This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC 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: [libstdc++] testsuite failures on sparc biarch using -m64: tr1_impl/boost_shared_ptr.h error:


2008/1/21, Jonathan Wakely <jwakely.gcc@gmail.com>:
> On 21/01/2008, Paolo Carlini <pcarlini@suse.de> wrote:
> > Christian Joensson wrote:
> > > Now, is there some funny stuff going on here that I simply miss or is
> > > this what to expect currently?
> > >
> > I would suggest compiling the testcase outside the testsuite and having
> > a look to the pre-processed output. On all the other targets I have
> > available things are fine: <memory> includes
> > <bits/boost_sp_shared_count.h>, which defines _Sp_make_shared_tag in
> > namespace std.
> >
> > I'm assuming you are seeing that when running the tests in 20_util,
> > because the version of shared_ptr in tr1 doesn't use that.
>
> Paolo's right, _Sp_make_shared_tag is only defined in C++0x mode, but
> should always be defined there.
>
> Christian, please let me know which testsuites fail, and send me the
> rest of the diagnostic output (the line you've shown is no use on its
> own, it doesn't tell us which testcase failed or if it was compiled in
> C++0x mode or not.)
>
> My first guess would be that you've somehow got the C++0x and TR1
> versions of boost_sp_shared_count.h mixed up and you're including the
> wrong one.

well, the testsuite results are posted at, e.g.,

http://gcc.gnu.org/ml/gcc-testresults/2008-01/msg00956.html

and the libstdc++.log file is attached, compressed using bzip2 -9



-- 
Cheers,

/ChJ

Attachment: libstdc++.log.bz2
Description: BZip2 compressed data


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