ARM: 2.6.3[45] PCI regression (IXP4xx and PXA?)

[FUJITA Tomonori]

On Wed, 11 Aug 2010 08:25:32 +0100
Russell King - ARM Linux <linux at arm.linux.org.uk> wrote:

> On Wed, Aug 11, 2010 at 11:06:00AM +0900, FUJITA Tomonori wrote:
> > On Tue, 10 Aug 2010 22:36:21 +0200
> > Krzysztof Halasa <khc at pm.waw.pl> wrote:
> > 
> > > 6fee48cd330c68332f9712bc968d934a1a84a32a broke
> > > pci_set_consistent_dma_mask() on IXP4xx and most probably PXA. Affected
> > > devices are e.g. IDE controller (CS5536-based: disk inaccessible) and
> > > e1000 ethernet ("Detected Tx Unit Hang").
> > 
> > Sorry about that.
> > 
> > > The attached patch makes it work again, though I'm not sure it's the
> > > best solution.
> > 
> > I think that we should avoid adding "#ifdef CONFIG_DMABOUNCE" to a
> > generic place.
> > 
> > Why the above patch breaks dmabounce.c? We can't set dev->coherent_dma_mask?
> 
> It doesn't break dmabounce.
> 
> What it breaks is the fact that a PCI device which can do 32-bit DMA is
> connected to a PCI bus which can only access the first 64MB of memory
> through the host bridge, but the system has more than 64MB available.
> 
> Allowing a 32-bit DMA mask means that dmabounce can't detect that memory
> above 64MB needs to be bounced to memory below the 64MB boundary.

But dmabounce doesn't look at dev->coherent_dma_mask.

The change breaks __dma_alloc_buffer()? If we set dev->coherent_dma_mask
to DMA_BIT_MASK(32) for ixp4xx's pci devices, __dma_alloc_buffer()
doesn't use GFP_DMA.