This is the mail archive of the libstdc++@gcc.gnu.org mailing list for the libstdc++ project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [3.4 PATCH] Fix IRIX bootstrap failure in libstdc++


Roger Sayle wrote:

mips-sgi-irix6.5 currently fails to bootstrap on both mainline and
the gcc 3.4 branch.  The patch below restores bootstrap on the
gcc-3_4-branch, and should resolve one of the multiple breakages
on mainline (for example, Richard Sandiford's fixproto/fix-headers
patch http://gcc.gnu.org/ml/gcc-patches/2004-09/msg03008.html is
also required).

The failures addressed below are the unresolved references to
strtof and strtold whilst linking libstdc++.  The problem is that
config/locale/generic/c_locale.cc currently guards its uses of
strtof and strtold with #ifdef _GLIBCXX_USE_C99.  Unfortunately on
MIPS/IRIX, _GLIBCXX_USE_C99 is defined, even though the corresponding
_GLIBCXX_HAVE_STRTOF and _GLIBC_HAVE_STRTOLD aren't.  The proposed
fix is to use the finer granularity configure HAVE_foo tests instead
of the more generic _GLIBCXX_USE_C99.


The following patch has been tested on i686-pc-linux-gnu with a full "make bootstrap", all default languages, and regression tested with a top-level "make -k check" with no new failures. As mentioned above, this patch also restores bootstrap of the gcc-3_4-branch for all default languages on mips-sgi-irix6.5.

Ok for mainline and the gcc-3_4-branch? Given that mips-sgi-irix6.5
is a primary evaluation platform in gcc-3.4/criteria.html, I think
this needs to be fixed for gcc 3.4.3.


Yes, this is OK for 3.4. For the mainline, one of the V3 people should sign off.

Thanks,

--
Mark Mitchell
CodeSourcery, LLC
(916) 791-8304
mark@codesourcery.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]