Re: PI BUG with -rt13

[Dinakar Guniguntala <dino@in.ibm.com>]

On Mon, Nov 21, 2005 at 07:24:55PM -0800, david singleton wrote:
> 
> 
> Dinakar,
> 	can you try the attached patch?  I believe it has the fix you 
> 	require. It has the
> correct locking fix I just sent, the patch-2.6.4-rt13-rf2, and has the 
> correct
> lock order in the exit path that was causing the circular deadlock in 
> Dave Carlson's
> application.

David,

This fixes the problem that I was noticing !!
Thank you and Ingo for fixing this

	-Dinakar

> 
> On Nov 21, 2005, at 1:26 PM, Ingo Molnar wrote:
> >* David Singleton <dsingleton@mvista.com> wrote:
> >
> >>Ingo,
> >>   here is a patch that provides the correct locking for the rt_mutex
> >>backing the robust pthread_mutex.   The patch also unifies the locking
> >>for all the robust functions and adds support for pthread_mutexes on
> >>the heap.
> >
> >thanks. Could you split up the patch into a fix and a 'heap' patch (at 
> >a
> >minimum)?
> >
> >it's this portion of the 'heap' patch that looks problematic:
> >
> >>--- base/linux-2.6.14/include/linux/mm.h	2005-11-18 
> >>20:36:53.000000000 -0800
> >>+++ wip/linux-2.6.14/include/linux/mm.h	2005-11-21 
> >>10:51:19.000000000 -0800
> >>@@ -109,6 +109,11 @@
> >> #ifdef CONFIG_NUMA
> >> 	struct mempolicy *vm_policy;	/* NUeMA policy for the VMA */
> >> #endif
> >>+#ifdef CONFIG_FUTEX
> >>+	int robust_init;		/* robust initialized? */
> >>+	struct list_head robust_list;	/* list of robust futexes in this 
> >>vma */
> >>+	struct semaphore robust_sem;	/* semaphore to protect the list */
> >>+#endif
> >> };
> >
> >why is there per-vma info needed?
> >
> >Also, what testing did this patch have - should it solve Dinakar's
> >problem(s)?
> >
> >	Ingo