workqueue: WQ_MEM_RECLAIM nvmet-wq:nvmet_rdma_release_queue_work [nvmet_rdma] is flushing !WQ_MEM_RECLAIM irdma-cleanup-wq:irdma_flush_worker [irdma]

Fri Dec 13 11:30:01 PST 2024

在 2024/12/13 19:55, Zhu Yanjun 写道:
> 在 2024/12/13 10:40, Honggang LI 写道:
>> It is 100% reproducible. The NVMEoRDMA client side is running RXE.
>> To reproduce it, the clinet side repeat to connect and disconnect
>> to the NVMEoRDMA target.
>>
>> [ 685.757357] ------------[ cut here ]------------
>> [ 685.758725] workqueue: WQ_MEM_RECLAIM nvmet- 
>> wq:nvmet_rdma_release_queue_work [nvmet_rdma] is flushing ! 
>> WQ_MEM_RECLAIM irdma-cleanup-wq:irdma_flush_worker [irdma]

I delved into this problem. It seems that it is a known problem.
Can you apply the following to make tests again?

diff --git a/drivers/infiniband/hw/irdma/hw.c 
b/drivers/infiniband/hw/irdma/hw.c
index ad50b77282f8..31501ff9f282 100644
--- a/drivers/infiniband/hw/irdma/hw.c
+++ b/drivers/infiniband/hw/irdma/hw.c
@@ -1872,7 +1872,7 @@ int irdma_rt_init_hw(struct irdma_device *iwdev,
                  * free cq bufs
                  */
                 iwdev->cleanup_wq = alloc_workqueue("irdma-cleanup-wq",
-                                       WQ_UNBOUND, WQ_UNBOUND_MAX_ACTIVE);
+                                       WQ_UNBOUND|WQ_MEM_RECLAIM, 
WQ_UNBOUND_MAX_ACTIVE);
                 if (!iwdev->cleanup_wq)
                         return -ENOMEM;
                 irdma_get_used_rsrc(iwdev);

Zhu Yanjun

>> [ 685.758809] WARNING: CPU: 16 PID: 1897 at kernel/workqueue.c:2966 
>> check_flush_dependency+0x11f/0x140
>> [ 685.762880] Modules linked in: nvmet_rdma nvmet nvme_keyring 
>> tcm_loop target_core_user uio target_core_pscsi target_core_file 
>> target_core_iblock rpcrdma qrtr rdma_ucm ib_srpt ib_isert 
>> iscsi_target_mod target_core_mod rfkill ib_iser libiscsi 
>> scsi_transport_iscsi rdma_cm iw_cm ib_cm intel_rapl_msr 
>> intel_rapl_common sb_edac x86_pkg_temp_thermal intel_powerclamp 
>> coretemp sunrpc kvm_intel kvm irqbypass binfmt_misc rapl intel_cstate 
>> irdma ipmi_ssif i40e iTCO_wdt intel_pmc_bxt iTCO_vendor_support 
>> ib_uverbs acpi_ipmi intel_uncore joydev ipmi_si pcspkr mxm_wmi ib_core 
>> mei_me ipmi_devintf i2c_i801 mei i2c_smbus lpc_ich ioatdma 
>> ipmi_msghandler loop dm_multipath nfnetlink zram ice crct10dif_pclmul 
>> crc32_pclmul crc32c_intel polyval_clmulni polyval_generic nvme isci 
>> nvme_core ghash_clmulni_intel sha512_ssse3 igb sha256_ssse3 libsas 
>> sha1_ssse3 nvme_auth mgag200 scsi_transport_sas dca gnss i2c_algo_bit 
>> wmi scsi_dh_rdac scsi_dh_emc scsi_dh_alua ip6_tables ip_tables fuse
>> [ 685.773891] CPU: 16 PID: 1897 Comm: kworker/16:2 Kdump: loaded 
>> Tainted: G S 6.8.4-300.patched.fc40.x86_64 #1
>> [ 685.775267] Hardware name: Sugon I620-G10/X9DR3-F, BIOS 3.0b 07/22/2014
>> [ 685.776627] Workqueue: nvmet-wq nvmet_rdma_release_queue_work 
>> [nvmet_rdma]
>> [ 685.777993] RIP: 0010:check_flush_dependency+0x11f/0x140
> 
> Maybe it is related with this line. What is the above line?
> 
> Zhu Yanjun
> 
>> [ 685.779331] Code: 8b 45 18 48 8d b2 b0 00 00 00 49 89 e8 48 8d 8b b0 
>> 00 00 00 48 c7 c7 28 fe b1 b2 c6 05 4f 97 59 02 01 48 89 c2 e8 a1 91 
>> fd ff <0f> 0b e9 fc fe ff ff 80 3d 3a 97 59 02 00 75 93 e9 2a ff ff ff 66
>> [ 685.782050] RSP: 0018:ffffb31348793cc8 EFLAGS: 00010082
>> [ 685.783398] RAX: 0000000000000000 RBX: ffff96c705754800 RCX: 
>> 0000000000000027
>> [ 685.784744] RDX: ffff96ce5fca18c8 RSI: 0000000000000001 RDI: 
>> ffff96ce5fca18c0
>> [ 685.786077] RBP: ffffffffc0d217f0 R08: 0000000000000000 R09: 
>> ffffb31348793b38
>> [ 685.787390] R10: ffffffffb3516808 R11: 0000000000000003 R12: 
>> ffff96c70d2aa8c0
>> [ 685.788688] R13: ffff96c7043c6a80 R14: 0000000000000001 R15: 
>> ffff96c704147400
>> [ 685.789970] FS: 0000000000000000(0000) GS:ffff96ce5fc80000(0000) 
>> knlGS:0000000000000000
>> [ 685.791239] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>> [ 685.792495] CR2: 00007f8207151000 CR3: 0000000d15422006 CR4: 
>> 00000000001706f0
>> [ 685.793745] Call Trace:
>> [ 685.794973] <TASK>
>> [ 685.796179] ? check_flush_dependency+0x11f/0x140
>> [ 685.797382] ? __warn+0x81/0x130
>> [ 685.798563] ? check_flush_dependency+0x11f/0x140
>> [ 685.799732] ? report_bug+0x16f/0x1a0
>> [ 685.800882] ? handle_bug+0x3c/0x80
>> [ 685.802003] ? exc_invalid_op+0x17/0x70
>> [ 685.803107] ? asm_exc_invalid_op+0x1a/0x20
>> [ 685.804200] ? __pfx_irdma_flush_worker+0x10/0x10 [irdma]
>> [ 685.805315] ? check_flush_dependency+0x11f/0x140
>> [ 685.806373] ? check_flush_dependency+0x11f/0x140
>> [ 685.807407] __flush_work.isra.0+0x10d/0x290
>> [ 685.808420] __cancel_work_timer+0x103/0x1a0
>> [ 685.809418] irdma_destroy_qp+0xd4/0x180 [irdma]
>> [ 685.810437] ib_destroy_qp_user+0x93/0x1a0 [ib_core]
>> [ 685.811474] nvmet_rdma_free_queue+0x35/0xc0 [nvmet_rdma]
>> [ 685.812437] nvmet_rdma_release_queue_work+0x1d/0x50 [nvmet_rdma]
>> [ 685.813385] process_one_work+0x170/0x330
>> [ 685.814300] worker_thread+0x280/0x3d0
>> [ 685.815201] ? __pfx_worker_thread+0x10/0x10
>> [ 685.816090] kthread+0xe8/0x120
>> [ 685.816956] ? __pfx_kthread+0x10/0x10
>> [ 685.817801] ret_from_fork+0x34/0x50
>> [ 685.818633] ? __pfx_kthread+0x10/0x10
>> [ 685.819439] ret_from_fork_asm+0x1b/0x30
>> [ 685.820232] </TASK>
>> [ 685.820994] Kernel panic - not syncing: kernel: panic_on_warn set ...
>> [ 685.821749] CPU: 16 PID: 1897 Comm: kworker/16:2 Kdump: loaded 
>> Tainted: G S 6.8.4-300.patched.fc40.x86_64 #1
>> [ 685.822513] Hardware name: Sugon I620-G10/X9DR3-F, BIOS 3.0b 07/22/2014
>> [ 685.823259] Workqueue: nvmet-wq nvmet_rdma_release_queue_work 
>> [nvmet_rdma]
>> [ 685.824002] Call Trace:
>> [ 685.824706] <TASK>
>> [ 685.825386] dump_stack_lvl+0x4d/0x70
>> [ 685.826060] panic+0x33e/0x370
>> [ 685.826724] ? check_flush_dependency+0x11f/0x140
>> [ 685.827383] check_panic_on_warn+0x44/0x60
>> [ 685.828021] __warn+0x8d/0x130
>> [ 685.828629] ? check_flush_dependency+0x11f/0x140
>> [ 685.829229] report_bug+0x16f/0x1a0
>> [ 685.829819] handle_bug+0x3c/0x80
>> [ 685.830396] exc_invalid_op+0x17/0x70
>> [ 685.830972] asm_exc_invalid_op+0x1a/0x20
>> [ 685.831548] RIP: 0010:check_flush_dependency+0x11f/0x140
>> [ 685.832129] Code: 8b 45 18 48 8d b2 b0 00 00 00 49 89 e8 48 8d 8b b0 
>> 00 00 00 48 c7 c7 28 fe b1 b2 c6 05 4f 97 59 02 01 48 89 c2 e8 a1 91 
>> fd ff <0f> 0b e9 fc fe ff ff 80 3d 3a 97 59 02 00 75 93 e9 2a ff ff ff 66
>> [ 685.833341] RSP: 0018:ffffb31348793cc8 EFLAGS: 00010082
>> [ 685.833954] RAX: 0000000000000000 RBX: ffff96c705754800 RCX: 
>> 0000000000000027
>> [ 685.834569] RDX: ffff96ce5fca18c8 RSI: 0000000000000001 RDI: 
>> ffff96ce5fca18c0
>> [ 685.835196] RBP: ffffffffc0d217f0 R08: 0000000000000000 R09: 
>> ffffb31348793b38
>> [ 685.835823] R10: ffffffffb3516808 R11: 0000000000000003 R12: 
>> ffff96c70d2aa8c0
>> [ 685.836450] R13: ffff96c7043c6a80 R14: 0000000000000001 R15: 
>> ffff96c704147400
>> [ 685.837079] ? __pfx_irdma_flush_worker+0x10/0x10 [irdma]
>> [ 685.837755] ? check_flush_dependency+0x11f/0x140
>> [ 685.838394] __flush_work.isra.0+0x10d/0x290
>> [ 685.839037] __cancel_work_timer+0x103/0x1a0
>> [ 685.839679] irdma_destroy_qp+0xd4/0x180 [irdma]
>> [ 685.840354] ib_destroy_qp_user+0x93/0x1a0 [ib_core]
>> [ 685.841049] nvmet_rdma_free_queue+0x35/0xc0 [nvmet_rdma]
>> [ 685.841707] nvmet_rdma_release_queue_work+0x1d/0x50 [nvmet_rdma]
>> [ 685.842367] process_one_work+0x170/0x330
>> [ 685.843020] worker_thread+0x280/0x3d0
>> [ 685.843670] ? __pfx_worker_thread+0x10/0x10
>> [ 685.844316] kthread+0xe8/0x120
>> [ 685.844955] ? __pfx_kthread+0x10/0x10
>> [ 685.845590] ret_from_fork+0x34/0x50
>> [ 685.846223] ? __pfx_kthread+0x10/0x10
>> [ 685.846853] ret_from_fork_asm+0x1b/0x30
>> [ 685.847485] </TASK>
>>
> 

-- 
Best Regards,
Yanjun.Zhu


