There were 27 test failures, all due to calls to "abort()"
that the tests were not expecting. It still includes a
bunch of iconv_*.h (and other) files because I'm not fixing
the iconv-h module. I'll be back on after the new year,
so Happy New Year (really! :)
Cheers - Bruce
$ find * -type d -o
Hi,
I'm short on time. I am sure if I hunted, I could find the
way, but the machinery is too convoluted for quick perusal.
Can someone quickly tell me, please? Thank you! - Bruce
../../tests/test-xalloc-die.sh: line 21: 6672 Aborted \
test-xalloc-die${EXEEXT} > out 2> err
FAIL:
On 12/30/10 16:19, Bruce Korb wrote:
> On 12/12/10 03:49, Bruno Haible wrote:
>> Very good point. The dependency comes from
>>
>> openat --> openat-die --> error
>> fdopendir
>
> fdopendir --> openat --> ...
Woops:
--> ??? --> xalloc-die --> error
I'm ifdef-ing the error code for xal
On 12/30/10 17:11, Bruno Haible wrote:
> Bruce Korb wrote:
>> libposix now installs the files listed below.
[...]
>> include/libposix/iconv_open-aix.h
>> include/libposix/iconv_open-hpux.h
>> include/libposix/iconv_open-irix.h
>> include/libposix/iconv_open-osf.h
>> include/libposix/iconv_open-sola
Hi,
I have another test ready.
Things of note:
* tests/test-fprintf-posix3.c and tests/test-dprintf-posix2.c
still have a spurious "free( malloc(0x88));" I haven't gotten
to doing a "dtrace" on the thing.
* openat-die.c will not call error() when GNULIB_LIBPOSIX is
defined. It is defined
Bruce Korb wrote:
> libposix now installs the files listed below.
> include/libposix/arg-nonnull.h
> include/libposix/c++defs.h
> include/libposix/iconv_open-aix.h
> include/libposix/iconv_open-hpux.h
> include/libposix/iconv_open-irix.h
> include/libposix/iconv_open-osf.h
> include/libposix/iconv
On 12/12/10 03:49, Bruno Haible wrote:
> Very good point. The dependency comes from
>
> openat --> openat-die --> error
> fdopendir
fdopendir --> openat --> ...
> Can you work on breaking this dependency?
I have hacked around it thus and am currently testing:
$ git diff|cat
diff --g
On 12/30/2010 01:24 PM, Eric Blake wrote:
> If you like it, I'll
> push to gnulib first and then synchronize coreutils to the latest.
Either order is fine, thanks, but we should try to keep them in sync.
On 12/26/10 00:54, Gary V. Vaughan wrote:
> That's great news, thanks for following up. As soon as I have
> time after the new year I will rerun the tests and let you know
> how I get on. I've seen quite a few patches go by in relation
> to my original reports; are you saying that (excepting the ab
* tests/init.sh (find_exe_basenames_): Exempt [.exe.
(create_exe_shims_): Return 0 when skipping.
---
This lets me get to the point where the few remaining test failures
of coreutils testsuite on cygwin are worth investigating individually,
rather than failing almost everything up front. If you l
10 matches
Mail list logo