This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Patch: assume-compiled fix
>>>>> "Andrew" == Andrew Haley <aph@redhat.com> writes:
Andrew> I assume by "tested" you mean "no regressions".
Yes, though that means little in this case (we don't ever use
-fno-assume-compiled in the test suite; so this only tests that
supers_all_compiled always returns true without that option). Also I
compiled libgcj from scratch, and I tried it on my test case.
Andrew> I suppose we'll need test cases for -fassume-compiled.
Yeah. That probably means writing more expect code :-(. We're going
to want to start running some tests in weird ways -- testing
compatibility of bytecode and java code, compatibility of the
interpreter and the compiler (right now at least method layout can
differ in the Miranda case, ugh). Lately I seem to find serious bugs
everywhere I look :-(
Tom