drivers:md:fix a potential use-after-free bug

In line 2884, "raid5_release_stripe(sh);" drops the reference to sh and
may cause sh to be released. However, sh is subsequently used in lines
2886 "if (sh->batch_head && sh != sh->batch_head)". This may result in an
use-after-free bug.

It can be fixed by moving "raid5_release_stripe(sh);" to the bottom of
the function.

Signed-off-by: Wentao_Liang <Wentao_Liang_g@163.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Ensure batch_last is released before sleeping for quiesce

A race condition exists where if raid5_quiesce() is called in the
middle of a request that has set batch_last, it will deadlock.

batch_last will hold a reference to a stripe when raid5_quiesce() is
called. This will cause the next raid5_get_active_stripe() call to
sleep waiting for the quiesce to finish, but the raid5_quiesce() thread
will wait for active_stripes to go to zero which will never happen
because request thread is waiting for the quiesce to stop.

Fix this by creating a special __raid5_get_active_stripe() function
which takes the request context and clears the last_batch before
sleeping.

While we're at it, change the arguments of raid5_get_active_stripe()
to bools.

Fixes: 3312e6c887fe ("md/raid5: Keep a reference to last stripe_head for batch")
Reported-by: David Sloan <David.Sloan@eideticom.com>
Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Move stripe_request_ctx up

Move stripe_request_ctx up. No functional changes intended.

This will be necessary in the next patch to release the batch_last
in the context before sleeping.

Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Drop unnecessary call to r5c_check_stripe_cache_usage()

Now that raid5_get_active_stripe() has been refactored it is appearant
that r5c_check_stripe_cache_usage() doesn't need to be called in
the wait_for_stripe branch.

r5c_check_stripe_cache_usage() will only conditionally call
r5l_wake_reclaim(), but that function is called two lines later.

Drop the call for cleanup.

Reported-by: Martin Oliveira <martin.oliveira@eideticom.com>
Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Make is_inactive_blocked() helper

The logic to wait_for_stripe is difficult to parse being on so many
lines and with confusing operator precedence. Move it to a helper
function to make it easier to read.

No functional changes intended.

Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Refactor raid5_get_active_stripe()

Refactor the raid5_get_active_stripe() to read more linearly in
the order it's typically executed.

The init_stripe() call is called if a free stripe is found and the
function is exited early which removes a lot of if (sh) checks and
unindents the following code.

Remove the while loop in favour of the 'goto retry' pattern, which
reduces indentation further. And use a 'goto wait_for_stripe' instead
of an additional indent seeing it is the unusual path and this makes
the code easier to read.

No functional changes intended. Will make subsequent changes
in patches easier to understand.

Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: pass struct queue_limits to the bio splitting helpers

Allow using the splitting helpers on just a queue_limits instead of
a full request_queue structure.  This will eventually allow file systems
or remapping drivers to split REQ_OP_ZONE_APPEND bios based on limits
calculated as the minimum common capabilities over multiple devices.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Link: https://lore.kernel.org/r/20220727162300.3089193-7-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: move bio_allowed_max_sectors to blk-merge.c

Move this helper into the only file where it is used.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Link: https://lore.kernel.org/r/20220727162300.3089193-6-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: move the call to get_max_io_size out of blk_bio_segment_split

Prepare for reusing blk_bio_segment_split for (file system controlled)
splits of REQ_OP_ZONE_APPEND bios by letting the caller control the
maximum size of the bio.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Link: https://lore.kernel.org/r/20220727162300.3089193-5-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: move ->bio_split to the gendisk

Only non-passthrough requests are split by the block layer and use the
->bio_split bio_set.  Move it from the request_queue to the gendisk.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Link: https://lore.kernel.org/r/20220727162300.3089193-4-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: change the blk_queue_bounce calling convention

The double indirect bio leads to somewhat suboptimal code generation.
Instead return the (original or split) bio, and make sure the
request_queue arguments to the lower level helpers is passed after the
bio to avoid constant reshuffling of the argument passing registers.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Link: https://lore.kernel.org/r/20220727162300.3089193-3-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: change the blk_queue_split calling convention

The double indirect bio leads to somewhat suboptimal code generation.
Instead return the (original or split) bio, and make sure the
request_queue arguments to the lower level helpers is passed after the
bio to avoid constant reshuffling of the argument passing registers.

Also give it and the helpers used to implement it more descriptive names.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220727162300.3089193-2-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: update MAINTAINERS for the new auth code

Add the common subdirectory and match all nvme* headers in
include/linux/.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet-tcp: fix lockdep complaint on nvmet_tcp_wq flush during queue teardown

We probably need nvmet_tcp_wq to have MEM_RECLAIM as we are
sending/receiving for the socket from works on this workqueue.
Also this eliminates lockdep complaints:
--
[ 6174.010200] workqueue: WQ_MEM_RECLAIM
nvmet-wq:nvmet_tcp_release_queue_work [nvmet_tcp] is flushing
!WQ_MEM_RECLAIM nvmet_tcp_wq:nvmet_tcp_io_work [nvmet_tcp]
[ 6174.010216] WARNING: CPU: 20 PID: 14456 at kernel/workqueue.c:2628
check_flush_dependency+0x110/0x14c

Reported-by: Yi Zhang <yi.zhang@redhat.com>
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: enable generic interface (/dev/ngXnY) for unknown command sets

Extend nvme_alloc_ns() and nvme_validate_ns() for unknown command-set as
well. Both are made to use a new helper (nvme_update_ns_info_cs_indep)
which is similar to nvme_update_ns_info but performs fewer operations
to get the generic interface up.

Suggested-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Joel Granados <j.granados@samsung.com>
Signed-off-by: Kanchan Joshi <joshi.k@samsung.com>
[hch: rebased on other refactoring patches]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Javier González <javier.gonz@samsung.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: factor out a nvme_ns_is_readonly helper

Add a little helper to check if a namespace should be marked read-only
that uses a new is_readonly flag in the nvme_ns_info structure.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Javier González <javier.gonz@samsung.com>
Reviewed-by: Joel Granados <j.granados@samsung.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Kanchan Joshi <joshi.k@samsung.com>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: refactor namespace probing

Change nvme_ns_scan to gather all information needed for generic
namespace setup into a nvme_ns_info structure.  This structure is filled
from the Command Set Idependent Identify Namespace data structure if
it is available or else the legacy Identify namespace structure.

With that everything related to the NVM command set (and the ZNS command
set derived from it) can be encapsulated in the nvme_update_ns_info_block
function while keeping the rest of the namespace probing generic.

The downside is that we now always issue two Identify Namespace calls for
each probed namespace instead of usually just a single one previously.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Javier González <javier.gonz@samsung.com>
Reviewed-by: Joel Granados <j.granados@samsung.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: generalize the nvme_multi_css check in nvme_scan_ns

Check for multiple command set support early on an error out if is
not supported when a !NVM command set namespace is found.  This
prepares for adding command set independent passthrough support.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Javier González <javier.gonz@samsung.com>
Reviewed-by: Joel Granados <j.granados@samsung.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Kanchan Joshi <joshi.k@samsung.com>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: rename nvme_validate_or_alloc_ns to nvme_scan_ns

This shorter name much better fits what this function does in
the scanning process.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Javier González <javier.gonz@samsung.com>
Reviewed-by: Joel Granados <j.granados@samsung.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Kanchan Joshi <joshi.k@samsung.com>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: catch -ENODEV from nvme_revalidate_zones again

nvme_revalidate_zones can also return -ENODEV if e.g. zone sizes aren't
constant or not a power of two.  In that case we should jump to marking
the gendisk hidden and only support pass through.

Fixes: 602e57c9799c ("nvme: also mark passthrough-only namespaces ready in nvme_update_ns_info")
Reported-by: Joel Granados <j.granados@samsung.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Joel Granados <j.granados@samsung.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet-auth: select the intended CRYPTO_DH_RFC7919_GROUPS

Commit 71ebe3842ebe ("nvmet-auth: Diffie-Hellman key exchange support")
intends to select 'Support for RFC 7919 FFDHE group parameters' for using
FFDHE groups for NVMe In-Band Authentication.

It however selects CRYPTO_DH_GROUPS_RFC7919, instead of the intended
CRYPTO_DH_RFC7919_GROUPS; notice the swapping of words here.

Correct the select to the intended config option.

Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet-auth: fix return value check in auth receive

nvmet_auth_challenge() return type is int and currently it uses status
variable that is of type u16 in nvmet_execute_auth_receive().

Catch the return value of nvmet_auth_challenge() into int and set the
NVME_SC_INTERNAL as status variable before we jump to error.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet-auth: fix return value check in auth send

nvmet_setup_auth() return type is int and currently it uses status
variable that is of type u16 in nvmet_execute_auth_send().

Catch the return value of nvmet_setup_auth() into int and set the
NVME_SC_INTERNAL as status variable before we jump to error.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet-auth: fix a couple of spelling mistakes

There are a couple of spelling mistakes in pr_warn and pr_debug messages.
Fix them.

Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet: fix a format specifier in nvmet_auth_ctrl_exponential

dh_keysize is a size_t, use the proper format specifier for printing it.

Reported-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Hannes Reinecke <hare@sues.de>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet: don't check for NULL pointer before kfree in nvmet_host_release

And add an empty line after the variable declaration.

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-apple: stop casting function pointer signatures

Casting function pointers breaks control flow enforcement and is
generally a horrible coding style.

Add two wrappers to get rid of these casts.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Sven Peter <sven@svenpeter.dev>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-tcp: split nvme_tcp_alloc_tagset

Split nvme_tcp_alloc_tagset into one helper for the admin tag_set and
one for the I/O tag set.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-rdma: split nvme_rdma_alloc_tagset

Split nvme_rdma_alloc_tagset into one helper for the admin tag_set and
one for the I/O tag set.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-pci: split nvme_dev_add

Split nvme_dev_add into a helper to actually allocate the tag set, and
one that just update the number of queues.  Add a local variable for
the tag_set to clean up the code a bit.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-pci: split nvme_alloc_admin_tags

Split nvme_alloc_admin_tags into a helper to actually allocate the
tag set, and one that just restarts the admin queue.  Add a local
variable for the tag_set to clean up the code a bit.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-pci: print the command name of aborted commands

To allow for slightly better debugging, print the command name when
aborting an command.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-pci: remove useless assignment in nvme_pci_setup_prps

If prp_list is NULL, nvme_unmap_sg will be performed, and the assignment
to first_dma is meaningless, so remove it.

Signed-off-by: Liu Song <liusong@linux.alibaba.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-auth: uninitialized variable in nvme_auth_transform_key()

A couple of the early error gotos call kfree_sensitive(transformed_key);
before "transformed_key" has been initialized.

Fixes: db1312dd9548 ("nvmet: implement basic In-Band Authentication")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-auth: fix off by one checks

The > ARRAY_SIZE() checks need to be >= ARRAY_SIZE() to prevent reading
one element beyond the end of the arrays.

Fixes: db1312dd9548 ("nvmet: implement basic In-Band Authentication")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: define compat_ioctl again to unbreak 32-bit userspace.

Commit 89b3d6e60550 ("nvme: simplify the compat ioctl handling") removed
the initialization of compat_ioctl from the nvme block_device_operations
structures.

Presumably the expectation was that 32-bit ioctls would be directed
through the regular handler but this is not the case: failing to assign
.compat_ioctl actually means that the compat case is disabled entirely,
and any attempt to submit nvme ioctls from 32-bit userspace fails
outright with -ENOTTY.

For example:

  % smartctl -x /dev/nvme0n1
  [...]
  Read NVMe Identify Controller failed: NVME_IOCTL_ADMIN_CMD: Inappropriate ioctl for device

The blkdev_compat_ptr_ioctl helper can be used to direct compat calls
through the main ioctl handler and makes things work again.

Fixes: 89b3d6e60550 ("nvme: simplify the compat ioctl handling")
Signed-off-by: Nick Bowler <nbowler@draconx.ca>
Reviewed-by: Guixin Liu <kanie@linux.alibaba.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: don't always build constants.o

The entire content of constants.c if guarded by an ifdef, so switch to
just building the file conditionally instead.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: use command_id instead of req->tag in trace_nvme_complete_rq()

Use command_id instead of req->tag in trace_nvme_complete_rq(),
because of commit e7006de6c238 ("nvme: code command_id with a genctr
for use authentication after release"), cmd->common.command_id is set to
((genctl & 0xf)< 12 | req->tag), no longer req->tag, which makes cid in
trace_nvme_complete_rq and trace_nvme_setup_cmd are not the same.

Fixes: e7006de6c238 ("nvme: code command_id with a genctr for use authentication after release")
Signed-off-by: Bean Huo <beanhuo@micron.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md-raid10: fix KASAN warning

There's a KASAN warning in raid10_remove_disk when running the lvm
test lvconvert-raid-reshape.sh. We fix this warning by verifying that the
value "number" is valid.

BUG: KASAN: slab-out-of-bounds in raid10_remove_disk+0x61/0x2a0 [raid10]
Read of size 8 at addr ffff889108f3d300 by task mdX_raid10/124682

CPU: 3 PID: 124682 Comm: mdX_raid10 Not tainted 5.19.0-rc6 #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014
Call Trace:
 <TASK>
 dump_stack_lvl+0x34/0x44
 print_report.cold+0x45/0x57a
 ? __lock_text_start+0x18/0x18
 ? raid10_remove_disk+0x61/0x2a0 [raid10]
 kasan_report+0xa8/0xe0
 ? raid10_remove_disk+0x61/0x2a0 [raid10]
 raid10_remove_disk+0x61/0x2a0 [raid10]
Buffer I/O error on dev dm-76, logical block 15344, async page read
 ? __mutex_unlock_slowpath.constprop.0+0x1e0/0x1e0
 remove_and_add_spares+0x367/0x8a0 [md_mod]
 ? super_written+0x1c0/0x1c0 [md_mod]
 ? mutex_trylock+0xac/0x120
 ? _raw_spin_lock+0x72/0xc0
 ? _raw_spin_lock_bh+0xc0/0xc0
 md_check_recovery+0x848/0x960 [md_mod]
 raid10d+0xcf/0x3360 [raid10]
 ? sched_clock_cpu+0x185/0x1a0
 ? rb_erase+0x4d4/0x620
 ? var_wake_function+0xe0/0xe0
 ? psi_group_change+0x411/0x500
 ? preempt_count_sub+0xf/0xc0
 ? _raw_spin_lock_irqsave+0x78/0xc0
 ? __lock_text_start+0x18/0x18
 ? raid10_sync_request+0x36c0/0x36c0 [raid10]
 ? preempt_count_sub+0xf/0xc0
 ? _raw_spin_unlock_irqrestore+0x19/0x40
 ? del_timer_sync+0xa9/0x100
 ? try_to_del_timer_sync+0xc0/0xc0
 ? _raw_spin_lock_irqsave+0x78/0xc0
 ? __lock_text_start+0x18/0x18
 ? _raw_spin_unlock_irq+0x11/0x24
 ? __list_del_entry_valid+0x68/0xa0
 ? finish_wait+0xa3/0x100
 md_thread+0x161/0x260 [md_mod]
 ? unregister_md_personality+0xa0/0xa0 [md_mod]
 ? _raw_spin_lock_irqsave+0x78/0xc0
 ? prepare_to_wait_event+0x2c0/0x2c0
 ? unregister_md_personality+0xa0/0xa0 [md_mod]
 kthread+0x148/0x180
 ? kthread_complete_and_exit+0x20/0x20
 ret_from_fork+0x1f/0x30
 </TASK>

Allocated by task 124495:
 kasan_save_stack+0x1e/0x40
 __kasan_kmalloc+0x80/0xa0
 setup_conf+0x140/0x5c0 [raid10]
 raid10_run+0x4cd/0x740 [raid10]
 md_run+0x6f9/0x1300 [md_mod]
 raid_ctr+0x2531/0x4ac0 [dm_raid]
 dm_table_add_target+0x2b0/0x620 [dm_mod]
 table_load+0x1c8/0x400 [dm_mod]
 ctl_ioctl+0x29e/0x560 [dm_mod]
 dm_compat_ctl_ioctl+0x7/0x20 [dm_mod]
 __do_compat_sys_ioctl+0xfa/0x160
 do_syscall_64+0x90/0xc0
 entry_SYSCALL_64_after_hwframe+0x46/0xb0

Last potentially related work creation:
 kasan_save_stack+0x1e/0x40
 __kasan_record_aux_stack+0x9e/0xc0
 kvfree_call_rcu+0x84/0x480
 timerfd_release+0x82/0x140
L __fput+0xfa/0x400
 task_work_run+0x80/0xc0
 exit_to_user_mode_prepare+0x155/0x160
 syscall_exit_to_user_mode+0x12/0x40
 do_syscall_64+0x42/0xc0
 entry_SYSCALL_64_after_hwframe+0x46/0xb0

Second to last potentially related work creation:
 kasan_save_stack+0x1e/0x40
 __kasan_record_aux_stack+0x9e/0xc0
 kvfree_call_rcu+0x84/0x480
 timerfd_release+0x82/0x140
 __fput+0xfa/0x400
 task_work_run+0x80/0xc0
 exit_to_user_mode_prepare+0x155/0x160
 syscall_exit_to_user_mode+0x12/0x40
 do_syscall_64+0x42/0xc0
 entry_SYSCALL_64_after_hwframe+0x46/0xb0

The buggy address belongs to the object at ffff889108f3d200
 which belongs to the cache kmalloc-256 of size 256
The buggy address is located 0 bytes to the right of
 256-byte region [ffff889108f3d200, ffff889108f3d300)

The buggy address belongs to the physical page:
page:000000007ef2a34c refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x1108f3c
head:000000007ef2a34c order:2 compound_mapcount:0 compound_pincount:0
flags: 0x4000000000010200(slab|head|zone=2)
raw: 4000000000010200 0000000000000000 dead000000000001 ffff889100042b40
raw: 0000000000000000 0000000080200020 00000001ffffffff 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
 ffff889108f3d200: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 ffff889108f3d280: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>ffff889108f3d300: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
                   ^
 ffff889108f3d380: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
 ffff889108f3d400: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md-raid: destroy the bitmap after destroying the thread

When we ran the lvm test "shell/integrity-blocksize-3.sh" on a kernel with
kasan, we got failure in write_page.

The reason for the failure is that md_bitmap_destroy is called before
destroying the thread and the thread may be waiting in the function
write_page for the bio to complete. When the thread finishes waiting, it
executes "if (test_bit(BITMAP_WRITE_ERROR, &bitmap->flags))", which
triggers the kasan warning.

Note that the commit 48df498daf62 that caused this bug claims that it is
neede for md-cluster, you should check md-cluster and possibly find
another bugfix for it.

BUG: KASAN: use-after-free in write_page+0x18d/0x680 [md_mod]
Read of size 8 at addr ffff889162030c78 by task mdX_raid1/5539

CPU: 10 PID: 5539 Comm: mdX_raid1 Not tainted 5.19.0-rc2 #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014
Call Trace:
 <TASK>
 dump_stack_lvl+0x34/0x44
 print_report.cold+0x45/0x57a
 ? __lock_text_start+0x18/0x18
 ? write_page+0x18d/0x680 [md_mod]
 kasan_report+0xa8/0xe0
 ? write_page+0x18d/0x680 [md_mod]
 kasan_check_range+0x13f/0x180
 write_page+0x18d/0x680 [md_mod]
 ? super_sync+0x4d5/0x560 [dm_raid]
 ? md_bitmap_file_kick+0xa0/0xa0 [md_mod]
 ? rs_set_dev_and_array_sectors+0x2e0/0x2e0 [dm_raid]
 ? mutex_trylock+0x120/0x120
 ? preempt_count_add+0x6b/0xc0
 ? preempt_count_sub+0xf/0xc0
 md_update_sb+0x707/0xe40 [md_mod]
 md_reap_sync_thread+0x1b2/0x4a0 [md_mod]
 md_check_recovery+0x533/0x960 [md_mod]
 raid1d+0xc8/0x2a20 [raid1]
 ? var_wake_function+0xe0/0xe0
 ? psi_group_change+0x411/0x500
 ? preempt_count_sub+0xf/0xc0
 ? _raw_spin_lock_irqsave+0x78/0xc0
 ? __lock_text_start+0x18/0x18
 ? raid1_end_read_request+0x2a0/0x2a0 [raid1]
 ? preempt_count_sub+0xf/0xc0
 ? _raw_spin_unlock_irqrestore+0x19/0x40
 ? del_timer_sync+0xa9/0x100
 ? try_to_del_timer_sync+0xc0/0xc0
 ? _raw_spin_lock_irqsave+0x78/0xc0
 ? __lock_text_start+0x18/0x18
 ? __list_del_entry_valid+0x68/0xa0
 ? finish_wait+0xa3/0x100
 md_thread+0x161/0x260 [md_mod]
 ? unregister_md_personality+0xa0/0xa0 [md_mod]
 ? _raw_spin_lock_irqsave+0x78/0xc0
 ? prepare_to_wait_event+0x2c0/0x2c0
 ? unregister_md_personality+0xa0/0xa0 [md_mod]
 kthread+0x148/0x180
 ? kthread_complete_and_exit+0x20/0x20
 ret_from_fork+0x1f/0x30
 </TASK>

Allocated by task 5522:
 kasan_save_stack+0x1e/0x40
 __kasan_kmalloc+0x80/0xa0
 md_bitmap_create+0xa8/0xe80 [md_mod]
 md_run+0x777/0x1300 [md_mod]
 raid_ctr+0x249c/0x4a30 [dm_raid]
 dm_table_add_target+0x2b0/0x620 [dm_mod]
 table_load+0x1c8/0x400 [dm_mod]
 ctl_ioctl+0x29e/0x560 [dm_mod]
 dm_compat_ctl_ioctl+0x7/0x20 [dm_mod]
 __do_compat_sys_ioctl+0xfa/0x160
 do_syscall_64+0x90/0xc0
 entry_SYSCALL_64_after_hwframe+0x46/0xb0

Freed by task 5680:
 kasan_save_stack+0x1e/0x40
 kasan_set_track+0x21/0x40
 kasan_set_free_info+0x20/0x40
 __kasan_slab_free+0xf7/0x140
 kfree+0x80/0x240
 md_bitmap_free+0x1c3/0x280 [md_mod]
 __md_stop+0x21/0x120 [md_mod]
 md_stop+0x9/0x40 [md_mod]
 raid_dtr+0x1b/0x40 [dm_raid]
 dm_table_destroy+0x98/0x1e0 [dm_mod]
 __dm_destroy+0x199/0x360 [dm_mod]
 dev_remove+0x10c/0x160 [dm_mod]
 ctl_ioctl+0x29e/0x560 [dm_mod]
 dm_compat_ctl_ioctl+0x7/0x20 [dm_mod]
 __do_compat_sys_ioctl+0xfa/0x160
 do_syscall_64+0x90/0xc0
 entry_SYSCALL_64_after_hwframe+0x46/0xb0

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
Fixes: 48df498daf62 ("md: move bitmap_destroy to the beginning of __md_stop")
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: return the allocated devices from md_alloc

Two callers of md_alloc want to use the newly allocated devices, so
return it instead of letting them find it cumbersomely after the
allocation.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-and-tested-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: open code md_probe in autorun_devices

autorun_devices should not be limited to the controls for the legacy
probe on open, so just call md_alloc directly.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-and-tested-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: remove unneeded semicolon

Eliminate the following coccicheck warning:
./drivers/md/md.c:8208:2-3: Unneeded semicolon

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

remove the sx8 block driver

This driver is for fairly obscure hardware, and has only seen random
drive-by changes after the maintainer stopped working on it in 2005
(about a year and a half after it was introduced).  It has some
"interesting" block layer interactions, so let's just drop it unless
anyone complains.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220721064102.1715460-1-hch@lst.de
[axboe: fix date typo, it was in 2005, not 2015]
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: fix build failure for !MODULE

After merging the block tree, today's linux-next build (x86_64
allmodconfig) failed like this:

drivers/md/md.c:717:22: error: 'mddev_find' defined but not used [-Werror=unused-function]
  717 | static struct mddev *mddev_find(dev_t unit)
      |                      ^~~~~~~~~~
cc1: all warnings being treated as errors

Caused by commit

  4500d5c17910 ("md: simplify md_open")

Make mddev_find() available only for non-modular builds.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220721131132.070be166@canb.auug.org.au
Signed-off-by: Jens Axboe <axboe@kernel.dk>

raid5: fix duplicate checks for rdev->saved_raid_disk

'first' will always be greater than or equal to 0, it is unnecessary to
repeat the 0 check, clean it up.

Signed-off-by: Jackie Liu <liuyun01@kylinos.cn>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: simplify md_open

Now that devices are on the all_mddevs list until the gendisk is freed,
there can't be any duplicates.  Remove the global list lookup and just
grab a reference.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: only delete entries from all_mddevs when the disk is freed

This ensures device names don't get prematurely reused.  Instead add a
deleted flag to skip already deleted devices in mddev_get and other
places that only want to see live mddevs.

Reported-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: stop using for_each_mddev in md_exit

Just do a simple list_for_each_entry_safe on all_mddevs, and only grab a
reference when we drop the lock and delete the now unused for_each_mddev
macro.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: stop using for_each_mddev in md_notify_reboot

Just do a simple list_for_each_entry_safe on all_mddevs, and only grab a
reference when we drop the lock.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: stop using for_each_mddev in md_do_sync

Just do a plain list_for_each that only grabs a mddev reference in
the case where the thread sleeps and restarts the list iteration.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: factor out the rdev overlaps check from rdev_size_store

This splits the code into nicely readable chunks and also avoids
the refcount inc/dec manipulations.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: rename md_free to md_kobj_release

The md_free name is rather misleading, so pick a better one.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: implement ->free_disk

Ensure that all private data is only freed once all accesses are done.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: fix error handling in md_alloc

Error handling in md_alloc is a mess.  Untangle it to just free the mddev
directly before add_disk is called and thus the gendisk is globally
visible.  After that clear the hold flag and let the mddev_put take care
of cleaning up the mddev through the usual mechanisms.

Fixes: 5e55e2f5fc95 ("[PATCH] md: convert compile time warnings into runtime warnings")
Fixes: 9be68dd7ac0e ("md: add error handling support for add_disk()")
Fixes: 7ad1069166c0 ("md: properly unwind when failing to add the kobject in md_alloc")
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: fix mddev->kobj lifetime

Once a kobject is initialized, the containing object should not be
directly freed.  So delay initialization until it is added.  Also
remove the kobject_del call as the last put will remove the kobject as
well.  The explicitly delete isn't needed here, and dropping it will
simplify further fixes.

With this md_free now does not need to check that ->gendisk is non-NULL
as it is always set by the time that kobject_init is called on
mddev->kobj.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Convert prepare_to_wait() to wait_woken() api

raid5_get_active_stripe() can sleep in various situations and it
is called by make_stripe_request() while inside the
prepare_to_wait()/finish_wait() section. Nested waits like this are
not supported.

This was noticed while making other changes that add different sleeps
to raid5_get_active_stripe() that caused a WARNING with
CONFIG_DEBUG_ATOMIC_SLEEP.

No ill effects have been noticed with the code as is, but theoretically
a nested and here could cause a dead lock so it should be fixed.

To fix this, convert the prepare_to_wait() call to use wake_woken()
which supports nested sleeps.

Link: https://lwn.net/Articles/628628/
Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Fix sectors_to_do bitmap overflow in raid5_make_request()

For unaligned IO that have nearly maximum sectors, the number of stripes
will end up being one greater than the size of the bitmap. When this
happens, the last stripe in the IO will not be processed as it should
be, resulting in data corruption.

However, this is not normally seen when the backing block devices have
4K physical block sizes since the block layer will split the request
before that happens.

To fix this increase the bitmap size by one bit and ensure the full
number of stripes are checked when calling find_first_bit().

Reported-by: David Sloan <David.Sloan@eideticom.com>
Fixes: 7e55c60acfbb ("md/raid5: Pivot raid5_make_request()")
Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

bcache: remove EXPERIMENTAL for Kconfig option 'Asynchronous device registration'

The "Asynchronous device registration (EXPERIMENTAL)" Kconfig option is
for 2+ years, it is used when registration takes too much time for
massive amount of cached data, to avoid udev task timeout during boot
time.

Many users and products enable this Kconfig option for quite long time
(e.g. SUSE Linux) and it works as expected and no issue reported.

It is time to remove the "EXPERIMENTAL" tag from this Kconfig item.

Signed-off-by: Coly Li <colyli@suse.de>
Link: https://lore.kernel.org/r/20220719042724.8498-2-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nbd: add missing definition of pr_fmt

commit 1243172d5894 ("nbd: use pr_err to output error message") tries
to define pr_fmt and use short pr_err() to output error message,
however, the definition is missed.

This patch also remove existing "nbd:" inside pr_err().

Fixes: 1243172d5894 ("nbd: use pr_err to output error message")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Link: https://lore.kernel.org/r/20220723082427.3890655-1-yukuai1@huaweicloud.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

null_blk: fix ida error handling in null_add_dev()

There needs to be some error checking if ida_simple_get() fails.
Also call ida_free() if there are errors later.

Fixes: 94bc02e30fb8 ("nullb: use ida to manage index")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Link: https://lore.kernel.org/r/YtEhXsr6vJeoiYhd@kili
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-multipath: refactor nvme_mpath_add_disk

Pass anagrpid as second argument. This is prep patch that allows reusing
this function for supporting unknown command sets.

Signed-off-by: Joel Granados <j.granados@samsung.com>
Signed-off-by: Kanchan Joshi <joshi.k@samsung.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-apple: use nvme core helper to cancel requests in tagset

Use nvme core helper nvme_cancel_tagset and nvme_cancel_admin_tagset
instead of same logic code.

Signed-off-by: Guixin Liu <kanie@linux.alibaba.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Ruozhu Li <liruozhu@huawei.com>
Reviewed-by: Sven Peter <sven@svenpeter.dev>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-pci: use nvme core helper to cancel requests in tagset

Use nvme core helper nvme_cancel_tagset and nvme_cancel_admin_tagset
instead of same logic code.

Signed-off-by: Guixin Liu <kanie@linux.alibaba.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Ruozhu Li <liruozhu@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-tcp: use in-capsule data for I/O connect

Currently, command data is only sent in-capsule on the for admin or I/O
commands on queues that indicate support for it.  Send fabrics command
data in-capsule for I/O queues as well to avoid needing a separate
H2CData PDU for the connect command.

This is optimization. Without this change, we send the connect command
capsule and data in separate PDUs (CapsuleCmd and H2CData), and must wait
for the controller to respond with an R2T PDU before sending the H2CData.

With the change, we send a single CapsuleCmd PDU that includes the data.
This reduces the number of bytes (and likely packets) sent across the network,
and simplifies the send state machine handling in the driver.

Signed-off-by: Caleb Sander <csander@purestorage.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-rdma: remove timeout for getting RDMA-CM established event

In case many controllers start error recovery at the same time (i.e.,
when port is down and up), they may never succeed to reconnect again.
This is because the target can't handle all the connect requests at
three seconds (the arbitrary value set today). Even if some of the
connections are established, when a single queue fails to connect,
all the controller's queues are destroyed as well. So, on the
following reconnection attempts the number of connect requests may
remain the same. To fix this, remove the timeout and wait for RDMA-CM
event to abort/complete the connect request. RDMA-CM sends unreachable
event when a timeout of ~90 seconds is expired. This approach is used
at other RDMA-CM users like SRP and iSER at blocking mode. The commit
also renames NVME_RDMA_CONNECT_TIMEOUT_MS to NVME_RDMA_CM_TIMEOUT_MS.

Signed-off-by: Israel Rukshin <israelr@nvidia.com>
Reviewed-by: Max Gurtovoy <mgurtovoy@nvidia.com>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

null_blk: add configfs variables for 2 options

Allow setting via configfs these two options:

no_sched
shared_tag_bitmap

Previously these could only be activated as module parameters.

Still missing are:

shared_tags
timeout
requeue
init_hctx

Signed-off-by: Vincent Fu <vincent.fu@samsung.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20220708174943.87787-3-vincent.fu@samsung.com
[axboe: fold in nullb == NULL fix]
Signed-off-by: Jens Axboe <axboe@kernel.dk>

null_blk: add module parameters for 4 options

Add as module parameters these options:

memory_backed
discard
mbps
cache_size

Previously these could only be set via configfs.

Still missing is bad_blocks.

The kernel test robot found a documentation formatting issue in v1 of
this patch.

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Vincent Fu <vincent.fu@samsung.com>
Link: https://lore.kernel.org/r/20220708174943.87787-2-vincent.fu@samsung.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block/rnbd-srv: Replace sess_dev_list with index_idr

The structure rnbd_srv_session maintains a list and an xarray of
rnbd_srv_dev. There is no need to keep both as one of them can serve the
purpose.

Since one of the places where the lookup of rnbd_srv_dev using
rnbd_srv_session is IO path, an xarray would serve us better than a list
traversal. Hence remove sess_dev_list from rnbd_srv_session, and replace
its uses from xarray.

Signed-off-by: Md Haris Iqbal <haris.iqbal@ionos.com>
Reviewed-by: Aleksei Marov <aleksei.marov@ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@ionos.com>
Link: https://lore.kernel.org/r/20220707143122.460362-3-haris.iqbal@ionos.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block/rnbd-srv: Set keep_id to true after mutex_trylock

After setting keep_id if the mutex trylock fails, the keep_id stays set
for the rest of the sess_dev lifetime.

Therefore, set keep_id to true after mutex_trylock succeeds, so that a
failure of trylock does'nt touch keep_id.

Fixes: b168e1d85cf3 ("block/rnbd-srv: Prevent a deadlock generated by accessing sysfs in parallel")
Cc: gi-oh.kim@ionos.com
Signed-off-by: Md Haris Iqbal <haris.iqbal@ionos.com>
Signed-off-by: Jack Wang <jinpu.wang@ionos.com>
Link: https://lore.kernel.org/r/20220707143122.460362-2-haris.iqbal@ionos.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet-auth: expire authentication sessions

Each authentication step is required to be completed within the
KATO interval (or two minutes if not set). So add a workqueue function
to reset the transaction ID and the expected next protocol step;
this will automatically the next authentication command referring
to the terminated authentication.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet-auth: Diffie-Hellman key exchange support

Implement Diffie-Hellman key exchange using FFDHE groups for NVMe
In-Band Authentication.
This patch adds a new host configfs attribute 'dhchap_dhgroup' to
select the FFDHE group to use.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet: implement basic In-Band Authentication

Implement NVMe-oF In-Band authentication according to NVMe TPAR 8006.
This patch adds three additional configfs entries 'dhchap_key',
'dhchap_ctrl_key', and 'dhchap_hash' to the 'host' configfs directory.
The 'dhchap_key' and 'dhchap_ctrl_key' entries need to be in the ASCII
format as specified in NVMe Base Specification v2.0 section 8.13.5.8
'Secret representation'.
'dhchap_hash' defaults to 'hmac(sha256)', and can be written to to
switch to a different HMAC algorithm.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvmet: parse fabrics commands on io queues

Some fabrics commands can be sent via io queues, so add a new
function nvmet_parse_fabrics_io_cmd() and rename the existing
nvmet_parse_fabrics_cmd() to nvmet_parse_fabrics_admin_cmd().

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-auth: Diffie-Hellman key exchange support

Implement Diffie-Hellman key exchange using FFDHE groups
for NVMe In-Band Authentication.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: implement In-Band authentication

Implement NVMe-oF In-Band authentication according to NVMe TPAR 8006.
This patch adds two new fabric options 'dhchap_secret' to specify the
pre-shared key (in ASCII respresentation according to NVMe 2.0 section
8.13.5.8 'Secret representation') and 'dhchap_ctrl_secret' to specify
the pre-shared controller key for bi-directional authentication of both
the host and the controller.
Re-authentication can be triggered by writing the PSK into the new
controller sysfs attribute 'dhchap_secret' or 'dhchap_ctrl_secret'.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
[axboe: fold in clang build fix]
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-fabrics: decode 'authentication required' connect error

The 'connect' command might fail with NVME_SC_AUTH_REQUIRED, so we
should be decoding this error, too.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: add definitions for NVMe In-Band authentication

Add new definitions for NVMe In-band authentication as defined in
the NVMe Base Specification v2.0.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

lib/base64: RFC4648-compliant base64 encoding

Add RFC4648-compliant base64 encoding and decoding routines, based on
the base64url encoding in fs/crypto/fname.c.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Cc: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

crypto: add crypto_has_kpp()

Add helper function to determine if a given key-agreement protocol
primitive is supported.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

crypto: add crypto_has_shash()

Add helper function to determine if a given synchronous hash is supported.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme-loop: use nvme core helpers to cancel all requests in a tagset

A helper now exist, no need to open-code the same functionality.

Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: fix qid param blk_mq_alloc_request_hctx

Only caller of the __nvme_submit_sync_cmd() with qid value not equal to
NVME_QID_ANY is nvmf_connect_io_queues(), where qid value is alway set
to > 0.

[1] __nvme_submit_sync_cmd() callers with  qid parameter from :-

        Caller                  |   qid parameter
------------------------------------------------------
* nvme_fc_connect_io_queues()   |
   nvmf_connect_io_queue()      |      qid > 0
* nvme_rdma_start_io_queues()   |
   nvme_rdma_start_queue()      |
    nvmf_connect_io_queues()    |      qid > 0
* nvme_tcp_start_io_queues()    |
   nvme_tcp_start_queue()       |
    nvmf_connect_io_queues()    |      qid > 0
* nvme_loop_connect_io_queues() |
   nvmf_connect_io_queues()     |      qid > 0

When qid value of the function parameter __nvme_submit_sync_cmd() is > 0
from above callers, we use blk_mq_alloc_request_hctx(), where we pass
last parameter as 0 if qid functional parameter value is set to 0 with
conditional operators, see 1002 :-

991 int __nvme_submit_sync_cmd(struct request_queue *q, struct nvme_command *cmd,
 992                 union nvme_result *result, void *buffer, unsigned bufflen,
 993                 int qid, int at_head, blk_mq_req_flags_t flags)
 994 {
 995         struct request *req;
 996         int ret;
 997
 998         if (qid == NVME_QID_ANY)
 999                 req = blk_mq_alloc_request(q, nvme_req_op(cmd), flags);
1000         else
1001                 req = blk_mq_alloc_request_hctx(q, nvme_req_op(cmd), flags,
1002                                                 qid ? qid - 1 : 0);
1003

But qid function parameter value of the __nvme_submit_sync_cmd() will
never be 0 from above caller list see [1], and all the other callers of
__nvme_submit_sync_cmd() use NVME_QID_ANY as qid value :-
1. nvme_submit_sync_cmd()
2. nvme_features()
3. nvme_sec_submit()
4. nvmf_reg_read32()
5. nvmf_reg_read64()
6. nvmf_ref_write32()
7. nvmf_connect_admin_queue()

Remove the conditional operator to pass the qid as 0 in the call to
blk_mq_alloc_requst_hctx().

Signed-off-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: remove unused timeout parameter

The function __nvme_submit_sync_cmd() has following list of callers
that sets the timeout value to 0 :-

        Callers               |   Timeout value
------------------------------------------------
nvme_submit_sync_cmd()        |        0
nvme_features()               |        0
nvme_sec_submit()             |        0
nvmf_reg_read32()             |        0
nvmf_reg_read64()             |        0
nvmf_reg_write32()            |        0
nvmf_connect_admin_queue()    |        0
nvmf_connect_io_queue()       |        0

Remove the timeout function parameter from __nvme_submit_sync_cmd() and
adjust the rest of code accordingly.

Signed-off-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: handle the persistent internal error AER

In the NVM Express Revision 1.4 spec, Figure 145 describes possible
values for an AER with event type "Error" (value 000b). For a
Persistent Internal Error (value 03h), the host should perform a
controller reset.

Add support for this error using code that already exists for
doing a controller reset. As part of this support, introduce
two utility functions for parsing the AER type and subtype.

This new support was tested in a lab environment where we can
generate the persistent internal error on demand, and observe
both the Linux side and NVMe controller side to see that the
controller reset has been done.

Signed-off-by: Michael Kelley <mikelley@microsoft.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

nvme: remove a double word in a comment

Delete the redundant word 'be'.

Signed-off-by: Xiang wangx <wangxiang@cdjrlc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

rnbd-clt: make rnbd_clt_change_capacity return void

No need to checking the return value, make it return void.

Acked-by: Jack Wang <jinpu.wang@ionos.com>
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://lore.kernel.org/r/20220706133152.12058-9-guoqing.jiang@linux.dev
Signed-off-by: Jens Axboe <axboe@kernel.dk>

rnbd-clt: pass sector_t type for resize capacity

Let's change the parameter type to 'sector_t' then we don't need to cast
it from rnbd_clt_resize_dev_store, and update rnbd_clt_resize_disk too.

Acked-by: Jack Wang <jinpu.wang@ionos.com>
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://lore.kernel.org/r/20220706133152.12058-8-guoqing.jiang@linux.dev
Signed-off-by: Jens Axboe <axboe@kernel.dk>

rnbd-clt: check capacity inside rnbd_clt_change_capacity

Currently, process_msg_open_rsp checks if capacity changed or not before
call rnbd_clt_change_capacity while the checking also make sense for
rnbd_clt_resize_dev_store, let's move the checking into the function.

Acked-by: Jack Wang <jinpu.wang@ionos.com>
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://lore.kernel.org/r/20220706133152.12058-7-guoqing.jiang@linux.dev
Signed-off-by: Jens Axboe <axboe@kernel.dk>

rnbd-clt: adjust the layout of struct rnbd_clt_dev

While at it, let re-arrange the struct to remove holes.

Before, pahole reports

/* size: 232, cachelines: 4, members: 17 */
/* sum members: 224, holes: 2, sum holes: 8 */
/* last cacheline: 40 bytes */

After the change, the report changes to

/* size: 224, cachelines: 4, members: 17 */
/* last cacheline: 32 bytes */

Acked-by: Jack Wang <jinpu.wang@ionos.com>
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://lore.kernel.org/r/20220706133152.12058-6-guoqing.jiang@linux.dev
Signed-off-by: Jens Axboe <axboe@kernel.dk>

rnbd-clt: reduce the size of struct rnbd_clt_dev

Previously, both map and remap trigger rnbd_clt_set_dev_attr to set
some members in rnbd_clt_dev such as wc, fua and logical_block_size
etc, but those members are only useful for map scenario given the
setup_request_queue is only called from the path:

rnbd_clt_map_device -> rnbd_client_setup_device

Since rnbd_clt_map_device frees rsp after rnbd_client_setup_device,
we can pass rsp to rnbd_client_setup_device and it's callees, which
means queue's attributes can be set directly from relevant members
of rsp instead from rnbd_clt_dev.

After that, we can kill 11 members from rnbd_clt_dev, and we don't
need rnbd_clt_set_dev_attr either.

Acked-by: Jack Wang <jinpu.wang@ionos.com>
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://lore.kernel.org/r/20220706133152.12058-5-guoqing.jiang@linux.dev
Signed-off-by: Jens Axboe <axboe@kernel.dk>

rnbd-clt: kill read_only from struct rnbd_clt_dev

The member is not needed since we can call get_disk_ro to achieve the
same goal.

Acked-by: Jack Wang <jinpu.wang@ionos.com>
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://lore.kernel.org/r/20220706133152.12058-4-guoqing.jiang@linux.dev
Signed-off-by: Jens Axboe <axboe@kernel.dk>

rnbd-clt: don't free rsp in msg_open_conf for map scenario

For map scenario, rsp is freed in two places:

1. msg_open_conf frees rsp if rtrs_clt_request returns 0.

2. Otherwise, rsp is freed by the call sites of rtrs_clt_request.

Now, We'd like to control full lifecycle of rsp in rnbd_clt_map_device,
with that, it is feasible to pass rsp to rnbd_client_setup_device in
next commit.

For 1, it is possible to free rsp from the caller of send_usr_msg
because of the synchronization of iu->comp.wait. And we put iu later
in rnbd_clt_map_device to ensure order of release rsp and iu.

Acked-by: Jack Wang <jinpu.wang@ionos.com>
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://lore.kernel.org/r/20220706133152.12058-3-guoqing.jiang@linux.dev
Signed-off-by: Jens Axboe <axboe@kernel.dk>

rnbd-clt: open code send_msg_open in rnbd_clt_map_device

Let's open code it in rnbd_clt_map_device, then we can use information
from rsp to setup gendisk and request_queue in next commits. After that,
we can remove some members (wc, fua and max_hw_sectors etc) from struct
rnbd_clt_dev.

Acked-by: Jack Wang <jinpu.wang@ionos.com>
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://lore.kernel.org/r/20220706133152.12058-2-guoqing.jiang@linux.dev
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: null_blk: Use the bitmap API to allocate bitmaps

Use bitmap_zalloc()/bitmap_free() instead of hand-writing them.

It is less verbose and it improves the semantic.

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/7c4d3116ba843fc4a8ae557dd6176352a6cd0985.1656864320.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md: Fix spelling mistake in comments

There are 2 spelling mistakes in comments. Fix it.

Signed-off-by: Zhang Jiaming <jiaming@nfschina.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Increase restriction on max segments per request

The block layer defaults the maximum segments to 128, which means
requests tend to get split around the 512KB depending on how many
pages can be merged. There's no such restriction in the raid5 code
so increase the limit to USHRT_MAX so that larger requests can be
sent as one.

Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Improve debug prints

Add a debug print for raid5_make_request() so that each request is
printed and add the logical sector number to the debug print in
__add_stripe_bio().

Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Pivot raid5_make_request()

raid5_make_request() loops through every page in the request,
finds the appropriate stripe and adds the bio for that page in the
disk.

This causes a great deal of contention on the hash_lock and extra
work seeing each stripe must be found once for every data disk.

The number of times a stripe must be found can be reduced by pivoting
raid5_make_request() so that it loops through every stripe and then
loops through every disk in that stripe to see if the bio must be
added. This reduces the number of times the hash lock must be taken
by a factor equal to the number of data disks.

To accomplish this, the logical sectors that have already been added
must be tracked. Tracking them is done with a bitmap: the bits
for all pages are set at the start of the request and each bit
is cleared once the bio is added to a stripe.

Finding the next sector to be done is then just a call to
find_first_bit() so that sectors that have been done can simply be
skipped.

One minor downside is that the maximum sectors for a request must be
limited so that the bitmap can be appropriately sized on the stack.
This limit is arbitrarily chosen to be 256 stripe pages which works out
to 1MB if PAGE_SIZE == DEFAULT_STRIPE_SIZE. This doesn't actually
restrict the maximum request further seeing the default block queue
settings are used which restricts the number of segments to 128 (which
results in request sizes that are approximately 512KB).

Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

md/raid5: Check all disks in a stripe_head for reshape progress

When testing if a previous stripe has had reshape expand past it, use
the earliest or latest logical sector in all the disks for that stripe
head. This will allow adding multiple disks at a time in a subesquent
patch.

To do this cleaner, refactor the check into a helper function called
stripe_ahead_of_reshape().

Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Song Liu <song@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>