Messages in this thread | | | Date | Wed, 2 Dec 2009 09:53:15 +1000 | Subject | Re: | From | Dave Airlie <> |
| |
On Fri, Nov 20, 2009 at 11:29 PM, Jerome Glisse <jglisse@redhat.com> wrote: > This patch series add ttm range validation function. Aim is to > include this in 2.6.33 so i have time to iron out issue, comments.
I missed these first time around,
Thomas if you have any opinions on the TTM stuff please see if you can take a look.
> ttm: > I duplicated a bunch of ttm functions but now i think, best would > be to add range to all function and use free list if range cover > all the manager space. Doing so we might also be able to simplify > mem_space alocation into a simpler function like ttm_bo_mem_space_range > > radeon: > The second patch is a rework/cleanup of radeon object, it solves > few issues along the way (i can't remember them now after fews > days testing the patches). Biggest change is that we now rely > on BO being validated before doing any change to radeon bo structure. > As with any big patch i might introduce regressions, so far after > testing on AGP:R1XX,R2XX,R3XX,R6XX PCIE:R3XX,R4XX,R5XX,R6XX,R7XX > and RS480,RS690 i didn't found anythings obvious (test being X + > glxgears + compiz(on hw which support it) + suspend/resume).
I'll try and take a look at this here it doesn't seem to depend on the first patch so I can push it separate if needed.
Dave.
> > Last patch is smaller, it just use the interface introduced by > the first patch. > >
| |