[PATCH 2/2] jmpbuf: Add paddings for target specific usage

[Florian Weimer]

On 11/14/2017 12:22 AM, H.J. Lu wrote:
> diff --git a/setjmp/tst-sigsetjmp2.c b/setjmp/tst-sigsetjmp2.c
> new file mode 100644
> index 0000000000..131531feb0
> --- /dev/null
> +++ b/setjmp/tst-sigsetjmp2.c
> @@ -0,0 +1,50 @@
> +/* Test that sigprocmask does not read from the unused part of jmpbuf.
> +   Copyright (C) 2017 Free Software Foundation, Inc.
> +   This file is part of the GNU C Library.
> +
> +   The GNU C Library is free software; you can redistribute it and/or
> +   modify it under the terms of the GNU Lesser General Public
> +   License as published by the Free Software Foundation; either
> +   version 2.1 of the License, or (at your option) any later version.
> +
> +   The GNU C Library 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
> +   Lesser General Public License for more details.
> +
> +   You should have received a copy of the GNU Lesser General Public
> +   License along with the GNU C Library; if not, see
> +<http://www.gnu.org/licenses/>.  */
> +
> +#include <stdlib.h>
> +#include <signal.h>
> +#include <setjmp.h>
> +#include <stdio.h>
> +#include <string.h>
> +#include <errno.h>
> +#include <jmp_buf-macros.h>
> +#include <support/next_to_fault.h>
> +
> +static int
> +do_test (void)
> +{
> +  struct support_next_to_fault jmpbuf
> +    = support_next_to_fault_allocate (SAVED_MASK_OFFSET + (_NSIG / 8));
> +  struct __jmp_buf_tag *sj = (struct __jmp_buf_tag *) jmpbuf.buffer;
> +
> +  errno = 0;
> +  if (sigsetjmp (sj, 1) == 0)
> +    {
> +      siglongjmp (sj, 1);
> +      return EXIT_FAILURE;
> +    }
> +  if (errno != 0)
> +    {
> +      printf ("sigsetjmp: %s\n", strerror (errno));
> +      return EXIT_FAILURE;
> +    }
> +  return EXIT_SUCCESS;
> +}
> +
> +#define TEST_FUNCTION do_test ()
> +#include "../test-skeleton.c"

Please use <support/test-driver.c>.

Wouldn't the test start to fail if sigsetjmp/siglongjmp actually started 
using the padding you are freeing up for use?

I think you need test sigprocmask directly to ensure it has the desired 
property of not clobbering the padding.

Thanks,
Florian