[Bug target/38109] New: [avr] Wrong code with gcc.c-torture/execute/20081112-1.c

[Bug target/38109] New: [avr] Wrong code with gcc.c-torture/execute/20081112-1.c

[eric dot weddington at atmel dot com]

Using version 4.4.0 20081112 (experimental) [trunk revision 141785]
compiling gcc.c-torture/execute/20081112-1.c (new regression) test fails for:

atmega128-sim: gcc.c-torture/execute/20081112-1.c execution,  -O2 
atmega128-sim: gcc.c-torture/execute/20081112-1.c execution,  -O3
-fomit-frame-pointer 
atmega128-sim: gcc.c-torture/execute/20081112-1.c execution,  -O3 -g 
atmega128-sim: gcc.c-torture/execute/20081112-1.c execution,  -Os 

Wrong code is generated for -O[23s], but is generated correctly for -O[01].

Attachments to follow.


-- 
           Summary: [avr] Wrong code with gcc.c-torture/execute/20081112-1.c
           Product: gcc
           Version: 4.4.0
            Status: UNCONFIRMED
          Keywords: wrong-code
          Severity: normal
          Priority: P3
         Component: target
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: eric dot weddington at atmel dot com
GCC target triplet: avr-*-*


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38109

[Bug target/38109] [avr] Wrong code with gcc.c-torture/execute/20081112-1.c

[eric dot weddington at atmel dot com]

------- Comment #1 from eric dot weddington at atmel dot com  2008-11-14 00:38 -------
Created an attachment (id=16667)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16667&action=view)
Generated assembler listing showing good code at -O1.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38109

[Bug target/38109] [avr] Wrong code with gcc.c-torture/execute/20081112-1.c

[eric dot weddington at atmel dot com]

------- Comment #2 from eric dot weddington at atmel dot com  2008-11-14 00:41 -------
Created an attachment (id=16668)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16668&action=view)
Generated assembler listing showing bad code at -O2.

The problem in 20081112-1-bad.s is in the function foo. The compare and branch
have been optimized out, but it is wrongly calling abort, when it should just
ret. The wrong block is being generated.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38109

[Bug target/38109] [avr] Wrong code with gcc.c-torture/execute/20081112-1.c

[eric dot weddington at atmel dot com]

------- Comment #3 from eric dot weddington at atmel dot com  2008-11-15 19:30 -------
Seems to be fixed now on HEAD. Closing bug.


-- 

eric dot weddington at atmel dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |4.4.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38109