[8/8] New test about step over clone syscall

  This patch adds a new test for stepping over clone syscall.

2016-02-19  Yao Qi  <yao.qi@linaro.org>

	* gdb.base/step-over-syscall.exp (step_over_syscall): Kfail.
	Invoke step_over_syscall "clone".
	* gdb.base/step-over-clone.c: New file.
---
 gdb/testsuite/gdb.base/step-over-clone.c     | 53 ++++++++++++++++++++++++++++
 gdb/testsuite/gdb.base/step-over-syscall.exp | 17 +++++++++
 2 files changed, 70 insertions(+)
 create mode 100644 gdb/testsuite/gdb.base/step-over-clone.c
  

On 02/19/2016 12:36 PM, Yao Qi wrote:
> This patch adds a new test for stepping over clone syscall.
>
> 2016-02-19  Yao Qi  <yao.qi@linaro.org>
>
> 	* gdb.base/step-over-syscall.exp (step_over_syscall): Kfail.
> 	Invoke step_over_syscall "clone".
> 	* gdb.base/step-over-clone.c: New file.
> ---
>   gdb/testsuite/gdb.base/step-over-clone.c     | 53 ++++++++++++++++++++++++++++
>   gdb/testsuite/gdb.base/step-over-syscall.exp | 17 +++++++++
>   2 files changed, 70 insertions(+)
>   create mode 100644 gdb/testsuite/gdb.base/step-over-clone.c
>
> diff --git a/gdb/testsuite/gdb.base/step-over-clone.c b/gdb/testsuite/gdb.base/step-over-clone.c
> new file mode 100644
> index 0000000..43b1933
> --- /dev/null
> +++ b/gdb/testsuite/gdb.base/step-over-clone.c
> @@ -0,0 +1,53 @@
> +/* This testcase is part of GDB, the GNU debugger.
> +
> +   Copyright 2016 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/>.  */
> +
> +#define _GNU_SOURCE
> +#include <stdlib.h>
> +#include <unistd.h>
> +#include <sched.h>
> +
> +static void
> +marker () {}
> +
> +#define STACK_SIZE 0x1000
> +
> +static int
> +clone_fn (void *unused)
> +{
> +  return 0;
> +}
> +
> +int
> +main (void)
> +{
> +  int i, pid;
> +  unsigned char *stack[6];
> +
> +  for (i = 0; i < (sizeof (stack) / sizeof (stack[0])); i++)
> +    stack[i] = malloc (STACK_SIZE);
> +
> +  for (i = 0; i < (sizeof (stack) / sizeof (stack[0])); i++)
> +    {
> +      pid = clone (clone_fn, stack[i] + STACK_SIZE, CLONE_FILES | CLONE_VM,
> +		   NULL);
> +    }
> +
> +  for (i = 0; i < (sizeof (stack) / sizeof (stack[0])); i++)
> +    free (stack[i]);
> +
> +  marker ();
> +}
> diff --git a/gdb/testsuite/gdb.base/step-over-syscall.exp b/gdb/testsuite/gdb.base/step-over-syscall.exp
> index 863194b..e5a3d55 100644
> --- a/gdb/testsuite/gdb.base/step-over-syscall.exp
> +++ b/gdb/testsuite/gdb.base/step-over-syscall.exp
> @@ -122,6 +122,13 @@ proc step_over_syscall { syscall } {
>   	foreach displaced $disp_step_opts {
>   	    with_test_prefix "displaced $displaced" {
>
> +		if { $displaced == "on" && $syscall == "clone" } {
> +		    # GDB doesn't support step over clone syscall with

"...stepping over clone ..."?

> +		    # displaced stepping.
> +		    kfail "gdb/19675" "single step over clone"
> +		    continue
> +		}
> +
>   		gdb_test "continue" \
>   		    "Continuing\\..*Breakpoint \[0-9\]+, (.* in |__libc_|)$syscall \\(\\).*" \
>   		    "continue to $syscall (3rd time)"
> @@ -187,6 +194,15 @@ proc step_over_syscall { syscall } {
>   			}
>   		    }
>
> +		if { $syscall == "clone" } {
> +		    # Create a breakpoint in the child with the condition
> +		    # which is false, so that GDBserver can get the
> +		    # event from the child but GDB doesn't see it.  In
> +		    # this way, we don't have to adjust test flow for

"... to adjust the test flow..."?
  

On 02/19/2016 02:36 PM, Yao Qi wrote:
> +		if { $syscall == "clone" } {
> +		    # Create a breakpoint in the child with the condition
> +		    # which is false, so that GDBserver can get the
> +		    # event from the child but GDB doesn't see it.  In
> +		    # this way, we don't have to adjust test flow for
> +		    # "clone".

I think this comment should mention this is a regression test for
a previous bug.  I'd file the description of patch #1 as a PR,
and reference the PR here.  I had read that patch before this one,
and it took me a bit to see the connection.

> +		    gdb_test "break clone_fn if main == 0"
> +		}

Thanks,
Pedro Alves
  

On 02/19/2016 02:36 PM, Yao Qi wrote:
> +static void
> +marker () {}

static void
marker (void)
{
}

Thanks,
Pedro Alves