Re: optimization/3041: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

Re: optimization/3041: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

[mmitchel]

Synopsis: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

State-Changed-From-To: analyzed->feedback
State-Changed-By: mmitchel
State-Changed-When: Thu Jun  7 18:35:42 2001
State-Changed-Why:
    Probably I was unclear.
    
    I would like to know which particular loop is slow.  Also,
    what command-line flags were used to compile and on what
    target triplet?  I am going to try to understand the bug,
    but the less wading about in the source I have to do, the
    more likely I will be to have time to do it.  Thanks.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=3041&database=gcc

Re: optimization/3041: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

[mmitchel]

Synopsis: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

State-Changed-From-To: closed->analyzed
State-Changed-By: mmitchel
State-Changed-When: Tue Jun 12 20:24:14 2001
State-Changed-Why:
    The patch referenced below turned out to be unsafe.
    
    See this thread:
    
      http://gcc.gnu.org/ml/gcc/2001-06/msg00623.html
    
    for a discussion of the issues.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=3041&database=gcc

Re: optimization/3041: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

[mmitchel]

Synopsis: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

State-Changed-From-To: analyzed->closed
State-Changed-By: mmitchel
State-Changed-When: Sat Jun  9 16:25:32 2001
State-Changed-Why:
    Fixed.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=3041&database=gcc

Re: optimization/3041: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

[mmitchel]

Synopsis: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

Responsible-Changed-From-To: unassigned->mmitchel
Responsible-Changed-By: mmitchel
Responsible-Changed-When: Fri Jun  8 10:40:46 2001
Responsible-Changed-Why:
    Will investigate.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=3041&database=gcc

Re: optimization/3041: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

[dje]

Synopsis: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

State-Changed-From-To: feedback->analyzed
State-Changed-By: dje
State-Changed-When: Thu Jun  7 19:22:07 2001
State-Changed-Why:
    Feedback provided directly via email

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=3041&database=gcc

Re: optimization/3041: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

[dje]

Synopsis: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

State-Changed-From-To: feedback->analyzed
State-Changed-By: dje
State-Changed-When: Thu Jun  7 15:55:29 2001
State-Changed-Why:
    preprocessed source code attached

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=3041&database=gcc

optimization/3041: Stepanov Abstraction Penalty regression for powerpc-ibm-aix

[dje]

>Number:         3041
>Category:       optimization
>Synopsis:       Stepanov Abstraction Penalty regression for powerpc-ibm-aix
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    unassigned
>State:          open
>Class:          pessimizes-code
>Submitter-Id:   net
>Arrival-Date:   Sun Jun 03 14:16:00 PDT 2001
>Closed-Date:
>Last-Modified:
>Originator:     dje@watson.ibm.com
>Release:        gcc version 3.0 20010601 (prerelease)
>Organization:
>Environment:
AIX 4.3.3.0
>Description:
stepanov_v1p2 regression from gcc-2.95

gcc-2.95.3:
Total absolute time: 5.24 sec
Abstraction Penalty: 0.98

gcc-3.0 (prerelease):
Total absolute time: 13.44 sec
Abstraction Penalty: 2.34
>How-To-Repeat:

>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted: