This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: PATCH: Reverse SUPPORTS_INIT_PRIORITY Change to iris6.h
- To: jason at redhat dot com
- Subject: Re: PATCH: Reverse SUPPORTS_INIT_PRIORITY Change to iris6.h
- From: Mark Mitchell <mark at codesourcery dot com>
- Date: Thu, 01 Feb 2001 21:50:53 -0800
- Cc: pedwards at disaster dot jaj dot com, oldham at codesourcery dot com, gcc-patches at gcc dot gnu dot org
- Organization: CodeSourcery, LLC
- References: <u9u26eccg1.fsf@casey.cambridge.redhat.com>
>>>>> "Jason" == Jason Merrill <jason@redhat.com> writes:
Jason> It should. If the MIPS port uses it to mean something
Jason> else, it should be fixed.
It's not the MIPS port.
/* -fgnu-linker specifies use of the GNU linker for initializations.
(Or, more generally, a linker that handles initializations.)
-fno-gnu-linker says that collect2 will be used. */
#ifdef USE_COLLECT2
int flag_gnu_linker = 0;
#else
int flag_gnu_linker = 1;
#endif
This is really the linker_supports_init_fini flag.
Even the option description:
{"gnu-linker", &flag_gnu_linker, 1,
"Output GNU ld formatted global initialisers"},
Jeffrey, I'm not sure whether you should try to fix this -- doing so
means extracting configury information, and I don't think there's
mileage in that right at the minute.
My thinking is that we should just change the flag to be named
linker_supports_init_fini, and lets be done with it.
That is consistently how the flag is used.
Jason, if you think we need another flag for GNU-linkerness, in
general, we should add that separately.
It does makes sense for this flag to be true on IRIX because IRIX ld
does handle .init/.fini stuff.
--
Mark Mitchell mark@codesourcery.com
CodeSourcery, LLC http://www.codesourcery.com