https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122325
Thomas Schwinge <tschwinge at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |tschwinge at gcc dot gnu.org
--- Comment #4 from Thomas Schwinge <tschwinge at gcc dot gnu.org> ---
Right, so after commit r16-188-ge543eaa671d40868575385360d13ef37d87fb2a0
"middle-end/60779 - LTO vs. -fcx-fortran-rules and -fcx-limited-range" had
gotten in, I had made a TODO note the check whether we don't need something
like:
--- gcc/common.opt
+++ gcc/common.opt
@@ -1327,7 +1327,7 @@ Common Ignore
[...]
fcx-method=
-Common Joined RejectNegative Enum(complex_method) Var(flag_complex_method)
Optimization SetByCombined
+Common Joined RejectNegative Enum(complex_method) Var(flag_complex_method)
Optimization SetByCombined Init(1)
[...]
... to restore the default '-fcx-method=fortran', but that the patch doesn't
actually cause any changes in testsuite results, and that it's unclear to me
what the initial value of 'flag_complex_method' actually is used for. I never
got time allocated to continue looking into that.
If I got it right, you've now confirmed/resolved that, thanks. ;-)
