https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109960
--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
I happened to notice this because I am working on a match patch that transform
`a ? 1 : b` into `a | b`.
In the case of stmt_can_terminate_bb_p, I noticed we had:
<bb 31> [local count: 330920071]:
_48 = MEM[(const struct gasm *)t_22(D)].D.129035.D.128905.D.128890.subcode;
_49 = _48 & 2;
if (_49 != 0)
goto <bb 33>; [34.00%]
else
goto <bb 32>; [66.00%]
<bb 32> [local count: 218407246]:
_50 = (bool) _48;
<bb 33> [local count: 940291388]:
# _13 = PHI <0(14), _50(32), _12(29), 0(11), 0(30), 1(2), 1(31), 0(25)>
And the patch to match would do:
<bb 31> [local count: 330920071]:
_48 = MEM[(const struct gasm *)t_22(D)].D.129035.D.128905.D.128890.subcode;
_49 = _48 & 2;
_50 = (bool) _48;
_127 = _49 != 0;
_44 = _50 | _127;
<bb 32> [local count: 940291388]:
# _13 = PHI <0(14), 0(25), _12(29), 0(11), 0(30), 1(2), _44(31)>
Which is definitely better than before but I was like isn't that the same as:
_49 = _48 & 3;
_44 = _49 != 0;