Re: [PATCH][gdb/testsuite] Rewrite catch-follow-exec.exp

[Simon Marchi <simon dot marchi at polymtl dot ca>]

On 2018-10-11 03:47, Tom de Vries wrote:
> [gdb/testsuite] Rewrite catch-follow-exec.exp
>
> There are two problems with the current catch-follow-exec.exp:
> - INTERNAL_GDBFLAGS (containing the datadir setting) is not used
> - remote host testing doesn't work
>
> Fix the former by using gdb_spawn_with_cmdline_opts.  Fix the latter by
> requiring gdb-native.
>
> Build on x86_64-linux with and without ubsan, tested with and without
> --target_board=native-gdbserver.

My build of GDB (and probably some on the buildbot too?) uses 
-fsanitize=address, and on it the test does not pass.  On a build 
without -fsanitize=address, it does pass.  The failing test is:

FAIL: gdb.base/catch-follow-exec.exp: [lindex $result 3] == 1

and the value of $result is "17872 exp10 0 23".  This is because ASan 
exits with 23 if it detects leaks. If there's a way to set 
LSAN_OPTIONS='exitcode=1' in the environment GDB runs in, it would 
probably make it work...

> 2018-10-05  Tom de Vries  <tdevries@suse.de>
>
> 	PR gdb/23730
> 	* gdb.base/catch-follow-exec.c: Add copyright notice.
> 	* gdb.base/catch-follow-exec.exp: Rewrite to use
> 	gdb_spawn_with_cmdline_opts.  Require gdb-native.
>
> ---
>  gdb/testsuite/gdb.base/catch-follow-exec.c   | 17 ++++++++++++
>  gdb/testsuite/gdb.base/catch-follow-exec.exp | 41 
> ++++++++++++++++++++++------
>  2 files changed, 49 insertions(+), 9 deletions(-)
>
> diff --git a/gdb/testsuite/gdb.base/catch-follow-exec.c
> b/gdb/testsuite/gdb.base/catch-follow-exec.c
> index fa68a2a34e..1a59f58aa5 100644
> --- a/gdb/testsuite/gdb.base/catch-follow-exec.c
> +++ b/gdb/testsuite/gdb.base/catch-follow-exec.c
> @@ -1,3 +1,20 @@
> +/* This testcase is part of GDB, the GNU debugger.
> +
> +   Copyright 2018 Free Software Foundation, Inc.
> +
> +   This program is free software; you can redistribute it and/or 
> modify
> +   it under the terms of the GNU General Public License as published 
> by
> +   the Free Software Foundation; either version 3 of the License, or
> +   (at your option) any later version.
> +
> +   This program is distributed in the hope that it will be useful,
> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> +   GNU General Public License for more details.
> +
> +   You should have received a copy of the GNU General Public License
> +   along with this program.  If not, see 
> <http://www.gnu.org/licenses/>.  */
> +
>  #include <stdio.h>
>  #include <string.h>
>  #include <unistd.h>
> diff --git a/gdb/testsuite/gdb.base/catch-follow-exec.exp
> b/gdb/testsuite/gdb.base/catch-follow-exec.exp
> index 0e32ed4a6f..2ccd0115e7 100644
> --- a/gdb/testsuite/gdb.base/catch-follow-exec.exp
> +++ b/gdb/testsuite/gdb.base/catch-follow-exec.exp
> @@ -22,8 +22,8 @@ if {[build_executable "failed to prepare" $testfile
> $srcfile debug] == -1} {
>      return -1
>  }
>
> -if { ![remote_file target exists /bin/bash] } {
> -    unsupported "no bash"
> +if { [target_info gdb_protocol] != "" } {
> +    unsupported "not native"

Please add a comment here, something like:

# Even though the feature under features being tested are supported by 
gdbserver,
# the way this test is written doesn't make it easy with a remote 
target.

>      return
>  }
>
> @@ -34,7 +34,7 @@ if { ![remote_file target exists /bin/ls] } {
>
>  proc catch_follow_exec { } {
>      global binfile
> -    global GDB
> +    global gdb_spawn_id
>
>      set test "catch-follow-exec"
>
> @@ -45,13 +45,36 @@ proc catch_follow_exec { } {
>      append FLAGS " -ex \"run\""
>      append FLAGS " -ex \"info prog\""
>
> -    catch {exec /bin/bash -c "$GDB $FLAGS"} catchlog
> -    send_log "$catchlog\n"
> +    gdb_exit
> +    if {[gdb_spawn_with_cmdline_opts "$FLAGS"] != 0} {
> +	fail "spawn"
> +	return
> +    }
> +
> +    gdb_test_multiple "" "run til exit" {
> +	"runtime error:" {
> +	    # Error in case of --enable-ubsan
> +	    fail "No runtime error"
> +	}

Can you explain what this catches?  If this isn't there but UBSan is 
enabled, the crash isn't detected properly by the code below?

Please use a lower case letter for the test name.

> +	eof {
> +	    set result [wait -i $gdb_spawn_id]
> +	    verbose $result
> +
> +	    gdb_assert { [lindex $result 2] == 0 }
> +
> +	    # We suspect this will be zero instead of one after fixing 
> PR23368
> +	    # - "gdb goes to into background when hitting exec catchpoint 
> with
> +	    # follow-exec-mode new"
> +            gdb_assert { [lindex $result 3] == 1 }
> +
> +	    # Error in case of --disable-ubsan, we get
> +	    # "CHILDKILLED SIGSEGV {segmentation violation}" as extra
> +	    # argument(s).
> +	    gdb_assert { [llength $result] == 4 }
> +	}
>
> -    if { [regexp {No selected thread} $catchlog] } {
> -	pass $test
> -    } else {
> -	fail $test
> +	remote_close host
> +	clear_gdb_spawn_id
>      }
>  }

Simon