[RFC PATCH 2/6] drm/radeon: don't use ttm bo->offset

[Christian König]

Am 13.02.20 um 14:28 schrieb Nirmoy:
>
> On 2/13/20 2:00 PM, Christian König wrote:
>> Am 13.02.20 um 13:31 schrieb Nirmoy:
>>>
>>> On 2/13/20 1:18 PM, Christian König wrote:
>>>> Looks like most of the patch is missing?
>>>>
>>>> We should have quite a number of uses of the BO offset in radeon or 
>>>> do all of those go through radeon_bo_gpu_offset?
>>> Compilation worked so I think all those(bo->offset) accesses went 
>>> through radeon_bo_gpu_offset.
>>
>> Cool, than that is a lot easier to implement than I thought it would be.
>>
>> I assume you don't have Radeon hardware lying around to test this?
>>
>> If you can you give me a branch on the AMD (or public) servers I can 
>> give it at least a quick round of testing.
>
> huh it seems I have to rebase it a bit for drm/drm-next as it is based 
> on our internal branch :/
>
> You can access the branch at 
> http://gitlab1.amd.com/nirmodas/linux/tree/ttm_clean.1
>
> I will rebase this for drm-next in next version of the patch series.

Actually you should probably base it on drm-misc-next instead.

And additional to the already noted VRAM helpers the bochs driver fails 
to compile on your branch.

The good news is that my old HD5670 still works with the radeon patch 
applied. So this patch is Reviewed-and-tested-by: Christian König 
<christian.koenig at amd.com>.

Regards,
Christian.

>
>>
>> Christian.
>>
>>>>
>>>> If yes then the change is much smaller than I thought i needs to be.
>>>>
>>>> Christian.
>>>>
>>> Regards,
>>>
>>> Nirmoy
>>>
>>> _______________________________________________
>>> amd-gfx mailing list
>>> amd-gfx at lists.freedesktop.org
>>> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.freedesktop.org%2Fmailman%2Flistinfo%2Famd-gfx&data=02%7C01%7CNirmoy.Das%40amd.com%7C60aa17c05a2f474663f008d7b084b550%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637171956342728389&sdata=I2n0yMXK5CtrG5tY9Q47igxJv9Onb%2FA7PBeLwrpPb54%3D&reserved=0 
>>>
>>