[PATCH 2/6] drm/ttm: Store the bo_kmap_type in struct iosys_map

Christian König christian.koenig at amd.com
Mon Jun 17 12:56:17 UTC 2024


Hi,

Am 17.06.24 um 14:32 schrieb Thomas Zimmermann:
> Hi
>
> Am 14.06.24 um 16:31 schrieb Christian König:
>> Am 14.06.24 um 15:21 schrieb Thomas Zimmermann:
>>> For each instances of struct iosys_map set up by ttm_bo_vmap(), store
>>> the type of allocation in the instance. Use this information to unmap
>>> the memory in ttm_bo_vunmap(). This change simplifies the unmap code
>>> and puts the complicated logic entirely into the map code.
>>
>> I'm not sure that's a good idea.
>>
>> The mapping information should already be available in the resource 
>> and storing it in the iosys_map structures duplicates that information.
>>
>> So we might run into the issue that the resource has changed and so 
>> we need a different approach now, but the iosys_map will say that we 
>> should unmap things for example.
>
> Patches 1 and 2 are only here to make patch 4 (add the kmap special 
> case) work. How can I distinguish between vmap'ed and kmap'ed memory? 
> It's not clear to me, whether there is a benefit from patch 4. The xe 
> driver makes it sound like that, but the rest of the kernel appears to 
> disagree.

Yeah, exactly that's the point.

The last time we talked about that we came to the conclusion that the 
kmap approach of mapping only a single page or range of pages isn't that 
useful in general.

The only use case where you actually need this is the 
ttm_bo_vm_access_kmap() helper and that is static and internal to TTM.

So what exactly is the use case xe tries to handle here?

Regards,
Christian.

>
> Best regards
> Thomas
>
>>
>> Regards,
>> Christian.
>>
>>>
>>> Signed-off-by: Thomas Zimmermann <tzimmermann at suse.de>
>>> ---
>>>   drivers/gpu/drm/ttm/ttm_bo_util.c | 46 
>>> +++++++++++++++++++++----------
>>>   1 file changed, 31 insertions(+), 15 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c 
>>> b/drivers/gpu/drm/ttm/ttm_bo_util.c
>>> index 0b3f4267130c4..a9df0deff2deb 100644
>>> --- a/drivers/gpu/drm/ttm/ttm_bo_util.c
>>> +++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
>>> @@ -36,6 +36,7 @@
>>>   #include <drm/ttm/ttm_tt.h>
>>>     #include <drm/drm_cache.h>
>>> +#include <drm/drm_device.h>
>>>     struct ttm_transfer_obj {
>>>       struct ttm_buffer_object base;
>>> @@ -479,24 +480,29 @@ int ttm_bo_vmap(struct ttm_buffer_object *bo, 
>>> struct iosys_map *map)
>>>         if (mem->bus.is_iomem) {
>>>           void __iomem *vaddr_iomem;
>>> +        u16 alloc_flags;
>>>   -        if (mem->bus.addr)
>>> +        if (mem->bus.addr) {
>>>               vaddr_iomem = (void __iomem *)mem->bus.addr;
>>> -        else if (mem->bus.caching == ttm_write_combined)
>>> -            vaddr_iomem = ioremap_wc(mem->bus.offset,
>>> -                         bo->base.size);
>>> +            alloc_flags = ttm_bo_map_premapped;
>>> +        } else if (mem->bus.caching == ttm_write_combined) {
>>> +            vaddr_iomem = ioremap_wc(mem->bus.offset, bo->base.size);
>>> +            alloc_flags = ttm_bo_map_iomap;
>>>   #ifdef CONFIG_X86
>>> -        else if (mem->bus.caching == ttm_cached)
>>> -            vaddr_iomem = ioremap_cache(mem->bus.offset,
>>> -                          bo->base.size);
>>> +        } else if (mem->bus.caching == ttm_cached) {
>>> +            vaddr_iomem = ioremap_cache(mem->bus.offset, 
>>> bo->base.size);
>>> +            alloc_flags = ttm_bo_map_iomap;
>>>   #endif
>>> -        else
>>> +        } else {
>>>               vaddr_iomem = ioremap(mem->bus.offset, bo->base.size);
>>> +            alloc_flags = ttm_bo_map_iomap;
>>> +        }
>>>             if (!vaddr_iomem)
>>>               return -ENOMEM;
>>>             iosys_map_set_vaddr_iomem(map, vaddr_iomem);
>>> +        map->alloc_flags = alloc_flags;
>>>         } else {
>>>           struct ttm_operation_ctx ctx = {
>>> @@ -506,6 +512,7 @@ int ttm_bo_vmap(struct ttm_buffer_object *bo, 
>>> struct iosys_map *map)
>>>           struct ttm_tt *ttm = bo->ttm;
>>>           pgprot_t prot;
>>>           void *vaddr;
>>> +        u16 alloc_flags;
>>>             ret = ttm_tt_populate(bo->bdev, ttm, &ctx);
>>>           if (ret)
>>> @@ -519,8 +526,10 @@ int ttm_bo_vmap(struct ttm_buffer_object *bo, 
>>> struct iosys_map *map)
>>>           vaddr = vmap(ttm->pages, ttm->num_pages, 0, prot);
>>>           if (!vaddr)
>>>               return -ENOMEM;
>>> +        alloc_flags = ttm_bo_map_vmap;
>>>             iosys_map_set_vaddr(map, vaddr);
>>> +        map->alloc_flags = alloc_flags;
>>>       }
>>>         return 0;
>>> @@ -537,20 +546,27 @@ EXPORT_SYMBOL(ttm_bo_vmap);
>>>    */
>>>   void ttm_bo_vunmap(struct ttm_buffer_object *bo, struct iosys_map 
>>> *map)
>>>   {
>>> -    struct ttm_resource *mem = bo->resource;
>>> -
>>>       dma_resv_assert_held(bo->base.resv);
>>>         if (iosys_map_is_null(map))
>>>           return;
>>>   -    if (!map->is_iomem)
>>> -        vunmap(map->vaddr);
>>> -    else if (!mem->bus.addr)
>>> +    switch (map->alloc_flags) {
>>> +    case ttm_bo_map_iomap:
>>>           iounmap(map->vaddr_iomem);
>>> -    iosys_map_clear(map);
>>> -
>>> +        break;
>>> +    case ttm_bo_map_vmap:
>>> +        vunmap(map->vaddr);
>>> +        break;
>>> +    case ttm_bo_map_premapped:
>>> +        break;
>>> +    default:
>>> +        drm_err(bo->base.dev, "Unsupported alloc_flags 0x%x\n", 
>>> map->alloc_flags);
>>> +        return;
>>> +    }
>>>       ttm_mem_io_free(bo->bdev, bo->resource);
>>> +
>>> +    iosys_map_clear(map);
>>>   }
>>>   EXPORT_SYMBOL(ttm_bo_vunmap);
>>
>



More information about the dri-devel mailing list