On Wed, 2019-03-27 at 09:40 +0800, Ming Lei wrote:
> On Tue, Mar 26, 2019 at 11:17 PM Laurence Oberman <
> loberman@xxxxxxxxxx> wrote:
> >
> > On Thu, 2019-03-21 at 08:44 -0400, Laurence Oberman wrote:
> > > On Wed, 2019-03-20 at 16:35 -0400, Laurence Oberman wrote:
> > > > On Wed, 2019-03-20 at 13:28 -0700, Bart Van Assche wrote:
> > > > > On Wed, 2019-03-20 at 11:11 -0400, Laurence Oberman wrote:
> > > > > > On Wed, 2019-03-20 at 09:45 -0400, Laurence Oberman wrote:
> > > > > > > Hello Bart, I hope all is well with you.
> > > > > > >
> > > > > > > Quick question
> > > > > > > preparing to test v5.1-rc2 SRP, my usual method is first
> > > > > > > validate
> > > > > > > the
> > > > > > > prior kernel I had in place.
> > > > > > > This had passed tests previously (5.0.0-rc2) but I had
> > > > > > > not
> > > > > > > run
> > > > > > > the
> > > > > > > target server reboot test, just the disconnect tests.
> > > > > > >
> > > > > > > Today with mapped SRP devices I rebooted the target
> > > > > > > server
> > > > > > > and
> > > > > > > the
> > > > > > > client panicked.
> > > > > > >
> > > > > > > Its been a while and I have been so busy that have not
> > > > > > > kept
> > > > > > > up
> > > > > > > with
> > > > > > > all
> > > > > > > the fixes. Is this a known issue.
> > > > > > >
> > > > > > > Thanks
> > > > > > > Laurence
> > > > > > >
> > > > > > > 5414228.917507] scsi host2: ib_srp: Path record query
> > > > > > > failed:
> > > > > > > sgid
> > > > > > > fe80:0000:0000:0000:7cfe:9003:0072:6ed3, dgid
> > > > > > > fe80:0000:0000:0000:7cfe:9003:0072:6e4f, pkey 0xffff,
> > > > > > > service_id
> > > > > > > 0x7cfe900300726e4e
> > > > > > > [5414229.014355] scsi host2: reconnect attempt 7 failed
> > > > > > > (-
> > > > > > > 110)
> > > > > > > [5414239.318161] scsi host2: ib_srp: Sending CM DREQ
> > > > > > > failed
> > > > > > > [5414239.318165] scsi host2: ib_srp: Sending CM DREQ
> > > > > > > failed
> > > > > > > [5414239.318167] scsi host2: ib_srp: Sending CM DREQ
> > > > > > > failed
> > > > > > > [5414239.318168] scsi host2: ib_srp: Sending CM DREQ
> > > > > > > failed
> > > > > > > [5414239.318170] scsi host2: ib_srp: Sending CM DREQ
> > > > > > > failed
> > > > > > > [5414239.318172] scsi host2: ib_srp: Sending CM DREQ
> > > > > > > failed
> > > > > > > [5414239.318173] scsi host2: ib_srp: Sending CM DREQ
> > > > > > > failed
> > > > > > > [5414239.318175] scsi host2: ib_srp: Sending CM DREQ
> > > > > > > failed
> > > > > > > [5414243.670072] scsi host2: ib_srp: Got failed path rec
> > > > > > > status
> > > > > > > -110
> > > > > > > [5414243.702179] scsi host2: ib_srp: Path record query
> > > > > > > failed:
> > > > > > > sgid
> > > > > > > fe80:0000:0000:0000:7cfe:9003:0072:6ed3, dgid
> > > > > > > fe80:0000:0000:0000:7cfe:9003:0072:6e4f, pkey 0xffff,
> > > > > > > service_id
> > > > > > > 0x7cfe900300726e4e
> > > > > > > [5414243.799313] scsi host2: reconnect attempt 8 failed
> > > > > > > (-
> > > > > > > 110)
> > > > > > > [5414247.510115] scsi host1: ib_srp: Sending CM REQ
> > > > > > > failed
> > > > > > > [5414247.510140] scsi host1: reconnect attempt 1 failed
> > > > > > > (-
> > > > > > > 104)
> > > > > > > [5414247.849078] BUG: unable to handle kernel NULL
> > > > > > > pointer
> > > > > > > dereference
> > > > > > > at 00000000000000b8
> > > > > > > [5414247.893793] #PF error: [normal kernel read fault]
> > > > > > > [5414247.921839] PGD 0 P4D 0
> > > > > > > [5414247.937280] Oops: 0000 [#1] SMP PTI
> > > > > > > [5414247.958332] CPU: 4 PID: 7773 Comm: kworker/4:1H
> > > > > > > Kdump:
> > > > > > > loaded
> > > > > > > Tainted: G I 5.0.0-rc2+ #2
> > > > > > > [5414248.012856] Hardware name: HP ProLiant DL380 G7,
> > > > > > > BIOS
> > > > > > > P67
> > > > > > > 08/16/2015
> > > > > > > [5414248.026174] device-mapper: multipath: Failing path
> > > > > > > 8:48.
> > > > > > >
> > > > > > >
> > > > > > > [5414248.050003] Workqueue: kblockd blk_mq_run_work_fn
> > > > > > > [5414248.108378] RIP:
> > > > > > > 0010:blk_mq_dispatch_rq_list+0xc9/0x590
> > > > > > > [5414248.139724] Code: 0f 85 c2 04 00 00 83 44 24 28 01
> > > > > > > 48 8b
> > > > > > > 45
> > > > > > > 00
> > > > > > > 48
> > > > > > > 39 c5 0f 84 ea 00 00 00 48 8b 5d 00 80 3c 24 00 4c 8d 6b
> > > > > > > b8
> > > > > > > 4c
> > > > > > > 8b
> > > > > > > 63
> > > > > > > c8
> > > > > > > 75 25 <49> 8b 84 24 b8 00 00 00 48 8b 40 40 48 8b 40 10
> > > > > > > 48 85
> > > > > > > c0
> > > > > > > 74
> > > > > > > 10
> > > > > > > 4c
> > > > > > > [5414248.246176] RSP: 0018:ffffb1cd8760fd90 EFLAGS:
> > > > > > > 00010246
> > > > > > > [5414248.275599] RAX: ffffa049d67a1308 RBX:
> > > > > > > ffffa049d67a1308
> > > > > > > RCX:
> > > > > > > 0000000000000004
> > > > > > > [5414248.316090] RDX: 0000000000000000 RSI:
> > > > > > > ffffb1cd8760fe20
> > > > > > > RDI:
> > > > > > > ffffa0552ca08000
> > > > > > > [5414248.356884] RBP: ffffb1cd8760fe20 R08:
> > > > > > > 0000000000000000
> > > > > > > R09:
> > > > > > > 8080808080808080
> > > > > > > [5414248.397632] R10: 0000000000000001 R11:
> > > > > > > 0000000000000001
> > > > > > > R12:
> > > > > > > 0000000000000000
> > > > > > > [5414248.439323] R13: ffffa049d67a12c0 R14:
> > > > > > > 0000000000000000
> > > > > > > R15:
> > > > > > > ffffa0552ca08000
> > > > > > > [5414248.481743] FS: 0000000000000000(0000)
> > > > > > > GS:ffffa04a37880000(0000)
> > > > > > > knlGS:0000000000000000
> > > > > > > [5414248.528310] CS: 0010 DS: 0000 ES: 0000 CR0:
> > > > > > > 0000000080050033
> > > > > > > [5414248.561779] CR2: 00000000000000b8 CR3:
> > > > > > > 0000000e9d40e004
> > > > > > > CR4:
> > > > > > > 00000000000206e0
> > > > > > > [5414248.602420] Call Trace:
> > > > > > > [5414248.616660] blk_mq_sched_dispatch_requests+0x15c/0x
> > > > > > > 180
> > > > > > > [5414248.647066] __blk_mq_run_hw_queue+0x5f/0xf0
> > > > > > > [5414248.672633] process_one_work+0x171/0x370
> > > > > > > [5414248.695443] worker_thread+0x49/0x3f0
> > > > > > > [5414248.716730] kthread+0xf8/0x130
> > > > > > > [5414248.735085] ? max_active_store+0x80/0x80
> > > > > > > [5414248.758569] ? kthread_bind+0x10/0x10
> > > > > > > [5414248.779953] ret_from_fork+0x35/0x40
> > > > > > >
> > > > > > > [5414248.801005] Modules linked in: ib_isert
> > > > > > > iscsi_target_mod
> > > > > > > target_core_mod ib_srp rpcrdma scsi_transport_srp
> > > > > > > rdma_ucm
> > > > > > > ib_iser
> > > > > > > ib_ipoib ib_umad rdma_cm libiscsi iw_cm
> > > > > > > scsi_transport_iscsi
> > > > > > > ib_cm
> > > > > > > sunrpc mlx5_ib ib_uverbs ib_core intel_powerclamp
> > > > > > > coretemp
> > > > > > > kvm_intel
> > > > > > > kvm irqbypass ipmi_ssif crct10dif_pclmul crc32_pclmul
> > > > > > > ghash_clmulni_intel aesni_intel iTCO_wdt crypto_simd
> > > > > > > cryptd
> > > > > > > gpio_ich
> > > > > > > iTCO_vendor_support glue_helper joydev ipmi_si
> > > > > > > dm_service_time
> > > > > > > pcspkr
> > > > > > > ipmi_devintf hpilo hpwdt sg ipmi_msghandler
> > > > > > > acpi_power_meter
> > > > > > > lpc_ich
> > > > > > > i7core_edac pcc_cpufreq dm_multipath ip_tables xfs
> > > > > > > libcrc32c
> > > > > > > radeon
> > > > > > > sd_mod i2c_algo_bit drm_kms_helper syscopyarea
> > > > > > > sysfillrect
> > > > > > > sysimgblt
> > > > > > > fb_sys_fops ttm drm mlx5_core crc32c_intel serio_raw
> > > > > > > i2c_core
> > > > > > > hpsa
> > > > > > > bnx2
> > > > > > > scsi_transport_sas mlxfw devlink dm_mirror dm_region_hash
> > > > > > > dm_log
> > > > > > > dm_mod
> > > > > > > [5414249.199354] CR2: 00000000000000b8
> > > > > > >
> > > > > > >
> > > > > >
> > > > > > Looking at the vmcore
> > > > > >
> > > > > > PID: 7773 TASK: ffffa04a2c1e2b80 CPU: 4 COMMAND:
> > > > > > "kworker/4:1H"
> > > > > > #0 [ffffb1cd8760fab0] machine_kexec at ffffffffaaa6003f
> > > > > > #1 [ffffb1cd8760fb08] __crash_kexec at ffffffffaab373ed
> > > > > > #2 [ffffb1cd8760fbd0] crash_kexec at ffffffffaab385b9
> > > > > > #3 [ffffb1cd8760fbe8] oops_end at ffffffffaaa31931
> > > > > > #4 [ffffb1cd8760fc08] no_context at ffffffffaaa6eb59
> > > > > > #5 [ffffb1cd8760fcb0] do_page_fault at ffffffffaaa6feb2
> > > > > > #6 [ffffb1cd8760fce0] page_fault at ffffffffab2010ee
> > > > > > [exception RIP: blk_mq_dispatch_rq_list+201]
> > > > > > RIP: ffffffffaad90589 RSP: ffffb1cd8760fd90 RFLAGS:
> > > > > > 00010246
> > > > > > RAX: ffffa049d67a1308 RBX: ffffa049d67a1308 RCX:
> > > > > > 0000000000000004
> > > > > > RDX: 0000000000000000 RSI: ffffb1cd8760fe20 RDI:
> > > > > > ffffa0552ca08000
> > > > > > RBP: ffffb1cd8760fe20 R8: 0000000000000000 R9:
> > > > > > 8080808080808080
> > > > > > R10: 0000000000000001 R11: 0000000000000001 R12:
> > > > > > 0000000000000000
> > > > > > R13: ffffa049d67a12c0 R14: 0000000000000000 R15:
> > > > > > ffffa0552ca08000
> > > > > > ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018
> > > > > > #7 [ffffb1cd8760fe18] blk_mq_sched_dispatch_requests at
> > > > > > ffffffffaad9570c
> > > > > > #8 [ffffb1cd8760fe60] __blk_mq_run_hw_queue at
> > > > > > ffffffffaad8de3f
> > > > > > #9 [ffffb1cd8760fe78] process_one_work at ffffffffaaab0ab1
> > > > > > #10 [ffffb1cd8760feb8] worker_thread at ffffffffaaab11d9
> > > > > > #11 [ffffb1cd8760ff10] kthread at ffffffffaaab6758
> > > > > > #12 [ffffb1cd8760ff50] ret_from_fork at ffffffffab200215
> > > > > >
> > > > > > We were working on this request_queue for
> > > > > > blk_mq_sched_dispatch_requests
> > > > > >
> > > > > > crash> dev -d | grep ffffa0552ca08000
> > > > > > 8
> > > > > > ffffa055c81b5800 sdd ffffa0552ca08000 0
> > > > > > 0
> > > > > >
> > > > > > 0
> > > > > > ]
> > > > > >
> > > > > > That device was no longer accessible
> > > > > >
> > > > > > sdev_state = SDEV_TRANSPORT_OFFLINE,
> > > > > >
> > > > > > So it looks like we tried to process a no longer valid list
> > > > > > entry
> > > > > > in
> > > > > > blk_mq_dispatch_rq_list
> > > > > >
> > > > > > /home/loberman/rpmbuild/BUILD/kernel-5.0.0_rc2+/block/blk-
> > > > > > mq.h:
> > > > > > 211
> > > > > > 0xffffffffaad90589
> > > > > > <blk_mq_dispatch_rq_list+201>: mov 0xb8(%r12),%rax
> > > > > >
> > > > > > R12 is NULL
> > > > > >
> > > > > >
> > > > > > From
> > > > > > static inline bool blk_mq_get_dispatch_budget(struct
> > > > > > blk_mq_hw_ctx
> > > > > > *hctx)
> > > > > > {
> > > > > > struct request_queue *q = hctx->queue;
> > > > > >
> > > > > > if (q->mq_ops->get_budget)
> > > > > > return q->mq_ops->get_budget(hctx);
> > > > > > return true;
> > > > > > }
> > > > > >
> > > > > > Willw ait for a reply befaore i try the newer kernel, but
> > > > > > looks
> > > > > > like a
> > > > > > use after free to me
> > > > >
> > > > > Hi Laurence,
> > > > >
> > > > > I don't think that any of the recent SRP initiator changes
> > > > > can be
> > > > > the
> > > > > root
> > > > > cause of this crash. However, significant changes went
> > > > > upstream
> > > > > in
> > > > > the block
> > > > > layer core during the v5.1-rc1 merge window, e.g. multi-page
> > > > > bvec
> > > > > support.
> > > > > Is it possible for you to bisect this kernel oops?
> > > > >
> > > > > Thanks,
> > > > >
> > > > > Bart.
> > > > >
> > > >
> > > > OK, I will see I can reproduce on the fly and I will bisect.
> > > > I do agree its not SRP, more likley some block layer race.
> > > > I was just able to reproduce it using SRP
> > > >
> > > > Note that this was on 5.0.0-rc2+, prior to me trying 5.1.
> > > >
> > > > I usually reboot the target server as part of my test series
> > > > but
> > > > when
> > > > I
> > > > last tested 5.0.0-rc2+ I only reset the SRP interfaces and had
> > > > devices
> > > > get rediscovered.
> > > >
> > > > I did not see it during those tests.
> > > >
> > > > Back when I have more to share.
> > > >
> > > > Many Thanks for your time as always
> > > > Laurence
> > > >
> > > >
> > >
> > > Something crept in, in the block layer causing a use after free
> > > 4.19.0-rc1+ does not have the issue, so I will narrow the bisect
> > > Thanks
> > > Laurence
> > >
> >
> > This took a long time to bisect.
> > Repeating the issue seen. We have changes that when the target is
> > rebooted with mapped srp devices the client then experiences a ist
> > corruption and panic as already shown.
> >
> > Some stacks
> >
> > [ 222.631998] scsi host1: ib_srp: Path record query failed: sgid
> > fe80:0000:0000:0000:7cfe:9003:0072:6ed2, dgid
> > fe80:0000:0000:0000:7cfe:9003:0072:6e4e, pkey 0xffff, service_id
> > 0x7cfe900300726e4e
> > [ 222.729639] scsi host1: reconnect attempt 1 failed (-110)
> > [ 223.176766] fast_io_fail_tmo expired for SRP port-1:1 / host1.
> >
> > [ 223.518759] BUG: unable to handle kernel NULL pointer
> > dereference at
> > 00000000000000b8
> > [ 223.519736] sd 1:0:0:0: rejecting I/O to offline device
> > [ 223.563769] #PF error: [normal kernel read fault]
> > [ 223.563770] PGD 0 P4D 0
> > [ 223.563774] Oops: 0000 [#1] SMP PTI
> > [ 223.563778] CPU: 3 PID: 9027 Comm: kworker/3:1H Tainted:
> > G I 5.0.0-rc1 #22
> > [ 223.563779] Hardware name: HP ProLiant DL380 G7, BIOS P67
> > 08/16/2015
> > [ 223.563787] Workqueue: kblockd blk_mq_run_work_fn
> > [ 223.593723] device-mapper: multipath: Failing path 8:48.
> > [ 223.620801] RIP: 0010:blk_mq_dispatch_rq_list+0xc9/0x590
> > [ 223.635266] print_req_error: I/O error, dev dm-6, sector 8191872
> > flags 80700
> > [ 223.655565] Code: 0f 85 c2 04 00 00 83 44 24 28 01 48 8b 45 00
> > 48 39
> > c5 0f 84 ea 00 00 00 48 8b 5d 00 80 3c 24 00 4c 8d 6b b8 4c 8b 63
> > c8 75
> > 25 <49> 8b 84 24 b8 00 00 00 48 8b 40 40 48 8b 40 10 48 85 c0 74 10
> > 4c
> > [ 223.655566] RSP: 0018:ffffa65b4c43fd90 EFLAGS: 00010246
> > [ 223.655570] RAX: ffff93ed9bfdbbc8 RBX: ffff93ed9bfdbbc8 RCX:
> > 0000000000000004
> > [ 223.702351] print_req_error: I/O error, dev dm-6, sector 8191872
> > flags 0
> > [ 223.737640] RDX: 0000000000000000 RSI: ffffa65b4c43fe20 RDI:
> > ffff93ed9b838000
> > [ 223.737641] RBP: ffffa65b4c43fe20 R08: 0000000000000000 R09:
> > 8080808080808080
> > [ 223.737642] R10: 0000000000000001 R11: 071c71c71c71c71c R12:
> > 0000000000000000
> > [ 223.737643] R13: ffff93ed9bfdbb80 R14: 0000000000000000 R15:
> > ffff93ed9b838000
> > [ 223.737645] FS: 0000000000000000(0000)
> > GS:ffff93ee33840000(0000)
> > knlGS:0000000000000000
> > [ 223.737646] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > [ 223.737646] CR2: 00000000000000b8 CR3: 000000059da0e006 CR4:
> > 00000000000206e0
> > [ 223.737648] Call Trace:
> >
> > [ 223.737657] blk_mq_sched_dispatch_requests+0x15c/0x180 ***
> > Freed
> > already
> >
> > [ 223.737660] __blk_mq_run_hw_queue+0x5f/0xf0
> > [ 223.737665] process_one_work+0x171/0x370
> > [ 223.737667] worker_thread+0x49/0x3f0
> > [ 223.737670] kthread+0xf8/0x130
> > [ 223.737673] ? max_active_store+0x80/0x80
> >
> > And:
> >
> > [ 643.425005] device-mapper: multipath: Failing path 67:0.
> > [ 643.696365] ------------[ cut here ]------------
> > [ 643.722927] list_add corruption. prev->next should be next
> > (ffffc8c0c3bd9448), but was ffff93b971965e08.
> > (prev=ffff93b971965e08).
> > [ 643.787089] WARNING: CPU: 14 PID: 6533 at lib/list_debug.c:28
> > __list_add_valid+0x6a/0x70
> > [ 643.830951] Modules linked in: ib_isert iscsi_target_mod
> > target_core_mod ib_srp scsi_transport_srp rpcrdma ib_ipoib ib_umad
> > rdma_ucm ib_iser rdma_cm iw_cm libiscsi sunrpc scsi_transport_iscsi
> > ib_cm mlx5_ib ib_uverbs ib_core intel_powerclamp coretemp kvm_intel
> > kvm
> > irqbypass crct10dif_pclmul crc32_pclmul iTCO_wdt
> > ghash_clmulni_intel
> > ipmi_ssif aesni_intel iTCO_vendor_support gpio_ich crypto_simd
> > dm_service_time cryptd glue_helper joydev ipmi_si ipmi_devintf
> > pcspkr
> > hpilo hpwdt sg ipmi_msghandler acpi_power_meter pcc_cpufreq lpc_ich
> > i7core_edac dm_multipath ip_tables xfs libcrc32c radeon sd_mod
> > i2c_algo_bit drm_kms_helper syscopyarea sysfillrect sysimgblt
> > fb_sys_fops ttm drm crc32c_intel serio_raw mlx5_core i2c_core bnx2
> > hpsa
> > mlxfw scsi_transport_sas dm_mirror dm_region_hash dm_log dm_mod
> > [ 644.224150] CPU: 14 PID: 6533 Comm: kworker/14:1H Tainted:
> > G I 4.20.0+ #26
> > [ 644.269637] Hardware name: HP ProLiant DL380 G7, BIOS P67
> > 08/16/2015
> > [ 644.305110] Workqueue: kblockd blk_mq_run_work_fn
> > [ 644.330984] RIP: 0010:__list_add_valid+0x6a/0x70
> > [ 644.357601] Code: 31 c0 48 c7 c7 70 e4 ab b6 e8 22 a6 cc ff 0f
> > 0b 31
> > c0 c3 48 89 d1 48 c7 c7 20 e4 ab b6 48 89 f2 48 89 c6 31 c0 e8 06
> > a6 cc
> > ff <0f> 0b 31 c0 c3 90 48 8b 07 48 b9 00 01 00 00 00 00 ad de 48 8b
> > 57
> > [ 644.462542] RSP: 0018:ffffa8ccc72bfc00 EFLAGS: 00010286
> > [ 644.491594] RAX: 0000000000000000 RBX: ffff93b971965dc0 RCX:
> > 0000000000000000
> > [ 644.532745] RDX: 0000000000000001 RSI: ffff93b9b79d67b8 RDI:
> > ffff93b9b79d67b8
> > [ 644.573533] RBP: ffffc8c0c3bd9448 R08: 0000000000000000 R09:
> > 000000000000072c
> > [ 644.614180] R10: 0000000000000000 R11: ffffa8ccc72bf968 R12:
> > ffff93b96d454c00
> > [ 644.654683] R13: ffffc8c0c3bd9440 R14: ffff93b971965e08 R15:
> > ffff93b971965e08
> > [ 644.694275] FS: 0000000000000000(0000)
> > GS:ffff93b9b79c0000(0000)
> > knlGS:0000000000000000
> > [ 644.739906] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > [ 644.771879] CR2: 00007fd566c10000 CR3: 0000000253e0e002 CR4:
> > 00000000000206e0
> > [ 644.811438] Call Trace:
> > [ 644.824809] __blk_mq_insert_request+0x62/0x130
> > [ 644.849886] blk_mq_sched_insert_request+0x13c/0x1b0
> > [ 644.877402] blk_mq_try_issue_directly+0x105/0x2c0
> > [ 644.904452] blk_insert_cloned_request+0x9a/0x130
> > [ 644.931146] ? ktime_get+0x37/0x90
> > [ 644.950545] dm_mq_queue_rq+0x21c/0x3f0 [dm_mod]
> > [ 644.977064] ? blk_mq_get_driver_tag+0xa1/0x120
> > [ 645.003002] blk_mq_dispatch_rq_list+0x8e/0x590
> > [ 645.029812] ? __switch_to_asm+0x40/0x70
> > [ 645.052059] ? __switch_to_asm+0x34/0x70
> > [ 645.074664] ? __switch_to_asm+0x40/0x70
> > [ 645.097111] ? __switch_to_asm+0x34/0x70
> > [ 645.119948] ? __switch_to_asm+0x34/0x70
> > [ 645.143101] ? __switch_to_asm+0x40/0x70
> > [ 645.165273] ? syscall_return_via_sysret+0x10/0x7f
> > [ 645.192161] blk_mq_sched_dispatch_requests+0xe8/0x180
> > [ 645.221460] __blk_mq_run_hw_queue+0x5f/0xf0
> > [ 645.244766] process_one_work+0x171/0x370
> > [ 645.267164] worker_thread+0x49/0x3f0
> > [ 645.287860] kthread+0xf8/0x130
> > [ 645.304894] ? max_active_store+0x80/0x80
> > [ 645.327748] ? kthread_bind+0x10/0x10
> > [ 645.347898] ret_from_fork+0x35/0x40
> > [ 645.368298] ---[ end trace afa70bf68ffb006b ]---
> > [ 645.397356] ------------[ cut here ]------------
> > [ 645.423878] list_add corruption. prev->next should be next
> > (ffffc8c0c3bd9448), but was ffff93b971965e08.
> > (prev=ffff93b971965e08).
> > [ 645.488009] WAR
> > NING: CPU: 14 PID: 6533 at lib/list_debug.c:28
> > __list_add_valid+0x6a/0x70
> >
> >
> > I started just looking at block but nothing made sense so re-ran
> > the
> > bisect with the entire kernel.
> >
> > $ git bisect start
> > $ git bisect good v4.20
> > $ git bisect bad v5.0-rc1
> >
> > Bisecting: 5477 revisions left to test after this (roughly 13
> > steps)
> > *** Groan
> >
> > I got to here and the problem is its an entire merge
> >
> > [loberman@ibclient linux_torvalds]$ git bisect bad
> > 4b9254328254bed12a4ac449cdff2c332e630837 is the first bad commit
> > [loberman@ibclient linux_torvalds]$ git show
> > 4b9254328254bed12a4ac449cdff2c332e630837
> > commit 4b9254328254bed12a4ac449cdff2c332e630837
> > Merge: 1a9430d cd19181
> > Author: Jens Axboe <axboe@xxxxxxxxx>
> > Date: Tue Dec 18 08:29:53 2018 -0700
> >
> > Merge branch 'for-4.21/block' into for-4.21/aio
> >
> > * for-4.21/block: (351 commits)
> > blk-mq: enable IO poll if .nr_queues of type poll > 0
>
> You may set 4b9254328254b or 'cd19181bf9ad' as 'git bad' and start
> the
> 'git bisect'
> again.
>
> thanks,
> Ming Lei
Hello Ming, OK, but that means starting good at v4.20. long bisect
again.
>From the vmcore I got to this analysis
[ 382.412285] fast_io_fail_tmo expired for SRP port-2:1 / host2.
[ 382.604347] scsi host2: ib_srp: Got failed path rec status -110
[ 382.638622] scsi host2: ib_srp: Path record query failed: sgid
fe80:0000:0000:0000:7cfe:9003:0072:6ed2, dgid
fe80:0000:0000:0000:7cfe:9003:0072:6e4e, pkey 0xffff, service_id
0x7cfe900300726e4e
[ 382.736300] scsi host2: reconnect attempt 1 failed (-110)
[ 383.239347] BUG: unable to handle kernel NULL pointer dereference at
00000000000000b8
[ 383.239349] sd 2:0:0:0: rejecting I/O to offline device
[ 383.239370] device-mapper: multipath: Failing path 8:64.
[ 383.241278] sd 2:0:0:28: rejecting I/O to offline device
[ 383.241284] sd 2:0:0:27: rejecting I/O to offline device
[ 383.241289] device-mapper: multipath: Failing path 8:96.
[ 383.241291] device-mapper: multipath: Failing path 8:160.
[ 383.241301] print_req_error: I/O error, dev dm-8, sector 8191872
flags 80700
[ 383.241303] print_req_error: I/O error, dev dm-7, sector 8191872
flags 80700
[ 383.241335] print_req_error: I/O error, dev dm-8, sector 8191872
flags 0
[ 383.241338] Buffer I/O error on dev dm-8, logical block 8191872,
async page read
[ 383.241355] print_req_error: I/O error, dev dm-7, sector 8191872
flags 0
[ 383.241357] Buffer I/O error on dev dm-7, logical block 8191872,
async page read
[ 383.241367] print_req_error: I/O error, dev dm-7, sector 8191873
flags 0
[ 383.241368] Buffer I/O error on dev dm-7, logical block 8191873,
async page read
[ 383.241377] print_req_error: I/O error, dev dm-7, sector 8191874
flags 0
[ 383.241378] Buffer I/O error on dev dm-7, logical block 8191874,
async page read
[ 383.241387] print_req_error: I/O error, dev dm-7, sector 8191875
flags 0
[ 383.241388] Buffer I/O error on dev dm-7, logical block 8191875,
async page read
[ 383.241399] print_req_error: I/O error, dev dm-8, sector 8191873
flags 0
[ 383.241400] Buffer I/O error on dev dm-8, logical block 8191873,
async page read
[ 383.241409] print_req_error: I/O error, dev dm-7, sector 8191876
flags 0
[ 383.241410] Buffer I/O error on dev dm-7, logical block 8191876,
async page read
[ 383.241421] print_req_error: I/O error, dev dm-8, sector 8191874
flags 0
[ 383.241422] Buffer I/O error on dev dm-8, logical block 8191874,
async page read
[ 383.241431] Buffer I/O error on dev dm-7, logical block 8191877,
async page read
[ 383.241440] Buffer I/O error on dev dm-8, logical block 8191875,
async page read
[ 383.282566] #PF error: [normal kernel read fault]
[ 384.290154] PGD 0 P4D 0
[ 384.303791] Oops: 0000 [#1] SMP PTI
[ 384.323687] CPU: 1 PID: 9191 Comm: kworker/1:1H Kdump: loaded
Tainted: G W I 5.1.0-rc2+ #39
[ 384.375898] Hardware name: HP ProLiant DL380 G7, BIOS P67 08/16/2015
[ 384.411812] Workqueue: kblockd blk_mq_run_work_fn
[ 384.438498] RIP: 0010:blk_mq_dispatch_rq_list+0x72/0x570
[ 384.468539] Code: 08 84 d2 0f 85 cf 03 00 00 45 31 f6 c7 44 24 38 00
00 00 00 c7 44 24 3c 00 00 00 00 80 3c 24 00 4c 8d 6b b8 48 8b 6b c8 75
24 <48> 8b 85 b8 00 00 00 48 8b 40 40 48 8b 40 10 48 85 c0 74 10 48 89
[ 384.573942] RSP: 0018:ffffa9fe0759fd90 EFLAGS: 00010246
[ 384.604181] RAX: ffff9de9c4d3bbc8 RBX: ffff9de9c4d3bbc8 RCX:
0000000000000004
[ 384.645173] RDX: 0000000000000000 RSI: ffffa9fe0759fe20 RDI:
ffff9dea0dad87f0
[ 384.686668] RBP: 0000000000000000 R08: 0000000000000000 R09:
8080808080808080
[ 384.727154] R10: ffff9dea33827660 R11: ffffee9d9e097a00 R12:
ffffa9fe0759fe20
[ 384.767777] R13: ffff9de9c4d3bb80 R14: 0000000000000000 R15:
ffff9dea0dad87f0
[ 384.807728] FS: 0000000000000000(0000) GS:ffff9dea33800000(0000)
knlGS:0000000000000000
[ 384.852776] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 384.883825] CR2: 00000000000000b8 CR3: 000000092560e002 CR4:
00000000000206e0
[ 384.922960] Call Trace:
[ 384.936550] ? blk_mq_flush_busy_ctxs+0xca/0x120
[ 384.962425] blk_mq_sched_dispatch_requests+0x15c/0x180
[ 384.992025] __blk_mq_run_hw_queue+0x5f/0x100
[ 385.016980] process_one_work+0x171/0x380
[ 385.040065] worker_thread+0x49/0x3f0
[ 385.060971] kthread+0xf8/0x130
[ 385.079101] ? max_active_store+0x80/0x80
[ 385.102005] ? kthread_bind+0x10/0x10
[ 385.122531] ret_from_fork+0x35/0x40
[ 385.142477] Modules linked in: ib_isert iscsi_target_mod
target_core_mod ib_srp scsi_transport_srp rpcrdma rdma_ucm ib_iser
ib_umad ib_ipoib sunrpc rdma_cm iw_cm libiscsi scsi_transport_iscsi
ib_cm mlx5_ib ib_uverbs ib_core intel_powerclamp coretemp kvm_intel
ipmi_ssif kvm iTCO_wdt gpio_ich iTCO_vendor_support irqbypass
crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel
dm_service_time crypto_simd cryptd glue_helper ipmi_si pcspkr lpc_ich
joydev ipmi_devintf hpilo hpwdt sg ipmi_msghandler i7core_edac
acpi_power_meter pcc_cpufreq dm_multipath ip_tables xfs libcrc32c
radeon sd_mod i2c_algo_bit drm_kms_helper syscopyarea sysfillrect
sysimgblt fb_sys_fops ttm serio_raw crc32c_intel drm mlx5_core i2c_core
bnx2 hpsa mlxfw scsi_transport_sas dm_mirror dm_region_hash dm_log
dm_mod
[ 385.534331] CR2: 00000000000000b8
crash>
kblockd blk_mq_run_work_fn then blk_mq_dispatch_rq_list and we have the
panic
crash> bt
PID: 9191 TASK: ffff9dea0a8395c0 CPU: 1 COMMAND: "kworker/1:1H"
#0 [ffffa9fe0759fab0] machine_kexec at ffffffff938606cf
#1 [ffffa9fe0759fb08] __crash_kexec at ffffffff9393a48d
#2 [ffffa9fe0759fbd0] crash_kexec at ffffffff9393b659
#3 [ffffa9fe0759fbe8] oops_end at ffffffff93831c41
#4 [ffffa9fe0759fc08] no_context at ffffffff9386ecb9
#5 [ffffa9fe0759fcb0] do_page_fault at ffffffff93870012
#6 [ffffa9fe0759fce0] page_fault at ffffffff942010ee
[exception RIP: blk_mq_dispatch_rq_list+114]
RIP: ffffffff93b9f202 RSP: ffffa9fe0759fd90 RFLAGS: 00010246
RAX: ffff9de9c4d3bbc8 RBX: ffff9de9c4d3bbc8 RCX: 0000000000000004
RDX: 0000000000000000 RSI: ffffa9fe0759fe20 RDI: ffff9dea0dad87f0
RBP: 0000000000000000 R8: 0000000000000000 R9: 8080808080808080
R10: ffff9dea33827660 R11: ffffee9d9e097a00 R12: ffffa9fe0759fe20
R13: ffff9de9c4d3bb80 R14: 0000000000000000 R15: ffff9dea0dad87f0
ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018
#7 [ffffa9fe0759fe18] blk_mq_sched_dispatch_requests at
ffffffff93ba455c
#8 [ffffa9fe0759fe60] __blk_mq_run_hw_queue at ffffffff93b9e3cf
#9 [ffffa9fe0759fe78] process_one_work at ffffffff938b0c21
#10 [ffffa9fe0759feb8] worker_thread at ffffffff938b18d9
#11 [ffffa9fe0759ff10] kthread at ffffffff938b6ee8
#12 [ffffa9fe0759ff50] ret_from_fork at ffffffff94200215
crash> whatis blk_mq_sched_dispatch_requests
void blk_mq_sched_dispatch_requests(struct blk_mq_hw_ctx *);
void blk_mq_sched_dispatch_requests(struct blk_mq_hw_ctx *hctx)
{
struct request_queue *q = hctx->queue;
struct elevator_queue *e = q->elevator;
const bool has_sched_dispatch = e && e->type-
>ops.dispatch_request;
***** Should have panicked here
LIST_HEAD(rq_list);
..
..
/*
* Only ask the scheduler for requests, if we didn't have
residual
* requests from the dispatch list. This is to avoid the case
where
* we only ever dispatch a fraction of the requests available
because
* of low device queue depth. Once we pull requests out of the
IO
* scheduler, we can no longer merge or sort them. So it's best
to
* leave them there for as long as we can. Mark the hw queue as
* needing a restart in that case.
*
* We want to dispatch from the scheduler if there was nothing
* on the dispatch list or we were able to dispatch from the
* dispatch list.
*/
if (!list_empty(&rq_list)) {
blk_mq_sched_mark_restart_hctx(hctx);
if (blk_mq_dispatch_rq_list(q, &rq_list, false)) {
if (has_sched_dispatch)
blk_mq_do_dispatch_sched(hctx);
else
blk_mq_do_dispatch_ctx(hctx);
}
} else if (has_sched_dispatch) {
blk_mq_do_dispatch_sched(hctx);
} else if (hctx->dispatch_busy) {
/* dequeue request one by one from sw queue if queue is
busy */
blk_mq_do_dispatch_ctx(hctx);
} else {
blk_mq_flush_busy_ctxs(hctx, &rq_list);
***** Called here
blk_mq_dispatch_rq_list(q, &rq_list,
false);
}
}
crash> request_queue.elevator 0xffff9dea0dad87f0
elevator = 0x0
Should have panicked earlier as elevator is NULL
request was for 2:0:0:0
sdev_state = SDEV_TRANSPORT_OFFLINE,
crash> dis -l blk_mq_dispatch_rq_list+114
/home/loberman/git/linux_torvalds/block/blk-mq.h: 210
0xffffffff93b9f202
<blk_mq_dispatch_rq_list+114>: mov 0xb8(%rbp),%rax
RBP: 0000000000000000
static inline bool blk_mq_get_dispatch_budget(struct blk_mq_hw_ctx
*hctx)
{
struct request_queue *q = hctx->queue;
if (q->mq_ops->get_budget) ******* This
Line
return q->mq_ops->get_budget(hctx);
return true;
}
crash> blk_mq_hw_ctx.queue ffff9de9c4d11000
queue = 0xffff9dea0dad87f0
crash> request_queue.mq_ops 0xffff9dea0dad87f0
mq_ops = 0xffffffff946afdc0
crash> blk_mq_ops 0xffffffff946afdc0
struct blk_mq_ops {
queue_rq = 0xffffffff93d619e0,
commit_rqs = 0x0,
get_budget = 0xffffffff93d60510,
put_budget = 0xffffffff93d5efd0,
timeout = 0xffffffff93d5fa80,
poll = 0x0,
complete = 0xffffffff93d60f70,
init_hctx = 0x0,
exit_hctx = 0x0,
init_request = 0xffffffff93d5f9e0,
exit_request = 0xffffffff93d5f9b0,
initialize_rq_fn = 0xffffffff93d5f850,
busy = 0xffffffff93d60900,
map_queues = 0xffffffff93d5f980,
show_rq = 0xffffffff93d68b30
}[ 382.412285] fast_io_fail_tmo expired for SRP port-2:1 / host2.
[ 382.604347] scsi host2: ib_srp: Got failed path rec status -110
[ 382.638622] scsi host2: ib_srp: Path record query failed: sgid
fe80:0000:0000:0000:7cfe:9003:0072:6ed2, dgid
fe80:0000:0000:0000:7cfe:9003:0072:6e4e, pkey 0xffff, service_id
0x7cfe900300726e4e
[ 382.736300] scsi host2: reconnect attempt 1 failed (-110)
[ 383.239347] BUG: unable to handle kernel NULL pointer dereference at
00000000000000b8
[ 383.239349] sd 2:0:0:0: rejecting I/O to offline device
[ 383.239370] device-mapper: multipath: Failing path 8:64.
[ 383.241278] sd 2:0:0:28: rejecting I/O to offline device
[ 383.241284] sd 2:0:0:27: rejecting I/O to offline device
[ 383.241289] device-mapper: multipath: Failing path 8:96.
[ 383.241291] device-mapper: multipath: Failing path 8:160.
[ 383.241301] print_req_error: I/O error, dev dm-8, sector 8191872
flags 80700
[ 383.241303] print_req_error: I/O error, dev dm-7, sector 8191872
flags 80700
[ 383.241335] print_req_error: I/O error, dev dm-8, sector 8191872
flags 0
[ 383.241338] Buffer I/O error on dev dm-8, logical block 8191872,
async page read
[ 383.241355] print_req_error: I/O error, dev dm-7, sector 8191872
flags 0
[ 383.241357] Buffer I/O error on dev dm-7, logical block 8191872,
async page read
[ 383.241367] print_req_error: I/O error, dev dm-7, sector 8191873
flags 0
[ 383.241368] Buffer I/O error on dev dm-7, logical block 8191873,
async page read
[ 383.241377] print_req_error: I/O error, dev dm-7, sector 8191874
flags 0
[ 383.241378] Buffer I/O error on dev dm-7, logical block 8191874,
async page read
[ 383.241387] print_req_error: I/O error, dev dm-7, sector 8191875
flags 0
[ 383.241388] Buffer I/O error on dev dm-7, logical block 8191875,
async page read
[ 383.241399] print_req_error: I/O error, dev dm-8, sector 8191873
flags 0
[ 383.241400] Buffer I/O error on dev dm-8, logical block 8191873,
async page read
[ 383.241409] print_req_error: I/O error, dev dm-7, sector 8191876
flags 0
[ 383.241410] Buffer I/O error on dev dm-7, logical block 8191876,
async page read
[ 383.241421] print_req_error: I/O error, dev dm-8, sector 8191874
flags 0
[ 383.241422] Buffer I/O error on dev dm-8, logical block 8191874,
async page read
[ 383.241431] Buffer I/O error on dev dm-7, logical block 8191877,
async page read
[ 383.241440] Buffer I/O error on dev dm-8, logical block 8191875,
async page read
[ 383.282566] #PF error: [normal kernel read fault]
[ 384.290154] PGD 0 P4D 0
[ 384.303791] Oops: 0000 [#1] SMP PTI
[ 384.323687] CPU: 1 PID: 9191 Comm: kworker/1:1H Kdump: loaded
Tainted: G W I 5.1.0-rc2+ #39
[ 384.375898] Hardware name: HP ProLiant DL380 G7, BIOS P67 08/16/2015
[ 384.411812] Workqueue: kblockd blk_mq_run_work_fn
[ 384.438498] RIP: 0010:blk_mq_dispatch_rq_list+0x72/0x570
[ 384.468539] Code: 08 84 d2 0f 85 cf 03 00 00 45 31 f6 c7 44 24 38 00
00 00 00 c7 44 24 3c 00 00 00 00 80 3c 24 00 4c 8d 6b b8 48 8b 6b c8 75
24 <48> 8b 85 b8 00 00 00 48 8b 40 40 48 8b 40 10 48 85 c0 74 10 48 89
[ 384.573942] RSP: 0018:ffffa9fe0759fd90 EFLAGS: 00010246
[ 384.604181] RAX: ffff9de9c4d3bbc8 RBX: ffff9de9c4d3bbc8 RCX:
0000000000000004
[ 384.645173] RDX: 0000000000000000 RSI: ffffa9fe0759fe20 RDI:
ffff9dea0dad87f0
[ 384.686668] RBP: 0000000000000000 R08: 0000000000000000 R09:
8080808080808080
[ 384.727154] R10: ffff9dea33827660 R11: ffffee9d9e097a00 R12:
ffffa9fe0759fe20
[ 384.767777] R13: ffff9de9c4d3bb80 R14: 0000000000000000 R15:
ffff9dea0dad87f0
[ 384.807728] FS: 0000000000000000(0000) GS:ffff9dea33800000(0000)
knlGS:0000000000000000
[ 384.852776] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 384.883825] CR2: 00000000000000b8 CR3: 000000092560e002 CR4:
00000000000206e0
[ 384.922960] Call Trace:
[ 384.936550] ? blk_mq_flush_busy_ctxs+0xca/0x120
[ 384.962425] blk_mq_sched_dispatch_requests+0x15c/0x180
[ 384.992025] __blk_mq_run_hw_queue+0x5f/0x100
[ 385.016980] process_one_work+0x171/0x380
[ 385.040065] worker_thread+0x49/0x3f0
[ 385.060971] kthread+0xf8/0x130
[ 385.079101] ? max_active_store+0x80/0x80
[ 385.102005] ? kthread_bind+0x10/0x10
[ 385.122531] ret_from_fork+0x35/0x40
[ 385.142477] Modules linked in: ib_isert iscsi_target_mod
target_core_mod ib_srp scsi_transport_srp rpcrdma rdma_ucm ib_iser
ib_umad ib_ipoib sunrpc rdma_cm iw_cm libiscsi scsi_transport_iscsi
ib_cm mlx5_ib ib_uverbs ib_core intel_powerclamp coretemp kvm_intel
ipmi_ssif kvm iTCO_wdt gpio_ich iTCO_vendor_support irqbypass
crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel
dm_service_time crypto_simd cryptd glue_helper ipmi_si pcspkr lpc_ich
joydev ipmi_devintf hpilo hpwdt sg ipmi_msghandler i7core_edac
acpi_power_meter pcc_cpufreq dm_multipath ip_tables xfs libcrc32c
radeon sd_mod i2c_algo_bit drm_kms_helper syscopyarea sysfillrect
sysimgblt fb_sys_fops ttm serio_raw crc32c_intel drm mlx5_core i2c_core
bnx2 hpsa mlxfw scsi_transport_sas dm_mirror dm_region_hash dm_log
dm_mod
[ 385.534331] CR2: 00000000000000b8
crash>
kblockd blk_mq_run_work_fn then blk_mq_dispatch_rq_list and we have the
panic
crash> bt
PID: 9191 TASK: ffff9dea0a8395c0 CPU: 1 COMMAND: "kworker/1:1H"
#0 [ffffa9fe0759fab0] machine_kexec at ffffffff938606cf
#1 [ffffa9fe0759fb08] __crash_kexec at ffffffff9393a48d
#2 [ffffa9fe0759fbd0] crash_kexec at ffffffff9393b659
#3 [ffffa9fe0759fbe8] oops_end at ffffffff93831c41
#4 [ffffa9fe0759fc08] no_context at ffffffff9386ecb9
#5 [ffffa9fe0759fcb0] do_page_fault at ffffffff93870012
#6 [ffffa9fe0759fce0] page_fault at ffffffff942010ee
[exception RIP: blk_mq_dispatch_rq_list+114]
RIP: ffffffff93b9f202 RSP: ffffa9fe0759fd90 RFLAGS: 00010246
RAX: ffff9de9c4d3bbc8 RBX: ffff9de9c4d3bbc8 RCX: 0000000000000004
RDX: 0000000000000000 RSI: ffffa9fe0759fe20 RDI: ffff9dea0dad87f0
RBP: 0000000000000000 R8: 0000000000000000 R9: 8080808080808080
R10: ffff9dea33827660 R11: ffffee9d9e097a00 R12: ffffa9fe0759fe20
R13: ffff9de9c4d3bb80 R14: 0000000000000000 R15: ffff9dea0dad87f0
ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018
#7 [ffffa9fe0759fe18] blk_mq_sched_dispatch_requests at
ffffffff93ba455c
#8 [ffffa9fe0759fe60] __blk_mq_run_hw_queue at ffffffff93b9e3cf
#9 [ffffa9fe0759fe78] process_one_work at ffffffff938b0c21
#10 [ffffa9fe0759feb8] worker_thread at ffffffff938b18d9
#11 [ffffa9fe0759ff10] kthread at ffffffff938b6ee8
#12 [ffffa9fe0759ff50] ret_from_fork at ffffffff94200215
crash> whatis blk_mq_sched_dispatch_requests
void blk_mq_sched_dispatch_requests(struct blk_mq_hw_ctx *);
void blk_mq_sched_dispatch_requests(struct blk_mq_hw_ctx *hctx)
{
struct request_queue *q = hctx->queue;
struct elevator_queue *e = q->elevator;
const bool has_sched_dispatch = e && e->type-
>ops.dispatch_request;
***** Should have panicked here
LIST_HEAD(rq_list);
..
..
/*
* Only ask the scheduler for requests, if we didn't have
residual
* requests from the dispatch list. This is to avoid the case
where
* we only ever dispatch a fraction of the requests available
because
* of low device queue depth. Once we pull requests out of the
IO
* scheduler, we can no longer merge or sort them. So it's best
to
* leave them there for as long as we can. Mark the hw queue as
* needing a restart in that case.
*
* We want to dispatch from the scheduler if there was nothing
* on the dispatch list or we were able to dispatch from the
* dispatch list.
*/
if (!list_empty(&rq_list)) {
blk_mq_sched_mark_restart_hctx(hctx);
if (blk_mq_dispatch_rq_list(q, &rq_list, false)) {
if (has_sched_dispatch)
blk_mq_do_dispatch_sched(hctx);
else
blk_mq_do_dispatch_ctx(hctx);
}
} else if (has_sched_dispatch) {
blk_mq_do_dispatch_sched(hctx);
} else if (hctx->dispatch_busy) {
/* dequeue request one by one from sw queue if queue is
busy */
blk_mq_do_dispatch_ctx(hctx);
} else {
blk_mq_flush_busy_ctxs(hctx, &rq_list);
***** Called here
blk_mq_dispatch_rq_list(q, &rq_list,
false);
}
}
crash> request_queue.elevator 0xffff9dea0dad87f0
elevator = 0x0
Should have panicked earlier as elevator is NULL
request was for 2:0:0:0
sdev_state = SDEV_TRANSPORT_OFFLINE,
crash> dis -l blk_mq_dispatch_rq_list+114
/home/loberman/git/linux_torvalds/block/blk-mq.h: 210
0xffffffff93b9f202
<blk_mq_dispatch_rq_list+114>: mov 0xb8(%rbp),%rax
RBP: 0000000000000000
static inline bool blk_mq_get_dispatch_budget(struct blk_mq_hw_ctx
*hctx)
{
struct request_queue *q = hctx->queue;
if (q->mq_ops->get_budget) ******* This
Line
return q->mq_ops->get_budget(hctx);
return true;
}
crash> blk_mq_hw_ctx.queue ffff9de9c4d11000
queue = 0xffff9dea0dad87f0
crash> request_queue.mq_ops 0xffff9dea0dad87f0
mq_ops = 0xffffffff946afdc0
crash> blk_mq_ops 0xffffffff946afdc0
struct blk_mq_ops {
queue_rq = 0xffffffff93d619e0,
commit_rqs = 0x0,
get_budget = 0xffffffff93d60510,
put_budget = 0xffffffff93d5efd0,
timeout = 0xffffffff93d5fa80,
poll = 0x0,
complete = 0xffffffff93d60f70,
init_hctx = 0x0,
exit_hctx = 0x0,
init_request = 0xffffffff93d5f9e0,
exit_request = 0xffffffff93d5f9b0,
initialize_rq_fn = 0xffffffff93d5f850,
busy = 0xffffffff93d60900,
map_queues = 0xffffffff93d5f980,
show_rq = 0xffffffff93d68b30
}
