linux-kernel - Re: SATA powersave patches

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 for Android: free password hash cracker in your pocket

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

Date:	Tue, 21 Nov 2006 13:30:54 +0900
From:	Tejun Heo <htejun@...il.com>
To:	Pavel Machek <pavel@...e.cz>
CC:	kernel list <linux-kernel@...r.kernel.org>, axboe@...e.de
Subject: Re: SATA powersave patches

Hello,

Pavel Machek wrote:
>> If I understood correctly, the high power consumption of 
>> ahci controller can be solved by dynamically turning off 
>> command processing while the controller is idle, which 
>> fits nicely into link powersaving, right?  So, I think 
>> full-fledged leveled dynamic PM would be an overkill for 
>> this particular problem, but then again, maybe the 
> 
> It is single bit, and it should not even need a timeout, AFAICT, so
> perhaps we should just fix it (no need for dynamic PM layers). It
> probably does not even need to be configurable...

I think this has been discussed in linux-ide recently but just to add my 
2 cents.  ALPE and ASP can cause quite some problems.  Many devices 
don't implement link powermanagement mode properly and some locks up 
completely (recent LG dvd-rams, only physical power removal/reapply can 
recover it) while others (wd raptors) spin down on slumber.  If you turn 
on ALPE and ASP on such devices, all hell will break loose.

Even for devices which implement partial and slumber modes properly, I'm 
not very comfortable with ALPE.  It puts the link into power save mode 
as soon as there is no command pending.  IMHO, that's way too 
aggressive.  It's not like being that aggressive will bring any 
noticeable difference in power consumption.  Disk idle period is too 
long for that level of aggressiveness to be effective.

Thanks.

-- 
tejun
-
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/