This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [RFA] gdb.trace, replace wildcard regexp
- From: Michael Snyder <msnyder at vmware dot com>
- To: Ulrich Weigand <uweigand at de dot ibm dot com>
- Cc: "gdb-patches at sourceware dot org" <gdb-patches at sourceware dot org>, Pedro Alves <pedro at codesourcery dot com>
- Date: Fri, 11 Jun 2010 15:47:03 -0700
- Subject: Re: [RFA] gdb.trace, replace wildcard regexp
- References: <201006111701.o5BH1tlT007147@d12av02.megacenter.de.ibm.com>
Ulrich Weigand wrote:
Michael Snyder wrote:
--- gdb.trace/packetlen.exp 2 Jun 2010 21:55:28 -0000 1.14
+++ gdb.trace/packetlen.exp 4 Jun 2010 18:02:31 -0000
@@ -34,8 +34,8 @@ if { [gdb_compile "$srcdir/$subdir/$srcf
return -1
}
gdb_load $binfile
-gdb_test "tstop" "" ""
-gdb_test "tfind none" "" ""
+gdb_test "tstop" ".*" ""
+gdb_test "tfind none" ".*" ""
runto_main
gdb_reinitialize_dir $srcdir/$subdir
@@ -48,7 +48,7 @@ gdb_reinitialize_dir $srcdir/$subdir
#
gdb_delete_tracepoints
-gdb_test "trace gdb_c_test" "" ""
+gdb_test "trace gdb_c_test" ".*" ""
gdb_trace_setactions "setup collect actions" \
"" \
"collect parm\[0\], parm\[1\], parm\[2\], parm\[3\]" "^$" \
@@ -61,7 +61,7 @@ gdb_trace_setactions "setup collect acti
"end" ""
gdb_test_no_output "tstart" "survive the long packet send"
-gdb_test "break end" "" ""
+gdb_test "break end" ".*" ""
gdb_test "continue" \
"Continuing.*Breakpoint $decimal, end.*" \
"run trace experiment"
This change also causes new FAILs for me:
(gdb) PASS: gdb.trace/packetlen.exp: setup collect actions
tstart^M
You can't do that when your target is `child'^M
(gdb) FAIL: gdb.trace/packetlen.exp: survive the long packet send
break end^M
Breakpoint 3 at 0x760: file /home/uweigand/fsf/gdb-head/gdb/testsuite/gdb.trace/actions.c, line 110.^M
(gdb) continue^M
Continuing.^M
^M
Breakpoint 3, end () at /home/uweigand/fsf/gdb-head/gdb/testsuite/gdb.trace/actions.c:110^M
110 }^M
(gdb) PASS: gdb.trace/packetlen.exp: run trace experiment
tstop^M
Trace is not running.^M
(gdb) FAIL: gdb.trace/packetlen.exp: confirm: survived the long packet send
Yeah, but these ones really do fail, when the target can't do
tracepoints. They need to be run against gdbserver.
Odd, though, most of the trace tests bail early if the target
can't do tracepoints. They don't give arbitrary failures...