This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
[patch] testsuite: sigall.exp: Fix false FAILs
- From: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- To: gdb-patches at sourceware dot org
- Date: Sun, 12 Sep 2010 15:31:13 +0200
- Subject: [patch] testsuite: sigall.exp: Fix false FAILs
Hi,
if your sourcetree topdir contains a substring "kill" then while still:
make check RUNTESTFLAGS=gdb.base/sigall.exp
# of expected passes 380
the other case starts FAILing:
make check//unix RUNTESTFLAGS=gdb.base/sigall.exp
# of expected passes 290
# of unexpected failures 90
It is due to a false catch in the second case:
Breakpoint 61, gen_WAITING () at /home/jkratoch/redhat/gdb-test-dir-kill/gdb/testsuite/gdb.base/sigall.c:1136^M
1136 handle_WAITING (0);^M
which is not a problem in the first case:
Breakpoint 61, gen_WAITING () at ./gdb.base/sigall.c:1136^M
1136 handle_WAITING (0);^M
The two lines being matched look always like:
826 kill (getpid (), SIGILL);^M
839 handle_EMT (0);^M
Tested on x86_64-fedora14snapshot-linux-gnu.
Going to check it in in some days as obvious.
Thanks,
Jan
gdb/testsuite/
2010-09-12 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix false FAILs on sourcetree topdir directory containing "kill".
* gdb.base/sigall.exp (test_one_sig) <advance to $nextsig>: Extend the
source line matching regexp.
--- a/gdb/testsuite/gdb.base/sigall.exp
+++ b/gdb/testsuite/gdb.base/sigall.exp
@@ -94,11 +94,11 @@ proc test_one_sig {nextsig} {
if { $missed_handler == "0" } then {
gdb_test_multiple "signal 0" "advance to $nextsig" {
- -re "Breakpoint.*gen_$nextsig.*kill.*$gdb_prompt $" {
+ -re "Breakpoint.*gen_$nextsig.*\r\n\[0-9\]+\[ \t\]+kill \\(.*\r\n$gdb_prompt $" {
pass "advance to $nextsig"
set sig_supported 1
}
- -re "Breakpoint.*gen_$nextsig.*handle.*$gdb_prompt $" {
+ -re "Breakpoint.*gen_$nextsig.*\r\n\[0-9\]+\[ \t\]+handle_.*\r\n$gdb_prompt $" {
pass "advance to $nextsig"
set sig_supported 0
}