This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH] Skip argv0-symlink.exp on target argv[0] isn't available
- From: Pedro Alves <palves at redhat dot com>
- To: Yao Qi <yao at codesourcery dot com>, gdb-patches at sourceware dot org
- Date: Thu, 16 Oct 2014 18:48:53 +0100
- Subject: Re: [PATCH] Skip argv0-symlink.exp on target argv[0] isn't available
- Authentication-results: sourceware.org; auth=none
- References: <1413440712-3645-1-git-send-email-yao at codesourcery dot com>
On 10/16/2014 07:25 AM, Yao Qi wrote:
>
> When the program is running with gdbserver, argv[1] to argv[N] aren't
> available, but argv[0] is. Fortunately, argv0-symlink.exp only
> requires argv[0]. argv0-symlink.exp can be run with gdbserver board
> file, as what we do now.
>
> What we need to check is whether argv[0] is available, so I add a new
> proc gdb_has_argv0 to do so by starting a program, and check
> argc/argv[0] to see whether argv[0] is available.
Sounds good.
> --- a/gdb/testsuite/gdb.base/argv0-symlink.exp
> +++ b/gdb/testsuite/gdb.base/argv0-symlink.exp
> @@ -13,6 +13,11 @@
> # You should have received a copy of the GNU General Public License
> # along with this program. If not, see <http://www.gnu.org/licenses/>.
>
> +if { ![gdb_has_argv0] } {
> + unsupported "Can't get argv\[0\]."
> + return
> +}
> +
I think we shouldn't skip the whole test though; only the
printing argv[0] test. The test file also makes sure
that "info inferiors" shows the symlink name, not the target
of the symlink, and that is host-dependent, not target
dependent. See git 4856b6bc.
>
> # Check that the dereferenced value is sane
> - if { ! [target_info exists noargs] } {
> + global has_argv0
> + if { $has_argv0 } {
> gdb_test "python print (arg0)" "0x.*$testfile\"" "verify dereferenced value"
> }
>
> diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp
> index 7b2a402..dd525dc 100644
> --- a/gdb/testsuite/lib/gdb.exp
> +++ b/gdb/testsuite/lib/gdb.exp
> @@ -4227,6 +4227,85 @@ gdb_caching_proc gdb_skip_xml_test {
> return $xml_missing
> }
>
> +# Return true if argv[0] is available.
> +
> +gdb_caching_proc gdb_has_argv0 {
> + set result 0
> +
> + # Set up, compile, and execute a test program to check whether
> + # argv[0] is available.
> + set src [standard_temp_file has_argv0[pid].c]
> + set exe [standard_temp_file has_argv0[pid].x]
> +
> + gdb_produce_source $src {
> + int main (int argc, char **argv) {
> + return 0;
> + }
> + }
> +
> + gdb_compile $src $exe executable {debug}
> +
> + # Helper proc.
> + proc gdb_has_argv0_1 { exe } {
> + global srcdir subdir
> + global gdb_prompt hex decimal
> +
> + gdb_exit
> + gdb_start
> + gdb_reinitialize_dir $srcdir/$subdir
> + gdb_load "$exe"
> +
> + # Set breakpoint on main.
> + send_gdb "break main\n"
> + gdb_expect {
> + -re "Breakpoint.*${gdb_prompt} $" {
> + }
> + -re "${gdb_prompt} $" {
> + return 0
> + }
> + }
> +
> + # Run to main.
> + gdb_run_cmd
> + gdb_expect {
> + -re "Breakpoint.*${gdb_prompt} $" {
> + }
> + -re "${gdb_prompt} $" {
> + return 0
> + }
> + }
> +
> + # Check whether argc is 1.
> + send_gdb "p argc\n"
> + gdb_expect {
> + -re " = 1\r\n${gdb_prompt} $" {
> +
> + send_gdb "p argv\[0\]\n"
> + gdb_expect {
> + -re " = $hex \".*$exe\"\r\n${gdb_prompt} $" {
I suspect this may break if remote (host|target) testing,
and not sharing the filesystem between build/host/target.
Isn't $exe here a full path on the build?
> + return 1
> + }
> + -re "${gdb_prompt} $" {
> + return 0
> + }
> + }
> + }
> + -re "${gdb_prompt} $" {
> + return 0
> + }
> + }
> + return 0
> + }
> +
I think these gdb_expect's should be gdb_test_multiple's instead.
Thanks,
Pedro Alves