This is the mail archive of the
gdb@sources.redhat.com
mailing list for the GDB project.
Some notes on hpux regressions
- From: Randolph Chung <randolph at tausq dot org>
- To: gdb-testers at sources dot redhat dot com, gdb at sources dot redhat dot com
- Date: Sat, 10 Jul 2004 23:46:08 -0700
- Subject: Some notes on hpux regressions
- Reply-to: Randolph Chung <randolph at tausq dot org>
Just a quick look at some of the regressions....
working through Michael C's list at:
http://sources.redhat.com/ml/gdb-testers/2004-q3/msg00012.html
gdb.base/siginfo.exp:
6.1.1 has 3 PASS 1 FAIL 1 KFAIL
HEAD has 3 PASS 2 FAIL
One of the fails in HEAD is a known problem with the way signal frames
are now handled; because of the export stub one frame appears twice in
the backtrace.
The KFAIL->FAIL is because it used to just exit when doing a step from
the signal handler, and now it prints an error message about unwind
failures.
I don't think this is really a regression. It's not working correctly in
either case.
gdb.base/callfuncs.exp:
6.1.1 has 6 PASS 93 FAIL 3 KFAIL
HEAD has 73 PASS 26 FAIL 3 KPASS 1 UNRESOLVED
New FAILs are:
+FAIL: gdb.base/callfuncs.exp: gdb function calls preserve register contents
+FAIL: gdb.base/callfuncs.exp: continue after stop in call dummy preserves register contents
+FAIL: gdb.base/callfuncs.exp: nested call dummies preserve register contents
The 1st and 3rd cases are false PASSes in 6.1.1 -- the test program was
exiting prematurely in 6.1.1, so a "info all-registers" was always
reporting "The program has no registers now." and so it passed the
register check. The mismatched registers in HEAD are ior/isr, which
i believe cannot be restored from userspace.
The 2nd test was not being run with 6.1.1, probably also
because the test program failed prematurely.
so again i don't think these are real regressions.
gdb.base/sigstep.exp:
6.1.1 has 15 PASS 10 FAIL 1 KFAIL
HEAD has 19 PASS 7 FAIL
I don't see any new FAILs in HEAD that are not in 6.1.1 on this one.
gdb.base/ena-dis-br.exp:
6.1.1 has 47 PASS 1 FAIL
HEAD has 47 PASS 1 XFAIL
We actually have 1 FAIL->PASS and 1 PASS->XFAIL here
The PASS->XFAIL looks like a test script problem to me, but not very
sure.
gdb.base/annota3.exp:
19 PASS 6 FAIL 1 XFAIL in both 6.1.1 and HEAD, don't see any regressions
haven't looked at the other ones yet.
randolph
--
Randolph Chung
Debian GNU/Linux Developer, hppa/ia64 ports
http://www.tausq.org/