Re: BUG in drivers/dma/ioat/dma_v2.c:314

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

On 6/30/2010 11:43 AM, Woodhouse, David wrote:
> On Wed, 2010-06-30 at 19:26 +0100, Chris Li wrote:
>>
>> The delta seems to be this line:
>> ioatdma 0000:00:0f.0: ioat2_set_chainaddr: chainaddr: ffffe000
>
> That's a reasonable address if the IOMMU is enabled. We start at 4GiB
> and work down, so that's the second page given out (or the first 8KiB
> chunk).
>
> It looks like the DMA is going AWOL causing the initialisation to
> fail... but it's interesting that there are no DMA faults reported by
> the IOMMU.
>

It seems the 5400 has a dedicated remapping engine just for the DMA 
device [1], is Linux only setting up the iommus per root port (which the 
DMA bypasses)??

 From the dmesg:
> IOMMU 0: reg_base_addr fe710000 ver 1:0 cap 900000c2f0462 ecap e01
> DMAR: DRHD base: 0x000000fe714000 flags: 0x0
> IOMMU 1: reg_base_addr fe714000 ver 1:0 cap 900000c2f0462 ecap e01
> DMAR: DRHD base: 0x000000fe719000 flags: 0x0
> IOMMU 2: reg_base_addr fe719000 ver 1:0 cap 900000c2f0462 ecap e01
> DMAR: DRHD base: 0x000000fe718000 flags: 0x1
> IOMMU 3: reg_base_addr fe718000 ver 1:0 cap 900000c2f0462 ecap e01

Where we expect bit 54 to be set for the DMA iommu, and it does not 
appear to show up.

--
Dan

[1]: http://www.intel.com/Assets/PDF/datasheet/318610.pdf (Section 
3.11.2 page 256)