[Intel-gfx] [PATCH 2/5] drm/i915/ttm: Adjust gem flags and caching settings after a move
Thomas Hellström
thomas.hellstrom at linux.intel.com
Thu Jun 3 13:40:38 UTC 2021
Thanks for reviewing, Matthew.
On 6/3/21 12:35 PM, Matthew Auld wrote:
> On Wed, 2 Jun 2021 at 18:08, Thomas Hellström
> <thomas.hellstrom at linux.intel.com> wrote:
>> After a TTM move we need to update the i915 gem flags and caching
>> settings to reflect the new placement.
>> Also introduce gpu_binds_iomem() and cpu_maps_iomem() to clean up the
>> various ways we previously used to detect this.
>> Finally, initialize the TTM object reserved to be able to update
>> flags and caching before anyone else gets hold of the object.
>>
>> Signed-off-by: Thomas Hellström <thomas.hellstrom at linux.intel.com>
>> ---
>> drivers/gpu/drm/i915/gem/i915_gem_ttm.c | 112 +++++++++++++++++++-----
>> 1 file changed, 90 insertions(+), 22 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/gem/i915_gem_ttm.c b/drivers/gpu/drm/i915/gem/i915_gem_ttm.c
>> index ae12a2be11a2..c73c51755c20 100644
>> --- a/drivers/gpu/drm/i915/gem/i915_gem_ttm.c
>> +++ b/drivers/gpu/drm/i915/gem/i915_gem_ttm.c
>> @@ -70,6 +70,17 @@ static struct ttm_placement i915_sys_placement = {
>> .busy_placement = &lmem0_sys_placement_flags[1],
>> };
>>
>> +static bool gpu_binds_iomem(struct ttm_resource *mem)
> What does gpu_binds mean here? It's mapped though the GTT?
gpu_bind_XX would here translate to "make xx visible to the gpu", We
could of course use the term "gpu_maps" if you think that fits better.
>
>> +{
>> + return (mem->mem_type != TTM_PL_SYSTEM);
>> +}
>> +
>> +static bool cpu_maps_iomem(struct ttm_resource *mem)
>> +{
>> + /* Once / if we support GGTT, this is also false for cached ttm_tts */
>> + return (mem->mem_type != TTM_PL_SYSTEM);
>> +}
> Can drop the extra brackets.
>
>> +
>> static void i915_ttm_adjust_lru(struct drm_i915_gem_object *obj);
>>
>> static struct ttm_tt *i915_ttm_tt_create(struct ttm_buffer_object *bo,
>> @@ -175,6 +186,41 @@ static void i915_ttm_free_cached_io_st(struct drm_i915_gem_object *obj)
>> obj->ttm.cached_io_st = NULL;
>> }
>>
>> +static void
>> +i915_ttm_adjust_domains_after_cpu_move(struct drm_i915_gem_object *obj)
>> +{
>> + struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
>> +
>> + if (cpu_maps_iomem(&bo->mem) || bo->ttm->caching != ttm_cached) {
>> + obj->write_domain = I915_GEM_DOMAIN_WC;
>> + obj->read_domains = I915_GEM_DOMAIN_WC;
>> + } else {
>> + obj->write_domain = I915_GEM_DOMAIN_CPU;
>> + obj->read_domains = I915_GEM_DOMAIN_CPU;
>> + }
>> +}
>> +
>> +static void i915_ttm_adjust_gem_after_move(struct drm_i915_gem_object *obj)
>> +{
>> + struct drm_i915_private *i915 = to_i915(obj->base.dev);
>> + struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
>> + unsigned int cache_level;
>> +
>> + obj->mem_flags &= ~(I915_BO_FLAG_STRUCT_PAGE | I915_BO_FLAG_IOMEM);
>> +
>> + obj->mem_flags |= cpu_maps_iomem(&bo->mem) ? I915_BO_FLAG_IOMEM :
>> + I915_BO_FLAG_STRUCT_PAGE;
>> +
>> + if ((HAS_LLC(i915) || HAS_SNOOP(i915)) && !gpu_binds_iomem(&bo->mem) &&
> I think all modern hw has support for snooping or the fast shared LLC.
> Are we expecting to hit this path for non-dgpu?
Not initially, no. Possibly later.
> Also by default we
> will choose ttm_cached at least for platforms like dg1?
Not for evicted LMEM. They still have ttm_write_combine. User-space can
still access those through mmap() while evicted.
> Also pin_map()
> is still a separate interface at this point?
Yes, I had a comment on that in your caching patch that I reviewed, we
need to make sure even for pin_map that we only map using the single
supported caching attributes for the BO.
>> + bo->ttm->caching == ttm_cached) {
>> + cache_level = I915_CACHE_LLC;
>> + } else {
>> + cache_level = I915_CACHE_NONE;
>> + }
>> +
>> + i915_gem_object_set_cache_coherency(obj, cache_level);
>> +}
>> +
>> static void i915_ttm_purge(struct drm_i915_gem_object *obj)
>> {
>> struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
>> @@ -190,8 +236,10 @@ static void i915_ttm_purge(struct drm_i915_gem_object *obj)
>>
>> /* TTM's purge interface. Note that we might be reentering. */
>> ret = ttm_bo_validate(bo, &place, &ctx);
>> -
>> if (!ret) {
>> + obj->write_domain = 0;
>> + obj->read_domains = 0;
>> + i915_ttm_adjust_gem_after_move(obj);
>> i915_ttm_free_cached_io_st(obj);
>> obj->mm.madv = __I915_MADV_PURGED;
>> }
>> @@ -273,12 +321,15 @@ i915_ttm_resource_get_st(struct drm_i915_gem_object *obj,
>> struct ttm_resource *res)
>> {
>> struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
>> - struct ttm_resource_manager *man =
>> - ttm_manager_type(bo->bdev, res->mem_type);
>>
>> - if (man->use_tt)
>> + if (!gpu_binds_iomem(res))
>> return i915_ttm_tt_get_st(bo->ttm);
>>
>> + /*
>> + * If CPU mapping differs, we need to add the ttm_tt pages to
>> + * the resulting st. Might make sense for GGTT.
>> + */
>> + GEM_WARN_ON(!cpu_maps_iomem(res));
>> return intel_region_ttm_node_to_st(obj->mm.region, res->mm_node);
>> }
>>
>> @@ -290,8 +341,6 @@ static int i915_ttm_move(struct ttm_buffer_object *bo, bool evict,
>> struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
>> struct ttm_resource_manager *dst_man =
>> ttm_manager_type(bo->bdev, dst_mem->mem_type);
>> - struct ttm_resource_manager *src_man =
>> - ttm_manager_type(bo->bdev, bo->mem.mem_type);
>> struct intel_memory_region *dst_reg, *src_reg;
>> union {
>> struct ttm_kmap_iter_tt tt;
>> @@ -332,34 +381,36 @@ static int i915_ttm_move(struct ttm_buffer_object *bo, bool evict,
>> if (IS_ERR(dst_st))
>> return PTR_ERR(dst_st);
>>
>> - /* If we start mapping GGTT, we can no longer use man::use_tt here. */
>> - dst_iter = dst_man->use_tt ?
>> + dst_iter = !cpu_maps_iomem(dst_mem) ?
>> ttm_kmap_iter_tt_init(&_dst_iter.tt, bo->ttm) :
>> ttm_kmap_iter_iomap_init(&_dst_iter.io, &dst_reg->iomap,
>> dst_st, dst_reg->region.start);
>>
>> - src_iter = src_man->use_tt ?
>> + src_iter = !cpu_maps_iomem(&bo->mem) ?
>> ttm_kmap_iter_tt_init(&_src_iter.tt, bo->ttm) :
>> ttm_kmap_iter_iomap_init(&_src_iter.io, &src_reg->iomap,
>> obj->ttm.cached_io_st,
>> src_reg->region.start);
>>
>> ttm_move_memcpy(bo, dst_mem->num_pages, dst_iter, src_iter);
>> + /* Below dst_mem becomes bo->mem. */
>> ttm_bo_move_sync_cleanup(bo, dst_mem);
>> + i915_ttm_adjust_domains_after_cpu_move(obj);
>> i915_ttm_free_cached_io_st(obj);
>>
>> - if (!dst_man->use_tt) {
>> + if (gpu_binds_iomem(dst_mem) || cpu_maps_iomem(dst_mem)) {
>> obj->ttm.cached_io_st = dst_st;
>> obj->ttm.get_io_page.sg_pos = dst_st->sgl;
>> obj->ttm.get_io_page.sg_idx = 0;
>> }
>>
>> + i915_ttm_adjust_gem_after_move(obj);
>> return 0;
>> }
>>
>> static int i915_ttm_io_mem_reserve(struct ttm_device *bdev, struct ttm_resource *mem)
>> {
>> - if (mem->mem_type < I915_PL_LMEM0)
>> + if (!cpu_maps_iomem(mem))
>> return 0;
>>
>> mem->bus.caching = ttm_write_combined;
>> @@ -421,6 +472,16 @@ static int i915_ttm_get_pages(struct drm_i915_gem_object *obj)
>> if (ret)
>> return ret == -ENOSPC ? -ENXIO : ret;
>>
>> + i915_ttm_adjust_lru(obj);
>> + if (bo->ttm && !ttm_tt_is_populated(bo->ttm)) {
>> + ret = ttm_tt_populate(bo->bdev, bo->ttm, &ctx);
>> + if (ret)
>> + return ret;
>> +
>> + i915_ttm_adjust_domains_after_cpu_move(obj);
>> + i915_ttm_adjust_gem_after_move(obj);
>> + }
>> +
>> /* Object either has a page vector or is an iomem object */
>> st = bo->ttm ? i915_ttm_tt_get_st(bo->ttm) : obj->ttm.cached_io_st;
>> if (IS_ERR(st))
>> @@ -428,8 +489,6 @@ static int i915_ttm_get_pages(struct drm_i915_gem_object *obj)
>>
>> __i915_gem_object_set_pages(obj, st, i915_sg_dma_sizes(st->sgl));
>>
>> - i915_ttm_adjust_lru(obj);
>> -
>> return ret;
>> }
>>
>> @@ -563,6 +622,7 @@ static u64 i915_ttm_mmap_offset(struct drm_i915_gem_object *obj)
>>
>> const struct drm_i915_gem_object_ops i915_gem_ttm_obj_ops = {
>> .name = "i915_gem_object_ttm",
>> + .flags = I915_GEM_OBJECT_IS_SHRINKABLE,
>>
>> .get_pages = i915_ttm_get_pages,
>> .put_pages = i915_ttm_put_pages,
>> @@ -599,6 +659,10 @@ int __i915_gem_ttm_object_init(struct intel_memory_region *mem,
>> {
>> static struct lock_class_key lock_class;
>> struct drm_i915_private *i915 = mem->i915;
>> + struct ttm_operation_ctx ctx = {
>> + .interruptible = true,
>> + .no_wait_gpu = false,
>> + };
>> enum ttm_bo_type bo_type;
>> size_t alignment = 0;
>> int ret;
>> @@ -618,15 +682,14 @@ int __i915_gem_ttm_object_init(struct intel_memory_region *mem,
>> i915_gem_object_init(obj, &i915_gem_ttm_obj_ops, &lock_class, flags);
>> i915_gem_object_init_memory_region(obj, mem);
>> i915_gem_object_make_unshrinkable(obj);
>> - obj->read_domains = I915_GEM_DOMAIN_WC | I915_GEM_DOMAIN_GTT;
>> - obj->mem_flags |= I915_BO_FLAG_IOMEM;
>> - i915_gem_object_set_cache_coherency(obj, I915_CACHE_NONE);
>> INIT_RADIX_TREE(&obj->ttm.get_io_page.radix, GFP_KERNEL | __GFP_NOWARN);
>> mutex_init(&obj->ttm.get_io_page.lock);
>>
>> bo_type = (obj->flags & I915_BO_ALLOC_USER) ? ttm_bo_type_device :
>> ttm_bo_type_kernel;
>>
>> + obj->base.vma_node.driver_private = i915_gem_to_ttm(obj);
>> +
>> /*
>> * If this function fails, it will call the destructor, but
>> * our caller still owns the object. So no freeing in the
>> @@ -634,14 +697,19 @@ int __i915_gem_ttm_object_init(struct intel_memory_region *mem,
>> * Similarly, in delayed_destroy, we can't call ttm_bo_put()
>> * until successful initialization.
>> */
>> - obj->base.vma_node.driver_private = i915_gem_to_ttm(obj);
>> - ret = ttm_bo_init(&i915->bdev, i915_gem_to_ttm(obj), size,
>> - bo_type, &i915_sys_placement, alignment,
>> - true, NULL, NULL, i915_ttm_bo_destroy);
>> + ret = ttm_bo_init_reserved(&i915->bdev, i915_gem_to_ttm(obj), size,
>> + bo_type, &i915_sys_placement, alignment,
>> + &ctx, NULL, NULL, i915_ttm_bo_destroy);
>> +
>> + if (ret)
>> + goto out;
>>
>> - if (!ret)
>> - obj->ttm.created = true;
>> + obj->ttm.created = true;
>> + i915_ttm_adjust_domains_after_cpu_move(obj);
>> + i915_ttm_adjust_gem_after_move(obj);
>> + i915_gem_object_unlock(obj);
>>
>> +out:
>> /* i915 wants -ENXIO when out of memory region space. */
>> return (ret == -ENOSPC) ? -ENXIO : ret;
>> }
>> --
>> 2.31.1
>>
>> _______________________________________________
>> Intel-gfx mailing list
>> Intel-gfx at lists.freedesktop.org
>> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
More information about the Intel-gfx
mailing list