This is the mail archive of the egcs-testresults@egcs.cygnus.com mailing list for the EGCS project. See the EGCS home page for more information.


[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index] [Subject Index] [Author Index] [Thread Index]

Re: Results for egcs-2.93.12 19990315 (gcc2 ss-980929 experimental) testsuite on i686-pc-linux-gnu



>Craig,
>    I haven't inspected my mail-report.log files by hand for awhile now.
>I believe the answer to your questions:
>
>craig@jcb-sc.com said:
>> Are you *sure* you're using a fully-up-to-date checked-out copy of the
>> source tree??  You might want to try checking out a fresh copy next
>> time, right after updating whatever local tree you're using, then
>> compare them and see how much they differ.
>
>> If you are sure you're using a consistent snapshot of the source
>> repository, could you try investigating why the tests below are
>> failing?  They're very short, and should be fairly self-explanatory
>> (unless you find Fortran extremely hard to read, which I could
>> understand).  It's possible my patch isn't a complete fix for the
>> problem on certain combinations of OS/libc/x86 targets, I guess. 
>
>are really quite simple, and they are contained at the end of the report:
>I'm building the snapshots with the configure option:
>'--enable-languages=c,c++'.  I'm surprised this doesn't omit the g77 and objc 
>tests entirely, but 'make check' seems to run the tests against the older 
>installed version of g77.

Indeed, that would explain the results.

>I'm firing up a build of the latest CVS sources without 
>--enable-languages=c,c++ and I'll send the results when they're done.

If I'm right that I've already seen the results, they do now seem
normal.

>Sorry for the confusion.

No problem -- I figured it was better to ask than to assume it was SEP
(Somebody Else's Problem).

And, you might be right that testing should be done of only enabled
languages.  But I don't know much about the test infrastructure -- there
might be good reasons for it to test anything it finds sitting around.

In the meantime, if you configure with --prefix pointing to a fresh
directory each time, that might avoid the problem.  I don't know.
(I usually configure a snapshot dated YYYYMMDD with
--prefix /develop/egcs/YYYYMMDD, it makes my life a lot easier.
But I don't enable/disable specific languages.)

        tq vm, (burley)