- To: Jeff Garzik <jgarzik@xxxxxxxxx>, Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>, Lin Ming <minggr@xxxxxxxxx>, James Bottomley <James.Bottomley@xxxxxxxxxxxxxxxxxxxxx>, Jeff Wu <jeff.wu@xxxxxxx>
- Subject: [PATCH v2 3/7] scsi: sr: use scsi layer autopm in sr driver
- From: Aaron Lu <aaron.lu@xxxxxxx>
- Date: Tue, 24 Jul 2012 17:37:41 +0800
- Cc: <linux-kernel@xxxxxxxxxxxxxxx>, <linux-pm@xxxxxxxxxxxxxxx>, <linux-scsi@xxxxxxxxxxxxxxx>, <linux-ide@xxxxxxxxxxxxxxx>, Aaron Lu <aaron.lwe@xxxxxxxxx>, Aaron Lu <aaron.lu@xxxxxxx>
- In-reply-to: <1343122665-18711-1-git-send-email-aaron.lu@amd.com>
Use scsi_autopm_* whenever possible instead of the pm_runtime_* in sr
driver as suggested by Alan Stern.
The reason pm_runtime_get_noresume in sr_suspend is used is that we
can't do sync resume call in suspend callback, or rpm_resume will wait
forever for the suspend request to finish.
Signed-off-by: Aaron Lu <aaron.lu@xxxxxxx>
---
V2:
Use scsi_autopm_* instead of pm_runtime_* in sr driver as suggested
by Alan Stern.
drivers/scsi/sr.c | 16 +++++-----------
1 file changed, 5 insertions(+), 11 deletions(-)
diff --git a/drivers/scsi/sr.c b/drivers/scsi/sr.c
index 225848c..e54945a 100644
--- a/drivers/scsi/sr.c
+++ b/drivers/scsi/sr.c
@@ -377,10 +377,8 @@ do_tur:
* thread (in-kernel polling) and old versions of udisks,
* to avoid put the device twice, an atomic operation is used.
*/
- if (poweroff && atomic_add_unless(&cd->suspend_count, -1, 0)) {
- pm_runtime_mark_last_busy(&cd->device->sdev_gendev);
- pm_runtime_put_autosuspend(&cd->device->sdev_gendev);
- }
+ if (poweroff && atomic_add_unless(&cd->suspend_count, -1, 0))
+ scsi_autopm_put_device(cd->device);
}
if (cd->ignore_get_event)
@@ -655,11 +653,8 @@ static int sr_block_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd,
int ret;
/* Make sure the ODD is not suspended */
- ret = pm_runtime_get_sync(&sdev->sdev_gendev);
- if (ret < 0) {
- pm_runtime_put_noidle(&sdev->sdev_gendev);
+ if (scsi_autopm_get_device(sdev))
return -EACCES;
- }
mutex_lock(&sr_mutex);
@@ -692,8 +687,7 @@ static int sr_block_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd,
out:
mutex_unlock(&sr_mutex);
- pm_runtime_mark_last_busy(&cd->device->sdev_gendev);
- pm_runtime_put_autosuspend(&cd->device->sdev_gendev);
+ scsi_autopm_put_device(sdev);
return ret;
}
@@ -1121,7 +1115,7 @@ static int sr_remove(struct device *dev)
/* disable runtime pm and possibly resume the device */
if (!atomic_dec_and_test(&cd->suspend_count))
- pm_runtime_get_sync(dev);
+ scsi_autopm_get_device(cd->device);
blk_queue_prep_rq(cd->device->request_queue, scsi_prep_fn);
del_gendisk(cd->disk);
--
1.7.11.3
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
[SCSI Target Devel]
[Linux SCSI Target Infrastructure]
[Kernel Newbies]
[Share Photos]
[IDE]
[Security]
[Git]
[Netfilter]
[Bugtraq]
[Photos]
[Yosemite]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Linux ATA RAID]
[Linux IIO]
[Samba]
[Video 4 Linux]
[Device Mapper]
[Linux Resources]