kernel/linux.git/drivers/md/dm.c, branch v4.14.286

dm: interlock pending dm_io and dm_wait_for_bios_completion

2022-05-12T10:17:10+00:00

commit 9f6dc633761006f974701d4c88da71ab68670749 upstream. Commit d208b89401e0 ("dm: fix mempool NULL pointer race when completing IO") didn't go far enough. When bio_end_io_acct ends the count of in-flight I/Os may reach zero and the DM device may be suspended. There is a possibility that the suspend races with dm_stats_account_io. Fix this by adding percpu "pending_io" counters to track outstanding dm_io. Move kicking of suspend queue to dm_io_dec_pending(). Also, rename md_in_flight_bios() to dm_in_flight_bios() and update it to iterate all pending_io counters. Fixes: d208b89401e0 ("dm: fix mempool NULL pointer race when completing IO") Cc: stable@vger.kernel.org Co-developed-by: Mikulas Patocka Signed-off-by: Mikulas Patocka Signed-off-by: Mike Snitzer Signed-off-by: Mikulas Patocka Reviewed-by: Mike Snitzer Signed-off-by: Greg Kroah-Hartman

dm: fix mempool NULL pointer race when completing IO

2022-05-12T10:17:10+00:00

commit d208b89401e073de986dc891037c5a668f5d5d95 upstream. dm_io_dec_pending() calls end_io_acct() first and will then dec md in-flight pending count. But if a task is swapping DM table at same time this can result in a crash due to mempool->elements being NULL: task1 task2 do_resume ->do_suspend ->dm_wait_for_completion bio_endio ->clone_endio ->dm_io_dec_pending ->end_io_acct ->wakeup task1 ->dm_swap_table ->__bind ->__bind_mempools ->bioset_exit ->mempool_exit ->free_io [ 67.330330] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000 ...... [ 67.330494] pstate: 80400085 (Nzcv daIf +PAN -UAO) [ 67.330510] pc : mempool_free+0x70/0xa0 [ 67.330515] lr : mempool_free+0x4c/0xa0 [ 67.330520] sp : ffffff8008013b20 [ 67.330524] x29: ffffff8008013b20 x28: 0000000000000004 [ 67.330530] x27: ffffffa8c2ff40a0 x26: 00000000ffff1cc8 [ 67.330535] x25: 0000000000000000 x24: ffffffdada34c800 [ 67.330541] x23: 0000000000000000 x22: ffffffdada34c800 [ 67.330547] x21: 00000000ffff1cc8 x20: ffffffd9a1304d80 [ 67.330552] x19: ffffffdada34c970 x18: 000000b312625d9c [ 67.330558] x17: 00000000002dcfbf x16: 00000000000006dd [ 67.330563] x15: 000000000093b41e x14: 0000000000000010 [ 67.330569] x13: 0000000000007f7a x12: 0000000034155555 [ 67.330574] x11: 0000000000000001 x10: 0000000000000001 [ 67.330579] x9 : 0000000000000000 x8 : 0000000000000000 [ 67.330585] x7 : 0000000000000000 x6 : ffffff80148b5c1a [ 67.330590] x5 : ffffff8008013ae0 x4 : 0000000000000001 [ 67.330596] x3 : ffffff80080139c8 x2 : ffffff801083bab8 [ 67.330601] x1 : 0000000000000000 x0 : ffffffdada34c970 [ 67.330609] Call trace: [ 67.330616] mempool_free+0x70/0xa0 [ 67.330627] bio_put+0xf8/0x110 [ 67.330638] dec_pending+0x13c/0x230 [ 67.330644] clone_endio+0x90/0x180 [ 67.330649] bio_endio+0x198/0x1b8 [ 67.330655] dec_pending+0x190/0x230 [ 67.330660] clone_endio+0x90/0x180 [ 67.330665] bio_endio+0x198/0x1b8 [ 67.330673] blk_update_request+0x214/0x428 [ 67.330683] scsi_end_request+0x2c/0x300 [ 67.330688] scsi_io_completion+0xa0/0x710 [ 67.330695] scsi_finish_command+0xd8/0x110 [ 67.330700] scsi_softirq_done+0x114/0x148 [ 67.330708] blk_done_softirq+0x74/0xd0 [ 67.330716] __do_softirq+0x18c/0x374 [ 67.330724] irq_exit+0xb4/0xb8 [ 67.330732] __handle_domain_irq+0x84/0xc0 [ 67.330737] gic_handle_irq+0x148/0x1b0 [ 67.330744] el1_irq+0xe8/0x190 [ 67.330753] lpm_cpuidle_enter+0x4f8/0x538 [ 67.330759] cpuidle_enter_state+0x1fc/0x398 [ 67.330764] cpuidle_enter+0x18/0x20 [ 67.330772] do_idle+0x1b4/0x290 [ 67.330778] cpu_startup_entry+0x20/0x28 [ 67.330786] secondary_start_kernel+0x160/0x170 Fix this by: 1) Establishing pointers to 'struct dm_io' members in dm_io_dec_pending() so that they may be passed into end_io_acct() _after_ free_io() is called. 2) Moving end_io_acct() after free_io(). Cc: stable@vger.kernel.org Signed-off-by: Jiazi Li Signed-off-by: Mike Snitzer Signed-off-by: Mikulas Patocka Reviewed-by: Mike Snitzer Signed-off-by: Greg Kroah-Hartman

dm: eliminate potential source of excessive kernel log noise

2021-01-23T14:48:44+00:00

commit 0378c625afe80eb3f212adae42cc33c9f6f31abf upstream. There wasn't ever a real need to log an error in the kernel log for ioctls issued with insufficient permissions. Simply return an error and if an admin/user is sufficiently motivated they can enable DM's dynamic debugging to see an explanation for why the ioctls were disallowed. Reported-by: Nir Soffer Fixes: e980f62353c6 ("dm: don't allow ioctls to targets that don't map to whole devices") Signed-off-by: Mike Snitzer Signed-off-by: Greg Kroah-Hartman

dm: use noio when sending kobject event

2020-07-22T07:22:19+00:00

commit 6958c1c640af8c3f40fa8a2eee3b5b905d95b677 upstream. kobject_uevent may allocate memory and it may be called while there are dm devices suspended. The allocation may recurse into a suspended device, causing a deadlock. We must set the noio flag when sending a uevent. The observed deadlock was reported here: https://www.redhat.com/archives/dm-devel/2020-March/msg00025.html Reported-by: Khazhismel Kumykov Reported-by: Tahsin Erdogan Reported-by: Gabriel Krisman Bertazi Signed-off-by: Mikulas Patocka Cc: stable@vger.kernel.org Signed-off-by: Mike Snitzer Signed-off-by: Greg Kroah-Hartman

dm: fix potential for q->make_request_fn NULL pointer

2020-02-14T21:32:23+00:00

commit 47ace7e012b9f7ad71d43ac9063d335ea3d6820b upstream. Move blk_queue_make_request() to dm.c:alloc_dev() so that q->make_request_fn is never NULL during the lifetime of a DM device (even one that is created without a DM table). Otherwise generic_make_request() will crash simply by doing: dmsetup create -n test mount /dev/dm-N /mnt While at it, move ->congested_data initialization out of dm.c:alloc_dev() and into the bio-based specific init method. Reported-by: Stefan Bader BugLink: https://bugs.launchpad.net/bugs/1860231 Fixes: ff36ab34583a ("dm: remove request-based logic from make_request_fn wrapper") Depends-on: c12c9a3c3860c ("dm: various cleanups to md->queue initialization code") Cc: stable@vger.kernel.org Signed-off-by: Mike Snitzer [smb: adjusted for context and dm_init_md_queue() exitsting in older kernels] Signed-off-by: Stefan Bader Signed-off-by: Greg Kroah-Hartman

dm: fix report zone remapping to account for partition offset

2018-10-18T07:16:24+00:00

commit 9864cd5dc54cade89fd4b0954c2e522841aa247c upstream. If dm-linear or dm-flakey are layered on top of a partition of a zoned block device, remapping of the start sector and write pointer position of the zones reported by a report zones BIO must be modified to account for the target table entry mapping (start offset within the device and entry mapping with the dm device). If the target's backing device is a partition of a whole disk, the start sector on the physical device of the partition must also be accounted for when modifying the zone information. However, dm_remap_zone_report() was not considering this last case, resulting in incorrect zone information remapping with targets using disk partitions. Fix this by calculating the target backing device start sector using the position of the completed report zones BIO and the unchanged position and size of the original report zone BIO. With this value calculated, the start sector and write pointer position of the target zones can be correctly remapped. Fixes: 10999307c14e ("dm: introduce dm_remap_zone_report()") Cc: stable@vger.kernel.org Signed-off-by: Damien Le Moal Signed-off-by: Mike Snitzer Signed-off-by: Greg Kroah-Hartman

dm: prevent DAX mounts if not supported

2018-07-11T14:29:23+00:00

commit dbc626597c39b24cefce09fbd8e9dea85869a801 upstream. Currently device_supports_dax() just checks to see if the QUEUE_FLAG_DAX flag is set on the device's request queue to decide whether or not the device supports filesystem DAX. Really we should be using bdev_dax_supported() like filesystems do at mount time. This performs other tests like checking to make sure the dax_direct_access() path works. We also explicitly clear QUEUE_FLAG_DAX on the DM device's request queue if any of the underlying devices do not support DAX. This makes the handling of QUEUE_FLAG_DAX consistent with the setting/clearing of most other flags in dm_table_set_restrictions(). Now that bdev_dax_supported() explicitly checks for QUEUE_FLAG_DAX, this will ensure that filesystems built upon DM devices will only be able to mount with DAX if all underlying devices also support DAX. Signed-off-by: Ross Zwisler Fixes: commit 545ed20e6df6 ("dm: add infrastructure for DAX support") Cc: stable@vger.kernel.org Acked-by: Dan Williams Reviewed-by: Toshi Kani Signed-off-by: Mike Snitzer Signed-off-by: Greg Kroah-Hartman

dm: set QUEUE_FLAG_DAX accordingly in dm_table_set_restrictions()

2018-07-11T14:29:23+00:00

commit ad3793fc3945173f64d82d05d3ecde41f6c0435c upstream. Rather than having DAX support be unique by setting it based on table type in dm_setup_md_queue(). Signed-off-by: Mike Snitzer Signed-off-by: Ross Zwisler Signed-off-by: Greg Kroah-Hartman

dm: correctly handle chained bios in dec_pending()

2018-02-22T14:42:29+00:00

commit 8dd601fa8317243be887458c49f6c29c2f3d719f upstream. dec_pending() is given an error status (possibly 0) to be recorded against a bio. It can be called several times on the one 'struct dm_io', and it is careful to only assign a non-zero error to io->status. However when it then assigned io->status to bio->bi_status, it is not careful and could overwrite a genuine error status with 0. This can happen when chained bios are in use. If a bio is chained beneath the bio that this dm_io is handling, the child bio might complete and set bio->bi_status before the dm_io completes. This has been possible since chained bios were introduced in 3.14, and has become a lot easier to trigger with commit 18a25da84354 ("dm: ensure bio submission follows a depth-first tree walk") as that commit caused dm to start using chained bios itself. A particular failure mode is that if a bio spans an 'error' target and a working target, the 'error' fragment will complete instantly and set the ->bi_status, and the other fragment will normally complete a little later, and will clear ->bi_status. The fix is simply to only assign io_error to bio->bi_status when io_error is not zero. Reported-and-tested-by: Milan Broz Cc: stable@vger.kernel.org (v3.14+) Signed-off-by: NeilBrown Signed-off-by: Mike Snitzer Signed-off-by: Greg Kroah-Hartman

dm: fix race between dm_get_from_kobject() and __dm_destroy()

2017-11-30T08:40:43+00:00

commit b9a41d21dceadf8104812626ef85dc56ee8a60ed upstream. The following BUG_ON was hit when testing repeat creation and removal of DM devices: kernel BUG at drivers/md/dm.c:2919! CPU: 7 PID: 750 Comm: systemd-udevd Not tainted 4.1.44 Call Trace: [] dm_get_from_kobject+0x34/0x3a [] dm_attr_show+0x2b/0x5e [] ? mutex_lock+0x26/0x44 [] sysfs_kf_seq_show+0x83/0xcf [] kernfs_seq_show+0x23/0x25 [] seq_read+0x16f/0x325 [] kernfs_fop_read+0x3a/0x13f [] __vfs_read+0x26/0x9d [] ? security_file_permission+0x3c/0x44 [] ? rw_verify_area+0x83/0xd9 [] vfs_read+0x8f/0xcf [] ? __fdget_pos+0x12/0x41 [] SyS_read+0x4b/0x76 [] system_call_fastpath+0x12/0x71 The bug can be easily triggered, if an extra delay (e.g. 10ms) is added between the test of DMF_FREEING & DMF_DELETING and dm_get() in dm_get_from_kobject(). To fix it, we need to ensure the test of DMF_FREEING & DMF_DELETING and dm_get() are done in an atomic way, so _minor_lock is used. The other callers of dm_get() have also been checked to be OK: some callers invoke dm_get() under _minor_lock, some callers invoke it under _hash_lock, and dm_start_request() invoke it after increasing md->open_count. Signed-off-by: Hou Tao Signed-off-by: Mike Snitzer Signed-off-by: Greg Kroah-Hartman