[PATCH] D69224: [InstCombine] Fold uadd.sat(a, b) == 0 and usub.sat(a, b) == 0

[Roman Lebedev via Phabricator via llvm-commits]

lebedev.ri accepted this revision.
lebedev.ri added a comment.
This revision is now accepted and ready to land.

  ----------------------------------------
  Name: uadd.sat(a, b) == 0  ->  (a | b) == 0
    %o0 = uadd_sat i8 %x, %y
    %r = icmp eq i8 %o0, 0
    ret i1 %r
  =>
    %n0 = or i8 %x, %y
    %r = icmp eq i8 %n0, 0
    ret i1 %r
    %o0 = uadd_sat i8 %x, %y
  
  Done: 1
  Optimization is correct!
  
  ----------------------------------------
  Name: uadd.sat(a, b) != 0  ->  (a | b) != 0
    %o0 = uadd_sat i8 %x, %y
    %r = icmp ne i8 %o0, 0
    ret i1 %r
  =>
    %n0 = or i8 %x, %y
    %r = icmp ne i8 %n0, 0
    ret i1 %r
    %o0 = uadd_sat i8 %x, %y
  
  Done: 1
  Optimization is correct!
  
  ----------------------------------------
  Name: usub.sat(a, b) == 0  ->  a <= b
    %o0 = usub_sat i8 %x, %y
    %r = icmp eq i8 %o0, 0
    ret i1 %r
  =>
    %r = icmp ule i8 %x, %y
    ret i1 %r
    %o0 = usub_sat i8 %x, %y
  
  Done: 1
  Optimization is correct!
  
  ----------------------------------------
  Name: usub.sat(a, b) != 0  ->  a > b
    %o0 = usub_sat i8 %x, %y
    %r = icmp ne i8 %o0, 0
    ret i1 %r
  =>
    %r = icmp ugt i8 %x, %y
    ret i1 %r
    %o0 = usub_sat i8 %x, %y
  
  Done: 1
  Optimization is correct!

LG


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D69224/new/

https://reviews.llvm.org/D69224