From: Javed Hasan Date: Thu, 24 Jun 2021 17:18:02 +0000 (-0700) Subject: scsi: qedf: Add check to synchronize abort and flush X-Git-Tag: v5.15~733^2~4 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=df99446d5c2a63dc6e6920c8090da0e9da6539d5;p=platform%2Fkernel%2Flinux-starfive.git scsi: qedf: Add check to synchronize abort and flush A race condition was observed between qedf_cleanup_fcport() and qedf_process_error_detect()->qedf_initiate_abts(): [2069091.203145] BUG: unable to handle kernel NULL pointer dereference at 0000000000000030 [2069091.213100] IP: [] qedf_process_error_detect+0x96/0x130 [qedf] [2069091.223391] PGD 1943049067 PUD 194304e067 PMD 0 [2069091.233420] Oops: 0000 [#1] SMP [2069091.361820] CPU: 1 PID: 14751 Comm: kworker/1:46 Kdump: loaded Tainted: P OE ------------ 3.10.0-1160.25.1.el7.x86_64 #1 [2069091.388474] Hardware name: HPE Synergy 480 Gen10/Synergy 480 Gen10 Compute Module, BIOS I42 04/08/2020 [2069091.402148] Workqueue: qedf_io_wq qedf_fp_io_handler [qedf] [2069091.415780] task: ffff9bb9f5190000 ti: ffff9bacaef9c000 task.ti: ffff9bacaef9c000 [2069091.429590] RIP: 0010:[] [] qedf_process_error_detect+0x96/0x130 [qedf] [2069091.443666] RSP: 0018:ffff9bacaef9fdb8 EFLAGS: 00010246 [2069091.457692] RAX: 0000000000000000 RBX: ffff9bbbbbfb18a0 RCX: ffffffffc0672310 [2069091.471997] RDX: 00000000000005de RSI: ffffffffc066e7f0 RDI: ffff9beb3f4538d8 [2069091.486130] RBP: ffff9bacaef9fdd8 R08: 0000000000006000 R09: 0000000000006000 [2069091.500321] R10: 0000000000001551 R11: ffffb582996ffff8 R12: ffffb5829b39cc18 [2069091.514779] R13: ffff9badab380c28 R14: ffffd5827f643900 R15: 0000000000000040 [2069091.529472] FS: 0000000000000000(0000) GS:ffff9beb3f440000(0000) knlGS:0000000000000000 [2069091.543926] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [2069091.558942] CR2: 0000000000000030 CR3: 000000193b9a2000 CR4: 00000000007607e0 [2069091.573424] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [2069091.587876] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [2069091.602007] PKRU: 00000000 [2069091.616010] Call Trace: [2069091.629902] [] qedf_process_cqe+0x109/0x2e0 [qedf] [2069091.643941] [] qedf_fp_io_handler+0x26/0x60 [qedf] [2069091.657948] [] process_one_work+0x17f/0x440 [2069091.672111] [] worker_thread+0x126/0x3c0 [2069091.686057] [] ? manage_workers.isra.26+0x2a0/0x2a0 [2069091.700033] [] kthread+0xd1/0xe0 [2069091.713891] [] ? insert_kthread_work+0x40/0x40 Add check in qedf_process_error_detect(). When flush is active, let the cmds be completed from the cleanup contex. Link: https://lore.kernel.org/r/20210624171802.598-1-jhasan@marvell.com Signed-off-by: Javed Hasan Signed-off-by: Martin K. Petersen --- diff --git a/drivers/scsi/qedf/qedf_io.c b/drivers/scsi/qedf/qedf_io.c index 6184bc4..6b5b6a7 100644 --- a/drivers/scsi/qedf/qedf_io.c +++ b/drivers/scsi/qedf/qedf_io.c @@ -1515,9 +1515,19 @@ void qedf_process_error_detect(struct qedf_ctx *qedf, struct fcoe_cqe *cqe, { int rval; + if (io_req == NULL) { + QEDF_INFO(NULL, QEDF_LOG_IO, "io_req is NULL.\n"); + return; + } + + if (io_req->fcport == NULL) { + QEDF_INFO(NULL, QEDF_LOG_IO, "fcport is NULL.\n"); + return; + } + if (!cqe) { QEDF_INFO(&qedf->dbg_ctx, QEDF_LOG_IO, - "cqe is NULL for io_req %p\n", io_req); + "cqe is NULL for io_req %p\n", io_req); return; } @@ -1533,6 +1543,16 @@ void qedf_process_error_detect(struct qedf_ctx *qedf, struct fcoe_cqe *cqe, le32_to_cpu(cqe->cqe_info.err_info.rx_buf_off), le32_to_cpu(cqe->cqe_info.err_info.rx_id)); + /* When flush is active, let the cmds be flushed out from the cleanup context */ + if (test_bit(QEDF_RPORT_IN_TARGET_RESET, &io_req->fcport->flags) || + (test_bit(QEDF_RPORT_IN_LUN_RESET, &io_req->fcport->flags) && + io_req->sc_cmd->device->lun == (u64)io_req->fcport->lun_reset_lun)) { + QEDF_ERR(&qedf->dbg_ctx, + "Dropping EQE for xid=0x%x as fcport is flushing", + io_req->xid); + return; + } + if (qedf->stop_io_on_error) { qedf_stop_all_io(qedf); return;