From: Nick Piggin <piggin@cyberone.com.au>
To: John Bradford <john@grabjohn.com>
Cc: alan@lxorguk.ukuu.org.uk, davidsen@tmr.com,
linux-kernel@vger.kernel.org, zwane@linuxpower.ca
Subject: Re: [PATCH] 2.6 workaround for Athlon/Opteron prefetch errata
Date: Mon, 15 Sep 2003 19:58:39 +1000 [thread overview]
Message-ID: <3F658D4F.1020409@cyberone.com.au> (raw)
In-Reply-To: <200309150939.h8F9d13D000943@81-2-122-30.bradfords.org.uk>
John Bradford wrote:
>>>>>That's a non-issue. 300 bytes matters a lot on some systems. The
>>>>>fact that there are drivers that are bloated is nothing to do with
>>>>>it.
>>>>>
>>>>>
>>>>Its kind of irrelevant when by saying "Athlon" you've added 128 byte
>>>>alignment to all the cache friendly structure padding.
>>>>
>>>>
>>>My intention is that we won't have done 128 byte alignments just by
>>>'supporting' Athlons, only if we want to run fast on Athlons. A
>>>distribution kernel that is intended to boot on all CPUs needs
>>>workarounds for Athlon bugs, but it doesn't need 128 byte alignment.
>>>
>>>Obviously using such a kernel for anything other than getting a system
>>>up and running to compile a better kernel is a Bad Thing, but the
>>>distributions could supply separate Athlon, PIV, and 386 _optimised_
>>>kernels.
>>>
>>>
>>Why bother with that complexity? Just use 128 byte lines. This allows
>>a decent generic kernel. The people who have space requirements would
>>only compile what they need anyway.
>>
>
>So, basically, if you compile a kernel for a 386, but think that maybe
>one day you might need to run it on an Athlon for debugging purposes,
>you use 128 byte padding, because it's not too bad on the 386? Seems
>pretty wasteful to me when the obvious, simple, elegant solution is to
>allow independent selection of workaround inclusion and optimisation.
>Especially since half of the work has already been done.
>
I missed the "simple, elegant" part. Conceptually elegant maybe.
If you mean to use the optimise option only to set cache line size, then
that might be a bit saner.
As far as the case study goes though: if you were worried about being
wasteful, why wouldn't you compile just for the 386 and debug from that?
next prev parent reply other threads:[~2003-09-15 9:59 UTC|newest]
Thread overview: 122+ messages / expand[flat|nested] mbox.gz Atom feed top
2003-09-15 9:39 [PATCH] 2.6 workaround for Athlon/Opteron prefetch errata John Bradford
2003-09-15 9:58 ` Nick Piggin [this message]
-- strict thread matches above, loose matches on Subject: below --
2003-09-16 2:23 richard.brunner
2003-09-15 20:20 Nakajima, Jun
2003-09-15 20:03 Nakajima, Jun
2003-09-15 19:51 richard.brunner
2003-09-16 0:01 ` David Lang
2003-09-16 0:20 ` Bill Davidsen
2003-09-16 11:50 ` Jamie Lokier
2003-09-16 13:46 ` Bill Davidsen
2003-09-16 17:21 ` Jamie Lokier
2003-09-15 19:34 John Bradford
2003-09-15 19:25 ` Bill Davidsen
2003-09-15 22:28 ` Alan Cox
2003-09-15 19:19 John Bradford
[not found] <200309150632.h8F6WnHb000589@81-2-122-30.bradfords.org.uk.suse.lists.linux.kernel>
[not found] ` <1063611650.2674.1.camel@dhcp23.swansea.linux.org.uk.suse.lists.linux.kernel>
2003-09-15 18:29 ` Andi Kleen
2003-09-15 16:21 richard.brunner
2003-09-15 19:15 ` Bill Davidsen
2003-09-16 11:46 ` Jamie Lokier
2003-09-16 13:30 ` Dave Jones
2003-09-16 13:52 ` Bill Davidsen
2003-09-16 15:25 ` Timothy Miller
2003-09-16 16:53 ` Bill Davidsen
2003-09-16 17:22 ` Jamie Lokier
2003-09-16 17:23 ` Jamie Lokier
2003-09-18 7:43 ` Pavel Machek
2003-09-18 14:05 ` Dave Jones
2003-09-18 15:56 ` Jamie Lokier
2003-09-18 17:34 ` Bill Davidsen
2003-09-15 14:21 John Bradford
2003-09-15 12:43 John Bradford
2003-09-15 18:21 ` Bill Davidsen
2003-09-15 12:28 Mikael Pettersson
2003-09-15 18:13 ` Bill Davidsen
2003-09-16 11:54 ` Jamie Lokier
2003-09-15 11:46 John Bradford
2003-09-15 12:38 ` Nick Piggin
2003-09-15 13:46 ` Chris Meadors
2003-09-15 14:00 ` Nick Piggin
2003-09-16 15:06 ` Bill Davidsen
2003-09-16 15:24 ` Nick Piggin
2003-09-15 10:54 John Bradford
2003-09-15 10:52 ` Nick Piggin
2003-09-15 13:45 ` Alan Cox
2003-09-15 18:44 ` Bill Davidsen
2003-09-15 8:31 John Bradford
2003-09-15 8:32 ` Nick Piggin
2003-09-15 20:51 ` Adrian Bunk
2003-09-15 6:32 John Bradford
2003-09-15 7:40 ` Alan Cox
2003-09-15 12:02 ` Bill Davidsen
2003-09-15 11:48 ` Bill Davidsen
2003-09-15 20:55 ` Adrian Bunk
2003-09-16 0:26 ` Bill Davidsen
2003-09-12 21:24 John Bradford
2003-09-11 17:17 richard.brunner
2003-09-13 16:54 ` Pavel Machek
2003-09-11 17:14 richard.brunner
2003-09-11 17:09 richard.brunner
2003-09-11 4:55 richard.brunner
2003-09-11 16:55 ` Jamie Lokier
2003-09-12 14:14 ` Martin Schlemmer
[not found] <uqD5.3BI.3@gated-at.bofh.it>
2003-09-11 4:14 ` Andi Kleen
2003-09-11 4:58 ` dada1
2003-09-11 5:11 ` Andi Kleen
2003-09-11 5:58 ` dada1
2003-09-11 3:43 Nakajima, Jun
2003-09-11 4:03 ` Valdis.Kletnieks
2003-09-11 0:56 Update on AMD Athlon/Opteron/Athlon64 Prefetch Errata richard.brunner
2003-09-11 1:27 ` [PATCH] 2.6 workaround for Athlon/Opteron prefetch errata Andi Kleen
2003-09-11 1:44 ` Andrew Morton
2003-09-11 1:47 ` Andi Kleen
2003-09-11 14:15 ` Jeff Garzik
2003-09-11 14:26 ` Andi Kleen
2003-09-11 14:34 ` Jeff Garzik
2003-09-11 14:58 ` Andi Kleen
2003-09-11 15:06 ` Jeff Garzik
2003-09-11 20:08 ` bill davidsen
2003-09-11 19:56 ` bill davidsen
2003-09-11 20:44 ` Alan Cox
2003-09-11 21:29 ` Mike Fedyk
2003-09-11 21:38 ` bill davidsen
2003-09-12 17:32 ` Eric W. Biederman
2003-09-12 17:56 ` Andi Kleen
2003-09-12 17:59 ` Jeff Garzik
2003-09-12 18:22 ` Adrian Bunk
2003-09-12 18:28 ` Andi Kleen
2003-09-12 18:39 ` Jeff Garzik
2003-09-12 18:45 ` Jeff Garzik
2003-09-12 18:48 ` Adrian Bunk
2003-09-12 19:07 ` Andi Kleen
2003-09-12 19:05 ` Martin Schlemmer
2003-09-12 19:30 ` Andi Kleen
2003-09-12 19:58 ` Martin Schlemmer
2003-09-12 20:00 ` Adrian Bunk
2003-09-15 0:15 ` bill davidsen
2003-09-14 23:51 ` bill davidsen
2003-09-14 23:49 ` bill davidsen
2003-09-14 23:47 ` bill davidsen
2003-09-15 1:02 ` Zwane Mwaikambo
2003-09-15 2:08 ` Nick Piggin
2003-09-15 3:55 ` Bill Davidsen
2003-09-15 7:45 ` Alan Cox
2003-09-15 12:11 ` Bill Davidsen
2003-09-15 13:48 ` Alan Cox
2003-09-15 18:50 ` Bill Davidsen
2003-09-12 18:03 ` Mike Fedyk
2003-09-11 13:54 ` Linus Torvalds
2003-09-11 14:01 ` Andi Kleen
2003-09-11 14:14 ` Dave Jones
2003-09-11 14:29 ` Andi Kleen
2003-09-11 14:14 ` Alan Cox
2003-09-11 14:24 ` Andi Kleen
2003-09-11 14:28 ` Dave Jones
2003-09-11 14:32 ` Andi Kleen
2003-09-11 20:14 ` bill davidsen
2003-09-11 16:58 ` Jamie Lokier
2003-09-11 17:05 ` Andi Kleen
2003-09-11 17:32 ` Jamie Lokier
2003-09-11 17:39 ` Andi Kleen
2003-09-11 17:48 ` Jamie Lokier
2003-09-11 18:18 ` Andi Kleen
2003-09-11 18:59 ` Jamie Lokier
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=3F658D4F.1020409@cyberone.com.au \
--to=piggin@cyberone.com.au \
--cc=alan@lxorguk.ukuu.org.uk \
--cc=davidsen@tmr.com \
--cc=john@grabjohn.com \
--cc=linux-kernel@vger.kernel.org \
--cc=zwane@linuxpower.ca \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).