From: Mark Salyzyn <mark_salyzyn@xyratex.com>
Date: Tue, 17 Jan 2012 17:56:45 +0000 (-0500)
Subject: [SCSI] pm8001: fix lockup on phy_control hard reset.
X-Git-Tag: v3.4-rc1~145^2~175
X-Git-Url: https://git.karo-electronics.de/?a=commitdiff_plain;h=5c4fb76af31e9dabcd132a0e69ed3799df1304c3;p=karo-tx-linux.git

[SCSI] pm8001: fix lockup on phy_control hard reset.

pm8001_phy_control PHY_FUNC_HARD_RESET locks up on second try via
smp_phy_control because response HW_EVENT_PHY_START_STATUS fails to complete
previous command. The PM8001F_RUN_TIME flag is not treated as a bit, but a
state in all readers, yet once we are operational or in the run time state,
the flags use a bit-set operation.

Signed-off-by: Mark Salyzyn <mark_salyzyn@xyratex.com>
Acked-by: Jack Wang <jack_wang@usish.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
---

diff --git a/drivers/scsi/pm8001/pm8001_sas.c b/drivers/scsi/pm8001/pm8001_sas.c
index fb3dc9978861..7ae22a67bd31 100644
--- a/drivers/scsi/pm8001/pm8001_sas.c
+++ b/drivers/scsi/pm8001/pm8001_sas.c
@@ -615,7 +615,7 @@ static int pm8001_dev_found_notify(struct domain_device *dev)
 	wait_for_completion(&completion);
 	if (dev->dev_type == SAS_END_DEV)
 		msleep(50);
-	pm8001_ha->flags |= PM8001F_RUN_TIME ;
+	pm8001_ha->flags = PM8001F_RUN_TIME;
 	return 0;
 found_out:
 	spin_unlock_irqrestore(&pm8001_ha->lock, flags);