linux/drivers/scsi, branch v5.11

scsi: scsi_debug: Fix a memory leak

2021-02-09T02:51:25Z

The sdebug_q_arr pointer must be freed when the module is unloaded. $ cat /sys/kernel/debug/kmemleak unreferenced object 0xffff888e1cfb0000 (size 4096): comm "modprobe", pid 165555, jiffies 4325987516 (age 685.194s) hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ backtrace: [<00000000458f4f5d>] 0xffffffffc06702d9 [<000000003edc4b1f>] do_one_initcall+0xe9/0x57d [<00000000da7d518c>] do_init_module+0x1d1/0x6f0 [<000000009a6a9248>] load_module+0x36bd/0x4f50 [<00000000ddb0c3ce>] __do_sys_init_module+0x1db/0x260 [<000000009532db57>] do_syscall_64+0xa5/0x420 [<000000002916b13d>] entry_SYSCALL_64_after_hwframe+0x6a/0xdf Fixes: 87c715dcde63 ("scsi: scsi_debug: Add per_host_store option") Link: https://lore.kernel.org/r/20210208111734.34034-1-mlombard@redhat.com Acked-by: Douglas Gilbert Signed-off-by: Maurizio Lombardi Signed-off-by: Martin K. Petersen

scsi: lpfc: Fix EEH encountering oops with NVMe traffic

2021-01-29T18:41:39Z

In testing, in a configuration with Redfish and native NVMe multipath when an EEH is injected, a kernel oops is being encountered: (unreliable) lpfc_nvme_ls_req+0x328/0x720 [lpfc] __nvme_fc_send_ls_req.constprop.13+0x1d8/0x3d0 [nvme_fc] nvme_fc_create_association+0x224/0xd10 [nvme_fc] nvme_fc_reset_ctrl_work+0x110/0x154 [nvme_fc] process_one_work+0x304/0x5d the NBMe transport is issuing a Disconnect LS request, which the driver receives and tries to post but the work queue used by the driver is already being torn down by the eeh. Fix by validating the validity of the work queue before proceeding with the LS transmit. Link: https://lore.kernel.org/r/20210127221601.84878-1-jsmart2021@gmail.com Reviewed-by: Ewan D. Milne Signed-off-by: James Smart Signed-off-by: Martin K. Petersen

scsi: qla2xxx: Fix description for parameter ql2xenforce_iocb_limit

2021-01-21T02:43:31Z

Parameter ql2xenforce_iocb_limit is enabled by default. Link: https://lore.kernel.org/r/20210118184922.23793-1-ematsumiya@suse.de Fixes: 89c72f4245a8 ("scsi: qla2xxx: Add IOCB resource tracking") Reviewed-by: Daniel Wagner Reviewed-by: Himanshu Madhani Signed-off-by: Enzo Matsumiya Signed-off-by: Martin K. Petersen

scsi: ibmvfc: Set default timeout to avoid crash during migration

2021-01-15T03:02:59Z

While testing live partition mobility, we have observed occasional crashes of the Linux partition. What we've seen is that during the live migration, for specific configurations with large amounts of memory, slow network links, and workloads that are changing memory a lot, the partition can end up being suspended for 30 seconds or longer. This resulted in the following scenario: CPU 0 CPU 1 ------------------------------- ---------------------------------- scsi_queue_rq migration_store -> blk_mq_start_request -> rtas_ibm_suspend_me -> blk_add_timer -> on_each_cpu(rtas_percpu_suspend_me _______________________________________V | V -> IPI from CPU 1 -> rtas_percpu_suspend_me -> __rtas_suspend_last_cpu -- Linux partition suspended for > 30 seconds -- -> for_each_online_cpu(cpu) plpar_hcall_norets(H_PROD -> scsi_dispatch_cmd -> scsi_times_out -> scsi_abort_command -> queue_delayed_work -> ibmvfc_queuecommand_lck -> ibmvfc_send_event -> ibmvfc_send_crq - returns H_CLOSED <- returns SCSI_MLQUEUE_HOST_BUSY -> __blk_mq_requeue_request -> scmd_eh_abort_handler -> scsi_try_to_abort_cmd - returns SUCCESS -> scsi_queue_insert Normally, the SCMD_STATE_COMPLETE bit would protect against the command completion and the timeout, but that doesn't work here, since we don't check that at all in the SCSI_MLQUEUE_HOST_BUSY path. In this case we end up calling scsi_queue_insert on a request that has already been queued, or possibly even freed, and we crash. The patch below simply increases the default I/O timeout to avoid this race condition. This is also the timeout value that nearly all IBM SAN storage recommends setting as the default value. Link: https://lore.kernel.org/r/1610463998-19791-1-git-send-email-brking@linux.vnet.ibm.com Signed-off-by: Brian King Signed-off-by: Martin K. Petersen

scsi: fnic: Fix memleak in vnic_dev_init_devcmd2

2021-01-13T04:32:53Z

When ioread32() returns 0xFFFFFFFF, we should execute cleanup functions like other error handling paths before returning. Link: https://lore.kernel.org/r/20201225083520.22015-1-dinghao.liu@zju.edu.cn Acked-by: Karan Tilak Kumar Signed-off-by: Dinghao Liu Signed-off-by: Martin K. Petersen

scsi: libfc: Avoid invoking response handler twice if ep is already completed

2021-01-13T04:07:32Z

A race condition exists between the response handler getting called because of exchange_mgr_reset() (which clears out all the active XIDs) and the response we get via an interrupt. Sequence of events: rport ba0200: Port timeout, state PLOGI rport ba0200: Port entered PLOGI state from PLOGI state xid 1052: Exchange timer armed : 20000 msecs  xid timer armed here rport ba0200: Received LOGO request while in state PLOGI rport ba0200: Delete port rport ba0200: work event 3 rport ba0200: lld callback ev 3 bnx2fc: rport_event_hdlr: event = 3, port_id = 0xba0200 bnx2fc: ba0200 - rport not created Yet!! /* Here we reset any outstanding exchanges before freeing rport using the exch_mgr_reset() */ xid 1052: Exchange timer canceled /* Here we got two responses for one xid */ xid 1052: invoking resp(), esb 20000000 state 3 xid 1052: invoking resp(), esb 20000000 state 3 xid 1052: fc_rport_plogi_resp() : ep->resp_active 2 xid 1052: fc_rport_plogi_resp() : ep->resp_active 2 Skip the response if the exchange is already completed. Link: https://lore.kernel.org/r/20201215194731.2326-1-jhasan@marvell.com Signed-off-by: Javed Hasan Signed-off-by: Martin K. Petersen

scsi: scsi_transport_srp: Don't block target in failfast state

2021-01-13T03:56:49Z

If the port is in SRP_RPORT_FAIL_FAST state when srp_reconnect_rport() is entered, a transition to SDEV_BLOCK would be illegal, and a kernel WARNING would be triggered. Skip scsi_target_block() in this case. Link: https://lore.kernel.org/r/20210111142541.21534-1-mwilck@suse.com Reviewed-by: Bart Van Assche Signed-off-by: Martin Wilck Signed-off-by: Martin K. Petersen

scsi: ufs: Fix tm request when non-fatal error happens

2021-01-08T03:50:48Z

When non-fatal error like line-reset happens, ufshcd_err_handler() starts to abort tasks by ufshcd_try_to_abort_task(). When it tries to issue a task management request, we hit two warnings: WARNING: CPU: 7 PID: 7 at block/blk-core.c:630 blk_get_request+0x68/0x70 WARNING: CPU: 4 PID: 157 at block/blk-mq-tag.c:82 blk_mq_get_tag+0x438/0x46c After fixing the above warnings we hit another tm_cmd timeout which may be caused by unstable controller state: __ufshcd_issue_tm_cmd: task management cmd 0x80 timed-out Then, ufshcd_err_handler() enters full reset, and kernel gets stuck. It turned out ufshcd_print_trs() printed too many messages on console which requires CPU locks. Likewise hba->silence_err_logs, we need to avoid too verbose messages. This is actually not an error case. Link: https://lore.kernel.org/r/20210107185316.788815-3-jaegeuk@kernel.org Fixes: 69a6c269c097 ("scsi: ufs: Use blk_{get,put}_request() to allocate and free TMFs") Reviewed-by: Can Guo Signed-off-by: Jaegeuk Kim Signed-off-by: Martin K. Petersen

scsi: ufs: Fix livelock of ufshcd_clear_ua_wluns()

2021-01-08T03:50:48Z

When gate_work/ungate_work experience an error during hibern8_enter or exit we can livelock: ufshcd_err_handler() ufshcd_scsi_block_requests() ufshcd_reset_and_restore() ufshcd_clear_ua_wluns() -> stuck ufshcd_scsi_unblock_requests() In order to avoid this, ufshcd_clear_ua_wluns() can be called per recovery flows such as suspend/resume, link_recovery, and error_handler. Link: https://lore.kernel.org/r/20210107185316.788815-2-jaegeuk@kernel.org Fixes: 1918651f2d7e ("scsi: ufs: Clear UAC for RPMB after ufshcd resets") Reviewed-by: Can Guo Signed-off-by: Jaegeuk Kim Signed-off-by: Martin K. Petersen

scsi: ibmvfc: Fix missing cast of ibmvfc_event pointer to u64 handle

2021-01-08T03:32:56Z

Commit 2aa0102c6688 ("scsi: ibmvfc: Use correlation token to tag commands") sets the vfcFrame correlation token to the pointer handle of the associated ibmvfc_event. However, that commit failed to cast the pointer to an appropriate type which in this case is a u64. As such sparse warnings are generated for both correlation token assignments. ibmvfc.c:2375:36: sparse: incorrect type in argument 1 (different base types) ibmvfc.c:2375:36: sparse: expected unsigned long long [usertype] val ibmvfc.c:2375:36: sparse: got struct ibmvfc_event *[assigned] evt Add the appropriate u64 casts when assigning an ibmvfc_event as a correlation token. Link: https://lore.kernel.org/r/20210106203721.1054693-1-tyreld@linux.ibm.com Fixes: 2aa0102c6688 ("scsi: ibmvfc: Use correlation token to tag commands") Reported-by: kernel test robot Signed-off-by: Tyrel Datwyler Signed-off-by: Martin K. Petersen