This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: Bootstrap error in stage1 libgcc2.a on mips-sgi-irix6.2
- To: aoliva at cygnus dot com
- Subject: Re: Bootstrap error in stage1 libgcc2.a on mips-sgi-irix6.2
- From: Mark Mitchell <mark at codesourcery dot com>
- Date: Tue, 04 Apr 2000 23:16:13 -0700
- Cc: ghazi at caip dot rutgers dot edu, egcs-bugs at egcs dot cygnus dot com, gcc-patches at gcc dot gnu dot org
- Organization: CodeSourcery, LLC
- References: <200004031641.MAA16820@caip.rutgers.edu><orpus5umhx.fsf@zecarneiro.lsd.ic.unicamp.br>
>>>>> "Alexandre" == Alexandre Oliva <aoliva@cygnus.com> writes:
Alexandre> Here's an attempt to fix this bug. Ok to install?
Alexandre> (assuming it bootstraps; it will take a long while for
Alexandre> it to complete :-(
I think a better fix is to not call RETURN_IN_MEMORY when the type to
be returned is `void'. There's no point in asking: `void' things are
never returned in memory. This is only used in aggregate_value_p, so
it's an easy fix. There's no reason machine-descriptions should have
to worry about `void' being passed in here.
If that bootstraps/checks, please consider that change pre-approved.
There should probably also be a change to the TeXinfo documentation
indicating that RETURN_IN_MEMORY is only called with complete types.
--
Mark Mitchell mark@codesourcery.com
CodeSourcery, LLC http://www.codesourcery.com