From: "H.J. Lu" <hjl.tools@gmail.com>
To: "Aurelien Jarno" <aurelien@aurel32.net>,
linux-kernel@vger.kernel.org, gcc@gcc.gnu.org
Subject: Re: Linux doesn't follow x86/x86-64 ABI wrt direction flag
Date: Wed, 5 Mar 2008 08:56:11 -0800 [thread overview]
Message-ID: <6dc9ffc80803050856l1f69d262he3115fffa0df5fbe@mail.gmail.com> (raw)
In-Reply-To: <20080305153020.GA24631@volta.aurel32.net>
Hi,
According to i386 psABI,
---
The direction flag must be set "forward" direction before entry and
upon exit from
a function.
---
So, asm statement should make sure that the direction flag is cleared before
function returns and kernel should make sure that the direction flag is cleared
when calling a signal handler.
H.J.
On Wed, Mar 5, 2008 at 7:30 AM, Aurelien Jarno <aurelien@aurel32.net> wrote:
> Hi all,
>
> Since version 4.3, gcc changed its behaviour concerning the x86/x86-64
> ABI and the direction flag, that is it now assumes that the direction
> flag is cleared at the entry of a function and it doesn't clear once
> more if needed.
>
> This causes some problems with the Linux kernel which does not clear
> the direction flag when entering a signal handler. The small code below
> (for x86-64) demonstrates that.
>
> If the signal handler is using code that need the direction flag cleared
> (for example bzero() or memset()), the code is incorrectly executed.
>
> I guess this has to be fixed on the kernel side, but also gcc-4.3 could
> revert back to the old behaviour, that is clearing the direction flag
> when entering a routine that touches it until most people are running a
> fixed kernel.
>
> Kind regards,
> Aurelien
>
> [1] http://gcc.gnu.org/gcc-4.3/changes.html
>
>
> #include <stdint.h>
> #include <stdlib.h>
> #include <stdio.h>
> #include <signal.h>
>
> void handler(int signal) {
> uint64_t rflags;
>
> asm volatile("pushfq ; popq %0" : "=g" (rflags));
>
> if (rflags & (1 << 10))
> printf("DF = 1\n");
> else
> printf("DF = 0\n");
> }
>
> int main() {
> signal(SIGUSR1, handler);
>
> while(1)
> {
> asm volatile("std\r\n");
> }
>
> return 0;
> }
>
> --
> .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73
> : :' : Debian developer | Electrical Engineer
> `. `' aurel32@debian.org | aurelien@aurel32.net
> `- people.debian.org/~aurel32 | www.aurel32.net
>
next prev parent reply other threads:[~2008-03-05 16:56 UTC|newest]
Thread overview: 98+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-03-05 15:30 Aurelien Jarno
2008-03-05 16:00 ` H. Peter Anvin
2008-03-05 19:58 ` Joe Buck
2008-03-05 20:23 ` Aurelien Jarno
2008-03-05 20:38 ` Michael Matz
2008-03-05 20:42 ` Joe Buck
2008-03-05 20:49 ` Jan Hubicka
2008-03-05 21:02 ` Michael Matz
2008-03-05 21:20 ` RELEASE BLOCKER: " Joe Buck
2008-03-05 21:32 ` Richard Guenther
2008-03-05 21:34 ` H. Peter Anvin
2008-03-05 21:40 ` Richard Guenther
2008-03-05 22:16 ` David Miller
2008-03-05 22:37 ` Joe Buck
2008-03-05 22:51 ` Michael Matz
2008-03-05 22:58 ` H. Peter Anvin
2008-03-05 23:07 ` Michael Matz
2008-03-05 23:10 ` David Miller
2008-03-05 23:16 ` Joe Buck
2008-03-05 23:12 ` Olivier Galibert
2008-03-05 21:43 ` Joe Buck
2008-03-05 21:44 ` Richard Guenther
[not found] ` <738B72DB-A1D6-43F8-813A-E49688D05771@apple.com>
2008-03-05 21:59 ` Michael Matz
2008-03-05 22:13 ` Adrian Bunk
2008-03-05 22:21 ` David Miller
2008-03-05 23:13 ` Olivier Galibert
2008-03-06 0:36 ` Chris Lattner
2008-03-06 0:47 ` H. Peter Anvin
[not found] ` <578FCA7D-D7A6-44F6-9310-4A97C13CDCBE@apple.com>
2008-03-06 1:12 ` H. Peter Anvin
2008-03-06 9:17 ` Jakub Jelinek
2008-03-06 13:51 ` Olivier Galibert
2008-03-06 14:03 ` Paolo Bonzini
2008-03-06 14:12 ` Olivier Galibert
2008-03-06 14:15 ` Andrew Haley
2008-03-06 17:58 ` Joe Buck
2008-03-06 18:10 ` Olivier Galibert
2008-03-06 18:13 ` Paolo Bonzini
2008-03-06 18:31 ` Jack Lloyd
2008-03-06 18:35 ` Andrew Pinski
2008-03-06 19:44 ` Paolo Bonzini
2008-03-06 19:43 ` Paolo Bonzini
2008-03-06 20:16 ` Jack Lloyd
2008-03-06 21:37 ` Artur Skawina
2008-03-06 15:09 ` Robert Dewar
2008-03-06 15:37 ` NightStrike
2008-03-06 15:43 ` H.J. Lu
2008-03-06 15:50 ` H. Peter Anvin
2008-03-06 16:23 ` Jakub Jelinek
2008-03-06 16:27 ` İsmail Dönmez
2008-03-06 16:58 ` H.J. Lu
2008-03-06 17:06 ` H. Peter Anvin
2008-03-06 17:14 ` H.J. Lu
2008-03-06 17:17 ` H. Peter Anvin
2008-03-06 17:34 ` H.J. Lu
2008-03-06 19:35 ` Robert Dewar
2008-03-06 17:18 ` Robert Dewar
2008-03-06 17:19 ` H. Peter Anvin
2008-03-06 19:25 ` Robert Dewar
2008-03-06 20:37 ` H. Peter Anvin
2008-03-07 8:28 ` Florian Weimer
2008-03-07 8:00 ` Andreas Jaeger
2008-03-06 15:57 ` Robert Dewar
2008-03-06 16:29 ` Paolo Bonzini
2008-03-06 17:18 ` H. Peter Anvin
2008-03-06 16:14 ` Artur Skawina
2008-03-06 0:49 ` Aurelien Jarno
2008-03-05 22:05 ` H. Peter Anvin
2008-03-06 2:11 ` Krzysztof Halasa
2008-03-06 8:44 ` Andi Kleen
2008-03-06 9:01 ` Jakub Jelinek
2008-03-06 15:20 ` H. Peter Anvin
2008-03-05 21:45 ` Aurelien Jarno
2008-03-05 21:43 ` Andrew Pinski
2008-03-05 21:43 ` Michael Matz
2008-03-05 22:12 ` Joe Buck
2008-03-05 22:17 ` David Miller
2008-03-05 23:17 ` Olivier Galibert
2008-03-05 23:21 ` David Daney
2008-03-06 14:06 ` Olivier Galibert
2008-03-08 19:10 ` Alexandre Oliva
2008-03-05 21:07 ` H. Peter Anvin
2008-03-05 20:44 ` H. Peter Anvin
2008-03-05 20:52 ` Aurelien Jarno
2008-03-05 21:23 ` David Miller
2008-03-06 9:53 ` Andrew Haley
2008-03-06 11:45 ` Andi Kleen
2008-03-06 12:06 ` Richard Guenther
2008-03-06 17:34 ` Joe Buck
2008-03-06 20:54 ` Richard Guenther
2008-03-06 20:56 ` H. Peter Anvin
2008-03-06 22:06 ` Andi Kleen
2008-03-07 4:56 ` Chris Lattner
2008-03-07 14:09 ` Michael Matz
2008-03-06 9:45 ` Mikael Pettersson
2008-03-05 16:56 ` H.J. Lu [this message]
2008-03-05 18:14 ` [PATCH] x86: Clear DF before calling signal handler Aurelien Jarno
2008-03-05 18:17 ` H. Peter Anvin
2008-03-06 9:21 ` Ingo Molnar
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=6dc9ffc80803050856l1f69d262he3115fffa0df5fbe@mail.gmail.com \
--to=hjl.tools@gmail.com \
--cc=aurelien@aurel32.net \
--cc=gcc@gcc.gnu.org \
--cc=linux-kernel@vger.kernel.org \
/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
Powered by JetHome