[Bug tree-optimization/88842] missing optimization CSE, reassociation

2021-12-23 Thread pinskia at gcc dot gnu.org via Gcc-bugs 
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88842

Andrew Pinski  changed:

   What|Removed |Added

   Severity|normal  |enhancement

[Bug tree-optimization/88842] missing optimization CSE, reassociation

2020-05-14 Thread luoxhu at gcc dot gnu.org 
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88842

luoxhu at gcc dot gnu.org changed:

   What|Removed |Added

 CC||luoxhu at gcc dot gnu.org

--- Comment #3 from luoxhu at gcc dot gnu.org ---
One more case of missed optimization CSE, reassoc:

void foo(unsigned int a, unsigned int b, unsigned int c, unsigned int d, int
*res1, int*res2, int *res3)
{
  *res1 = a + b + c + d;
  *res2 = b + c;
  *res3 = a + d;
}

cat foo.s
.file   "foo.c"
.machine power8
.abiversion 2
.section".text"
.align 2
.p2align 4,,15
.globl foo
.type   foo, @function
foo:
.LFB0:
.cfi_startproc
add 10,5,6
add 10,10,4
add 4,4,5
add 10,10,3
add 3,3,6
stw 10,0(7)
stw 4,0(8)
stw 3,0(9)
blr

[Bug tree-optimization/88842] missing optimization CSE, reassociation

2019-01-15 Thread rguenth at gcc dot gnu.org 
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88842

Richard Biener  changed:

   What|Removed |Added

   Keywords||missed-optimization
 Status|UNCONFIRMED |NEW
   Last reconfirmed||2019-01-15
  Component|other   |tree-optimization
 Ever confirmed|0   |1

--- Comment #2 from Richard Biener  ---
void foo(int a, int b, int c, int d, int e, int *res) {
  res[0] = (e * a) * d;
  res[1] = (e * b) * d;
  res[2] = (e * c) * d;
}

associating (e * d) allows CSE.  Only possible when there's no overflow
(-fwrapv or unsigned types).

The issue is that reassoc is a "local" optimization, associating single
chains rather than looking at multiple chains at once.

And of course reassoc isn't "integrated" with CSE.