On Wed, Nov 04, 2020 at 10:50:49AM +0100, Christoph Hellwig wrote: > +int ib_dma_virt_map_sg(struct ib_device *dev, struct scatterlist *sg, int > nents) > +{ > + struct scatterlist *s; > + int i; > + > + for_each_sg(sg, s, nents, i) { > + sg_dma_address(s) = (uintptr_t)sg_virt(s); > + sg_dma_len(s) = s->length;
Hmm.. There is nothing ensuring the page is mapped here for this sg_virt(). Before maybe some of the kconfig stuff prevented highmem systems indirectly, I wonder if we should add something more direct to exclude highmem for these drivers? Sigh. I think the proper fix is to replace addr/length with a scatterlist pointer in the struct ib_sge, then have SW drivers directly use the page pointer properly. Then just delete this stuff, all drivers need is a noop dmaops Looks a lot hard though, so we should probably go ahead with this. Jason _______________________________________________ iommu mailing list iommu@lists.linux-foundation.org https://lists.linuxfoundation.org/mailman/listinfo/iommu