How does this patch look? Seems pretty uncontroversial, maybe we can get it
The pciBusAddrToHostAddr changes should be no-ops until we introduce that
function in arch code.
Created attachment 2460 [details] [review]
dri fixes for 64 bit machines
This'll probably have to hit the dri tree as well...
Created attachment 2524 [details] [review]
since frameBufferPhysicalAddress is in a shared structure between libdri and
the 2d drivers, we need to bump the libdri version number.
the radeon pointer printf changes have been dropped for the moment, i tried the
naive method of s/0x%016lx/%p/g but it caused more warnings than it cured.
Created attachment 2707 [details] [review]
the changes to bug #3163 make this trivial.
applied to head, closing.