[PATCH 8/8] drm/ttm: do not keep GPU dependent addresses

Christian König ckoenig.leichtzumerken at gmail.com
Tue Feb 18 18:37:44 UTC 2020 

Am 18.02.20 um 19:28 schrieb Thomas Zimmermann:
> Hi
>
> Am 18.02.20 um 19:23 schrieb Christian König:
>> Am 18.02.20 um 19:16 schrieb Thomas Zimmermann:
>>> Hi
>>>
>>> Am 18.02.20 um 18:13 schrieb Nirmoy:
>>>> On 2/18/20 1:44 PM, Christian König wrote:
>>>>> Am 18.02.20 um 13:40 schrieb Thomas Zimmermann:
>>>>>> Hi
>>>>>>
>>>>>> Am 17.02.20 um 16:04 schrieb Nirmoy Das:
>>>>>>> GPU address handling is device specific and should be handle by its
>>>>>>> device
>>>>>>> driver.
>>>>>>>
>>>>>>> Signed-off-by: Nirmoy Das <nirmoy.das at amd.com>
>>>>>>> ---
>>>>>>>     drivers/gpu/drm/ttm/ttm_bo.c    | 7 -------
>>>>>>>     include/drm/ttm/ttm_bo_api.h    | 2 --
>>>>>>>     include/drm/ttm/ttm_bo_driver.h | 1 -
>>>>>>>     3 files changed, 10 deletions(-)
>>>>>>>
>>>>>>> diff --git a/drivers/gpu/drm/ttm/ttm_bo.c
>>>>>>> b/drivers/gpu/drm/ttm/ttm_bo.c
>>>>>>> index 151edfd8de77..d5885cd609a3 100644
>>>>>>> --- a/drivers/gpu/drm/ttm/ttm_bo.c
>>>>>>> +++ b/drivers/gpu/drm/ttm/ttm_bo.c
>>>>>>> @@ -85,7 +85,6 @@ static void ttm_mem_type_debug(struct
>>>>>>> ttm_bo_device *bdev, struct drm_printer *p
>>>>>>>         drm_printf(p, "    has_type: %d\n", man->has_type);
>>>>>>>         drm_printf(p, "    use_type: %d\n", man->use_type);
>>>>>>>         drm_printf(p, "    flags: 0x%08X\n", man->flags);
>>>>>>> -    drm_printf(p, "    gpu_offset: 0x%08llX\n", man->gpu_offset);
>>>>>>>         drm_printf(p, "    size: %llu\n", man->size);
>>>>>>>         drm_printf(p, "    available_caching: 0x%08X\n",
>>>>>>> man->available_caching);
>>>>>>>         drm_printf(p, "    default_caching: 0x%08X\n",
>>>>>>> man->default_caching);
>>>>>>> @@ -345,12 +344,6 @@ static int ttm_bo_handle_move_mem(struct
>>>>>>> ttm_buffer_object *bo,
>>>>>>>     moved:
>>>>>>>         bo->evicted = false;
>>>>>>>     -    if (bo->mem.mm_node)
>>>>>>> -        bo->offset = (bo->mem.start << PAGE_SHIFT) +
>>>>>>> -            bdev->man[bo->mem.mem_type].gpu_offset;
>>>>>>> -    else
>>>>>>> -        bo->offset = 0;
>>>>>>> -
>>>>>> After moving this into users, the else branch has been lost. Is
>>>>>> 'bo->mem.mm_node' always true?
>>>>> At least for the amdgpu and radeon use cases, yes.
>>>>>
>>>>> But that is a rather good question I mean for it is illegal to get the
>>>>> GPU BO address if it is inaccessible (e.g. in the system domain).
>>>>>
>>>>> Could be that some driver relied on the behavior to get 0 for the
>>>>> system domain here.
>>>> I wonder how to verify that ?
>>>>
>>>> If I understand correctly:
>>>>
>>>> 1 qxl uses bo->offset only in qxl_bo_physical_address() which is not in
>>>> system domain.
>>>>
>>>> 2 unfortunately I can't say the same for bochs but it works with this
>>>> patch series so I think bochs is fine as well.
>>>>
>>>> 3 vmwgfx uses bo->offset only when bo->mem.mem_type == TTM_PL_VRAM so
>>>> vmwgfx should be fine.
>>>>
>>>> 4 amdgpu and radeon runs with 'bo->mem.mm_node' always true
>>>>
>>>> I am not sure about  nouveau as bo->offset is being used in many places.
>>>>
>>>> I could probably mirror the removed logic to nouveau as
>>> I suggest to introduce a ttm helper that contains the original branching
>>> and use it everywhere. Something like
>>>
>>>     s64 ttm_bo_offset(struct ttm_buffer_object *bo)
>>>     {
>>>        if (WARN_ON_ONCE(!bo->mem.mm_node))
>>>            return 0;
>>>        return bo->mem.start << PAGE_SHIFT;
>>>     }
>>>
>>> Could be static inline. The warning should point to broken drivers. This
>>> also gets rid of the ugly shift in the drivers.
>> Big NAK on this. That is exactly what we should NOT do.
>>
>> See the shift belongs into the driver, because it is driver dependent if
>> we work with page or byte offsets.
>>
>> For amdgpu we for example want to work with byte offsets and TTM should
>> not make any assumption about what bo->mem.start actually contains.
> OK. What about something like ttm_bo_pg_offset()? Same code without the
> shift. Would also make it clear that it's a page offset.

That is a rather good idea. We could name that ttm_bo_man_offset() and 
put it into ttm_bo_manager.c next to the manager which allocates them.

It's just that this manager is not used by everybody, so amdgpu, radeon 
and nouveau would still need a separate function.

Christian.

>
> Best regards
> Thomas
>
>> Regards,
>> Christian.
>>
>>> Best regards
>>> Thomas
>>>
>>>
>>>> diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c
>>>> b/drivers/gpu/drm/nouveau/nouveau_bo.c
>>>> index f8015e0318d7..5a6a2af91318 100644
>>>> --- a/drivers/gpu/drm/nouveau/nouveau_bo.c
>>>> +++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
>>>> @@ -1317,6 +1317,10 @@ nouveau_bo_move_ntfy(struct ttm_buffer_object
>>>> *bo, bool evict,
>>>>                   list_for_each_entry(vma, &nvbo->vma_list, head) {
>>>>                           nouveau_vma_map(vma, mem);
>>>>                   }
>>>> +               if (bo->mem.mm_node)
>>>> +                       nvbo->offset = (new_reg->start << PAGE_SHIFT);
>>>> +               else
>>>> +                       nvbo->offset = 0;
>>>>           } else {
>>>>                   list_for_each_entry(vma, &nvbo->vma_list, head) {
>>>>                           WARN_ON(ttm_bo_wait(bo, false, false));
>>>>
>>>> Regards,
>>>>
>>>> Nirmoy
>>>>
>>>>
>>>>> Regards,
>>>>> Christian.
>>>>>
>>>>>> Best regards
>>>>>> Thomas
>>>>>>
>>>>>>>         ctx->bytes_moved += bo->num_pages << PAGE_SHIFT;
>>>>>>>         return 0;
>>>>>>>     diff --git a/include/drm/ttm/ttm_bo_api.h
>>>>>>> b/include/drm/ttm/ttm_bo_api.h
>>>>>>> index b9bc1b00142e..d6f39ee5bf5d 100644
>>>>>>> --- a/include/drm/ttm/ttm_bo_api.h
>>>>>>> +++ b/include/drm/ttm/ttm_bo_api.h
>>>>>>> @@ -213,8 +213,6 @@ struct ttm_buffer_object {
>>>>>>>          * either of these locks held.
>>>>>>>          */
>>>>>>>     -    uint64_t offset; /* GPU address space is independent of CPU
>>>>>>> word size */
>>>>>>> -
>>>>>>>         struct sg_table *sg;
>>>>>>>     };
>>>>>>>     diff --git a/include/drm/ttm/ttm_bo_driver.h
>>>>>>> b/include/drm/ttm/ttm_bo_driver.h
>>>>>>> index c9e0fd09f4b2..c8ce6c181abe 100644
>>>>>>> --- a/include/drm/ttm/ttm_bo_driver.h
>>>>>>> +++ b/include/drm/ttm/ttm_bo_driver.h
>>>>>>> @@ -177,7 +177,6 @@ struct ttm_mem_type_manager {
>>>>>>>         bool has_type;
>>>>>>>         bool use_type;
>>>>>>>         uint32_t flags;
>>>>>>> -    uint64_t gpu_offset; /* GPU address space is independent of CPU
>>>>>>> word size */
>>>>>>>         uint64_t size;
>>>>>>>         uint32_t available_caching;
>>>>>>>         uint32_t default_caching;
>>>>>>>
>>>> _______________________________________________
>>>> dri-devel mailing list
>>>> dri-devel at lists.freedesktop.org
>>>> https://lists.freedesktop.org/mailman/listinfo/dri-devel
>> _______________________________________________
>> dri-devel mailing list
>> dri-devel at lists.freedesktop.org
>> https://lists.freedesktop.org/mailman/listinfo/dri-devel
>
> _______________________________________________
> amd-gfx mailing list
> amd-gfx at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/amd-gfx

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/amd-gfx/attachments/20200218/14952a69/attachment.htm>

More information about the amd-gfx mailing list