[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCHv2 net-next] xen-netback: fix race between napi_complete() and interrupt handler



On Thu, 2014-05-15 at 18:17 +0100, David Vrabel wrote:
> When the NAPI budget was not all used, xenvif_poll() would call
> napi_complete() /after/ enabling the interrupt.  This resulted in a
> race between the napi_complete() and the napi_schedule() in the
> interrupt handler.  The use of local_irq_save/restore() avoided by
> race iff the handler is running on the same CPU but not if it was
> running on a different CPU.
> 
> Fix this properly by calling napi_complete() before reenabling
> interrupts (in the xenvif_napi_schedule_or_enable_irq() call).
> 
> Signed-off-by: David Vrabel <david.vrabel@xxxxxxxxxx>
> ---
> v2:
> - Rename xenvif_check_rx_xenvif() to
>   xenvif_napi_schedule_or_enable_irq() to make it more obvious what it
>   does.

Which is to update the event pointer such that IRQs are generated again
rather than enabling IRQs as such (enable_irq makes me think of
EFLAGS.IF). But given that I can't think of a better way to describe it:
Acked-by: Ian Campbell <ian.campbell@xxxxxxxxxx>
(although in this area Wei's ack is more important than mine IMHO).

I think this is intended as a replacement to Wei's "xen-netback: don't
move event pointer in TX credit timeout callback" rather than in
addition to, correct?

Ian.


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.