Re: optimization/7269: Inlining may not eliminate unnecessary tests and branches.

Re: optimization/7269: Inlining may not eliminate unnecessary tests and branches.

[Andrew Pinski]

The following reply was made to PR optimization/7269; it has been noted by GNATS.

From: Andrew Pinski <pinskia@physics.uc.edu>
To: gcc-gnats@gcc.gnu.org, gcc-prs@gcc.gnu.org, anamax@earthlink.net,
   gcc-bugs@gcc.gnu.org, nobody@gcc.gnu.org
Cc:  
Subject: Re: optimization/7269: Inlining may not eliminate unnecessary tests and branches.
Date: Sun, 6 Oct 2002 21:35:24 -0700

 Do you have a simple sample which shows this?
 
 Thanks,
 Andrew Pinski
 
 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit- 
 trail&database=gcc&pr=7269
 

Re: optimization/7269: Inlining may not eliminate unnecessary tests and branches.

[Andy Freeman]

The following reply was made to PR optimization/7269; it has been noted by GNATS.

From: "Andy Freeman" <anamax@earthlink.net>
To: <ebotcazou@gcc.gnu.org>,
	<anamax@earthlink.net>,
	<gcc-bugs@gcc.gnu.org>,
	<gcc-prs@gcc.gnu.org>,
	<nobody@gcc.gnu.org>,
	<gcc-gnats@gcc.gnu.org>
Cc:  
Subject: Re: optimization/7269: Inlining may not eliminate unnecessary tests and branches.
Date: Wed, 19 Feb 2003 06:43:40 -0800

 I'm no longer in a position to produce the code.
 
 The specific example looked like:
 
 struct foo {
   int f(int i) {
     ... // no branches or returns
     if( 0 = i ) {
         return 17;
     }
     ...
   }
 };
 
 int bar(...) {
     foo f;
     if( f.f(0) ) {        // the branch in f was eliminated by constant propagation
         ....                  // but the branch testing f's return value was not.
     }
 }
 ----- Original Message ----- 
 From: <ebotcazou@gcc.gnu.org>
 To: <anamax@earthlink.net>; <gcc-bugs@gcc.gnu.org>; <gcc-prs@gcc.gnu.org>; <nobody@gcc.gnu.org>
 Sent: Wednesday, February 19, 2003 1:26 AM
 Subject: Re: optimization/7269: Inlining may not eliminate unnecessary tests and branches.
 
 
 > Synopsis: Inlining may not eliminate unnecessary tests and branches.
 > 
 > State-Changed-From-To: open->feedback
 > State-Changed-By: ebotcazou
 > State-Changed-When: Wed Feb 19 09:26:14 2003
 > State-Changed-Why:
 >     Sample code requested for 5 months.
 > 
 > http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=7269

Re: optimization/7269: Inlining may not eliminate unnecessary tests and branches.

[ebotcazou]

Synopsis: Inlining may not eliminate unnecessary tests and branches.

State-Changed-From-To: open->feedback
State-Changed-By: ebotcazou
State-Changed-When: Wed Feb 19 09:26:14 2003
State-Changed-Why:
    Sample code requested for 5 months.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=7269

optimization/7269: Inlining may not eliminate unnecessary tests and branches.

[anamax]

>Number:         7269
>Category:       optimization
>Synopsis:       Inlining may not eliminate unnecessary tests and branches.
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          pessimizes-code
>Submitter-Id:   net
>Arrival-Date:   Wed Jul 10 22:36:00 PDT 2002
>Closed-Date:
>Last-Modified:
>Originator:     Andy Freeman
>Release:        gcc 3.1
>Organization:
>Environment:
Redhat 7.3
>Description:
Inlined procedures may have unnnecessary tests and branches.

Using built-in specs.
Configured with: ../gcc-3.1/configure 
Thread model: single
gcc version 3.1
>How-To-Repeat:
The -S -O3 output has a branch to a compare&branch whose behavior is invariant.  The "je .L4" should/can be "je .L1".
>Fix:
Better constant value propagation
>Release-Note:
>Audit-Trail:
>Unformatted: