Re: c/2308: x86_64 opcode generated for i686

Re: c/2308: x86_64 opcode generated for i686

[Ulrich Drepper]

The following reply was made to PR c/2308; it has been noted by GNATS.

From: Ulrich Drepper <drepper@redhat.com>
To: Neil Booth <neil@daikokuya.demon.co.uk>
Cc: gcc-gnats@gcc.gnu.org
Subject: Re: c/2308: x86_64 opcode generated for i686
Date: 17 Mar 2001 08:37:11 -0800

 Neil Booth <neil@daikokuya.demon.co.uk> writes:
 
 > Things like this should go in the category "target"; I doubt it
 > has anything to do with the C front end.
 
 Then renamed "target" to "backend".  What do I know what you mean by
 "target".
 
 -- 
 ---------------.                          ,-.   1325 Chesapeake Terrace
 Ulrich Drepper  \    ,-------------------'   \  Sunnyvale, CA 94089 USA
 Red Hat          `--' drepper at redhat.com   `------------------------

Re: c/2308: x86_64 opcode generated for i686

[Joseph S. Myers]

The following reply was made to PR c/2308; it has been noted by GNATS.

From: "Joseph S. Myers" <jsm28@cam.ac.uk>
To: Ulrich Drepper <drepper@redhat.com>
Cc: <gcc-gnats@gcc.gnu.org>,  <neil@daikokuya.demon.co.uk>
Subject: Re: c/2308: x86_64 opcode generated for i686
Date: Sat, 17 Mar 2001 16:58:16 +0000 (GMT)

 On 17 Mar 2001, Ulrich Drepper wrote:
 
 >  Neil Booth <neil@daikokuya.demon.co.uk> writes:
 >  
 >  > Things like this should go in the category "target"; I doubt it
 >  > has anything to do with the C front end.
 >  
 >  Then renamed "target" to "backend".  What do I know what you mean by
 >  "target".
 
 The meanings of the categories are documented in both gnats.html and
 gcc.texi (though the latter is duplicate documentation that I think ought
 to go away in favour of a copy of gnats.html in release tarballs).
 
 It has been suggested, and would make sense, that there should be target-*
 categories for at least the more common targets (target-alpha,
 target-i386, ...).
 
 -- 
 Joseph S. Myers
 jsm28@cam.ac.uk
 

Re: c/2308: x86_64 opcode generated for i686

[Neil Booth]

The following reply was made to PR c/2308; it has been noted by GNATS.

From: Neil Booth <neil@daikokuya.demon.co.uk>
To: drepper@redhat.com
Cc: gcc-gnats@gcc.gnu.org
Subject: Re: c/2308: x86_64 opcode generated for i686
Date: Sat, 17 Mar 2001 11:57:08 +0000

 Ulrich Drepper wrote:-
 
 > >Synopsis:       x86_64 opcode generated for i686
 
 Things like this should go in the category "target"; I doubt it
 has anything to do with the C front end.
 
 Neil.

c/2308: x86_64 opcode generated for i686

[drepper]

>Number:         2308
>Category:       c
>Synopsis:       x86_64 opcode generated for i686
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Mar 16 22:36:00 PST 2001
>Closed-Date:
>Last-Modified:
>Originator:     Ulrich Drepper
>Release:        gcc version 3.1 20010316 (experimental)
>Organization:
>Environment:
Linux/x86
>Description:
Compiling the attached file with

  gcc -c -O2 ffsll.i

with today's (2000-3-16) gcc leads to the message

/tmp/ccdIDgSe.s:26: Error: suffix or operands invalid for `bsf'
/tmp/ccdIDgSe.s:43: Error: suffix or operands invalid for `bsf'

What happens is that the registers in the asm statements are substituted inccorectly:

        bsfl %rsi,%ebx

This seems to be the syntax for the extended x86_64 registers but of course is very wrong on i686.
>How-To-Repeat:
Compile attached file with

  gcc -c -O2 ffsll.i
>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted:
----gnatsweb-attachment----
Content-Type: application/octet-stream; name="ffsll.i"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="ffsll.i"
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