Re: 2.6.32-rc2: Intel(R) I/OAT DMA Engine init failed

[Dan Williams <dan.j.williams@intel.com>]

On Tue, Oct 6, 2009 at 3:54 AM, Alexander Beregalov
<a.beregalov@gmail.com> wrote:
> but when I commented DEBUG definition, it failed again:
>
> ioatdma: Intel(R) QuickData Technology Driver 4.00
> ioatdma 0000:00:08.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
> ioatdma 0000:00:08.0: setting latency timer to 64
>  alloc irq_desc for 33 on node -1
>  alloc kstat_irqs on node -1
> ioatdma 0000:00:08.0: irq 33 for MSI/MSI-X
> ioatdma 0000:00:08.0: Self-test copy timed out, disabling 0:0
> ioatdma 0000:00:08.0: Intel(R) I/OAT DMA Engine init failed
> ioatdma 0000:00:08.0: PCI INT A disabled

Odd, it looks like neither a dma channel interrupt nor the timer fired
while this operation was in flight.  However, when you turn on debug
we see timer events firing as expected (ioat1_timer_event).  The
operation completes successfully ( dma->device_is_tx_complete returns
DMA_SUCCESS), but its not clear what is causing us to violate the
timeout.  Before bisecting can you try commit 1f6672d4?  This is the
state of the aysnc_tx.git tree before it was merged for 2.6.32.  If
that fails then a full bisect is the only recourse I can see to get
closer to root cause.

Thanks,
Dan