[ofa-general] Re: mthca use of dma_sync_single is bogus

[Roland Dreier]

 > Hmm. This means there's no way to sync a range within
 > mapping created with map_sg?

It doesn't seem that there is one right now at least.

 > > It actually doesn't look too bad to replace our use of pci_map_sg()
 > > with dma_map_single(), at least at first glance.  I'll try to write a
 > > patch later.
 > 
 > Well, the reason map_sg is there is presumably because on some
 > architectures it's worth it to try and make the region contigious in DMA space.
 > But I agree this seems the lesser evil at this point ...

Given that we're already trying to allocate big chunks of physically
contiguous memory, I think that any virtual merging we get is likely
to be of very small benefit.

It is kind of a shame to give this up though.

 - R.