> > Sorry... I've been remiss in posting about this. I would actually > > prefer to see an extension to the dma_map_sg() interface (a new flag > > perhaps?) that would set the right magic bit in the DMA address on > > altix. The refactoring of ib_umem_get() to be called by low-level > > drivers makes this a fairly clean approach, and it avoids the problems > > with using dma_alloc_coherent() to allocate userspace buffers (for > > example, dma_alloc_coherent() uses up kernel virtual addresses, which > > may be scarce on 32 bit architectures).
> While this make sense it would be hard to push into the kernel proper. > Or no? Are you going to do that? I don't think I'm willing to merge a fix that uses dma_alloc_coherent() inside the kernel so this alternate fix is probably easier to merge. Yes, it does mean an extension to the DMA mapping API but I think getting that right will be useful in terms of making sure what we're doing really makes sense. - R. _______________________________________________ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg