: Daniel Wagner <dwagner@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
E3C/E4C SSDs do support the Write Zeroes command in theory, but have very
bad performance when using it. As the firmware has been frozen for these
products we can not expect firmware improvements for it, so disable
Write Zeroes.
Signed-off-by: Tina Hsu <tina_hsu@phison.corp-partner.google.com>
[hch: update the commit message]
Signed-off-by: Christoph Hellwig <hch@lst.de>
The IOC_PR_CLEAR and IOC_PR_RELEASE ioctls are
non-functional on NVMe devices because the nvme_pr_clear()
and nvme_pr_release() functions set the IEKEY field incorrectly.
The IEKEY field should be set only when the key is zero (i.e,
not specified). The current code does it backwards.
Furthermore, the NVMe spec describes the persistent
reservation "clear" function as an option on the reservation
release command. The current implementation of nvme_pr_clear()
erroneously uses the reservation register command.
Fix these errors. Note that NVMe version 1.3 and later specify
that setting the IEKEY field will return an error of Invalid
Field in Command. The fix will set IEKEY when the key is zero,
which is appropriate as these ioctls consider a zero key to
be "unspecified", and the intention of the spec change is
to require a valid key.
Tested on a version 1.4 PCI NVMe device in an Azure VM.
Fixes: 1673f1f08c ("nvme: move block_device_operations and ns/ctrl freeing to common code")
Fixes: 1d277a637a ("NVMe: Add persistent reservation ops")
Signed-off-by: Michael Kelley <mikelley@microsoft.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
The subsystem reset writes to a register, so we have to ensure the
device state is capable of handling that otherwise the driver may access
unmapped registers. Use the state machine to ensure the subsystem reset
doesn't try to write registers on a device already undergoing this type
of reset.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=214771
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
The passthrough commands already have this restriction, but the other
operations do not. Require the same capabilities for all users as all of
these operations, which include resets and rescans, can be disruptive.
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
The firmware revision can change on after a reset so copy the most
recent info each time instead of just the first time, otherwise the
sysfs firmware_rev entry may contain stale data.
Reported-by: Jeff Lien <jeff.lien@wdc.com>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Chao Leng <lengchao@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
If a reset occurs after the scan work attempts to issue a command, the
reset may quisce the admin queue, which blocks the scan work's command
from dispatching. The scan work will not be able to complete while the
queue is quiesced.
Meanwhile, the reset work will cancel all outstanding admin tags and
wait until all requests have transitioned to idle, which includes the
passthrough request. But the passthrough request won't be set to idle
until after the scan_work flushes, so we're deadlocked.
Fix this by handling the end effects after the request has been freed.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=216354
Reported-by: Jonathan Derrick <Jonathan.Derrick@solidigm.com>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chao Leng <lengchao@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
We need the poll_flags to know how to poll for the IO, and we should
have the batch structure in preparation for supporting batched
completions with iopoll.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Store a cookie during submission, and use that to implement
completion-polling inside the ->uring_cmd_iopoll handler.
This handler makes use of existing bio poll facility.
Signed-off-by: Kanchan Joshi <joshi.k@samsung.com>
Signed-off-by: Anuj Gupta <anuj20.g@samsung.com>
Link: https://lore.kernel.org/r/20220823161443.49436-5-joshi.k@samsung.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
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.
$ nvme list-subsys
nvme-subsys0 - NQN=nqn.2014-08.org.nvmexpress.discovery
\
+- nvme0 tcp traddr=192.168.56.1,trsvcid=8009,src_addr=192.168.56.101 live
Signed-off-by: Martin Belanger <martin.belanger@dell.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
The 32-bit field, dma_len, packs better in the iod struct above the
dma_addr_t on 64-bit systems.
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
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>
It's only true or false, so make this a bool to reflect that and save
some space in nvme_iod.
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Serge Semin
Christoph Hellwig
Russell King (Oracle)
Xander Li
Dan Carpenter
Sagi Grimberg
Xi Ruoyao
Abhijit
Linus Torvalds
Kanchan Joshi
Anuj Gupta
Jens Axboe
Keith Busch
Rishabh Bhatnagar
Tina Hsu
Michael Kelley
Martin Belanger