scsi: qla2xxx: Implement ref count for SRB

The timeout handler and the done function are racing. When
qla2x00_async_iocb_timeout() starts to run it can be preempted by the
normal response path (via the firmware?). qla24xx_async_gpsc_sp_done()
releases the SRB unconditionally. When scheduling back to
qla2x00_async_iocb_timeout() qla24xx_async_abort_cmd() will access an freed
sp->qpair pointer:

  qla2xxx [0000:83:00.0]-2871:0: Async-gpsc timeout - hdl=63d portid=234500 50:06:0e:80:08:77:b6:21.
  qla2xxx [0000:83:00.0]-2853:0: Async done-gpsc res 0, WWPN 50:06:0e:80:08:77:b6:21
  qla2xxx [0000:83:00.0]-2854:0: Async-gpsc OUT WWPN 20:45:00:27:f8:75:33:00 speeds=2c00 speed=0400.
  qla2xxx [0000:83:00.0]-28d8:0: qla24xx_handle_gpsc_event 50:06:0e:80:08:77:b6:21 DS 7 LS 6 rc 0 login 1|1 rscn 1|0 lid 5
  BUG: unable to handle kernel NULL pointer dereference at 0000000000000004
  IP: qla24xx_async_abort_cmd+0x1b/0x1c0 [qla2xxx]

Obvious solution to this is to introduce a reference counter. One reference
is taken for the normal code path (the 'good' case) and one for the timeout
path. As we always race between the normal good case and the timeout/abort
handler we need to serialize it. Also we cannot assume any order between
the handlers. Since this is slow path we can use proper synchronization via
locks.

When we are able to cancel a timer (del_timer returns 1) we know there
can't be any error handling in progress because the timeout handler hasn't
expired yet, thus we can safely decrement the refcounter by one.

If we are not able to cancel the timer, we know an abort handler is
running. We have to make sure we call sp->done() in the abort handlers
before calling kref_put().

Link: https://lore.kernel.org/r/20220110050218.3958-3-njavali@marvell.com
Cc: stable@vger.kernel.org
Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>
Co-developed-by: Daniel Wagner <dwagner@suse.de>
Signed-off-by: Daniel Wagner <dwagner@suse.de>
Signed-off-by: Saurav Kashyap <skashyap@marvell.com>
Signed-off-by: Nilesh Javali <njavali@marvell.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

1 files changed, 49 insertions, 21 deletions

diff --git a/drivers/scsi/qla2xxx/qla_init.c b/drivers/scsi/qla2xxx/qla_init.c
index e6f13cb6fa28..38c11b75f644 100644
--- a/drivers/scsi/qla2xxx/qla_init.c
+++ b/drivers/scsi/qla2xxx/qla_init.c
@@ -51,6 +51,9 @@ qla2x00_sp_timeout(struct timer_list *t)
 	WARN_ON(irqs_disabled());
 	iocb = &sp->u.iocb_cmd;
 	iocb->timeout(sp);
+
+	/* ref: TMR */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 }
 
 void qla2x00_sp_free(srb_t *sp)
@@ -125,8 +128,13 @@ static void qla24xx_abort_iocb_timeout(void *data)
 	}
 	spin_unlock_irqrestore(qpair->qp_lock_ptr, flags);
 
-	if (sp->cmd_sp)
+	if (sp->cmd_sp) {
+		/*
+		 * This done function should take care of
+		 * original command ref: INIT
+		 */
 		sp->cmd_sp->done(sp->cmd_sp, QLA_OS_TIMER_EXPIRED);
+	}
 
 	abt->u.abt.comp_status = cpu_to_le16(CS_TIMEOUT);
 	sp->done(sp, QLA_OS_TIMER_EXPIRED);
@@ -140,11 +148,11 @@ static void qla24xx_abort_sp_done(srb_t *sp, int res)
 	if (orig_sp)
 		qla_wait_nvme_release_cmd_kref(orig_sp);
 
-	del_timer(&sp->u.iocb_cmd.timer);
 	if (sp->flags & SRB_WAKEUP_ON_COMP)
 		complete(&abt->u.abt.comp);
 	else
-		sp->free(sp);
+		/* ref: INIT */
+		kref_put(&sp->cmd_kref, qla2x00_sp_release);
 }
 
 int qla24xx_async_abort_cmd(srb_t *cmd_sp, bool wait)
@@ -154,6 +162,7 @@ int qla24xx_async_abort_cmd(srb_t *cmd_sp, bool wait)
 	srb_t *sp;
 	int rval = QLA_FUNCTION_FAILED;
 
+	/* ref: INIT for ABTS command */
 	sp = qla2xxx_get_qpair_sp(cmd_sp->vha, cmd_sp->qpair, cmd_sp->fcport,
 				  GFP_ATOMIC);
 	if (!sp)
@@ -181,7 +190,8 @@ int qla24xx_async_abort_cmd(srb_t *cmd_sp, bool wait)
 
 	rval = qla2x00_start_sp(sp);
 	if (rval != QLA_SUCCESS) {
-		sp->free(sp);
+		/* ref: INIT */
+		kref_put(&sp->cmd_kref, qla2x00_sp_release);
 		return rval;
 	}
 
@@ -189,7 +199,8 @@ int qla24xx_async_abort_cmd(srb_t *cmd_sp, bool wait)
 		wait_for_completion(&abt_iocb->u.abt.comp);
 		rval = abt_iocb->u.abt.comp_status == CS_COMPLETE ?
 			QLA_SUCCESS : QLA_ERR_FROM_FW;
-		sp->free(sp);
+		/* ref: INIT */
+		kref_put(&sp->cmd_kref, qla2x00_sp_release);
 	}
 
 	return rval;
@@ -287,7 +298,8 @@ static void qla2x00_async_login_sp_done(srb_t *sp, int res)
 		qla24xx_handle_plogi_done_event(vha, &ea);
 	}
 
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 }
 
 int
@@ -306,6 +318,7 @@ qla2x00_async_login(struct scsi_qla_host *vha, fc_port_t *fcport,
 		return rval;
 	}
 
+	/* ref: INIT */
 	sp = qla2x00_get_sp(vha, fcport, GFP_KERNEL);
 	if (!sp)
 		goto done;
@@ -354,7 +367,8 @@ qla2x00_async_login(struct scsi_qla_host *vha, fc_port_t *fcport,
 	return rval;
 
 done_free_sp:
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 	fcport->flags &= ~FCF_ASYNC_SENT;
 done:
 	fcport->flags &= ~FCF_ASYNC_ACTIVE;
@@ -366,7 +380,8 @@ static void qla2x00_async_logout_sp_done(srb_t *sp, int res)
 	sp->fcport->flags &= ~(FCF_ASYNC_SENT | FCF_ASYNC_ACTIVE);
 	sp->fcport->login_gen++;
 	qlt_logo_completion_handler(sp->fcport, sp->u.iocb_cmd.u.logio.data[0]);
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 }
 
 int
@@ -376,6 +391,7 @@ qla2x00_async_logout(struct scsi_qla_host *vha, fc_port_t *fcport)
 	int rval = QLA_FUNCTION_FAILED;
 
 	fcport->flags |= FCF_ASYNC_SENT;
+	/* ref: INIT */
 	sp = qla2x00_get_sp(vha, fcport, GFP_KERNEL);
 	if (!sp)
 		goto done;
@@ -397,7 +413,8 @@ qla2x00_async_logout(struct scsi_qla_host *vha, fc_port_t *fcport)
 	return rval;
 
 done_free_sp:
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 done:
 	fcport->flags &= ~(FCF_ASYNC_SENT | FCF_ASYNC_ACTIVE);
 	return rval;
@@ -423,7 +440,8 @@ static void qla2x00_async_prlo_sp_done(srb_t *sp, int res)
 	if (!test_bit(UNLOADING, &vha->dpc_flags))
 		qla2x00_post_async_prlo_done_work(sp->fcport->vha, sp->fcport,
 		    lio->u.logio.data);
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 }
 
 int
@@ -433,6 +451,7 @@ qla2x00_async_prlo(struct scsi_qla_host *vha, fc_port_t *fcport)
 	int rval;
 
 	rval = QLA_FUNCTION_FAILED;
+	/* ref: INIT */
 	sp = qla2x00_get_sp(vha, fcport, GFP_KERNEL);
 	if (!sp)
 		goto done;
@@ -454,7 +473,8 @@ qla2x00_async_prlo(struct scsi_qla_host *vha, fc_port_t *fcport)
 	return rval;
 
 done_free_sp:
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 done:
 	fcport->flags &= ~FCF_ASYNC_ACTIVE;
 	return rval;
@@ -539,8 +559,8 @@ static void qla2x00_async_adisc_sp_done(srb_t *sp, int res)
 	ea.sp = sp;
 
 	qla24xx_handle_adisc_event(vha, &ea);
-
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 }
 
 int
@@ -555,6 +575,7 @@ qla2x00_async_adisc(struct scsi_qla_host *vha, fc_port_t *fcport,
 		return rval;
 
 	fcport->flags |= FCF_ASYNC_SENT;
+	/* ref: INIT */
 	sp = qla2x00_get_sp(vha, fcport, GFP_KERNEL);
 	if (!sp)
 		goto done;
@@ -582,7 +603,8 @@ qla2x00_async_adisc(struct scsi_qla_host *vha, fc_port_t *fcport,
 	return rval;
 
 done_free_sp:
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 done:
 	fcport->flags &= ~(FCF_ASYNC_SENT | FCF_ASYNC_ACTIVE);
 	qla2x00_post_async_adisc_work(vha, fcport, data);
@@ -1063,7 +1085,8 @@ static void qla24xx_async_gnl_sp_done(srb_t *sp, int res)
 	}
 	spin_unlock_irqrestore(&vha->hw->tgt.sess_lock, flags);
 
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 }
 
 int qla24xx_async_gnl(struct scsi_qla_host *vha, fc_port_t *fcport)
@@ -1093,6 +1116,7 @@ int qla24xx_async_gnl(struct scsi_qla_host *vha, fc_port_t *fcport)
 	vha->gnl.sent = 1;
 	spin_unlock_irqrestore(&vha->hw->tgt.sess_lock, flags);
 
+	/* ref: INIT */
 	sp = qla2x00_get_sp(vha, fcport, GFP_KERNEL);
 	if (!sp)
 		goto done;
@@ -1125,7 +1149,8 @@ int qla24xx_async_gnl(struct scsi_qla_host *vha, fc_port_t *fcport)
 	return rval;
 
 done_free_sp:
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 done:
 	fcport->flags &= ~(FCF_ASYNC_ACTIVE | FCF_ASYNC_SENT);
 	return rval;
@@ -1171,7 +1196,7 @@ done:
 	dma_pool_free(ha->s_dma_pool, sp->u.iocb_cmd.u.mbx.in,
 		sp->u.iocb_cmd.u.mbx.in_dma);
 
-	sp->free(sp);
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 }
 
 int qla24xx_post_prli_work(struct scsi_qla_host *vha, fc_port_t *fcport)
@@ -1216,7 +1241,7 @@ static void qla2x00_async_prli_sp_done(srb_t *sp, int res)
 		qla24xx_handle_prli_done_event(vha, &ea);
 	}
 
-	sp->free(sp);
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 }
 
 int
@@ -1274,7 +1299,8 @@ qla24xx_async_prli(struct scsi_qla_host *vha, fc_port_t *fcport)
 	return rval;
 
 done_free_sp:
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 	fcport->flags &= ~FCF_ASYNC_SENT;
 	return rval;
 }
@@ -1359,7 +1385,7 @@ done_free_sp:
 	if (pd)
 		dma_pool_free(ha->s_dma_pool, pd, pd_dma);
 
-	sp->free(sp);
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 	fcport->flags &= ~FCF_ASYNC_SENT;
 done:
 	fcport->flags &= ~FCF_ASYNC_ACTIVE;
@@ -1945,6 +1971,7 @@ qla2x00_async_tm_cmd(fc_port_t *fcport, uint32_t flags, uint32_t lun,
 	srb_t *sp;
 	int rval = QLA_FUNCTION_FAILED;
 
+	/* ref: INIT */
 	sp = qla2x00_get_sp(vha, fcport, GFP_KERNEL);
 	if (!sp)
 		goto done;
@@ -1988,7 +2015,8 @@ qla2x00_async_tm_cmd(fc_port_t *fcport, uint32_t flags, uint32_t lun,
 	}
 
 done_free_sp:
-	sp->free(sp);
+	/* ref: INIT */
+	kref_put(&sp->cmd_kref, qla2x00_sp_release);
 	fcport->flags &= ~FCF_ASYNC_SENT;
 done:
 	return rval;