gcc comptability: silent -fdefer-pop & -fno-defer-pop

[sylvestre.ledru]

FYI, it is one of the most common failure of Unknown argument:
http://clang.debian.net/status.php?version=3.4.2&key=UNKNOWN_ARG

[rnk]

lgtm

This seems like exactly the kind of optimization tweaking options that we ignore for gcc compatibility. If we implemented this, I guess it would control shrink-wrapping optimizations in the backend.

include/clang/Driver/Options.td

 def : Flag<["-"], "fexpensive-optimizations">, Group<clang_ignored_f_Group>;
 def : Flag<["-"], "fno-expensive-optimizations">, Group<clang_ignored_f_Group>;
 def fextdirs_EQ : Joined<["-"], "fextdirs=">, Group<f_Group>;
+def : Flag<["-"], "fdefer-pop">, Group<clang_ignored_f_Group>;
+def : Flag<["-"], "fno-defer-pop">, Group<clang_ignored_f_Group>;
 def : Flag<["-"], "fextended-identifiers">, Group<clang_ignored_f_Group>;
 def : Flag<["-"], "fno-extended-identifiers">, Group<f_Group>, Flags<[Unsupported]>;
 def fhosted : Flag<["-"], "fhosted">, Group<f_Group>;

test/Driver/clang_f_opts.c

 // RUN:     -fivopts -fno-ivopts                                              \
 // RUN:     -fnon-call-exceptions -fno-non-call-exceptions                    \
 // RUN:     -fpermissive -fno-permissive                                      \
+// RUN:     -fdefer-pop -fno-defer-pop                                        \
 // RUN:     -fprefetch-loop-arrays -fno-prefetch-loop-arrays                  \
 // RUN:     -fprofile-correction -fno-profile-correction                      \
 // RUN:     -fprofile-dir=bar                                                 \