Bug report - drivers/nvme/host/pci.c?

Bug report - drivers/nvme/host/pci.c?

[Fangjian (Turing)]

Hi,

Linux version 4.15.0-rc1.

I use fio tool to test my NVME card performace, I found kernel will dump 
call trace and error info.

test log:




$ fio -name=write -numjobs=10 -filename=/dev/nvme0n1 -rw=write 
-iodepth=128 -direct=1 -sync=0 -norandommap -group_reporting -runtime=30 
-time

_based -bs=1M
write: (g=0): rw=write, bs=1M-1M/1M-1M/1M-1M, ioengine=sync, iodepth=128
...
fio-2.1.11
Starting 10 processes
[330929.298425] WARNING: CPU: 27 PID: 2516 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.298815] WARNING: CPU: 28 PID: 2520 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
fio: pid=2520, err=5/file:engines[330929.298818] Modules linked in:
/sync.c:67, func=xfer, error=Inp[330929.298825] CPU: 28 PID: 2520 Comm: 
fio Tainted: G??????? W??????? 4.15.0-rc1-g1d739bd #18
ut/output error
[330929.298826] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.298827] task: ffff8413e6734880 task.stack: ffff00001b6a0000
[330929.298829] pstate: 20000005 (nzCv daif -PAN -UAO)
fio: pid=2513, err=5/file:engines[330929.298831] pc : 
nvme_queue_rq+0x660/0x9f8
/sync.c:67, func=xfer, error=Inp[330929.298833] lr : 
nvme_queue_rq+0x59c/0x9f8
ut/output error
[330929.298834] sp : ffff00001b6a3800
[330929.298835] x29: ffff00001b6a3800 x28: ffff8017e6f3d720
[330929.298838] x27: ffff8017e6f3d5c0 x26: 0000000000000000
[330929.298842] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.298845] x23: 0000000000000001 x22: ffff8013e2808000
[330929.298848] x21: ffff8013e3760000 x20: ffff8017e917ba00
[330929.298850] x19: 0000000000000001 x18: 0000000000000001
[330929.298854] x17: 000000000047c678 x16: ffff00000821d848
[330929.298857] x15: 0000000000000008 x14: 0000ffff7e235c1c
[330929.298860] x13: 000000005f9702c0 x12: 0000000000000000
[330929.298864] x11: 0000820000000000 x10: 0000000000000000
[330929.298867] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.298870] x7 : 0000000000000010 x6 : 000000000000c000
[330929.298873] x5 : 0000000000000003 x4 : ffff8413e0615000
[330929.298877] x3 : ffff8413e07d3d90 x2 : ffff7e005f604000
[330929.298880] x1 : 000000000000000a x0 : ffff8413e07d3d90
[330929.298884] Call trace:
[330929.298886]? nvme_queue_rq+0x660/0x9f8
[330929.298892]? __blk_mq_try_issue_directly+0x178/0x1d8
[330929.298894]? blk_mq_try_issue_directly+0x88/0xa0
[330929.298896]? blk_mq_make_request+0x294/0x488
[330929.298900]? generic_make_request+0xe4/0x258
[330929.298902]? submit_bio+0xa4/0x1b8
[330929.298906]? __blkdev_direct_IO_simple+0x120/0x298
[330929.298908]? blkdev_direct_IO+0x3b4/0x3b8
[330929.298911]? generic_file_direct_write+0xe8/0x188
[330929.298913]? __generic_file_write_iter+0xac/0x1b0
[330929.298915]? blkdev_write_iter+0x84/0xf8
[330929.298919]? __vfs_write+0xb0/0x118
[330929.298920]? vfs_write+0xa0/0x1b0
[330929.298922]? SyS_write+0x44/0xa0
[330929.298925]? el0_svc_naked+0x20/0x24
[330929.298927] ---[ end trace 7626222de6819d61 ]---
[330929.299262] WARNING: CPU: 28 PID: 2502 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.299263] Modules linked in:
[330929.299267] CPU: 28 PID: 2502 Comm: kworker/28:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330929.299269] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.299273] Workqueue: kblockd blk_mq_requeue_work
[330929.299275] task: ffff8017e6101d00 task.stack: ffff00001b458000
[330929.299277] pstate: 20000005 (nzCv daif -PAN -UAO)
[330929.299279] pc : nvme_queue_rq+0x660/0x9f8
[330929.299282] lr : nvme_queue_rq+0x59c/0x9f8
[330929.299283] sp : ffff00001b45bad0
[330929.299284] x29: ffff00001b45bad0 x28: ffff8017e6f3d720
[330929.299288] x27: ffff8017e6f3d5c0 x26: 0000000000000000
[330929.299292] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.299296] x23: 0000000000000001 x22: ffff8013e2808000
[330929.299299] x21: ffff8013e3760000 x20: ffff8017e917ba00
[330929.299302] x19: 0000000000000001 x18: 0000000000000400
[330929.299305] x17: 000000000047c678 x16: ffff00000821d848
[330929.299309] x15: 0000000000000000 x14: 0000000000000400
[330929.299313] x13: 0000000000000400 x12: 0000000000000000
[330929.299316] x11: 0000000000000000 x10: 0000000000000000
[330929.299320] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.299325] x7 : 0000000000000010 x6 : 000000000000c000
[330929.299328] x5 : 00000017d810c000 x4 : ffff8413e0615000
[330929.299331] x3 : ffff8413e07d3d90 x2 : ffff7e005f604000
[330929.299335] x1 : 000000000000000a x0 : ffff8413e07d3d90
[330929.299339] Call trace:
[330929.299341]? nvme_queue_rq+0x660/0x9f8
[330929.299343]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330929.299346]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330929.299349]? __blk_mq_run_hw_queue+0xec/0x100
[330929.299351]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330929.299353]? blk_mq_run_hw_queue+0x3c/0x98
[330929.299355]? blk_mq_run_hw_queues+0x50/0x70
[330929.299357]? blk_mq_requeue_work+0x140/0x158
[330929.299361]? process_one_work+0x138/0x2f0
[330929.299363]? worker_thread+0x144/0x408
[330929.299365]? kthread+0xf8/0x128
[330929.299368]? ret_from_fork+0x10/0x18
[330929.299370] ---[ end trace 7626222de6819d62 ]---
[330929.299772] WARNING: CPU: 28 PID: 2502 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.299774] Modules linked in:
[330929.299777] CPU: 28 PID: 2502 Comm: kworker/28:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330929.299778] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.299781] Workqueue: kblockd blk_mq_requeue_work
[330929.299784] task: ffff8017e6101d00 task.stack: ffff00001b458000
[330929.299787] pstate: 20000005 (nzCv daif -PAN -UAO)
[330929.299789] pc : nvme_queue_rq+0x660/0x9f8
[330929.299791] lr : nvme_queue_rq+0x59c/0x9f8
[330929.299792] sp : ffff00001b45bad0
[330929.299793] x29: ffff00001b45bad0 x28: ffff8017e6f3d720
[330929.299797] x27: ffff8017e6f3d5c0 x26: 0000000000000000
[330929.299801] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.299805] x23: 0000000000000001 x22: ffff8013e2808000
[330929.299809] x21: ffff8013e3760000 x20: ffff8017e917ba00
[330929.299812] x19: 0000000000000001 x18: 0000000000000400
[330929.299815] x17: 000000000047c678 x16: ffff00000821d848
[330929.299818] x15: 0000000000000000 x14: 0000000000000400
[330929.299821] x13: 0000000000000400 x12: 0000000000000001
[330929.299826] x11: 0000000000d4b5b1 x10: 0000000000000000
[330929.299830] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.299833] x7 : 0000000000000010 x6 : 000000000000c000
[330929.299837] x5 : 00000017d810c000 x4 : ffff8413e0615000
[330929.299840] x3 : ffff8413e07d3d90 x2 : ffff7e005f604000
[330929.299844] x1 : 000000000000000a x0 : ffff8413e07d3d90
[330929.299848] Call trace:
[330929.299850]? nvme_queue_rq+0x660/0x9f8
[330929.299852]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330929.299856]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330929.299859]? __blk_mq_run_hw_queue+0xec/0x100
[330929.299860]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330929.299863]? blk_mq_run_hw_queue+0x3c/0x98
[330929.299866]? blk_mq_run_hw_queues+0x50/0x70
[330929.299868]? blk_mq_requeue_work+0x140/0x158
[330929.299870]? process_one_work+0x138/0x2f0
[330929.299872]? worker_thread+0x144/0x408
[330929.299875]? kthread+0xf8/0x128
[330929.299878]? ret_from_fork+0x10/0x18
[330929.299879] ---[ end trace 7626222de6819d63 ]---
[330929.300186] WARNING: CPU: 28 PID: 2502 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.300187] Modules linked in:
[330929.300190] CPU: 28 PID: 2502 Comm: kworker/28:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330929.300193] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.300197] Workqueue: kblockd blk_mq_requeue_work
[330929.300199] task: ffff8017e6101d00 task.stack: ffff00001b458000
[330929.300201] pstate: 20000005 (nzCv daif -PAN -UAO)
[330929.300205] pc : nvme_queue_rq+0x660/0x9f8
[330929.300208] lr : nvme_queue_rq+0x59c/0x9f8
[330929.300209] sp : ffff00001b45bad0
[330929.300210] x29: ffff00001b45bad0 x28: ffff8017e6f3d720
[330929.300213] x27: ffff8017e6f3d5c0 x26: 0000000000000000
[330929.300217] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.300220] x23: 0000000000000001 x22: ffff8013e2808000
[330929.300223] x21: ffff8013e3760000 x20: ffff8017e917ba00
[330929.300226] x19: 0000000000000001 x18: 0000000000000400
[330929.300230] x17: 000000000047c678 x16: ffff00000821d848
[330929.300234] x15: 0000000000000000 x14: 0000000000000400
[330929.300237] x13: 0000000000000400 x12: 0000000000000001
[330929.300240] x11: 00000000014be4e4 x10: 0000000000000000
[330929.300243] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.300247] x7 : 0000000000000010 x6 : 000000000000c000
[330929.300251] x5 : 00000017d810c000 x4 : ffff8413e0615000
[330929.300255] x3 : ffff8413e07d3d90 x2 : ffff7e005f604000
[330929.300259] x1 : 000000000000000a x0 : ffff8413e07d3d90
[330929.300262] Call trace:
[330929.300265]? nvme_queue_rq+0x660/0x9f8
[330929.300269]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330929.300270]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330929.300273]? __blk_mq_run_hw_queue+0xec/0x100
[330929.300276]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330929.300278]? blk_mq_run_hw_queue+0x3c/0x98
[330929.300281]? blk_mq_run_hw_queues+0x50/0x70
[330929.300282]? blk_mq_requeue_work+0x140/0x158
[330929.300285]? process_one_work+0x138/0x2f0
[330929.300288]? worker_thread+0x144/0x408
[330929.300290]? kthread+0xf8/0x128
[330929.300292]? ret_from_fork+0x10/0x18
[330929.300294] ---[ end trace 7626222de6819d64 ]---
Jobs: 8 (f=8)
[330929.300599] WARNING: CPU: 28 PID: 2502 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.300600] Modules linked in:
[330929.300603] CPU: 28 PID: 2502 Comm: kworker/28:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330929.300605] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.300608] Workqueue: kblockd blk_mq_requeue_work
[330929.300610] task: ffff8017e6101d00 task.stack: ffff00001b458000
[330929.300612] pstate: 20000005 (nzCv daif -PAN -UAO)
[330929.300615] pc : nvme_queue_rq+0x660/0x9f8
[330929.300617] lr : nvme_queue_rq+0x59c/0x9f8
[330929.300617] sp : ffff00001b45bad0
[330929.300619] x29: ffff00001b45bad0 x28: ffff8017e6f3d720
[330929.300622] x27: ffff8017e6f3d5c0 x26: 0000000000000000
[330929.300624] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.300628] x23: 0000000000000001 x22: ffff8013e2808000
[330929.300632] x21: ffff8013e3760000 x20: ffff8017e917ba00
[330929.300635] x19: 0000000000000001 x18: 0000000000000400
[330929.300639] x17: 000000000047c678 x16: ffff00000821d848
[330929.300642] x15: 0000000000000000 x14: 0000000000000400
[330929.300647] x13: 0000000000000400 x12: 0000000000000001
[330929.300650] x11: 0000000001c46f35 x10: 0000000000000000
[330929.300653] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.300656] x7 : 0000000000000010 x6 : 000000000000c000
[330929.300661] x5 : 00000017d810c000 x4 : ffff8413e0615000
[330929.300664] x3 : ffff8413e07d3d90 x2 : ffff7e005f604000
[330929.300667] x1 : 000000000000000a x0 : ffff8413e07d3d90
[330929.300671] Call trace:
[330929.300673]? nvme_queue_rq+0x660/0x9f8
[330929.300675]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330929.300677]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330929.300681]? __blk_mq_run_hw_queue+0xec/0x100
[330929.300683]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330929.300686]? blk_mq_run_hw_queue+0x3c/0x98
[330929.300687]? blk_mq_run_hw_queues+0x50/0x70
[330929.300690]? blk_mq_requeue_work+0x140/0x158
[330929.300692]? process_one_work+0x138/0x2f0
[330929.300695]? worker_thread+0x144/0x408
[330929.300696]? kthread+0xf8/0x128
[330929.300699]? ret_from_fork+0x10/0x18
[330929.300700] ---[ end trace 7626222de6819d65 ]---
[330929.301006] WARNING: CPU: 28 PID: 2502 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.301008] Modules linked in:
[330929.301011] CPU: 28 PID: 2502 Comm: kworker/28:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330929.301012] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.301016] Workqueue: kblockd blk_mq_requeue_work
[330929.301021] task: ffff8017e6101d00 task.stack: ffff00001b458000
[330929.301023] pstate: 20000005 (nzCv daif -PAN -UAO)
[330929.301025] pc : nvme_queue_rq+0x660/0x9f8
[330929.301028] lr : nvme_queue_rq+0x59c/0x9f8
[330929.301031] sp : ffff00001b45bad0
[330929.301032] x29: ffff00001b45bad0 x28: ffff8017e6f3d720
[330929.301036] x27: ffff8017e6f3d5c0 x26: 0000000000000000
[330929.301039] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.301043] x23: 0000000000000001 x22: ffff8013e2808000
[330929.301048] x21: ffff8013e3760000 x20: ffff8017e917ba00
[330929.301052] x19: 0000000000000001 x18: 0000000000000400
[330929.301057] x17: 000000000047c678 x16: ffff00000821d848
[330929.301062] x15: 0000000000000000 x14: 0000000000000400
[330929.301065] x13: 0000000000000400 x12: 0000000000000001
[330929.301069] x11: 000000000234d930 x10: 0000000000000000
[330929.301074] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.301078] x7 : 0000000000000010 x6 : 000000000000c000
[330929.301083] x5 : 00000017d810c000 x4 : ffff8413e0615000
[330929.301086] x3 : ffff8413e07d3d90 x2 : ffff7e005f604000
[330929.301089] x1 : 000000000000000a x0 : ffff8413e07d3d90
[330929.301093] Call trace:
[330929.301095]? nvme_queue_rq+0x660/0x9f8
[330929.301097]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330929.301100]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330929.301103]? __blk_mq_run_hw_queue+0xec/0x100
[330929.301104]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330929.301108]? blk_mq_run_hw_queue+0x3c/0x98
[330929.301110]? blk_mq_run_hw_queues+0x50/0x70
[330929.301112]? blk_mq_requeue_work+0x140/0x158
[330929.301114]? process_one_work+0x138/0x2f0
[330929.301117]? worker_thread+0x144/0x408
[330929.301119]? kthread+0xf8/0x128
[330929.301121]? ret_from_fork+0x10/0x18
[330929.301124] ---[ end trace 7626222de6819d66 ]---
[330929.301385] print_req_error: I/O error, dev nvme0n1, sector 0
[330929.312726] WARNING: CPU: 13 PID: 2513 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.312729] Modules linked in:
[330929.312733] CPU: 13 PID: 2513 Comm: fio Tainted: G W??????? 
4.15.0-rc1-g1d739bd #18
[330929.312735] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.312736] task: ffff8413e65a8e80 task.stack: ffff000011c08000
[330929.312738] pstate: 20000005 (nzCv daif -PAN -UAO)
[330929.312740] pc : nvme_queue_rq+0x660/0x9f8
[330929.312742] lr : nvme_queue_rq+0x59c/0x9f8
[330929.312743] sp : ffff000011c0b800
[330929.312744] x29: ffff000011c0b800 x28: ffff8013e288d4e0
[330929.312748] x27: ffff8013e288d380 x26: 0000000000000000
[330929.312751] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.312754] x23: 0000000000000001 x22: ffff8013e2808000
[330929.312758] x21: ffff8013e3760000 x20: ffff8013e6734400
[330929.312760] x19: 0000000000000001 x18: 0000000000000001
[330929.312764] x17: 000000000047c678 x16: ffff00000821d848
[330929.312767] x15: 0000000000000008 x14: 0000ffff7e235c1c
[330929.312770] x13: 000000004ed809c0 x12: 0000000000000000
[330929.312773] x11: 0000820000000000 x10: 0000000000000000
[330929.312776] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.312780] x7 : 0000000000000010 x6 : 0000000000009000
[330929.312783] x5 : 0000000000000003 x4 : ffff8413e0615000
[330929.312786] x3 : ffff8413e0b73d68 x2 : ffff7e004ed9a9c0
[330929.312790] x1 : 0000000000000009 x0 : ffff8413e0b73d68
[330929.312793] Call trace:
[330929.312795]? nvme_queue_rq+0x660/0x9f8
[330929.312798]? __blk_mq_try_issue_directly+0x178/0x1d8
[330929.312800]? blk_mq_try_issue_directly+0x88/0xa0
[330929.312803]? blk_mq_make_request+0x294/0x488
[330929.312806]? generic_make_request+0xe4/0x258
[330929.312808]? submit_bio+0xa4/0x1b8
[330929.312810]? __blkdev_direct_IO_simple+0x120/0x298
[330929.312812]? blkdev_direct_IO+0x3b4/0x3b8
[330929.312815]? generic_file_direct_write+0xe8/0x188
[330929.312817]? __generic_file_write_iter+0xac/0x1b0
[330929.312819]? blkdev_write_iter+0x84/0xf8
[330929.312821]? __vfs_write+0xb0/0x118
[330929.312823]? vfs_write+0xa0/0x1b0
[330929.312825]? SyS_write+0x44/0xa0
[330929.312827]? el0_svc_naked+0x20/0x24
[330929.312828] ---[ end trace 7626222de6819d67 ]---
[330929.313418] WARNING: CPU: 12 PID: 2440 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.313419] Modules linked in:
[330929.313423] CPU: 12 PID: 2440 Comm: kworker/12:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330929.313424] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.313429] Workqueue: kblockd blk_mq_requeue_work
[330929.313432] task: ffff8013e6b78e80 task.stack: ffff000011b30000
[330929.313434] pstate: 20000005 (nzCv daif -PAN -UAO)
[330929.313437] pc : nvme_queue_rq+0x660/0x9f8
[330929.313441] lr : nvme_queue_rq+0x59c/0x9f8
[330929.313442] sp : ffff000011b33ad0
[330929.313443] x29: ffff000011b33ad0 x28: ffff8013e288d4e0
[330929.313448] x27: ffff8013e288d380 x26: 0000000000000000
[330929.313451] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.313454] x23: 0000000000000001 x22: ffff8013e2808000
[330929.313458] x21: ffff8013e3760000 x20: ffff8013e6734400
[330929.313461] x19: 0000000000000001 x18: 0000000000000400
[330929.313465] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330929.313469] x15: 0000000000000000 x14: 0000000000000000
[330929.313473] x13: 0000000000000007 x12: 0000000000000000
[330929.313477] x11: ffff000008a31880 x10: 0000000000000000
[330929.313479] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.313482] x7 : 0000000000000010 x6 : 0000000000009000
[330929.313486] x5 : 00000013b66b0000 x4 : ffff8413e0615000
[330929.313489] x3 : ffff8013e4ee8368 x2 : ffff7e004ed9a9c0
[330929.313493] x1 : 0000000000000009 x0 : ffff8013e4ee8368
[330929.313496] Call trace:
[330929.313498]? nvme_queue_rq+0x660/0x9f8
[330929.313500]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330929.313504]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330929.313506]? __blk_mq_run_hw_queue+0xec/0x100
[330929.313509]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330929.313512]? blk_mq_run_hw_queue+0x3c/0x98
[330929.313514]? blk_mq_run_hw_queues+0x50/0x70
[330929.313516]? blk_mq_requeue_work+0x140/0x158
[330929.313520]? process_one_work+0x138/0x2f0
[330929.313523]? worker_thread+0x144/0x408
[330929.313524]? kthread+0xf8/0x128
[330929.313528]? ret_from_fork+0x10/0x18
[330929.313529] ---[ end trace 7626222de6819d68 ]---
[330929.314452] WARNING: CPU: 12 PID: 2440 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.314453] Modules linked in:
[330929.314456] CPU: 12 PID: 2440 Comm: kworker/12:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330929.314458] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.314460] Workqueue: kblockd blk_mq_requeue_work
[330929.314462] task: ffff8013e6b78e80 task.stack: ffff000011b30000
[330929.314465] pstate: 20000005 (nzCv daif -PAN -UAO)
[330929.314467] pc : nvme_queue_rq+0x660/0x9f8
[330929.314469] lr : nvme_queue_rq+0x59c/0x9f8
[330929.314470] sp : ffff000011b33ad0
[330929.314471] x29: ffff000011b33ad0 x28: ffff8013e288d4e0
[330929.314474] x27: ffff8013e288d380 x26: 0000000000000000
[330929.314477] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.314480] x23: 0000000000000001 x22: ffff8013e2808000
[330929.314482] x21: ffff8013e3760000 x20: ffff8013e6734400
[330929.314485] x19: 0000000000000001 x18: 0000000000000400
[330929.314488] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330929.314491] x15: 0000000000000000 x14: 0000000000000400
[330929.314494] x13: 0000000000000400 x12: 0000000000000001
[330929.314497] x11: 0000000000820560 x10: 0000000000000000
[330929.314500] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.314503] x7 : 0000000000000010 x6 : 0000000000009000
[330929.314506] x5 : 00000013b66b0000 x4 : ffff8413e0615000
[330929.314509] x3 : ffff8013e4ee8368 x2 : ffff7e004ed9a9c0
[330929.314511] x1 : 0000000000000009 x0 : ffff8013e4ee8368
[330929.314515] Call trace:
[330929.314516]? nvme_queue_rq+0x660/0x9f8
[330929.314519]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330929.314520]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330929.314522]? __blk_mq_run_hw_queue+0xec/0x100
[330929.314524]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330929.314527]? blk_mq_run_hw_queue+0x3c/0x98
[330929.314528]? blk_mq_run_hw_queues+0x50/0x70
[330929.314530]? blk_mq_requeue_work+0x140/0x158
[330929.314532]? process_one_work+0x138/0x2f0
[330929.314534]? worker_thread+0x144/0x408
[330929.314536]? kthread+0xf8/0x128
[330929.314537]? ret_from_fork+0x10/0x18
[330929.314538] ---[ end trace 7626222de6819d69 ]---
[330929.315364] WARNING: CPU: 12 PID: 2440 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.315366] Modules linked in:
[330929.315370] CPU: 12 PID: 2440 Comm: kworker/12:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330929.315373] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
Jobs: 8 (f=8): [X(1),W(6),X(1),W([330929.315375] Workqueue: kblockd 
blk_mq_requeue_work
2)] [9.7% done] [0KB/2035MB/0KB /[330929.315378] task: ffff8013e6b78e80 
task.stack: ffff000011b30000
s] [0/2035/0 iops] [eta 00m:28s][330929.315382] pstate: 20000005 (nzCv 
daif -PAN -UAO)

[330929.315385] pc : nvme_queue_rq+0x660/0x9f8
[330929.315389] lr : nvme_queue_rq+0x59c/0x9f8
[330929.315390] sp : ffff000011b33ad0
[330929.315391] x29: ffff000011b33ad0 x28: ffff8013e288d4e0
[330929.315394] x27: ffff8013e288d380 x26: 0000000000000000
[330929.315398] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.315401] x23: 0000000000000001 x22: ffff8013e2808000
[330929.315404] x21: ffff8013e3760000 x20: ffff8013e6734400
[330929.315408] x19: 0000000000000001 x18: 0000000000000400
[330929.315411] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330929.315414] x15: 0000000000000000 x14: 0000000000000400
[330929.315419] x13: 0000000000000400 x12: 0000000000000001
[330929.315422] x11: 0000000000dcb000 x10: 0000000000000000
[330929.315426] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.315429] x7 : 0000000000000010 x6 : 0000000000009000
[330929.315432] x5 : 00000013b66b0000 x4 : ffff8413e0615000
[330929.315436] x3 : ffff8013e4ee8368 x2 : ffff7e004ed9a9c0
[330929.315439] x1 : 0000000000000009 x0 : ffff8013e4ee8368
[330929.315442] Call trace:
[330929.315446]? nvme_queue_rq+0x660/0x9f8
[330929.315448]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330929.315450]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330929.315453]? __blk_mq_run_hw_queue+0xec/0x100
[330929.315456]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330929.315458]? blk_mq_run_hw_queue+0x3c/0x98
[330929.315460]? blk_mq_run_hw_queues+0x50/0x70
[330929.315462]? blk_mq_requeue_work+0x140/0x158
[330929.315465]? process_one_work+0x138/0x2f0
[330929.315467]? worker_thread+0x144/0x408
[330929.315469]? kthread+0xf8/0x128
[330929.315471]? ret_from_fork+0x10/0x18
[330929.315473] ---[ end trace 7626222de6819d6a ]---
[330929.316402] WARNING: CPU: 12 PID: 2440 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.316403] Modules linked in:
[330929.316407] CPU: 12 PID: 2440 Comm: kworker/12:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330929.316407] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.316410] Workqueue: kblockd blk_mq_requeue_work
[330929.316413] task: ffff8013e6b78e80 task.stack: ffff000011b30000
[330929.316415] pstate: 20000005 (nzCv daif -PAN -UAO)
[330929.316417] pc : nvme_queue_rq+0x660/0x9f8
[330929.316418] lr : nvme_queue_rq+0x59c/0x9f8
[330929.316419] sp : ffff000011b33ad0
[330929.316420] x29: ffff000011b33ad0 x28: ffff8013e288d4e0
[330929.316423] x27: ffff8013e288d380 x26: 0000000000000000
[330929.316426] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.316429] x23: 0000000000000001 x22: ffff8013e2808000
[330929.316434] x21: ffff8013e3760000 x20: ffff8013e6734400
[330929.316437] x19: 0000000000000001 x18: 0000000000000400
[330929.316441] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330929.316444] x15: 0000000000000000 x14: 0000000000000400
[330929.316448] x13: 0000000000000400 x12: 0000000000000001
[330929.316451] x11: 000000000157efc3 x10: 0000000000000000
[330929.316454] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.316457] x7 : 0000000000000010 x6 : 0000000000009000
[330929.316460] x5 : 00000013b66b0000 x4 : ffff8413e0615000
[330929.316463] x3 : ffff8013e4ee8368 x2 : ffff7e004ed9a9c0
[330929.316467] x1 : 0000000000000009 x0 : ffff8013e4ee8368
[330929.316471] Call trace:
[330929.316472]? nvme_queue_rq+0x660/0x9f8
[330929.316474]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330929.316478]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330929.316482]? __blk_mq_run_hw_queue+0xec/0x100
[330929.316486]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330929.316487]? blk_mq_run_hw_queue+0x3c/0x98
[330929.316489]? blk_mq_run_hw_queues+0x50/0x70
[330929.316491]? blk_mq_requeue_work+0x140/0x158
[330929.316495]? process_one_work+0x138/0x2f0
[330929.316496]? worker_thread+0x144/0x408
[330929.316499]? kthread+0xf8/0x128
[330929.316501]? ret_from_fork+0x10/0x18
[330929.316502] ---[ end trace 7626222de6819d6b ]---
[330929.317400] WARNING: CPU: 12 PID: 2440 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330929.317401] Modules linked in:
[330929.317404] CPU: 12 PID: 2440 Comm: kworker/12:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330929.317405] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330929.317408] Workqueue: kblockd blk_mq_requeue_work
[330929.317410] task: ffff8013e6b78e80 task.stack: ffff000011b30000
[330929.317411] pstate: 20000005 (nzCv daif -PAN -UAO)
[330929.317414] pc : nvme_queue_rq+0x660/0x9f8
[330929.317416] lr : nvme_queue_rq+0x59c/0x9f8
[330929.317417] sp : ffff000011b33ad0
[330929.317418] x29: ffff000011b33ad0 x28: ffff8013e288d4e0
[330929.317421] x27: ffff8013e288d380 x26: 0000000000000000
[330929.317423] x25: ffff8413e0615a00 x24: 0000000000000010
[330929.317426] x23: 0000000000000001 x22: ffff8013e2808000
[330929.317429] x21: ffff8013e3760000 x20: ffff8013e6734400
[330929.317432] x19: 0000000000000001 x18: 0000000000000400
[330929.317434] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330929.317437] x15: 0000000000000000 x14: 0000000000000001
[330929.317440] x13: 0000000000000000 x12: 0000000000000001
[330929.317443] x11: ffff000008a31880 x10: 0000000000000000
[330929.317446] x9 : 0000000000000000 x8 : 00600000000008c4
[330929.317448] x7 : 0000000000000010 x6 : 0000000000009000
[330929.317451] x5 : 00000013b66b0000 x4 : ffff8413e0615000
[330929.317455] x3 : ffff8013e4ee8368 x2 : ffff7e004ed9a9c0
[330929.317457] x1 : 0000000000000009 x0 : ffff8013e4ee8368
[330929.317460] Call trace:
[330929.317462]? nvme_queue_rq+0x660/0x9f8
[330929.317464]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330929.317465]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330929.317467]? __blk_mq_run_hw_queue+0xec/0x100
[330929.317470]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330929.317471]? blk_mq_run_hw_queue+0x3c/0x98
[330929.317473]? blk_mq_run_hw_queues+0x50/0x70
[330929.317475]? blk_mq_requeue_work+0x140/0x158
[330929.317477]? process_one_work+0x138/0x2f0
[330929.317479]? worker_thread+0x144/0x408
[330929.317481]? kthread+0xf8/0x128
[330929.317483]? ret_from_fork+0x10/0x18
[330929.317484] ---[ end trace 7626222de6819d6c ]---
[330929.318364] print_req_error: I/O error, dev nvme0n1, sector 0
[330931.711078] Modules linked in:
[330931.714212] CPU: 27 PID: 2516 Comm: fio Tainted: G W??????? 
4.15.0-rc1-g1d739bd #18
[330931.722549] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330931.731754] task: ffff8413e65ad700 task.stack: ffff00001b680000
[330931.737748] pstate: 20000005 (nzCv daif -PAN -UAO)
[330931.742614] pc : nvme_queue_rq+0x660/0x9f8
[330931.746786] lr : nvme_queue_rq+0x59c/0x9f8
[330931.750956] sp : ffff00001b683800
[330931.754345] x29: ffff00001b683800 x28: ffff8017e6ef95e0
[330931.759735] x27: ffff8017e6ef9480 x26: 0000000000000000
[330931.765125] x25: ffff8413e0615500 x24: 0000000000000030
[330931.770515] x23: 0000000000000001 x22: ffff8013e2808000
[330931.775904] x21: ffff8013e3760000 x20: ffff8017e917bb00
[330931.781293] x19: 0000000000000003 x18: 0000000000000001
[330931.786681] x17: 000000000047c678 x16: ffff00000821d848
[330931.792070] x15: 0000000000000008 x14: 0000ffff7e235c1c
[330931.797459] x13: 000000005f9dc90 x12: 0000000000000000
[330931.802848] x11: 0000820000000000 x10: 0000000000000000
[330931.808237] x9 : 0000000000000000 x8 : 00600000000008c4
[330931.813627] x7 : 0000000000000010 x6 : 0000000000009000
[330931.819016] x5 : 0000000000000003 x4 : ffff8413e0615000
[330931.824405] x3 : ffff8413e04bbd68 x2 : ffff7e005f970c00
[330931.829794] x1 : 0000000000000009 x0 : ffff8413e04bbd68
[330931.835183] Call trace:
[330931.837705]? nvme_queue_rq+0x660/0x9f8
[330931.841532]? __blk_mq_try_issue_directly+0x178/0x1d8
[330931.846572]? blk_mq_try_issue_directly+0x88/0xa0
[330931.851265]? blk_mq_make_request+0x294/0x488
[330931.855611]? generic_make_request+0xe4/0x258
[330931.859955]? submit_bio+0xa4/0x1b8
[330931.863433]? __blkdev_direct_IO_simple+0x120/0x298
[330931.868298]? blkdev_direct_IO+0x3b4/0x3b8
[330931.872383]? generic_file_direct_write+0xe8/0x188
[330931.877162]? __generic_file_write_iter+0xac/0x1b0
[330931.881941]? blkdev_write_iter+0x84/0xf8
[330931.885940]? __vfs_write+0xb0/0x118
[330931.889504]? vfs_write+0xa0/0x1b0
[330931.892894]? SyS_write+0x44/0xa0
[330931.896197]? el0_svc_naked+0x20/0x24
[330931.899847] ---[ end trace 7626222de6819d6d ]---
[330931.904652] WARNING: CPU: 25 PID: 2443 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330931.913424] Modules linked in:
[330931.916555] CPU: 25 PID: 2443 Comm: kworker/25:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330931.925760] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330931.934966] Workqueue: kblockd blk_mq_run_work_fn
[330931.939747] task: ffff8017e62d0000 task.stack: ffff00001b508000
[330931.945740] pstate: 20000005 (nzCv daif -PAN -UAO)
[330931.950606] pc : nvme_queue_rq+0x660/0x9f8
[330931.954777] lr : nvme_queue_rq+0x59c/0x9f8
[330931.958947] sp : ffff00001b50bb90
[330931.962336] x29: ffff00001b50bb90 x28: ffff8017e6ef9820
[330931.967724] x27: ffff8017e6ef96c0 x26: 0000000000000000
[330931.973113] x25: ffff8413e0615900 x24: 0000000000000010
[330931.978501] x23: 0000000000000001 x22: ffff8013e2808000
[330931.983889] x21: ffff8013e3760000 x20: ffff8017e917bb00
[330931.989277] x19: 0000000000000001 x18: 0000000000000400
[330931.994666] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330932.000055] x15: 0000000000000000 x14: 0000000000000400
[330932.005443] x13: 0000000000000400 x12: 0000000000000000
[330932.010831] x11: 000000000004102b x10: 0000000000000000
[330932.016219] x9 : 0000000000000000 x8 : 00600000000008c4
[330932.021608] x7 : 0000000000000010 x6 : 0000000000005000
[330932.026996] x5 : 0000000000000003 x4 : ffff8413e0615000
[330932.032385] x3 : ffff8017e89d15e0 x2 : ffff7e005f9dc800
[330932.037773] x1 : 000000000000000c x0 : ffff8017e89d15e0
[330932.043161] Call trace:
[330932.045683]? nvme_queue_rq+0x660/0x9f8
[330932.049507]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330932.054198]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330932.059498]? __blk_mq_run_hw_queue+0xec/0x100
[330932.063929]? blk_mq_run_work_fn+0x50/0x68
[330932.068014]? process_one_work+0x138/0x2f0
[330932.072099]? worker_thread+0x144/0x408
[330932.075922]? kthread+0xf8/0x128
[330932.079138]? ret_from_fork+0x10/0x18
[330932.082787] ---[ end trace 7626222de6819d6e ]---
[330932.087685] WARNING: CPU: 25 PID: 2443 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330932.096455] Modules linked in:
[330932.099585] CPU: 25 PID: 2443 Comm: kworker/25:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330932.108790] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330932.117996] Workqueue: kblockd blk_mq_requeue_work
[330932.122862] task: ffff8017e62d0000 task.stack: ffff00001b508000
[330932.128856] pstate: 20000005 (nzCv daif -PAN -UAO)
[330932.133721] pc : nvme_queue_rq+0x660/0x9f8
[330932.137892] lr : nvme_queue_rq+0x59c/0x9f8
[330932.142062] sp : ffff00001b50bad0
[330932.145450] x29: ffff00001b50bad0 x28: ffff8017e6ef95e0
[330932.150838] x27: ffff8017e6ef9480 x26: 0000000000000000
Jobs: 8 (f=8): [X(1),W(6),X(1),W([330932.156226] x25: ffff8413e0615500 
x24: 0000000000000030
2)] [12.9% done] [0KB/2041MB/0KB [330932.164473] x23: 0000000000000001 
x22: ffff8013e2808000
/s] [0/2041/0 iops] [eta 00m:27s[330932.172721] x21: ffff8013e3760000 
x20: ffff8017e917bb00
]
[330932.180880] x19: 0000000000000003 x18: 0000000000000400
[330932.186437] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330932.191825] x15: 0000000000000000 x14: 0000000000000400
[330932.197213] x13: 0000000000000400 x12: 0000000000000000
[330932.202601] x11: 000000000004102b x10: 0000000000000000
[330932.207989] x9 : 0000000000000000 x8 : 00600000000008c4
[330932.213376] x7 : 0000000000000010 x6 : 0000000000009000
[330932.218764] x5 : 00000017d5fb9000 x4 : ffff8413e0615000
[330932.224152] x3 : ffff8017e89c8568 x2 : ffff7e005f970c00
[330932.229539] x1 : 0000000000000009 x0 : ffff8017e89c8568
[330932.234927] Call trace:
[330932.237448]? nvme_queue_rq+0x660/0x9f8
[330932.241272]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330932.245962]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330932.251261]? __blk_mq_run_hw_queue+0xec/0x100
[330932.255693]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330932.260644]? blk_mq_run_hw_queue+0x3c/0x98
[330932.264815]? blk_mq_run_hw_queues+0x50/0x70
[330932.269073]? blk_mq_requeue_work+0x140/0x158
[330932.273416]? process_one_work+0x138/0x2f0
[330932.277500]? worker_thread+0x144/0x408
[330932.281323]? kthread+0xf8/0x128
[330932.284539]? ret_from_fork+0x10/0x18
[330932.288188] ---[ end trace 7626222de6819d6f ]---
[330932.295917] WARNING: CPU: 25 PID: 2443 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330932.304687] Modules linked in:
[330932.307818] CPU: 25 PID: 2443 Comm: kworker/25:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330932.317023] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330932.326229] Workqueue: kblockd blk_mq_requeue_work
[330932.331095] task: ffff8017e62d0000 task.stack: ffff00001b508000
[330932.337088] pstate: 20000005 (nzCv daif -PAN -UAO)
[330932.341953] pc : nvme_queue_rq+0x660/0x9f8
[330932.346123] lr : nvme_queue_rq+0x59c/0x9f8
[330932.350292] sp : ffff00001b50bad0
[330932.353681] x29: ffff00001b50bad0 x28: ffff8017e6ef95e0
[330932.359069] x27: ffff8017e6ef9480 x26: 0000000000000000
[330932.364456] x25: ffff8413e0615500 x24: 0000000000000030
[330932.369844] x23: 0000000000000001 x22: ffff8013e2808000
[330932.375231] x21: ffff8013e3760000 x20: ffff8017e917bb00
[330932.380618] x19: 0000000000000003 x18: 0000000000000400
[330932.386005] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330932.391392] x15: 0000000000000000 x14: 0000000000000000
[330932.396780] x13: 0000000000000000 x12: 0000000000000000
[330932.402169] x11: ffff000008a31880 x10: 0000000000000000
[330932.407556] x9 : 0000000000000000 x8 : 00600000000008c4
[330932.412944] x7 : 0000000000000010 x6 : 0000000000009000
[330932.418332] x5 : 00000017d5fb9000 x4 : ffff8413e0615000
[330932.423719] x3 : ffff8017e89c8568 x2 : ffff7e005f970c00
[330932.429107] x1 : 0000000000000009 x0 : ffff8017e89c8568
[330932.434493] Call trace:
[330932.437015]? nvme_queue_rq+0x660/0x9f8
[330932.440838]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330932.445529]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330932.450829]? __blk_mq_run_hw_queue+0xec/0x100
[330932.455260]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330932.460211]? blk_mq_run_hw_queue+0x3c/0x98
[330932.464382]? blk_mq_run_hw_queues+0x50/0x70
[330932.468639]? blk_mq_requeue_work+0x140/0x158
[330932.472983]? process_one_work+0x138/0x2f0
[330932.477067]? worker_thread+0x144/0x408
[330932.480889]? kthread+0xf8/0x128
[330932.484105]? ret_from_fork+0x10/0x18
[330932.487754] ---[ end trace 7626222de6819d70 ]---
[330932.495186] WARNING: CPU: 25 PID: 2443 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330932.503957] Modules linked in:
[330932.507087] CPU: 25 PID: 2443 Comm: kworker/25:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330932.516291] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330932.525497] Workqueue: kblockd blk_mq_requeue_work
[330932.530364] task: ffff8017e62d0000 task.stack: ffff00001b508000
[330932.536356] pstate: 20000005 (nzCv daif -PAN -UAO)
[330932.541221] pc : nvme_queue_rq+0x660/0x9f8
[330932.545392] lr : nvme_queue_rq+0x59c/0x9f8
[330932.549563] sp : ffff00001b50bad0
[330932.552951] x29: ffff00001b50bad0 x28: ffff8017e6ef95e0
[330932.558338] x27: ffff8017e6ef9480 x26: 0000000000000000
[330932.563725] x25: ffff8413e0615500 x24: 0000000000000030
[330932.569113] x23: 0000000000000001 x22: ffff8013e2808000
[330932.574501] x21: ffff8013e3760000 x20: ffff8017e917bb00
[330932.579888] x19: 0000000000000003 x18: 0000000000000400
[330932.585275] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330932.590663] x15: 0000000000000000 x14: 0000000000000002
[330932.596051] x13: 0000000000000000 x12: 0000000000000002
[330932.601438] x11: ffff000008a31880 x10: 0000000000000000
[330932.606826] x9 : 0000000000000000 x8 : 00600000000008c4
[330932.612214] x7 : 0000000000000010 x6 : 0000000000009000
[330932.617601] x5 : 00000017d5fb9000 x4 : ffff8413e0615000
[330932.622988] x3 : ffff8017e89c8568 x2 : ffff7e005f970c00
[330932.628376] x1 : 0000000000000009 x0 : ffff8017e89c8568
[330932.633764] Call trace:
[330932.636285]? nvme_queue_rq+0x660/0x9f8
[330932.640109]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330932.644800]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330932.650099]? __blk_mq_run_hw_queue+0xec/0x100
[330932.654530]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330932.659482]? blk_mq_run_hw_queue+0x3c/0x98
[330932.663652]? blk_mq_run_hw_queues+0x50/0x70
[330932.667909]? blk_mq_requeue_work+0x140/0x158
[330932.672254]? process_one_work+0x138/0x2f0
[330932.676337]? worker_thread+0x144/0x408
[330932.680160]? kthread+0xf8/0x128
[330932.683375]? ret_from_fork+0x10/0x18
[330932.687024] ---[ end trace 7626222de6819d71 ]---
[330932.695046] WARNING: CPU: 25 PID: 2443 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330932.703816] Modules linked in:
[330932.706947] CPU: 25 PID: 2443 Comm: kworker/25:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330932.716151] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330932.725357] Workqueue: kblockd blk_mq_requeue_work
[330932.730222] task: ffff8017e62d0000 task.stack: ffff00001b508000
[330932.736215] pstate: 20000005 (nzCv daif -PAN -UAO)
[330932.741080] pc : nvme_queue_rq+0x660/0x9f8
[330932.745251] lr : nvme_queue_rq+0x59c/0x9f8
[330932.749420] sp : ffff00001b50bad0
[330932.752809] x29: ffff00001b50bad0 x28: ffff8017e6ef95e0
[330932.758196] x27: ffff8017e6ef9480 x26: 0000000000000000
[330932.763583] x25: ffff8413e0615500 x24: 0000000000000030
[330932.768970] x23: 0000000000000001 x22: ffff8013e2808000
[330932.774358] x21: ffff8013e3760000 x20: ffff8017e917bb00
[330932.779745] x19: 0000000000000003 x18: 0000000000000400
[330932.785132] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330932.790520] x15: 0000000000000000 x14: 0000000000000002
[330932.795908] x13: 0000000000000000 x12: 0000000000000002
[330932.801296] x11: ffff000008a31880 x10: 0000000000000000
[330932.806683] x9 : 0000000000000000 x8 : 00600000000008c4
[330932.812070] x7 : 0000000000000010 x6 : 0000000000009000
[330932.817458] x5 : 00000017d5fb9000 x4 : ffff8413e0615000
[330932.822846] x3 : ffff8017e89c8568 x2 : ffff7e005f970c00
[330932.828233] x1 : 0000000000000009 x0 : ffff8017e89c8568
[330932.833620] Call trace:
[330932.836141]? nvme_queue_rq+0x660/0x9f8
[330932.839964]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330932.844656]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330932.849956]? __blk_mq_run_hw_queue+0xec/0x100
[330932.854386]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330932.859339]? blk_mq_run_hw_queue+0x3c/0x98
[330932.863509]? blk_mq_run_hw_queues+0x50/0x70
[330932.867766]? blk_mq_requeue_work+0x140/0x158
[330932.872111]? process_one_work+0x138/0x2f0
[330932.876194]? worker_thread+0x144/0x408
[330932.880017]? kthread+0xf8/0x128
[330932.883233]? ret_from_fork+0x10/0x18
[330932.886882] ---[ end trace 7626222de6819d72 ]---
[330932.894867] WARNING: CPU: 25 PID: 2443 at 
drivers/nvme/host/pci.c:769 nvme_queue_rq+0x660/0x9f8
[330932.903638] Modules linked in:
[330932.906768] CPU: 25 PID: 2443 Comm: kworker/25:1H Tainted: G??????? 
W??????? 4.15.0-rc1-g1d739bd #18
[330932.915972] Hardware name: Huawei TaiShan 2280 /D05, BIOS Hisilicon 
D05 IT17 Nemo 2.0 RC0 11/29/2017
[330932.925178] Workqueue: kblockd blk_mq_requeue_work
[330932.930044] task: ffff8017e62d0000 task.stack: ffff00001b508000
[330932.936036] pstate: 20000005 (nzCv daif -PAN -UAO)
[330932.940901] pc : nvme_queue_rq+0x660/0x9f8
[330932.945072] lr : nvme_queue_rq+0x59c/0x9f8
[330932.949241] sp : ffff00001b50bad0
[330932.952630] x29: ffff00001b50bad0 x28: ffff8017e6ef95e0
[330932.958017] x27: ffff8017e6ef9480 x26: 0000000000000000
[330932.963406] x25: ffff8413e0615500 x24: 0000000000000030
[330932.968793] x23: 0000000000000001 x22: ffff8013e2808000
[330932.974180] x21: ffff8013e3760000 x20: ffff8017e917bb00
[330932.979567] x19: 0000000000000003 x18: 0000000000000400
[330932.984955] x17: 000000000047c2a0 x16: ffff0000080d07b8
[330932.990343] x15: 0000000000000000 x14: 0000000000000002
[330932.995730] x13: 0000000000000000 x12: 0000000000000002
[330933.001118] x11: ffff000008a31880 x10: 0000000000000000
[330933.006506] x9 : 0000000000000000 x8 : 00600000000008c4
[330933.011893] x7 : 0000000000000010 x6 : 0000000000009000
[330933.017281] x5 : 00000017d5fb9000 x4 : ffff8413e0615000
[330933.022669] x3 : ffff8017e89c8568 x2 : ffff7e005f970c00
[330933.028056] x1 : 0000000000000009 x0 : ffff8017e89c8568
[330933.033444] Call trace:
[330933.035965]? nvme_queue_rq+0x660/0x9f8
[330933.039788]? blk_mq_dispatch_rq_list+0x1a8/0x548
[330933.044480]? blk_mq_sched_dispatch_requests+0x1ac/0x1b0
[330933.049778]? __blk_mq_run_hw_queue+0xec/0x100
[330933.054209]? __blk_mq_delay_run_hw_queue+0xd8/0x130
[330933.059161]? blk_mq_run_hw_queue+0x3c/0x98
[330933.063332]? blk_mq_run_hw_queues+0x50/0x70
[330933.067589]? blk_mq_requeue_work+0x140/0x158
[330933.071933]? process_one_work+0x138/0x2f0
[330933.076016]? worker_thread+0x144/0x408
[330933.079838]? kthread+0xf8/0x128
[330933.083054]? ret_from_fork+0x10/0x18
[330933.086703] ---[ end trace 7626222de6819d73 ]---
[330933.094683] print_req_error: I/O error, dev nvme0n1, sector 0
fio: pid=2516, err=5/file:engines/sync.c:67, func=xfer, 
error=Input/output error
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [14.7% done] 
[0KB/2032MB/0KB /s] [0/2032/0 iops] [eta 00m:29s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [17.6% done] 
[0KB/2053MB/0KB /s] [0/2053/0 iops] [eta 00m:28s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [20.6% done] 
[0KB/2015MB/0KB /s] [0/2015/0 iops] [eta 00m:27s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [23.5% done] 
[0KB/1982MB/0KB /s] [0/1982/0 iops] [eta 00m:26s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [26.5% done] 
[0KB/1993MB/0KB /s] [0/1993/0 iops] [eta 00m:25s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [29.4% done] 
[0KB/1985MB/0KB /s] [0/1985/0 iops] [eta 00m:24s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [32.4% done] 
[0KB/1982MB/0KB /s] [0/1982/0 iops] [eta 00m:23s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [35.3% done] 
[0KB/1996MB/0KB /s] [0/1996/0 iops] [eta 00m:22s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [38.2% done] 
[0KB/1987MB/0KB /s] [0/1987/0 iops] [eta 00m:21s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [41.2% done] 
[0KB/1996MB/0KB /s] [0/1996/0 iops] [eta 00m:20s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [44.1% done] 
[0KB/1983MB/0KB /s] [0/1983/0 iops] [eta 00m:19s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [47.1% done] 
[0KB/1993MB/0KB /s] [0/1993/0 iops] [eta 00m:18s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [50.0% done] 
[0KB/1974MB/0KB /s] [0/1974/0 iops] [eta 00m:17s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [52.9% done] 
[0KB/1991MB/0KB /s] [0/1991/0 iops] [eta 00m:16s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [55.9% done] 
[0KB/2006MB/0KB /s] [0/2006/0 iops] [eta 00m:15s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [58.8% done] 
[0KB/1980MB/0KB /s] [0/1980/0 iops] [eta 00m:14s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [61.8% done] 
[0KB/1991MB/0KB /s] [0/1991/0 iops] [eta 00m:13s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [64.7% done] 
[0KB/1985MB/0KB /s] [0/1985/0 iops] [eta 00m:12s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [67.6% done] 
[0KB/1979MB/0KB /s] [0/1979/0 iops] [eta 00m:11s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [70.6% done] 
[0KB/1981MB/0KB /s] [0/1981/0 iops] [eta 00m:10s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [73.5% done] 
[0KB/1975MB/0KB /s] [0/1975/0 iops] [eta 00m:09s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [76.5% done] 
[0KB/1981MB/0KB /s] [0/1981/0 iops] [eta 00m:08s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [79.4% done] 
[0KB/1981MB/0KB /s] [0/1981/0 iops] [eta 00m:07s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [82.4% done] 
[0KB/1983MB/0KB /s] [0/1983/0 iops] [eta 00m:06s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [85.3% done] 
[0KB/1978MB/0KB /s] [0/1978/0 iops] [eta 00m:05s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [88.2% done] 
[0KB/1971MB/0KB /s] [0/1971/0 iops] [eta 00m:04s]
Jobs: 7 (f=7): [X(1),W(2),X(1),W(3),X(1),W(2)] [91.2% done] 
[0KB/1984MB/0KB /s] [0/1984/0 iops] [eta 00m:03s]

write: (groupid=0, jobs=10): err= 5 (file:engines/sync.c:67, func=xfer, 
error=Input/output error): pid=2513: Tue May 17 07:49:49 2016
 ? write: io=59914MB, bw=1996.9MB/s, iops=1996, runt= 30004msec
 ??? clat (usec): min=484, max=18385, avg=3368.54, stdev=792.34
 ???? lat (usec): min=513, max=19692, avg=3502.80, stdev=793.56
 ??? clat percentiles (usec):
 ???? |? 1.00th=[ 2640],? 5.00th=[ 2736], 10.00th=[ 2736], 20.00th=[ 2864],
 ???? | 30.00th=[ 2960], 40.00th=[ 3088], 50.00th=[ 3216], 60.00th=[ 3344],
 ???? | 70.00th=[ 3472], 80.00th=[ 3696], 90.00th=[ 4016], 95.00th=[ 4448],
 ???? | 99.00th=[ 7456], 99.50th=[ 8384], 99.90th=[ 9792], 99.95th=[10048],
 ???? | 99.99th=[10688]
 ??? bw (KB? /s): min=281498, max=323584, per=14.28%, avg=292075.33, 
stdev=4460.87
 ??? lat (usec) : 500=0.01%, 750=0.09%, 1000=0.02%
 ??? lat (msec) : 2=0.09%, 4=89.24%, 10=10.51%, 20=0.05%
 ? cpu????????? : usr=2.61%, sys=1.68%, ctx=59965, majf=0, minf=730
 ? IO depths??? : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, 
 >=64=0.0%
 ???? submit??? : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, 
 >=64=0.0%
 ???? complete? : 0=0.1%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, 
 >=64=0.0%
 ???? issued??? : total=r=0/w=59917/d=0, short=r=0/w=0/d=0
 ???? latency?? : target=0, window=0, percentile=100.00%, depth=128

Run status group 0 (all jobs):
 ? WRITE: io=59914MB, aggrb=1996.9MB/s, minb=1996.9MB/s, 
maxb=1996.9MB/s, mint=30004msec, maxt=30004msec

Disk stats (read/write):
 ? nvme0n1: ios=0/119093, merge=0/0, ticks=0/389736, in_queue=355852, 
util=99.44%



Thanks

FangJian

.

Bug report - drivers/nvme/host/pci.c?

[Keith Busch]

On Wed, Jan 03, 2018@07:46:05PM +0000, Chaitanya Kulkarni wrote:
>    HI Fangjian,
> 
>    I ran your test on the latest kernel and I did not see the problem.
<> 
>    Is there a particular reason why you are using Linux version 4.15.0-rc1?
> 
>    Can you please check if you are having the same problem on the latest
>    kernel as mentioned in my log?
> 
>    Regards,
> 
>    Chaitanya

I would guess the reporter has an IOMMU enabled that DMA maps into fewer
entries than the virtual table has. Since that's possible, I think the
WARN_ON is invalid unless you overwrite iod->nents to be the return of
dma_map_sg, I think.

Bug report - drivers/nvme/host/pci.c?

[Keith Busch]

On Wed, Jan 03, 2018@01:01:41PM -0700, Keith Busch wrote:
> On Wed, Jan 03, 2018@07:46:05PM +0000, Chaitanya Kulkarni wrote:
> >    HI Fangjian,
> > 
> >    I ran your test on the latest kernel and I did not see the problem.
> <> 
> >    Is there a particular reason why you are using Linux version 4.15.0-rc1?
> > 
> >    Can you please check if you are having the same problem on the latest
> >    kernel as mentioned in my log?
> > 
> >    Regards,
> > 
> >    Chaitanya
> 
> I would guess the reporter has an IOMMU enabled that DMA maps into fewer
> entries than the virtual table has. Since that's possible, I think the
> WARN_ON is invalid unless you overwrite iod->nents to be the return of
> dma_map_sg, I think.

Just fyi, the driver selecting when to use SGL vs PRP is done prior to
DMA mapping the scatter list. An IOMMU may map a virtually contiguous
buffer into a single address, which is probably where biggest win for
SGL over PRP exists. I think this driver handling ought to be reworked.

Bug report - drivers/nvme/host/pci.c?

[hch]

On Mon, Jan 08, 2018@04:25:52AM +0000, Chaitanya Kulkarni wrote:
> Hi Fangjian,
> 
> 
> In my config I don't have COMFIG_ARM_SMMU* enabled, the test system I have is x86_64.

Try intel_iommu=force if it's an Intel system, or look up the AMD option
for AMD systems.

Bug report - drivers/nvme/host/pci.c?

[hch]

On Wed, Jan 03, 2018@01:11:31PM -0700, Keith Busch wrote:
> Just fyi, the driver selecting when to use SGL vs PRP is done prior to
> DMA mapping the scatter list. An IOMMU may map a virtually contiguous
> buffer into a single address, which is probably where biggest win for
> SGL over PRP exists. I think this driver handling ought to be reworked.

True.  The virtually contiguous (vs physical) case is rather unusual
and can only happen on PARISC and HP IA64 systems, though.

Anyway, moving the decision to after dma_map makes sense.  What do you
think of the untested patch below?  I also (so far very crudely)
enforces we only use SGLs if they fit into the same allocation as the
equivalent PRPs, so that we don't need to check use_sgl at allocation
time:

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index d53550e612bc..b74031ce1e19 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -330,35 +330,17 @@ static int nvme_npages(unsigned size, struct nvme_dev *dev)
 	return DIV_ROUND_UP(8 * nprps, PAGE_SIZE - 8);
 }
 
-/*
- * Calculates the number of pages needed for the SGL segments. For example a 4k
- * page can accommodate 256 SGL descriptors.
- */
-static int nvme_pci_npages_sgl(unsigned int num_seg)
-{
-	return DIV_ROUND_UP(num_seg * sizeof(struct nvme_sgl_desc), PAGE_SIZE);
-}
-
 static unsigned int nvme_pci_iod_alloc_size(struct nvme_dev *dev,
-		unsigned int size, unsigned int nseg, bool use_sgl)
+		unsigned int size, unsigned int nseg)
 {
-	size_t alloc_size;
-
-	if (use_sgl)
-		alloc_size = sizeof(__le64 *) * nvme_pci_npages_sgl(nseg);
-	else
-		alloc_size = sizeof(__le64 *) * nvme_npages(size, dev);
-
-	return alloc_size + sizeof(struct scatterlist) * nseg;
+	return sizeof(__le64 *) * nvme_npages(size, dev) +
+		sizeof(struct scatterlist) * nseg;
 }
 
-static unsigned int nvme_pci_cmd_size(struct nvme_dev *dev, bool use_sgl)
+static unsigned int nvme_pci_cmd_size(struct nvme_dev *dev)
 {
-	unsigned int alloc_size = nvme_pci_iod_alloc_size(dev,
-				    NVME_INT_BYTES(dev), NVME_INT_PAGES,
-				    use_sgl);
-
-	return sizeof(struct nvme_iod) + alloc_size;
+	return sizeof(struct nvme_iod) + nvme_pci_iod_alloc_size(dev,
+			NVME_INT_BYTES(dev), NVME_INT_PAGES);
 }
 
 static int nvme_admin_init_hctx(struct blk_mq_hw_ctx *hctx, void *data,
@@ -448,20 +430,32 @@ static void **nvme_pci_iod_list(struct request *req)
 	return (void **)(iod->sg + blk_rq_nr_phys_segments(req));
 }
 
-static inline bool nvme_pci_use_sgls(struct nvme_dev *dev, struct request *req)
+#define SGL_FACTOR  (sizeof(struct nvme_sgl_desc) / sizeof(__le64))
+
+static inline bool nvme_pci_use_sgls(struct nvme_dev *dev, struct request *req,
+		int nr_mapped)
 {
 	struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
-	unsigned int avg_seg_size;
-
-	avg_seg_size = DIV_ROUND_UP(blk_rq_payload_bytes(req),
-			blk_rq_nr_phys_segments(req));
 
 	if (!(dev->ctrl.sgls & ((1 << 0) | (1 << 1))))
 		return false;
 	if (!iod->nvmeq->qid)
 		return false;
-	if (!sgl_threshold || avg_seg_size < sgl_threshold)
+	if (!sgl_threshold)
 		return false;
+	/*
+	 * Need to make sure we have enough space for the SGLs, which are twice
+	 * as large as the PRPs.
+	 *
+	 * XXX: probably should move to module parameter parsing time.
+	 */
+	if (WARN_ON_ONCE(sgl_threshold < SGL_FACTOR * PAGE_SIZE))
+		return false;
+
+	if (DIV_ROUND_UP(blk_rq_payload_bytes(req), nr_mapped) < sgl_threshold)
+		return false;
+
+	iod->use_sgl = true;
 	return true;
 }
 
@@ -471,11 +465,8 @@ static blk_status_t nvme_init_iod(struct request *rq, struct nvme_dev *dev)
 	int nseg = blk_rq_nr_phys_segments(rq);
 	unsigned int size = blk_rq_payload_bytes(rq);
 
-	iod->use_sgl = nvme_pci_use_sgls(dev, rq);
-
 	if (nseg > NVME_INT_PAGES || size > NVME_INT_BYTES(dev)) {
-		size_t alloc_size = nvme_pci_iod_alloc_size(dev, size, nseg,
-				iod->use_sgl);
+		size_t alloc_size = nvme_pci_iod_alloc_size(dev, size, nseg);
 
 		iod->sg = kmalloc(alloc_size, GFP_ATOMIC);
 		if (!iod->sg)
@@ -488,6 +479,7 @@ static blk_status_t nvme_init_iod(struct request *rq, struct nvme_dev *dev)
 	iod->npages = -1;
 	iod->nents = 0;
 	iod->length = size;
+	iod->use_sgl = false;
 
 	return BLK_STS_OK;
 }
@@ -722,25 +714,24 @@ static void nvme_pci_sgl_set_seg(struct nvme_sgl_desc *sge,
 }
 
 static blk_status_t nvme_pci_setup_sgls(struct nvme_dev *dev,
-		struct request *req, struct nvme_rw_command *cmd)
+		struct request *req, int nr_sgl, struct nvme_rw_command *cmd)
 {
 	struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
-	int length = blk_rq_payload_bytes(req);
 	struct dma_pool *pool;
 	struct nvme_sgl_desc *sg_list;
 	struct scatterlist *sg = iod->sg;
-	int entries = iod->nents, i = 0;
 	dma_addr_t sgl_dma;
+	int i = 0;
 
 	/* setting the transfer type as SGL */
 	cmd->flags = NVME_CMD_SGL_METABUF;
 
-	if (length == sg_dma_len(sg)) {
+	if (nr_sgl == 1) {
 		nvme_pci_sgl_set_data(&cmd->dptr.sgl, sg);
 		return BLK_STS_OK;
 	}
 
-	if (entries <= (256 / sizeof(struct nvme_sgl_desc))) {
+	if (nr_sgl <= 256 / sizeof(struct nvme_sgl_desc)) {
 		pool = dev->prp_small_pool;
 		iod->npages = 0;
 	} else {
@@ -757,7 +748,7 @@ static blk_status_t nvme_pci_setup_sgls(struct nvme_dev *dev,
 	nvme_pci_iod_list(req)[0] = sg_list;
 	iod->first_dma = sgl_dma;
 
-	nvme_pci_sgl_set_seg(&cmd->dptr.sgl, sgl_dma, entries);
+	nvme_pci_sgl_set_seg(&cmd->dptr.sgl, sgl_dma, nr_sgl);
 
 	do {
 		if (i == SGES_PER_PAGE) {
@@ -771,17 +762,13 @@ static blk_status_t nvme_pci_setup_sgls(struct nvme_dev *dev,
 			i = 0;
 			nvme_pci_iod_list(req)[iod->npages++] = sg_list;
 			sg_list[i++] = *link;
-			nvme_pci_sgl_set_seg(link, sgl_dma, entries);
+			nvme_pci_sgl_set_seg(link, sgl_dma, nr_sgl);
 		}
 
 		nvme_pci_sgl_set_data(&sg_list[i++], sg);
-
-		length -= sg_dma_len(sg);
 		sg = sg_next(sg);
-		entries--;
-	} while (length > 0);
+	} while (--nr_sgl > 0);
 
-	WARN_ON(entries > 0);
 	return BLK_STS_OK;
 }
 
@@ -793,6 +780,7 @@ static blk_status_t nvme_map_data(struct nvme_dev *dev, struct request *req,
 	enum dma_data_direction dma_dir = rq_data_dir(req) ?
 			DMA_TO_DEVICE : DMA_FROM_DEVICE;
 	blk_status_t ret = BLK_STS_IOERR;
+	int nr_mapped;
 
 	sg_init_table(iod->sg, blk_rq_nr_phys_segments(req));
 	iod->nents = blk_rq_map_sg(q, req, iod->sg);
@@ -800,15 +788,15 @@ static blk_status_t nvme_map_data(struct nvme_dev *dev, struct request *req,
 		goto out;
 
 	ret = BLK_STS_RESOURCE;
-	if (!dma_map_sg_attrs(dev->dev, iod->sg, iod->nents, dma_dir,
-				DMA_ATTR_NO_WARN))
+	nr_mapped = dma_map_sg_attrs(dev->dev, iod->sg, iod->nents, dma_dir,
+			DMA_ATTR_NO_WARN);
+	if (!nr_mapped)
 		goto out;
 
-	if (iod->use_sgl)
-		ret = nvme_pci_setup_sgls(dev, req, &cmnd->rw);
+	if (nvme_pci_use_sgls(dev, req, nr_mapped))
+		ret = nvme_pci_setup_sgls(dev, req, nr_mapped, &cmnd->rw);
 	else
 		ret = nvme_pci_setup_prps(dev, req, &cmnd->rw);
-
 	if (ret != BLK_STS_OK)
 		goto out_unmap;
 
@@ -1519,7 +1507,7 @@ static int nvme_alloc_admin_tags(struct nvme_dev *dev)
 		dev->admin_tagset.queue_depth = NVME_AQ_MQ_TAG_DEPTH;
 		dev->admin_tagset.timeout = ADMIN_TIMEOUT;
 		dev->admin_tagset.numa_node = dev_to_node(dev->dev);
-		dev->admin_tagset.cmd_size = nvme_pci_cmd_size(dev, false);
+		dev->admin_tagset.cmd_size = nvme_pci_cmd_size(dev);
 		dev->admin_tagset.flags = BLK_MQ_F_NO_SCHED;
 		dev->admin_tagset.driver_data = dev;
 
@@ -2047,11 +2035,7 @@ static int nvme_dev_add(struct nvme_dev *dev)
 		dev->tagset.numa_node = dev_to_node(dev->dev);
 		dev->tagset.queue_depth =
 				min_t(int, dev->q_depth, BLK_MQ_MAX_DEPTH) - 1;
-		dev->tagset.cmd_size = nvme_pci_cmd_size(dev, false);
-		if ((dev->ctrl.sgls & ((1 << 0) | (1 << 1))) && sgl_threshold) {
-			dev->tagset.cmd_size = max(dev->tagset.cmd_size,
-					nvme_pci_cmd_size(dev, true));
-		}
+		dev->tagset.cmd_size = nvme_pci_cmd_size(dev);
 		dev->tagset.flags = BLK_MQ_F_SHOULD_MERGE;
 		dev->tagset.driver_data = dev;
 

Bug report - drivers/nvme/host/pci.c?

[Keith Busch]

On Mon, Jan 08, 2018@10:49:51AM +0100, hch@lst.de wrote:
> Anyway, moving the decision to after dma_map makes sense.  What do you
> think of the untested patch below?  I also (so far very crudely)
> enforces we only use SGLs if they fit into the same allocation as the
> equivalent PRPs, so that we don't need to check use_sgl at allocation
> time:

Thanks, this looks pretty good. I'm not going to be able to test this
till next week, though. Would be good to hear if someone else with a
capable device can can verify this is successful.