mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Andreas Mohr <andi@rhlx01.fht-esslingen.de>
To: Jeremy Fitzhardinge <jeremy@goop.org>
Cc: linux-kernel@vger.kernel.org,
	Chuck Ebbert <76306.1226@compuserve.com>,
	Zachary Amsden <zach@vmware.com>,
	Jan Beulich <jbeulich@novell.com>, Andi Kleen <ak@suse.de>,
	Andrew Morton <akpm@osdl.org>
Subject: Re: [PATCH RFC 0/6] Implement per-processor data areas for i386.
Date: Sun, 27 Aug 2006 19:21:55 +0200	[thread overview]
Message-ID: <20060827172155.GA21724@rhlx01.fht-esslingen.de> (raw)
In-Reply-To: <20060827084417.918992193@goop.org>

Hi,

On Sun, Aug 27, 2006 at 01:44:17AM -0700, Jeremy Fitzhardinge wrote:
> This patch implements per-processor data areas by using %gs as the
> base segment of the per-processor memory.  This has two principle
> advantages:
> 
> - It allows very simple direct access to per-processor data by
>   effectively using an effective address of the form %gs:offset, where
>   offset is the offset into struct i386_pda.  These sequences are faster
>   and smaller than the current mechanism using current_thread_info().

Yess!!
Something like that had to be done eventually about the inefficient
current_thread_info() mechanism, but I wasn't sure what exactly.

> I haven't measured performance yet, but when using the PDA for "current"
> and "smp_processor_id", I see a 5715 byte reduction in .text segment
> size for my kernel.

This is interesting, since even by doing a non-elegant
current->... --> struct task_struct *tsk = current;
replacement for excessive uses of current, I was able to gain almost 10kB
within a single file already!
I guess it's due to having tried that on an older installation with gcc 3.2,
which probably does less efficient opcode merging of current_thread_info()
requests compared to a current gcc version.
IOW, .text segment reduction could be quite a bit higher for older gcc:s.

> This uses the x86 segmentation stuff in a way similar to NPTL's way of
> implementing Thread-Local Storage.  It relies on the fact that each CPU
> has its own Global Descriptor Table (GDT), which is basically an array
> of base-length pairs (with some extra stuff).  When a segment register
> is loaded with a descriptor (approximately, an index in the GDT), and
> you use that segment register for memory access, the address has the
> base added to it, and the resulting address is used.

Not a problem for more daring user-space apps (i.e. Wine), I hope?

Andreas Mohr

-- 
No programming skills!? Why not help translate many Linux applications! 
https://launchpad.ubuntu.com/rosetta
(or alternatively buy nicely packaged Linux distros/OSS software to help
support Linux developers creating shiny new things for you?)

  parent reply	other threads:[~2006-08-27 17:21 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-08-27  8:44 Jeremy Fitzhardinge
2006-08-27  8:44 ` [PATCH RFC 1/6] Basic definitions for i386-pda Jeremy Fitzhardinge
2006-08-27  8:44 ` [PATCH RFC 2/6] Initialize the per-CPU data area Jeremy Fitzhardinge
2006-08-27  8:44 ` [PATCH RFC 3/6] Use %gs as the PDA base-segment in the kernel Jeremy Fitzhardinge
2006-08-27  9:49   ` Keith Owens
2006-08-27 10:01     ` Jeremy Fitzhardinge
2006-08-27 15:57   ` Andi Kleen
2006-08-27 16:36     ` Jeremy Fitzhardinge
2006-08-27 17:20     ` Jeremy Fitzhardinge
2006-08-27 18:19       ` Andi Kleen
2006-08-27 20:03         ` Jan Engelhardt
2006-08-27 23:38         ` Jeremy Fitzhardinge
2006-08-28  9:51         ` Jan Beulich
2006-08-28 14:54           ` H. J. Lu
2006-08-28 17:24         ` H. Peter Anvin
2006-08-27  8:44 ` [PATCH RFC 4/6] Fix places where using %gs changes the usermode ABI Jeremy Fitzhardinge
2006-08-27 15:59   ` Andi Kleen
2006-08-27 16:37     ` Jeremy Fitzhardinge
2006-08-27  8:44 ` [PATCH RFC 5/6] Implement smp_processor_id() with the PDA Jeremy Fitzhardinge
2006-08-27  8:44 ` [PATCH RFC 6/6] Implement "current" " Jeremy Fitzhardinge
2006-08-27 16:01   ` Andi Kleen
2006-08-27 16:38     ` Jeremy Fitzhardinge
2006-08-27  9:47 ` [PATCH RFC 0/6] Implement per-processor data areas for i386 Arjan van de Ven
2006-08-27 16:46   ` Jeremy Fitzhardinge
2006-08-27 17:44     ` Arjan van de Ven
2006-08-27 18:07       ` Andi Kleen
2006-08-27 18:27         ` Jeremy Fitzhardinge
2006-08-27 16:01 ` Andi Kleen
2006-08-27 16:41   ` Jeremy Fitzhardinge
2006-08-27 17:21 ` Andreas Mohr [this message]
2006-08-27 17:34   ` Jeremy Fitzhardinge
2006-08-27 18:23     ` Andreas Mohr
2006-08-27 18:04   ` Andi Kleen
2006-08-27 18:27     ` Andreas Mohr
2006-08-27 18:35       ` Andi Kleen
2006-08-28  9:06 Chuck Ebbert
2006-08-30  9:00 Chuck Ebbert
2006-08-30  9:17 ` Jeremy Fitzhardinge
2006-08-30 12:33 Chuck Ebbert
2006-08-30 12:54 ` Andi Kleen
2006-08-30 16:39 ` Jeremy Fitzhardinge
2006-08-30 16:48   ` Andi Kleen
2006-08-30 17:13     ` Jeremy Fitzhardinge
2006-08-30 17:32       ` Andi Kleen

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=20060827172155.GA21724@rhlx01.fht-esslingen.de \
    --to=andi@rhlx01.fht-esslingen.de \
    --cc=76306.1226@compuserve.com \
    --cc=ak@suse.de \
    --cc=akpm@osdl.org \
    --cc=jbeulich@novell.com \
    --cc=jeremy@goop.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=zach@vmware.com \
    /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

all inboxes | Powered by JetHome®