Re: [RFA] fix crash when rerunning programs in non-stop mode

[Pedro Alves <pedro@codesourcery.com>]

On Friday 27 March 2009 00:46:01, Doug Evans wrote:

> One problem is that calling is_exited is problematic if
> ! target_has_execution (this one I'm tripping over).

Ooops.  FYI, I'm going to elimiminate those particular
target_has_execution checks in a near future.

> Another problem is if tp == NULL, then old->inferior_ptid
> is restored (this is just an observation).

Ooops 2.  Thank you for fixing this.

> 
> This patch reverses the sense of what's being tested,
> i.e. check for whether old->inferior_ptid restorable
> instead of whether it's dead.
> 
> Ok to check in?

Ok.  Thanks.

> Included is a testcase to exercise the problem.

Thanks, that's much appreciated.

> 2009-03-26  Doug Evans  <dje@google.com>
>
> 	* thread.c (do_restore_current_thread_cleanup): Redo test for
> 	whether to restore old->inferior_ptid.
> 
> 	* gdb.mi/mi-nonstop-exit.exp: New file.
> 	* gdb.mi/non-stop-exit.c: New file.

Remember to put the test logs in testsuite/ChangeLog.

I'll adjust this new test to be able run it against gdbserver
once this is in:

 http://sourceware.org/ml/gdb-patches/2009-03/msg00587.html

-- 
Pedro Alves