Skip to content

Commit

Permalink
scsi: qla2xxx: Fix fabric scan hang
Browse files Browse the repository at this point in the history 
On timeout, SRB pointer was cleared from outstanding command array and
dropped.  It was not allowed to go through the done process and cleanup.
This patch will abort the SRB where FW will return it with an error status
and resume the normal cleanup.

Link: https://lore.kernel.org/r/20191217220617.28084-3-hmadhani@marvell.com
Signed-off-by: Quinn Tran <qutran@marvell.com>
Signed-off-by: Himanshu Madhani <hmadhani@marvell.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
  • Loading branch information
@martinkpetersen
Quinn Tran authored and martinkpetersen committed Dec 20, 2019
1 parent 3c75ad1 commit f57a010
Show file tree
Hide file tree
Showing 3 changed files with 53 additions and 23 deletions.
1 change: 1 addition & 0 deletions drivers/scsi/qla2xxx/qla_gbl.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -255,6 +255,7 @@ extern char *qla2x00_get_fw_version_str(struct scsi_qla_host *, char *);

extern void qla2x00_mark_device_lost(scsi_qla_host_t *, fc_port_t *, int);
extern void qla2x00_mark_all_devices_lost(scsi_qla_host_t *);
extern int qla24xx_async_abort_cmd(srb_t *, bool);

extern struct fw_blob *qla2x00_request_firmware(scsi_qla_host_t *);

Expand Down
34 changes: 19 additions & 15 deletions drivers/scsi/qla2xxx/qla_init.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -49,16 +49,9 @@ qla2x00_sp_timeout(struct timer_list *t)
{
srb_t *sp = from_timer(sp, t, u.iocb_cmd.timer);
struct srb_iocb *iocb;
struct req_que *req;
unsigned long flags;
struct qla_hw_data *ha = sp->vha->hw;

WARN_ON_ONCE(irqs_disabled());
spin_lock_irqsave(&ha->hardware_lock, flags);
req = sp->qpair->req;
req->outstanding_cmds[sp->handle] = NULL;
WARN_ON(irqs_disabled());
iocb = &sp->u.iocb_cmd;
spin_unlock_irqrestore(&ha->hardware_lock, flags);
iocb->timeout(sp);
}

Expand Down Expand Up @@ -142,7 +135,7 @@ static void qla24xx_abort_sp_done(srb_t *sp, int res)
sp->free(sp);
}

static int qla24xx_async_abort_cmd(srb_t *cmd_sp, bool wait)
int qla24xx_async_abort_cmd(srb_t *cmd_sp, bool wait)
{
scsi_qla_host_t *vha = cmd_sp->vha;
struct srb_iocb *abt_iocb;
Expand Down Expand Up @@ -242,6 +235,7 @@ qla2x00_async_iocb_timeout(void *data)
case SRB_NACK_PRLI:
case SRB_NACK_LOGO:
case SRB_CTRL_VP:
default:
rc = qla24xx_async_abort_cmd(sp, false);
if (rc) {
spin_lock_irqsave(sp->qpair->qp_lock_ptr, flags);
Expand All @@ -258,10 +252,6 @@ qla2x00_async_iocb_timeout(void *data)
sp->done(sp, QLA_FUNCTION_TIMEOUT);
}
break;
default:
WARN_ON_ONCE(true);
sp->done(sp, QLA_FUNCTION_TIMEOUT);
break;
}
}

Expand Down Expand Up @@ -1758,9 +1748,23 @@ qla2x00_tmf_iocb_timeout(void *data)
{
srb_t *sp = data;
struct srb_iocb *tmf = &sp->u.iocb_cmd;
int rc, h;
unsigned long flags;

tmf->u.tmf.comp_status = CS_TIMEOUT;
complete(&tmf->u.tmf.comp);
rc = qla24xx_async_abort_cmd(sp, false);
if (rc) {
spin_lock_irqsave(sp->qpair->qp_lock_ptr, flags);
for (h = 1; h < sp->qpair->req->num_outstanding_cmds; h++) {
if (sp->qpair->req->outstanding_cmds[h] == sp) {
sp->qpair->req->outstanding_cmds[h] = NULL;
break;
}
}
spin_unlock_irqrestore(sp->qpair->qp_lock_ptr, flags);
tmf->u.tmf.comp_status = CS_TIMEOUT;
tmf->u.tmf.data = QLA_FUNCTION_FAILED;
complete(&tmf->u.tmf.comp);
}
}

static void qla2x00_tmf_sp_done(srb_t *sp, int res)
Expand Down
41 changes: 33 additions & 8 deletions drivers/scsi/qla2xxx/qla_iocb.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2537,13 +2537,32 @@ qla2x00_els_dcmd_iocb_timeout(void *data)
fc_port_t *fcport = sp->fcport;
struct scsi_qla_host *vha = sp->vha;
struct srb_iocb *lio = &sp->u.iocb_cmd;
unsigned long flags = 0;
int res, h;

ql_dbg(ql_dbg_io, vha, 0x3069,
"%s Timeout, hdl=%x, portid=%02x%02x%02x\n",
sp->name, sp->handle, fcport->d_id.b.domain, fcport->d_id.b.area,
fcport->d_id.b.al_pa);

complete(&lio->u.els_logo.comp);
/* Abort the exchange */
res = qla24xx_async_abort_cmd(sp, false);
if (res) {
ql_dbg(ql_dbg_io, vha, 0x3070,
"mbx abort_command failed.\n");
spin_lock_irqsave(sp->qpair->qp_lock_ptr, flags);
for (h = 1; h < sp->qpair->req->num_outstanding_cmds; h++) {
if (sp->qpair->req->outstanding_cmds[h] == sp) {
sp->qpair->req->outstanding_cmds[h] = NULL;
break;
}
}
spin_unlock_irqrestore(sp->qpair->qp_lock_ptr, flags);
complete(&lio->u.els_logo.comp);
} else {
ql_dbg(ql_dbg_io, vha, 0x3071,
"mbx abort_command success.\n");
}
}

static void qla2x00_els_dcmd_sp_done(srb_t *sp, int res)
Expand Down Expand Up @@ -2717,23 +2736,29 @@ qla2x00_els_dcmd2_iocb_timeout(void *data)
srb_t *sp = data;
fc_port_t *fcport = sp->fcport;
struct scsi_qla_host *vha = sp->vha;
struct qla_hw_data *ha = vha->hw;
unsigned long flags = 0;
int res;
int res, h;

ql_dbg(ql_dbg_io + ql_dbg_disc, vha, 0x3069,
"%s hdl=%x ELS Timeout, %8phC portid=%06x\n",
sp->name, sp->handle, fcport->port_name, fcport->d_id.b24);

/* Abort the exchange */
spin_lock_irqsave(&ha->hardware_lock, flags);
res = ha->isp_ops->abort_command(sp);
res = qla24xx_async_abort_cmd(sp, false);
ql_dbg(ql_dbg_io, vha, 0x3070,
"mbx abort_command %s\n",
(res == QLA_SUCCESS) ? "successful" : "failed");
spin_unlock_irqrestore(&ha->hardware_lock, flags);

sp->done(sp, QLA_FUNCTION_TIMEOUT);
if (res) {
spin_lock_irqsave(sp->qpair->qp_lock_ptr, flags);
for (h = 1; h < sp->qpair->req->num_outstanding_cmds; h++) {
if (sp->qpair->req->outstanding_cmds[h] == sp) {
sp->qpair->req->outstanding_cmds[h] = NULL;
break;
}
}
spin_unlock_irqrestore(sp->qpair->qp_lock_ptr, flags);
sp->done(sp, QLA_FUNCTION_TIMEOUT);
}
}

void qla2x00_els_dcmd2_free(scsi_qla_host_t *vha, struct els_plogi *els_plogi)
Expand Down

0 comments on commit f57a010

Please sign in to comment.