Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code

Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code

[Franz Sirl]

The following reply was made to PR optimization/2100; it has been noted by GNATS.

From: Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
To: Richard Henderson <rth@redhat.com>
Cc: rth@gcc.gnu.org,dllorens@inf.uji.es,gcc-bugs@gcc.gnu.org,
 nobody@gcc.gnu.org,gcc-gnats@gcc.gnu.org
Subject: Re: optimization/2100: -fstrength-reduce (is active with -O2)
  produces wrong code
Date: Wed, 03 Apr 2002 21:03:23 +0200

 At 23:43 02.04.2002, Richard Henderson wrote:
 >On Tue, Apr 02, 2002 at 11:32:12PM +0200, Franz Sirl wrote:
 > > Bad news, after reverting Jan's cost tweek the bug reappeared on
 > > i686-linux-gnu :-(. Should I add the testcase to the testsuite in mainline
 > > and branch even though it won't currently fail?
 >
 >Please.
 
 In case you haven't noticed yet, the testcase fails on x86-linux (mainline 
 and branch) if compiled with -Os.
 
 Franz.
 

Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code

[Richard Henderson]

The following reply was made to PR optimization/2100; it has been noted by GNATS.

From: Richard Henderson <rth@redhat.com>
To: Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
Cc: rth@gcc.gnu.org, dllorens@inf.uji.es, gcc-bugs@gcc.gnu.org,
   nobody@gcc.gnu.org, gcc-gnats@gcc.gnu.org
Subject: Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code
Date: Tue, 2 Apr 2002 13:43:42 -0800

 On Tue, Apr 02, 2002 at 11:32:12PM +0200, Franz Sirl wrote:
 > Bad news, after reverting Jan's cost tweek the bug reappeared on 
 > i686-linux-gnu :-(. Should I add the testcase to the testsuite in mainline 
 > and branch even though it won't currently fail?
 
 Please.
 
 
 r~

Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code

[Franz Sirl]

The following reply was made to PR optimization/2100; it has been noted by GNATS.

From: Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
To: Richard Henderson <rth@redhat.com>
Cc: rth@gcc.gnu.org,
 dllorens@inf.uji.es,
 gcc-bugs@gcc.gnu.org,
 nobody@gcc.gnu.org,
 gcc-gnats@gcc.gnu.org
Subject: Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code
Date: Tue, 2 Apr 2002 23:32:12 +0200

 On Tuesday 02 April 2002 13:24, Richard Henderson wrote:
 > On Tue, Apr 02, 2002 at 01:07:58PM +0200, Franz Sirl wrote:
 > > Is this really fixed now? See your comment in
 > > <http://gcc.gnu.org/ml/gcc-patches/2001-11/msg02150.html>, that's the
 > > reason I didn't close this one yet, though I should have added something
 > > to the PR :-(.
 >
 > Hum.  Wish I'd seen that earlier.
 >
 > It might be fixed anyway -- there have been several fixes in
 > this area over the last couple of months.  It'd be nice to
 > back out Jan's cost tweek and see what happens under the hood.
 >
 > Do you have time to do that?
 
 Bad news, after reverting Jan's cost tweek the bug reappeared on 
 i686-linux-gnu :-(. Should I add the testcase to the testsuite in mainline 
 and branch even though it won't currently fail?
 
 Franz.
 

Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code

[Franz Sirl]

The following reply was made to PR optimization/2100; it has been noted by GNATS.

From: Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
To: Richard Henderson <rth@redhat.com>
Cc: rth@gcc.gnu.org,dllorens@inf.uji.es,gcc-bugs@gcc.gnu.org,
 nobody@gcc.gnu.org,gcc-gnats@gcc.gnu.org
Subject: Re: optimization/2100: -fstrength-reduce (is active with -O2)
  produces wrong code
Date: Tue, 02 Apr 2002 13:49:42 +0200

 At 13:24 02.04.2002, Richard Henderson wrote:
 >On Tue, Apr 02, 2002 at 01:07:58PM +0200, Franz Sirl wrote:
 > > Is this really fixed now? See your comment in
 > > <http://gcc.gnu.org/ml/gcc-patches/2001-11/msg02150.html>, that's the
 > > reason I didn't close this one yet, though I should have added 
 > something to
 > > the PR :-(.
 >
 >Hum.  Wish I'd seen that earlier.
 >
 >It might be fixed anyway -- there have been several fixes in
 >this area over the last couple of months.  It'd be nice to
 >back out Jan's cost tweek and see what happens under the hood.
 >
 >Do you have time to do that?
 
 Yeah, I'll test it later today. Maybe at least the testcase should be 
 added, so we'll notice when it reappears.
 
 BTW, if it reappears with the cost tweek backed out, do you have a (branch) 
 preference on any of the suggested possible fixes in the PR?
 
 Franz.
 

Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code

[Richard Henderson]

The following reply was made to PR optimization/2100; it has been noted by GNATS.

From: Richard Henderson <rth@redhat.com>
To: Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
Cc: rth@gcc.gnu.org, dllorens@inf.uji.es, gcc-bugs@gcc.gnu.org,
   gcc-prs@gcc.gnu.org, nobody@gcc.gnu.org, gcc-gnats@gcc.gnu.org
Subject: Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code
Date: Tue, 2 Apr 2002 03:24:59 -0800

 On Tue, Apr 02, 2002 at 01:07:58PM +0200, Franz Sirl wrote:
 > Is this really fixed now? See your comment in 
 > <http://gcc.gnu.org/ml/gcc-patches/2001-11/msg02150.html>, that's the 
 > reason I didn't close this one yet, though I should have added something to 
 > the PR :-(.
 
 Hum.  Wish I'd seen that earlier.
 
 It might be fixed anyway -- there have been several fixes in
 this area over the last couple of months.  It'd be nice to 
 back out Jan's cost tweek and see what happens under the hood.
 
 Do you have time to do that?
 
 
 r~

Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code

[Franz Sirl]

The following reply was made to PR optimization/2100; it has been noted by GNATS.

From: Franz Sirl <Franz.Sirl-kernel@lauterbach.com>
To: rth@gcc.gnu.org,dllorens@inf.uji.es,gcc-bugs@gcc.gnu.org,
 gcc-prs@gcc.gnu.org,nobody@gcc.gnu.org,gcc-gnats@gcc.gnu.org
Cc: dllorens@inf.uji.es,gcc-bugs@gcc.gnu.org,gcc-prs@gcc.gnu.org,
 nobody@gcc.gnu.org
Subject: Re: optimization/2100: -fstrength-reduce (is active with -O2)
  produces wrong code
Date: Tue, 02 Apr 2002 13:07:58 +0200

 At 08:36 02.04.2002, rth@gcc.gnu.org wrote:
 >Synopsis: -fstrength-reduce (is active with -O2) produces wrong code
 >
 >State-Changed-From-To: open->closed
 >State-Changed-By: rth
 >State-Changed-When: Mon Apr  1 22:36:41 2002
 >State-Changed-Why:
 >     Fixed for gcc 3.1.
 >
 >http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=2100
 
 Is this really fixed now? See your comment in 
 <http://gcc.gnu.org/ml/gcc-patches/2001-11/msg02150.html>, that's the 
 reason I didn't close this one yet, though I should have added something to 
 the PR :-(.
 
 Would it still fail if we can come up with a more complex testcase?
 
 Franz.
 

Re: optimization/2100: -fstrength-reduce (is active with -O2) produces wrong code

[rth]

Synopsis: -fstrength-reduce (is active with -O2) produces wrong code

State-Changed-From-To: open->closed
State-Changed-By: rth
State-Changed-When: Mon Apr  1 22:36:41 2002
State-Changed-Why:
    Fixed for gcc 3.1.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=2100