On Son, 2012-10-07 at 15:06 +0200, Christian König wrote: > Based on Dmitries work, but splitting the code into page > directory and page table handling makes it far more > readable and (hopefully) more reliable. > > Allocations of page tables are made from the SA on demand, > that should still work fine since all page tables are of > the same size. > > Also using the fact that allocations from the SA are mostly > continuously (except for end of buffer wraps and under very > high memory pressure) to group updates send to the chipset > specific code into larger chunks. > > v3: mostly a rewrite of Dmitries previous patch. > > Signed-off-by: Dmitry Cherkasov <Dmitrii.Cherkasov@xxxxxxx> > Signed-off-by: Christian König <deathsimple@xxxxxxxxxxx> Tested-by: Michel Dänzer <michel.daenzer@xxxxxxx> There are some checkpatch errors and warnings though, e.g. trailing whitespace. Also, I noticed a few typos: > + /* initial clear the page directory */ 'Initially' > @@ -865,6 +906,155 @@ uint64_t radeon_vm_map_gart(struct radeon_device *rdev, uint64_t addr) > } > > /** > + * radeon_vm_update_pdes - make sure that page directory is valid > + * > + * @rdev: radeon_device pointer > + * @vm: requested vm > + * @start: start of GPU address range > + * @end: end of GPU address range > + * > + * Allocates new page tables if nessasary 'necessary' -- Earthling Michel Dänzer | http://www.amd.com Libre software enthusiast | Debian, X and DRI developer _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/dri-devel
