[lsb-bugs] [Bug 2641] several xts5 Xopen failures on newer distributions

[bugzilla-daemon at linux-foundation.org]

http://bugs.linuxbase.org/show_bug.cgi?id=2641


Stew Benedict <stewb at linux-foundation.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED




--- Comment #1 from Stew Benedict <stewb at linux-foundation.org>  2009-05-13 14:17:31 ---
Looking at the first of these, kysym/Test7, the failure is:

520|709 7 00019440 1 3|When the header file <X11/keysym.h> is included, then
the
520|709 7 00019440 1 4|symbols in the table are defined to have the hexadecimal
520|709 7 00019440 1 5|values given in the table.
520|709 7 00019440 1 6|METH: Include header file <X11/keysym.h>
520|709 7 00019440 1 7|METH: For each KeySym in table:
520|709 7 00019440 1 8|METH:    Verify that the symbol is defined using #ifdef.
520|709 7 00019440 1 9|METH:    Verify that the symbol has the expected value
using #if.
520|709 7 00019440 1 10|REPORT: Keysym "XK_VoidSymbol" is not defined.
220|709 7 1 05:40:45|FAIL
410|709 7 1 05:40:45|IC End

Looking at the code, I'm a little confused in that it looks like it's dropping
into a routine seemingly defined at compile time (reporterr). And in fact, the
snapshot build is also failing on SLES10 now too. Not sure we're really testing
anything here other than our headers. Odd thing is the build doesn't complain
about not finding keysym.h, but some values seem to be getting dropped (the
first 6 subtests pass).

#include        <X11/keysym.h>

kysym0()
{
int     pass = 0, fail = 0;
#ifdef XK_VoidSymbol
        if(test("XK_VoidSymbol", XK_VoidSymbol, 0xFFFFFF) == 0)
                FAIL;
        else
                CHECK;
#else
        reporterr("XK_VoidSymbol");
        FAIL;
#endif

Released 4.0 tests do not fail these tests, but rebuilding the pending 4.0
update with the new toolchain, I do seem to inherit the issue from devel.
Building the refresh package with a 4.0 toolchain the tests are also ok, so I
guess I've got some more xts5 mangling to do to accomodate headers.


-- 
Configure bugmail: http://bugs.linuxbase.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.