Corey,

I wonder if you are not simply seeing a side effect of how the
interrupt lazy masking is handled on PPC.

The reason regular spin_lock/spin_unlock are used is because the
__pfm_ctxsw() routine is expected to
be run with interrupts ALREADY masked by upper level code (likely
schedule()).  It may be that on PPC,
schedule does things differently at its tail (where switch_to()) is invoked.

You are trying adding a BUG_ON(!irqs_disabled()) and see if you catch
something. If not then I wonder
how this could be related to the lazy masking for which you had to add
the special pfm_spin_lock_irq.....



On Wed, Apr 9, 2008 at 11:49 PM, Corey Ashford <[EMAIL PROTECTED]> wrote:
> Hi,
>
>  We tried out this fix and it made our rapid sampling test case a lot more
> reliable.  Note that we are occasionally seeing another hang and will post
> about that later.  For now, here is the spin lock fix.
>
>  - Corey
>
>  --
>  Corey Ashford
>  Software Engineer
>  IBM Linux Technology Center, Linux Toolchain
>  Beaverton, OR
>  503-578-3507
>  [EMAIL PROTECTED]
>
>
> -------------------------------------------------------------------------
>  This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
>  Don't miss this year's exciting event. There's still time to save $100.
>  Use priority code J8TL2D2.
>
> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
> _______________________________________________
>  perfmon2-devel mailing list
>  [email protected]
>  https://lists.sourceforge.net/lists/listinfo/perfmon2-devel
>
>

-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
Don't miss this year's exciting event. There's still time to save $100. 
Use priority code J8TL2D2. 
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
perfmon2-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/perfmon2-devel

Reply via email to