Home » Mailing list archives » linux-kernel » 2008 » April » 24

Re: [RFC][PATCH 0/3] Skip I/O merges when disabled

Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
[view in full thread]
From: Andi Kleen
Subject: Re: [RFC][PATCH 0/3] Skip I/O merges when disabled
Date: Thursday, April 24, 2008 - 7:15 am

> Not a good idea IMHO, it's much better with an explicit setting. That

So you would be deterministically slower.

Another way to avoid this problem would be to keep the statistics per
IO context, then the same run of a program would always get the same
behaviour. Drawback is that if your non mergeable workload consists of
lots of short running processes (like a shell script) the optimization
wouldn't work. Not sure if it's really practical, but it would be an option.

I think in modern systems with caches etc. you typically have enough
non quite deterministic and other surprising and hard to analyze
behaviour anyways, so a little more doesn't make much difference.

-Andi



--
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
Messages in current thread:
[RFC][PATCH 0/3] Skip I/O merges when disabled, Alan D. Brunelle, (Wed Apr 23, 12:08 pm)
Re: [RFC][PATCH 1/3] Add flag and sysfs interfaces, Alan D. Brunelle, (Wed Apr 23, 12:12 pm)
[RFC][PATCH 2/3] Have __make_request skip merges when disabled, Alan D. Brunelle, (Wed Apr 23, 12:14 pm)
[RFC][PATCH 3/3] Do not use rqhash when merges disabled, Alan D. Brunelle, (Wed Apr 23, 12:15 pm)
Re: [RFC][PATCH 3/3] Do not use rqhash when merges disabled, Aaron Carroll, (Wed Apr 23, 5:37 pm)
Re: [RFC][PATCH 3/3] Do not use rqhash when merges disabled, Alan D. Brunelle, (Wed Apr 23, 5:59 pm)
Re: [RFC][PATCH 3/3] Do not use rqhash when merges disabled, Aaron Carroll, (Wed Apr 23, 7:07 pm)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Jens Axboe, (Thu Apr 24, 12:09 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Alan D. Brunelle, (Thu Apr 24, 5:09 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Andi Kleen, (Thu Apr 24, 6:29 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Alan D. Brunelle, (Thu Apr 24, 6:31 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Alan D. Brunelle, (Thu Apr 24, 6:43 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Jens Axboe, (Thu Apr 24, 6:59 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Alan D. Brunelle, (Thu Apr 24, 7:13 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Andi Kleen, (Thu Apr 24, 7:15 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Jens Axboe, (Thu Apr 24, 8:04 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Jens Axboe, (Thu Apr 24, 8:05 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, David Collier-Brown, (Thu Apr 24, 8:53 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Alan D. Brunelle, (Thu Apr 24, 9:29 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Alan D. Brunelle, (Thu Apr 24, 1:38 pm)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Carl Henrik Lunde, (Thu Apr 24, 3:04 pm)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Andi Kleen, (Fri Apr 25, 12:13 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Jens Axboe, (Fri Apr 25, 1:38 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Alan D. Brunelle, (Fri Apr 25, 4:17 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Jens Axboe, (Fri Apr 25, 4:25 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Aaron Carroll, (Fri Apr 25, 5:06 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Jens Axboe, (Fri Apr 25, 5:14 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Alan D. Brunelle, (Fri Apr 25, 5:17 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Alan D. Brunelle, (Mon Apr 28, 9:36 am)
Re: [RFC][PATCH 0/3] Skip I/O merges when disabled, Jens Axboe, (Tue Apr 29, 12:37 am)