Grant Grundler wrote: > I think pci_dma_supported() is obsoleted by pci_set_dma_mask(). nope, it hasn't. pci_set_dma_mask() in fact calls pci_dma_supported(). I haven't look at that code in a while... grant