From: Andi Kleen <andi@firstfloor.org>
To: Jakub Jelinek <jakub@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>,
James Cloos <cloos@jhcloos.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Andreas Schwab <schwab@redhat.com>,
Andi Kleen <andi@firstfloor.org>, Michael Matz <matz@suse.de>,
Dave Korn <dave.korn.cygwin@gmail.com>,
Richard Guenther <richard.guenther@gmail.com>,
gcc@gcc.gnu.org, Jim Bos <jim876@xs4all.nl>
Subject: Re: gcc 4.5.1 / as 2.20.51.0.11 miscompiling drivers/char/i8k.c ?
Date: Mon, 15 Nov 2010 10:12:19 +0100 [thread overview]
Message-ID: <20101115091219.GA7269@basil.fritz.box> (raw)
In-Reply-To: <20101115085605.GE2583@sunsite.ms.mff.cuni.cz>
> That said, changing the inline asm to just clobber one less register
> would be completely sufficient to make it work well with all gccs out there,
> just push/pop one of the register around the whole body. I doubt calling
> out SMM BIOS is actually so performance critical that one push and one pop
> would ruin it. Of course x86_64 version can stay as is, there are enough
> registers left...
Yes traditionally clobbering all registers has been dangerous
and it clearly can be done inside the asm too.
Here's a untested patch to do some manual push/pops too. Could someone with
the hardware please test it? (running a 32bit kernel)
-Andi
---
i8k: Clobber less registers
gcc doesn't like inline assembler statements that clobber nearly
all registers. Save a few registers manually on i386 to avoid this
problem.
Fix suggested by Jakub Jelinek
Signed-off-by: Andi Kleen <ak@linux.intel.com>
diff --git a/drivers/char/i8k.c b/drivers/char/i8k.c
index f0863be..a2da38b 100644
--- a/drivers/char/i8k.c
+++ b/drivers/char/i8k.c
@@ -146,7 +146,10 @@ static int i8k_smm(struct smm_regs *regs)
: "a"(regs)
: "%ebx", "%ecx", "%edx", "%esi", "%edi", "memory");
#else
- asm("pushl %%eax\n\t"
+ asm("pushl %%ebx\n\t"
+ "pushl %%ecx\n\t"
+ "pushl %%edx\n\t"
+ "pushl %%eax\n\t"
"movl 0(%%eax),%%edx\n\t"
"push %%edx\n\t"
"movl 4(%%eax),%%ebx\n\t"
@@ -167,10 +170,13 @@ static int i8k_smm(struct smm_regs *regs)
"movl %%edx,0(%%eax)\n\t"
"lahf\n\t"
"shrl $8,%%eax\n\t"
- "andl $1,%%eax\n"
+ "andl $1,%%eax\n\t"
+ "popl %%edx\n\t"
+ "popl %%ecx\n\t"
+ "popl %%ebx\n"
:"=a"(rc), "+m" (*regs)
: "a"(regs)
- : "%ebx", "%ecx", "%edx", "%esi", "%edi", "memory");
+ : "%esi", "%edi", "memory");
#endif
if (rc != 0 || (regs->eax & 0xffff) == 0xffff || regs->eax == eax)
return -EINVAL;
--
ak@linux.intel.com -- Speaking for myself only.
next prev parent reply other threads:[~2010-11-15 9:12 UTC|newest]
Thread overview: 129+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-11-06 11:15 gcc 4.5.1 / as 2.20.51.0.11 miscompiling drivers/char/i8k.c ? Jim
2010-11-07 21:31 ` Andi Kleen
2010-11-07 22:41 ` Andreas Schwab
2010-11-07 23:03 ` Andi Kleen
2010-11-08 10:49 ` Richard Guenther
2010-11-08 11:20 ` Andi Kleen
2010-11-08 11:20 ` Richard Guenther
2010-11-08 11:47 ` Paul Koning
2010-11-08 11:53 ` Jakub Jelinek
2010-11-08 12:20 ` Michael Matz
2010-11-08 18:39 ` Dave Korn
2010-11-09 13:00 ` Michael Matz
2010-11-09 13:48 ` Andi Kleen
2010-11-09 13:57 ` Andreas Schwab
2010-11-09 16:43 ` Jim
2010-11-13 11:13 ` [PATCH] i8k: Tell gcc that *regs gets clobbered Jim Bos
2010-11-15 0:52 ` gcc 4.5.1 / as 2.20.51.0.11 miscompiling drivers/char/i8k.c ? James Cloos
2010-11-15 3:21 ` Linus Torvalds
2010-11-15 8:56 ` Jakub Jelinek
2010-11-15 9:12 ` Andi Kleen [this message]
2010-11-15 9:20 ` Jakub Jelinek
2010-11-15 10:03 ` Jakub Jelinek
2010-11-15 10:54 ` Andi Kleen
2010-11-15 11:16 ` Jakub Jelinek
2010-11-15 11:37 ` Andi Kleen
2010-11-15 17:36 ` Jim Bos
2010-11-15 17:44 ` Jakub Jelinek
2010-11-15 18:17 ` Jim Bos
2010-11-15 18:26 ` Jakub Jelinek
2010-11-15 19:10 ` Jim Bos
2010-11-15 16:04 ` Linus Torvalds
2010-11-15 17:40 ` Jim Bos
2010-11-15 18:08 ` Linus Torvalds
2010-11-15 18:30 ` Jim Bos
2010-11-15 18:37 ` Jim Bos
2010-11-15 18:56 ` Linus Torvalds
2010-11-15 18:58 ` Jakub Jelinek
2010-11-15 19:12 ` Jakub Jelinek
2010-11-15 19:21 ` Linus Torvalds
2010-11-15 19:51 ` Jakub Jelinek
2010-11-15 20:22 ` Jim Bos
2011-06-03 13:05 ` 2.6.39.1 immediately reboots/resets on EFI system Jim Bos
2011-06-03 13:33 ` Matthew Garrett
2011-06-03 14:26 ` Jim Bos
2011-06-03 14:46 ` Matthew Garrett
2011-06-05 10:40 ` Jim Bos
2011-06-05 12:57 ` Maarten Lankhorst
2011-06-06 15:01 ` Maarten Lankhorst
2011-06-06 15:40 ` Jim Bos
2011-06-06 15:44 ` Matthew Garrett
2011-06-06 15:27 ` Maarten Lankhorst
2011-06-06 16:11 ` Jim Bos
2011-06-06 16:43 ` Maarten Lankhorst
2011-06-07 0:19 ` Yinghai Lu
2011-06-07 1:41 ` Matthew Garrett
2011-06-07 2:05 ` Yinghai Lu
2011-06-07 8:25 ` Maarten Lankhorst
2011-06-07 15:14 ` Yinghai Lu
2011-06-07 9:08 ` Maarten Lankhorst
2011-06-07 12:22 ` Maarten Lankhorst
2011-06-07 22:25 ` Yinghai Lu
2011-06-08 16:44 ` Jim Bos
2011-06-08 19:17 ` Yinghai Lu
2011-06-08 19:23 ` Matthew Garrett
2011-06-08 19:27 ` Yinghai Lu
2011-06-08 19:29 ` Matthew Garrett
2011-06-08 19:35 ` Yinghai Lu
2011-06-08 19:38 ` Matthew Garrett
2011-06-08 19:46 ` Yinghai Lu
2011-06-08 19:52 ` Matthew Garrett
2011-06-08 19:48 ` Yinghai Lu
2011-06-08 19:52 ` Matthew Garrett
2011-06-08 20:03 ` Yinghai Lu
2011-06-08 20:09 ` Matthew Garrett
2011-06-08 20:23 ` Yinghai Lu
2011-06-08 20:30 ` Matthew Garrett
2011-06-08 20:36 ` Yinghai Lu
2011-06-08 20:42 ` Matthew Garrett
2011-06-08 20:46 ` Yinghai Lu
2011-06-08 21:06 ` Matthew Garrett
2011-06-08 21:06 ` Linus Torvalds
2011-06-08 21:28 ` Matthew Garrett
2011-06-08 21:31 ` H. Peter Anvin
2011-06-08 21:36 ` Matthew Garrett
2011-06-08 21:31 ` Linus Torvalds
2011-06-08 21:42 ` Matthew Garrett
2011-06-08 21:51 ` H. Peter Anvin
2011-06-08 22:57 ` Linus Torvalds
2011-06-08 23:54 ` Maarten Lankhorst
2011-06-08 21:38 ` Yinghai Lu
2011-06-10 16:47 ` Matthew Garrett
2011-06-10 17:51 ` Maarten Lankhorst
2011-06-10 17:54 ` Matthew Garrett
2011-06-10 22:45 ` Maarten Lankhorst
2011-06-10 22:58 ` Yinghai Lu
2011-06-10 23:03 ` Matthew Garrett
2011-06-10 23:17 ` Greg KH
2011-06-10 23:22 ` Maarten Lankhorst
2011-06-10 23:25 ` H. Peter Anvin
2011-06-10 23:26 ` Yinghai Lu
2011-06-10 23:32 ` H. Peter Anvin
2011-06-10 23:55 ` Yinghai Lu
2011-06-11 0:00 ` H. Peter Anvin
2011-06-11 0:19 ` Yinghai Lu
2011-06-14 18:06 ` H. Peter Anvin
2011-06-11 15:29 ` Matthew Garrett
2011-06-10 23:00 ` Yinghai Lu
2011-06-13 16:47 ` Matthew Garrett
2011-06-13 17:52 ` Maarten Lankhorst
2011-06-13 18:00 ` Matthew Garrett
2011-06-13 18:14 ` Maarten Lankhorst
2011-06-13 18:17 ` Matthew Garrett
2011-06-13 18:23 ` Maarten Lankhorst
2011-06-13 18:33 ` Matthew Garrett
2011-06-13 18:45 ` Maarten Lankhorst
2011-06-14 14:34 ` Maarten Lankhorst
2011-06-14 14:50 ` Maarten Lankhorst
2011-06-14 14:55 ` Matthew Garrett
2010-11-15 22:43 ` gcc 4.5.1 / as 2.20.51.0.11 miscompiling drivers/char/i8k.c ? Andi Kleen
2010-11-15 22:46 ` Jakub Jelinek
2010-11-15 19:53 ` Richard Henderson
2010-11-15 10:24 ` Richard Guenther
2010-11-15 18:45 ` Jeff Law
2010-11-15 19:04 ` Linus Torvalds
2010-11-15 22:07 ` Richard Guenther
2010-11-15 22:58 ` Jeff Law
2010-11-15 23:07 ` Richard Guenther
2010-11-16 4:10 ` Jeff Law
2010-11-15 18:42 ` Jeff Law
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=20101115091219.GA7269@basil.fritz.box \
--to=andi@firstfloor.org \
--cc=cloos@jhcloos.com \
--cc=dave.korn.cygwin@gmail.com \
--cc=gcc@gcc.gnu.org \
--cc=jakub@redhat.com \
--cc=jim876@xs4all.nl \
--cc=linux-kernel@vger.kernel.org \
--cc=matz@suse.de \
--cc=richard.guenther@gmail.com \
--cc=schwab@redhat.com \
--cc=torvalds@linux-foundation.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;
as well as URLs for NNTP newsgroup(s).