After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 347944 - make check fails 2 test on Solaris 9
make check fails 2 test on Solaris 9
Status: RESOLVED FIXED
Product: glib
Classification: Platform
Component: general
2.12.x
Other All
: Normal normal
: ---
Assigned To: gtkdev
gtkdev
Depends on:
Blocks:
 
 
Reported: 2006-07-18 17:35 UTC by John Taisto
Modified: 2006-10-08 20:00 UTC
See Also:
GNOME target: ---
GNOME version: ---


Attachments
patch to fix the run-markup-tests.sh script (630 bytes, patch)
2006-10-03 02:21 UTC, Dan McMahill
none Details | Review

Description John Taisto 2006-07-18 17:35:43 UTC
Please describe the problem:
The first problem is with the "run-markup-tests.sh" for loop.  If I simply change the shebang line to use bash instead of bourne the script passes otherwise:

PASS: uri-test
./run-markup-tests.sh: syntax error at line 29: `(' unexpected
FAIL: run-markup-tests.sh

The second problem is with the "run-collate-tests.sh" test, not sure what is wrong, but it outputs the following:
2d1
< GTK+
8a8
> GTK+
Test failed: unexpected error when using g_utf8_collate() on ./collate/collate-1
.in
FAIL: run-collate-tests.sh
PASS: run-bookmark-test.sh




Steps to reproduce:
1. Use libiconv version 1.10 
2. Use gettext version 0.14.6
3. Use gcc 3.4.6
4. Used this configure:
./configure --prefix=my prefix \
            --libdir=my prefix/lib/sparcv9 \
            --enable-static --enable--shared \
            --with-libiconv=gnu \
            --with-threads=posix
            



Actual results:
Everything seems to build fine if you fix the first problem and remove the second problem script from the Makefiles.  Make install seems to work and all executables and shared libraries seem to work fine--so far.

Expected results:
make check to produce no errors.

Does this happen every time?
Yes.

Other information:
This is a 64-bit build ( all linked libraries are of course 64-bit too ).
Used environmental variables CC CXX CFLAGS CXXFLAGS LDFLAGS CPPFLAGS.
I also had to make a change to the configure script to properly find my xgettext binary.
*** configure   Tue Jul 18 11:04:25 2006
--- configure.orig      Tue Jul 18 11:02:15 2006
***************
*** 7403,7409 ****
  
          # Extract the first word of "xgettext", so it can be a program name with args.
  set dummy xgettext; ac_word=$2
- ac_word=g${2}
  echo "$as_me:$LINENO: checking for $ac_word" >&5
  echo $ECHO_N "checking for $ac_word... $ECHO_C" >&6
  if test "${ac_cv_path_XGETTEXT+set}" = set; then
--- 7403,7408 ----

I tried setting the environment variable XGETTEXT to see if it would override it or not.  I tried setting an alias to do this too, but to no avail.  This problem stems from me using the transform when building my gettext package.  I had used this option in my gettext build:

--program-transform-name='s/^gettext$/g&/;s/^msgfmt$/g&/;s/^xgettext$/g&/'
Comment 1 Daniel Macks 2006-07-19 02:59:39 UTC
The collate-tests test failure is likely a dup of Bug #333977 
Comment 2 Dan McMahill 2006-10-03 02:21:54 UTC
Created attachment 73916 [details] [review]
patch to fix the run-markup-tests.sh script

This patch makes the run-markup-tests.sh script a bit more portable and fixes
one of the two failures reported here.

-Dan
Comment 3 Matthias Clasen 2006-10-08 20:00:37 UTC
2006-10-08  Matthias Clasen  <mclasen@redhat.com>

	* tests/run-markup-tests.sh: Small portability fix.  (#347944,
	Dan McMahill)

Closing this bug, according to comment #1