Date: Fri, 25 Aug 2023 16:20:27 GMT From: Warner Losh <imp@FreeBSD.org> To: src-committers@FreeBSD.org, dev-commits-src-all@FreeBSD.org, dev-commits-src-main@FreeBSD.org Subject: git: 2a6b7055a980 - main - nvme: Timeout expired transactions Message-ID: <202308251620.37PGKRiJ076652@gitrepo.freebsd.org>
next in thread | raw e-mail | index | archive | help
The branch main has been updated by imp: URL: https://cgit.FreeBSD.org/src/commit/?id=2a6b7055a980f7e7543dfdbda4aa0c356133b77d commit 2a6b7055a980f7e7543dfdbda4aa0c356133b77d Author: Warner Losh <imp@FreeBSD.org> AuthorDate: 2023-08-25 16:09:50 +0000 Commit: Warner Losh <imp@FreeBSD.org> CommitDate: 2023-08-25 16:10:44 +0000 nvme: Timeout expired transactions When we went to having a shared timeout routine, failing the timed-out transaction code was inadvertantly dropped. Reinstate it. Fixes: 502dc84a8b670 Sponsored by: Netflix MFC After: 2 weeks Reviewed by: chuck, jhb Differential Revision: https://reviews.freebsd.org/D36921 --- sys/dev/nvme/nvme_qpair.c | 60 ++++++++++++++++++++++++++++++++++++----------- 1 file changed, 46 insertions(+), 14 deletions(-) diff --git a/sys/dev/nvme/nvme_qpair.c b/sys/dev/nvme/nvme_qpair.c index 45a999c17d89..0ad0b7cbe17f 100644 --- a/sys/dev/nvme/nvme_qpair.c +++ b/sys/dev/nvme/nvme_qpair.c @@ -944,6 +944,30 @@ nvme_io_qpair_destroy(struct nvme_qpair *qpair) nvme_qpair_destroy(qpair); } +static void +nvme_abort_complete(void *arg, const struct nvme_completion *status) +{ + struct nvme_tracker *tr = arg; + + /* + * If cdw0 == 1, the controller was not able to abort the command + * we requested. We still need to check the active tracker array, + * to cover race where I/O timed out at same time controller was + * completing the I/O. + */ + if (status->cdw0 == 1 && tr->qpair->act_tr[tr->cid] != NULL) { + /* + * An I/O has timed out, and the controller was unable to + * abort it for some reason. Construct a fake completion + * status, and then complete the I/O's tracker manually. + */ + nvme_printf(tr->qpair->ctrlr, + "abort command failed, aborting command manually\n"); + nvme_qpair_manual_complete_tracker(tr, + NVME_SCT_GENERIC, NVME_SC_ABORTED_BY_REQUEST, 0, ERROR_PRINT_ALL); + } +} + static void nvme_qpair_timeout(void *arg) { @@ -952,36 +976,44 @@ nvme_qpair_timeout(void *arg) struct nvme_tracker *tr; sbintime_t now; bool idle; + bool expired; uint32_t csts; uint8_t cfs; mtx_lock(&qpair->lock); idle = TAILQ_EMPTY(&qpair->outstanding_tr); + again: switch (qpair->recovery_state) { case RECOVERY_NONE: - if (idle) - break; + /* + * Check to see if we need to timeout any commands. If we do, then + * we also enter a recovery phase. + */ now = getsbinuptime(); - idle = true; + expired = false; TAILQ_FOREACH(tr, &qpair->outstanding_tr, tailq) { if (tr->deadline == SBT_MAX) continue; idle = false; if (now > tr->deadline) { - /* - * We're now passed our earliest deadline. We - * need to do expensive things to cope, but next - * time. Flag that and close the door to any - * further processing. - */ - qpair->recovery_state = RECOVERY_START; - nvme_printf(ctrlr, "RECOVERY_START %jd vs %jd\n", - (uintmax_t)now, (uintmax_t)tr->deadline); - break; + expired = true; + nvme_ctrlr_cmd_abort(ctrlr, tr->cid, qpair->id, + nvme_abort_complete, tr); } } - break; + if (!expired) + break; + + /* + * We're now passed our earliest deadline. We need to do + * expensive things to cope, but next time. Flag that + * and close the door to any further processing. + */ + qpair->recovery_state = RECOVERY_START; + nvme_printf(ctrlr, "RECOVERY_START %jd vs %jd\n", + (uintmax_t)now, (uintmax_t)tr->deadline); + /* FALLTHROUGH */ case RECOVERY_START: /* * Read csts to get value of cfs - controller fatal status.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?202308251620.37PGKRiJ076652>