IRIX failures [was: next m4 snapshot]

[Eric Blake]

On 08/31/2010 02:27 PM, Tom G. Christensen wrote:
> > > Here's the latest snapshot tarball, if nothing goes terribly wrong, then
> > > I hope to get 1.4.15 out the door this week.
> >
> > http://people.redhat.com/eblake/m4/m4-1.4.14.9-edf2d.tar.gz
>
> There are problems on IRIX 6.5(.30).

Thanks for testing.  Unfortunately, your email missed me by about 2 
hours, so the actual released m4 1.4.15 likely has the same bugs (I 
would have tried to fix them before the release, if things had happened 
in the opposite order).

> It fails to build with gcc 4.3.1 (and 4.5.0):
> gcc -std=gnu99  -I.     -g -O2 -MT printf-args.o -MD -MP -MF 
> .deps/printf-args.Tpo -c -o
>   printf-args.o printf-args.c
> In file included from /usr/include/wchar.h:9,
>                   from ./wchar.h:71,
>                   from printf-args.h:42,
>                   from printf-args.c:30:
> ./wctype.h:585: error: static declaration of 'iswblank' follows non-static 
> declaration
> gmake[3]: *** [printf-args.o] Error 1

Definitely something that can be fixed in gnulib.  Can you help figure 
out where the earlier iswblank declaration was coming from?  Also, that 
particular declaration at line 585 of wctype.h (line 282 of wctype.in.h) 
is guarded by:

#if ! @HAVE_ISWCNTRL@ || @REPLACE_ISWCNTRL@
#elif ! @HAVE_ISWBLANK@

So knowing the value of all the HAVE_IS* and REPLACE_IS* from config.log 
would be useful to help diagnose what is going wrong here.

Once you can get past the compile error, this particular issue has no 
bearing on the final m4 executable, since m4 isn't using iswblank.

> Building with MIPSpro works but there are testsuite failures:
> $ ./test-strtod
> test-strtod.c:377: assertion failed
> Abort (core dumped)

Phooey - a regression.  This is already a known bug ("1E 2" is 
mistakenly treated as 100.0 instead of 1.0 on HP-UX 11.11, IRIX 6.5, and 
OSF/1 4.0, per the comments on that line); so our rewrite to call the 
native strtod under the hood of rpl_strtod is not working around this 
particular bug.

This bug will only affect the 'format' builtin of m4, and is enough of a 
corner case that you probably don't have to worry about it much.

> $ ./test-update-copyright.sh
> --- -   2010-08-31 22:20:22.990686000 +0200
> +++ update-copyright.test-ex-stderr     2010-08-31 22:20:22.981772600 +0200
> @@ -1,2 +1 @@
> -update-copyright.test-ex.4: warning: FSF copyright statement not found
> -update-copyright.test-ex.5: warning: FSF copyright statement not found
> +./test-update-copyright.sh[95]: update-copyright:  not found

Regex gone wrong somewhere?  Maybe due to an older perl bug?  What 
version of perl are you using?

> Adding build-aux to the PATH and using nekoware perl this test passes.

Yep, more proof that this is a perl issue.  Fortunately, this test is 
for a maintainer program, and failure here has no bearing on m4 itself.

> $ ./test-vc-list-files-cvs.sh
> non-numeric argument
> init.sh: failed test: Cannot create directory 
> "/usr/people/tgc/tmp/m4-1.4.14.9-edf2d/tests/gt-init.sh.": File exists
> Cannot create directory "vc-cvs": File exists
> Cannot create directory "vc-cvs": File exists

Can you rerun this test as 'sh -vx ./test-vc-list-files-cvs.sh', so we 
can see which lines are failing?  Again, this test is for a maintainer 
program, and failure here has no bearing on m4 itself.

--
Eric Blake   address@hidden    +1-801-349-2682
Libvirt virtualization library http://libvirt.org