Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Daniel Jacobowitz <drow@false.org>
To: Jan Kratochvil <jan.kratochvil@redhat.com>
Cc: Mark Kettenis <mark.kettenis@xs4all.nl>, gdb-patches@sourceware.org
Subject: Re: [patch] Disable child VMA randomizations
Date: Thu, 26 Jun 2008 15:52:00 -0000	[thread overview]
Message-ID: <20080626144150.GL22726@caradoc.them.org> (raw)
In-Reply-To: <20080608164331.GA6358@host0.dyn.jankratochvil.net>

Please also add the new commands to NEWS.

On Sun, Jun 08, 2008 at 06:43:31PM +0200, Jan Kratochvil wrote:
> @@ -149,6 +156,19 @@ show_debug_linux_nat_async (struct ui_fi
>  		    value);
>  }
>  
> +#ifdef HAVE_PERSONALITY
> +static int disable_randomization = 1;
> +
> +static void
> +show_disable_randomization (struct ui_file *file, int from_tty,
> +			    struct cmd_list_element *c, const char *value)
> +{
> +  fprintf_filtered (file, _("\
> +Disabling randomization of debuggee's virtual address space is %s.\n"),
> +		    value);
> +}
> +#endif /* HAVE_PERSONALITY */
> +
>  static int linux_parent_pid;
>  
>  struct simple_pid_list

I think it would be better to register the command unconditionally,
but only call personality if it is available.  I'd like to keep
conditionally compiled code to a minimum.

> +	warning (_("Error setting the enabled disable-randomization: %s"),

Error disabling address space randomization:

> +		 safe_strerror (errno));
> +    }
> +#endif /* HAVE_PERSONALITY */
> +
>    linux_ops->to_create_inferior (exec_file, allargs, env, from_tty);
>  
> +#ifdef HAVE_PERSONALITY
> +  if (personality_set)
> +    {
> +      errno = 0;
> +      personality (personality_orig);
> +      if (errno != 0)
> +	warning (_("Problem restoring the disable-randomization state: %s"),

Error restoring address space randomization:

> +When this mode is on (which is the default), the randomization of\n\
> +the virtual address space is disabled.  Standalone programs run with the\n\
> +randomization enabled by default on some platforms."),

You don't need "the" before randomization in this case (two places).

> +misusing it.  Still even a prelinked shared library can get loaded a a new

at a

> +    gdb_test_multiple "continue" $testname {
> +	-re "address = (0x\[0-9a-f\]*).*Program exited normally..*$gdb_prompt $" {
> +	    pass $testname
> +	    return $expect_out(1,string)
> +	}

Please set a breakpoint and use GDB to print out the variable, instead
of relying on printf; some test configurations don't work with stdio.
They probably won't work with this test anyway, but no reason to
complicate things.

> +gdb_test "set disable-randomization off"
> +set addr1 [address_get "randomized first address"]
> +set addr2 [address_get "randomized second address"]
> +set test "randomized addresses should not match"
> +if {$addr1 eq $addr2} {
> +    fail $test
> +} else {
> +    pass $test
> +}

This test will only pass if the system supports address randomization
and has it enabled.  I don't know about you, but the first thing I do
on a new developer system is turn it off...

-- 
Daniel Jacobowitz
CodeSourcery


  reply	other threads:[~2008-06-26 14:42 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-06-07 19:54 Jan Kratochvil
2008-06-07 20:41 ` Andreas Schwab
2008-06-08  9:43   ` Jan Kratochvil
2008-06-08 10:39     ` Eli Zaretskii
2008-06-08 11:38       ` Jan Kratochvil
2008-06-07 20:43 ` Eli Zaretskii
2008-06-08 15:14 ` Mark Kettenis
2008-06-08 16:44   ` Jan Kratochvil
2008-06-26 15:52     ` Daniel Jacobowitz [this message]
2008-06-27 23:19       ` Jan Kratochvil
2008-07-09 21:15         ` Daniel Jacobowitz
2008-07-10  9:34           ` Jan Kratochvil
2008-07-12 21:16         ` Ulrich Weigand
2008-07-13  6:55           ` Jan Kratochvil
2008-07-15 18:41             ` Ulrich Weigand

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20080626144150.GL22726@caradoc.them.org \
    --to=drow@false.org \
    --cc=gdb-patches@sourceware.org \
    --cc=jan.kratochvil@redhat.com \
    --cc=mark.kettenis@xs4all.nl \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox