gdb.gdb/selftest.exp failures

[Ulrich Weigand]

FAIL: gdb.gdb/selftest.exp: unknown source line after step over instream initialization
FAIL: gdb.gdb/selftest.exp: step into xmalloc call

Needs further analysis.

[Yao Qi (yao-codesourcery)]

The first failure is caused by incorrect regex in selftest.exp on source code line matching.

Here is a patch to fix,
Index: gdb/testsuite/gdb.gdb/selftest.exp
===================================================================
RCS file: /cvs/src/src/gdb/testsuite/gdb.gdb/selftest.exp,v
retrieving revision 1.24
diff -u -p -r1.24 selftest.exp
--- gdb/testsuite/gdb.gdb/selftest.exp 26 Jun 2010 06:44:47 -0000 1.24
+++ gdb/testsuite/gdb.gdb/selftest.exp 16 Aug 2010 07:12:44 -0000
@@ -185,11 +185,15 @@ proc do_steps_and_nexts {} {
   set description "step over instream initialization"
   set command "step"
      }
- -re ".*getcwd .gdb_dirbuf, sizeof .gdb_dirbuf..;.*$gdb_prompt $" {
+ -re ".*getcwd .gdb_dirbuf, sizeof .gdb_dirbuf.*$gdb_prompt $" {
   set description "next over getcwd"
   set command "next"
      }
- -re ".*quit_flag = 0.*$gdb_prompt $" {
+ -re ".*gdb_program_name = xstrdup.*$gdb_prompt $" {
+ set description "next over xstrdup"
+ set command "next"
+ }
+ -re ".*quit_flag = 0.*$gdb_prompt $" {
   set description "step over quit_flag initialization"
   set command "step"
      }

After patch is applied, original two failures are disappeared, but there is a new failure. However looks like a separate issue.

FAIL: gdb.gdb/selftest.exp: backtrace through signal handler

  === gdb Summary ===

# of expected passes 38
# of unexpected failures 1
/home/yao/maverick/home/yao/cvs/build/gdb/testsuite/../../gdb/gdb version 7.2.50.20100811-cvs -nw -nx

(gdb) PASS: gdb.gdb/selftest.exp: send ^C to child process
signal SIGINT^M
Continuing with signal SIGINT.^M
^C^M
Program received signal SIGINT, Interrupt.^M
__libc_do_syscall () at ../ports/sysdeps/unix/sysv/linux/arm/eabi/libc-do-syscall.S:41^M
41 in ../ports/sysdeps/unix/sysv/linux/arm/eabi/libc-do-syscall.S^M
(gdb) PASS: gdb.gdb/selftest.exp: send SIGINT signal to child process
backtrace^M
#0 __libc_do_syscall () at ../ports/sysdeps/unix/sysv/linux/arm/eabi/libc-do-syscall.S:41^M
#1 0x403b85ee in __sigprocmask (how=2, set=0xbe8a4ee8, oset=0x0) at ../sysdeps/unix/sysv/linux/sigprocmask.c:66^M
#2 0x400392c4 in ?? () from /lib/libreadline.so.6^M
#3 0x400392c4 in ?? () from /lib/libreadline.so.6^M
Backtrace stopped: previous frame identical to this frame (corrupt stack?)^M
(gdb) FAIL: gdb.gdb/selftest.exp: backtrace through signal handler

[Ulrich Weigand (uweigand)]

The patch looks good to me.

The new failure seems to be caused by having no debug information for libreadline.so, since the fall-back prologue analyzer doesn't yet work correctly for Thumb-2 code (I'm working on that). For now, it might go away by installing the corresponding -dbg package containing stand-along debug info for libreadline.

[Yao Qi (yao-codesourcery)]

Install libreadline6-dgb, no failure any more,

Running /home/yao/maverick/home/yao/cvs/src/gdb/testsuite/gdb.gdb/selftest.exp ...

  === gdb Summary ===

# of expected passes 39
/home/yao/maverick/home/yao/cvs/build/gdb/testsuite/../../gdb/gdb version 7.2.50.20100811-cvs -nw -nx

[Yao Qi (yao-codesourcery)]

Sent patch to gdb-patches, http://www.cygwin.com/ml/gdb-patches/2010-08/msg00252.html

[Yao Qi (yao-codesourcery)]

Patch is committed to upstream gdb.
http://sourceware.org/ml/gdb-cvs/2010-08/msg00095.html

[Yao Qi (yao-codesourcery)]

Committed to GDB 7.2 branch.
http://www.cygwin.org/ml/gdb-cvs/2010-08/msg00159.html