[Bug rtl-optimization/38854] New: missed optimization with -m32: empty functions shouldn't contain push/mov/pop

[Bug rtl-optimization/38854] New: missed optimization with -m32: empty functions shouldn't contain push/mov/pop

[edwintorok at gmail dot com]

Testcase:
void noop(char* foo) {}

For 64-bit it does the right thing:
$ gcc -O2 foo.c -S

Output is:
noop:
.LFB2:
        rep
        ret

However for 32-bit it has useless extra instructions:
$ gcc -m32 -O2 foo.c -S

noop:
        pushl   %ebp
        movl    %esp, %ebp
        popl    %ebp
        ret

The push/mov/pop should be removed.

I tested this with gcc-4.2, 4.3, and 4.4(trunk revision 143382), all produce
same code.


-- 
           Summary: missed optimization with -m32: empty functions shouldn't
                    contain push/mov/pop
           Product: gcc
           Version: 4.3.3
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: rtl-optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: edwintorok at gmail dot com
 GCC build triplet: x86_64-linux-gnu
  GCC host triplet: x86_64-linux-gnu
GCC target triplet: x86_64-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38854

[Bug rtl-optimization/38854] missed optimization with -m32: empty functions shouldn't contain push/mov/pop

[rguenth at gcc dot gnu dot org]

------- Comment #1 from rguenth at gcc dot gnu dot org  2009-01-15 09:45 -------
This is the frame-pointer setup code.  You can get the same code as x86_64 with
-momit-leaf-frame-pointer.


-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |INVALID


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38854

[Bug rtl-optimization/38854] missed optimization with -m32: empty functions shouldn't contain push/mov/pop

[edwintorok at gmail dot com]

------- Comment #2 from edwintorok at gmail dot com  2009-01-15 09:50 -------
(In reply to comment #1)
> This is the frame-pointer setup code.  You can get the same code as x86_64 with
> -momit-leaf-frame-pointer.
> 

Ok. But can't the optimizers see that you push ebp, write something to it, and
then pop to it again, and remove that code automatically?
At least the movl is a dead store ...


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38854

[Bug rtl-optimization/38854] missed optimization with -m32: empty functions shouldn't contain push/mov/pop

[edwintorok at gmail dot com]

------- Comment #3 from edwintorok at gmail dot com  2009-01-15 09:51 -------
(In reply to comment #2)
> (In reply to comment #1)
> > This is the frame-pointer setup code.  You can get the same code as x86_64 with
> > -momit-leaf-frame-pointer.
> > 
> 
> Ok. But can't the optimizers see that you push ebp, write something to it, and
> then pop to it again, and remove that code automatically?
> At least the movl is a dead store ...
> 

Or is the prologue/epilogue code special, and not run through such
optimizations?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38854

[Bug rtl-optimization/38854] missed optimization with -m32: empty functions shouldn't contain push/mov/pop

[steven at gcc dot gnu dot org]

------- Comment #4 from steven at gcc dot gnu dot org  2009-01-15 11:15 -------
Re. #3
Yes, the pro/epilogues are special.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38854

[Bug rtl-optimization/38854] missed optimization with -m32: empty functions shouldn't contain push/mov/pop

[rguenth at gcc dot gnu dot org]

------- Comment #5 from rguenth at gcc dot gnu dot org  2009-01-15 14:37 -------
IMHO it's unwanted to remove the frame setup for empty functions without
-momit-leaf-frame-pointer.  Debuggers may need it to do backtraces from
within that function.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38854