[ewg] Re: [ofa-general] Toward next OFED release (1.3)

[Roland Dreier]

 > > 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).
 > > 
 > 
 > Check. 
 > 
 > Generating a patch for OFED 1.3 is on my to do list. 

That's great, but please let's not think about it as a patch "for OFED
1.3."  I think this sort of change to the user/kernel interface really
needs to go upstream before it goes into OFED, so just work on getting
the changes into the kernel and libmthca, and then we can worry about
getting them into an OFED release.

 - R.