From: Hannes Reinecke <hare@suse.de>
To: "Martin K. Petersen" <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>,
James Bottomley <james.bottomley@hansenpartnership.com>,
linux-scsi@vger.kernel.org, Hannes Reinecke <hare@suse.de>
Subject: [PATCH 11/16] snic: allocate device reset command
Date: Tue, 17 Oct 2023 12:07:24 +0200 [thread overview]
Message-ID: <20231017100729.123506-12-hare@suse.de> (raw)
In-Reply-To: <20231017100729.123506-1-hare@suse.de>
Allocate a command to send a device reset instead of relying
on using the command which triggered the device failure.
Signed-off-by: Hannes Reinecke <hare@suse.de>
---
drivers/scsi/snic/snic_scsi.c | 72 +++++++++++++++++++----------------
1 file changed, 40 insertions(+), 32 deletions(-)
diff --git a/drivers/scsi/snic/snic_scsi.c b/drivers/scsi/snic/snic_scsi.c
index f1ef781df837..1807f713504f 100644
--- a/drivers/scsi/snic/snic_scsi.c
+++ b/drivers/scsi/snic/snic_scsi.c
@@ -2091,70 +2091,78 @@ snic_unlink_and_release_req(struct snic *snic, struct scsi_cmnd *sc, int flag)
int
snic_device_reset(struct scsi_cmnd *sc)
{
- struct Scsi_Host *shost = sc->device->host;
+ struct scsi_device *sdev = sc->device;
+ struct Scsi_Host *shost = sdev->host;
struct snic *snic = shost_priv(shost);
+ struct request *req;
struct snic_req_info *rqi = NULL;
- int tag = snic_cmd_tag(sc);
int start_time = jiffies;
int ret = FAILED;
int dr_supp = 0;
- SNIC_SCSI_DBG(shost, "dev_reset:sc %p :0x%x :req = %p :tag = %d\n",
- sc, sc->cmnd[0], scsi_cmd_to_rq(sc),
- snic_cmd_tag(sc));
- dr_supp = snic_dev_reset_supported(sc->device);
+ SNIC_SCSI_DBG(shost, "dev_reset\n");
+ dr_supp = snic_dev_reset_supported(sdev);
if (!dr_supp) {
/* device reset op is not supported */
SNIC_HOST_INFO(shost, "LUN Reset Op not supported.\n");
- snic_unlink_and_release_req(snic, sc, SNIC_DEV_RST_NOTSUP);
-
- goto dev_rst_end;
+ return ret;
}
if (unlikely(snic_get_state(snic) != SNIC_ONLINE)) {
- snic_unlink_and_release_req(snic, sc, 0);
SNIC_HOST_ERR(shost, "Devrst: Parent Devs are not online.\n");
- goto dev_rst_end;
+ return ret;
}
- /* There is no tag when lun reset is issue through ioctl. */
- if (unlikely(tag <= SNIC_NO_TAG)) {
- SNIC_HOST_INFO(snic->shost,
- "Devrst: LUN Reset Recvd thru IOCTL.\n");
-
- rqi = snic_req_init(snic, 0);
- if (!rqi)
- goto dev_rst_end;
-
- memset(scsi_cmd_priv(sc), 0,
- sizeof(struct snic_internal_io_state));
- CMD_SP(sc) = (char *)rqi;
- CMD_FLAGS(sc) = SNIC_NO_FLAGS;
+ rqi = snic_req_init(snic, 0);
+ if (!rqi)
+ return ret;
- /* Add special tag for dr coming from user spc */
- rqi->tm_tag = SNIC_TAG_IOCTL_DEV_RST;
- rqi->sc = sc;
+ req = scsi_alloc_request(sdev->request_queue, REQ_OP_DRV_IN,
+ BLK_MQ_REQ_NOWAIT);
+ if (!req) {
+ /*
+ * Request allocation might fail, indicating that
+ * all tags are busy.
+ * But device reset will be called only from within
+ * SCSI EH, at which time all I/O is stopped. So the
+ * only active tags would be for failed I/O, but
+ * when all I/O is failed it'll be better to escalate
+ * to host reset anyway.
+ */
+ SNIC_HOST_ERR(snic->shost,
+ "Devrst: TMF busy, escalate to host reset\n");
+ goto dev_rst_end;
}
+ sc = blk_mq_rq_to_pdu(req);
+ memset(scsi_cmd_priv(sc), 0,
+ sizeof(struct snic_internal_io_state));
+ CMD_SP(sc) = (char *)rqi;
+ CMD_FLAGS(sc) = SNIC_NO_FLAGS;
+ /* Add special tag for dr coming from user spc */
+ rqi->tm_tag = SNIC_TAG_IOCTL_DEV_RST;
+ rqi->sc = sc;
+ WRITE_ONCE(req->state, MQ_RQ_IN_FLIGHT);
ret = snic_send_dr_and_wait(snic, sc);
if (ret) {
SNIC_HOST_ERR(snic->shost,
"Devrst: IO w/ Tag %x Failed w/ err = %d\n",
- tag, ret);
-
+ snic_cmd_tag(sc), ret);
+ blk_mq_set_request_complete(req);
snic_unlink_and_release_req(snic, sc, 0);
goto dev_rst_end;
}
-
+ blk_mq_set_request_complete(req);
ret = snic_dr_finish(snic, sc);
dev_rst_end:
- SNIC_TRC(snic->shost->host_no, tag, (ulong) sc,
+ SNIC_TRC(snic->shost->host_no, snic_cmd_tag(sc), (ulong) sc,
jiffies_to_msecs(jiffies - start_time),
0, SNIC_TRC_CMD(sc), SNIC_TRC_CMD_STATE_FLAGS(sc));
-
+ if (req)
+ blk_mq_free_request(req);
SNIC_SCSI_DBG(snic->shost,
"Devrst: Returning from Device Reset : %s\n",
(ret == SUCCESS) ? "SUCCESS" : "FAILED");
--
2.35.3
next prev parent reply other threads:[~2023-10-17 10:07 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-10-17 10:07 [PATCHv7 00/16] scsi: EH rework prep patches, part 2 Hannes Reinecke
2023-10-17 10:07 ` [PATCH 01/16] zfcp: do not wait for rports to become unblocked after host reset Hannes Reinecke
2023-10-19 17:44 ` Benjamin Block
2023-10-20 6:18 ` Hannes Reinecke
2023-10-17 10:07 ` [PATCH 02/16] bfa: Do not use scsi command to signal TMF status Hannes Reinecke
2023-10-17 10:07 ` [PATCH 03/16] aha152x: look for stuck command when resetting device Hannes Reinecke
2023-10-17 10:07 ` [PATCH 04/16] a1000u2w: do not rely on the command for inia100_device_reset() Hannes Reinecke
2023-10-17 10:07 ` [PATCH 05/16] fas216: Rework device reset to not rely on SCSI command pointer Hannes Reinecke
2023-10-17 10:07 ` [PATCH 06/16] xen-scsifront: add scsi device as argument to scsifront_do_request() Hannes Reinecke
2023-10-17 10:07 ` [PATCH 07/16] xen-scsifront: rework scsifront_action_handler() Hannes Reinecke
2023-10-17 10:07 ` [PATCH 08/16] libiscsi: use cls_session as argument for target and session reset Hannes Reinecke
2023-10-17 10:07 ` [PATCH 09/16] scsi_transport_iscsi: use session as argument for iscsi_block_scsi_eh() Hannes Reinecke
2023-10-17 10:07 ` [PATCH 10/16] snic: reserve tag for TMF Hannes Reinecke
2023-10-17 10:07 ` Hannes Reinecke [this message]
2023-10-18 5:47 ` [PATCH 11/16] snic: allocate device reset command Christoph Hellwig
2023-10-17 10:07 ` [PATCH 12/16] snic: Use scsi_host_busy_iter() to traverse commands Hannes Reinecke
2023-10-17 10:07 ` [PATCH 13/16] fnic: allocate device reset command on the fly Hannes Reinecke
2023-10-17 10:07 ` [PATCH 14/16] fnic: use fc_block_rport() correctly Hannes Reinecke
2023-10-17 10:07 ` [PATCH 15/16] csiostor: use separate TMF command Hannes Reinecke
2023-10-17 10:07 ` [PATCH 16/16] dc395x: Remove 'scmd' parameter from doing_srb_done() Hannes Reinecke
-- strict thread matches above, loose matches on Subject: below --
2023-10-23 9:14 [PATCHv8 00/16] scsi: EH rework prep patches, part 2 Hannes Reinecke
2023-10-23 9:15 ` [PATCH 11/16] snic: allocate device reset command Hannes Reinecke
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20231017100729.123506-12-hare@suse.de \
--to=hare@suse.de \
--cc=hch@lst.de \
--cc=james.bottomley@hansenpartnership.com \
--cc=linux-scsi@vger.kernel.org \
--cc=martin.petersen@oracle.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox