At Sat, 23 Aug 2008 19:39:58 +0000,
Joel Soete wrote:
> Hello Takashi,
> Takashi Iwai wrote:
> > At Fri, 22 Aug 2008 13:04:36 +0100,
> > Joel Soete wrote:
> >> Hello Takashi et al.,
> > ...
> >> I build and boot successfully kernel 32bit including your patch on 2
> >> systems
> >> (a b2k using sba and a d380 using ccio).
> > Thanks for testing!
> welcome ;-)
> >> I just noticed that the above code is ~ the same; otoh there is also a
> >> iommu-helpers.h containing also common code to those 2 drivers. So may be
> >> for
> >> easiest maintenance, could you merge and move this code in this 'helper' as
> >> follow:
> >> --- ./drivers/parisc/iommu-helpers.h.Orig 2008-08-01 12:57:22.000000000
> >> +0000
> >> +++ ./drivers/parisc/iommu-helpers.h 2008-08-22 08:07:26.000000000
> >> +0000
> > That sounds like a good idea.
> > One concern is to define a non-inline function in *.h. But,
> Yes (I thought too but didn't find any other good reason then avoiding
> useless duplicate code)
> > iommu-helper.h is included only by these two drivers, so there is no
> > problem as now, although a comment would be more helpful.
> Yes I hope it will be enough for this stuff to be accepted ;-)
I, too :)
Now updated my git tree:
I'll post each patch again if preferred.
Do you guys see any pending issues? I'd love to merge these patches
into the upstream for 2.6.28.
To get things clear -- I don't intend to fix the problem of mmap on
every non-coherent platform perfectly (yet). Instead, this patch
series is intended to fix the current behavior, at least, for the
sound drivers not to crash unconditionally. It provides a (minimal)
way to mmap the pages taken via dma_alloc_coherent().