linux-kernel - Re: [RFC PATCH v3 4/4] [SCSI] sd: change to auto suspend mode

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAF1ivSbUHpo-ia+xZY0EZ9vQ41f6NTN-Y-+FaO=AMom0t+Xkjw@mail.gmail.com>
Date:	Thu, 24 May 2012 00:46:42 +0800
From:	Lin Ming <ming.m.lin@...el.com>
To:	Alan Stern <stern@...land.harvard.edu>
Cc:	Jens Axboe <axboe@...nel.dk>, linux-kernel@...r.kernel.org,
	linux-pm@...r.kernel.org, linux-scsi@...r.kernel.org
Subject: Re: [RFC PATCH v3 4/4] [SCSI] sd: change to auto suspend mode

On Wed, May 23, 2012 at 10:43 PM, Alan Stern <stern@...land.harvard.edu> wrote:
> On Tue, 22 May 2012, Lin Ming wrote:
>
>> Uses block layer runtime pm helper functions in scsi_runtime_suspend/resume.
>> Remove scsi_autopm_* from sd open/release/remove path.
>
> Sorry I didn't have time to get to this yesterday...
>
>> --- a/drivers/scsi/scsi_pm.c
>> +++ b/drivers/scsi/scsi_pm.c
>
>> @@ -171,9 +183,10 @@ static int scsi_runtime_idle(struct device *dev)
>>
>>       /* Insert hooks here for targets, hosts, and transport classes */
>>
>> -     if (scsi_is_sdev_device(dev))
>> -             err = pm_schedule_suspend(dev, 100);
>> -     else
>> +     if (scsi_is_sdev_device(dev)) {
>> +             pm_runtime_mark_last_busy(dev);
>> +             err = pm_request_autosuspend(dev);
>
> This really should be pm_runtime_autosuspend(dev).  In practice there's
> very little difference; it's mostly a matter of style.
>
>> --- a/drivers/scsi/sd.c
>> +++ b/drivers/scsi/sd.c
>
>> @@ -2631,7 +2624,7 @@ static void sd_probe_async(void *data, async_cookie_t cookie)
>>
>>       sd_printk(KERN_NOTICE, sdkp, "Attached SCSI %sdisk\n",
>>                 sdp->removable ? "removable " : "");
>> -     scsi_autopm_put_device(sdp);
>> +     pm_runtime_put_sync_autosuspend(&sdp->sdev_gendev);
>
> This should be left the way it was.  scsi_autopm_put_device() does
> pm_runtime_put_sync(), which will call scsi_runtime_idle(), which will
> now call pm_runtime_autosuspend().
>
>>       put_device(&sdkp->dev);
>>  }
>>
>> @@ -2755,7 +2748,6 @@ static int sd_remove(struct device *dev)
>>       struct scsi_disk *sdkp;
>>
>>       sdkp = dev_get_drvdata(dev);
>> -     scsi_autopm_get_device(sdkp->device);
>
> This line should be kept as is.  The SCSI core uses the incremented
> usage count to prevent driverless devices from being runtime-suspended.

Will update.

Thanks a lot, Alan!

>
> Alan Stern
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/