Re: Regression: [Re: RFC: change "program exited" message]

[Jan Kratochvil <jan.kratochvil@redhat.com>]

On Wed, 09 Mar 2011 15:15:16 +0100, Tom Tromey wrote:
> I built and regtested this on the compile farm, but we already know that
> may have some issues.

There remains one regression:

-PASS: gdb.ada/catch_ex.exp: continuing to program completion
+FAIL: gdb.ada/catch_ex.exp: continue until exit (the program exited)

due to:

Catchpoint 6, unhandled CONSTRAINT_ERROR at 0x00000000004017c3 in foo () at /home/jkratoch/redhat/gdb-clean/gdb/testsuite/gdb.ada/catch_ex/foo.adb:41
41         raise Constraint_Error;  -- SPOT4
(gdb) PASS: gdb.ada/catch_ex.exp: continuing to unhandled exception
continue
Continuing.

raised CONSTRAINT_ERROR : foo.adb:41 explicit raise
[Inferior 1 (process 14730) exited with code 01]
(gdb) FAIL: gdb.ada/catch_ex.exp: continue until exit (the program exited)
testcase ./gdb.ada/catch_ex.exp completed in 1 seconds 

that the testcase does not return exit code 0.

While the .exp testfile does not have too straightforward way to accept the
non-zero exit code I do not think a testfile should return with exit code
non-zero.  But I failed to modify the .adb file such way, isn't there some
easy way, Joel?


> Some of the things in that log don't seem to be related to my patch.  E.g.,
> the attachstop-mt.exp failure.

Yes, attachstop-mt.exp has flipping results.  As all the code around SIGSTOP
should get improved/reworked/merged-with-Fedora and it is kernel specific and
the SIGSTOP issues are being discussed in upstream kernel these months I delay
it after it gets resolved for upstream kernel.



Thanks,
Jan