| Message ID | 20210115194101.1037430-1-imre.deak@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | None | expand |
Quoting Imre Deak (2021-01-15 19:41:00) > Add a simple helper to read data with the CPU from the page of a GEM > object. Do the read either via a kmap if the object has struct pages > or an iomap otherwise. This is needed by the next patch, reading a u64 > value from the object (w/o requiring the obj to be mapped to the GPU). > > Suggested by Chris. > > v2 (Chris): > - Sanitize the type and order of func params. > - Avoid consts requiring too many casts. > - Use BUG_ON instead of WARN_ON, simplify the conditions. > - Fix __iomem sparse errors. > - Leave locking/syncing/pinning up to the caller, require only that the > caller has pinned the object pages. > - Check for iomem backing store before reading via an iomap. > > Cc: Chris Wilson <chris@chris-wilson.co.uk> > Signed-off-by: Imre Deak <imre.deak@intel.com> > --- > drivers/gpu/drm/i915/gem/i915_gem_object.c | 64 ++++++++++++++++++++++ > drivers/gpu/drm/i915/gem/i915_gem_object.h | 8 +++ > 2 files changed, 72 insertions(+) > > diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.c b/drivers/gpu/drm/i915/gem/i915_gem_object.c > index 00d24000b5e8..67956a5f5fe3 100644 > --- a/drivers/gpu/drm/i915/gem/i915_gem_object.c > +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.c > @@ -32,6 +32,7 @@ > #include "i915_gem_mman.h" > #include "i915_gem_object.h" > #include "i915_globals.h" > +#include "i915_memcpy.h" > #include "i915_trace.h" > > static struct i915_global_object { > @@ -383,6 +384,69 @@ void __i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj, > } > } > > +static void > +i915_gem_object_read_from_page_kmap(struct drm_i915_gem_object *obj, u64 offset, void *dst, int size) > +{ > + void *src_map; > + void *src_ptr; > + > + src_map = kmap_atomic(i915_gem_object_get_page(obj, offset >> PAGE_SHIFT)); > + > + src_ptr = src_map + offset_in_page(offset); > + if (!(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_READ)) > + drm_clflush_virt_range(src_ptr, size); > + memcpy(dst, src_ptr, size); > + > + kunmap_atomic(src_map); > +} > + > +static void > +i915_gem_object_read_from_page_iomap(struct drm_i915_gem_object *obj, u64 offset, void *dst, int size) > +{ > + void __iomem *src_map; > + void __iomem *src_ptr; > + > + src_map = io_mapping_map_wc(&obj->mm.region->iomap, > + i915_gem_object_get_dma_address(obj, offset >> PAGE_SHIFT), I've been corrected in that one needs to use dma_addr_t dma = i915_gem_object_get_dma_address(obj, offset >> PAGE_SHIFT); src_map = io_mapping_map_wc(&obj->mm.region->iomap, dma - obj->mm.region->region.start, > + PAGE_SIZE); > + > + src_ptr = src_map + offset_in_page(offset); > + if (!i915_memcpy_from_wc(dst, (void __force *)src_ptr, size)) > + memcpy_fromio(dst, src_ptr, size); > + > + io_mapping_unmap(src_map); > +} > + > +/** > + * i915_gem_object_read_from_page - read data from the page of a GEM object > + * @obj: GEM object to read from > + * @offset: offset within the object > + * @dst: buffer to store the read data > + * @size: size to read > + * > + * Reads data from @obj at the specified offset. The requested region to read > + * from can't cross a page boundary. The caller must ensure that @obj pages > + * are pinned and that @obj is synced wrt. any related writes. > + * > + * Returns 0 on sucess or -ENODEV if the type of @obj's backing store is > + * unsupported. > + */ > +int i915_gem_object_read_from_page(struct drm_i915_gem_object *obj, u64 offset, void *dst, int size) > +{ > + GEM_BUG_ON(offset >= obj->base.size); > + GEM_BUG_ON(offset_in_page(offset) > PAGE_SIZE - size); > + GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj)); > + > + if (i915_gem_object_has_struct_page(obj)) > + i915_gem_object_read_from_page_kmap(obj, offset, dst, size); > + else if (i915_gem_object_has_iomem(obj)) > + i915_gem_object_read_from_page_iomap(obj, offset, dst, size); > + else > + return -ENODEV; Otherwise, that looks to be as simple as possible (offloading the setup to the caller where it is already done), so Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> with the dma offset before Matthew corrects me, again. -Chris
On Wed, Jan 20, 2021 at 12:02:49PM +0000, Chris Wilson wrote: > Quoting Imre Deak (2021-01-15 19:41:00) > > Add a simple helper to read data with the CPU from the page of a GEM > > object. Do the read either via a kmap if the object has struct pages > > or an iomap otherwise. This is needed by the next patch, reading a u64 > > value from the object (w/o requiring the obj to be mapped to the GPU). > > > > Suggested by Chris. > > > > v2 (Chris): > > - Sanitize the type and order of func params. > > - Avoid consts requiring too many casts. > > - Use BUG_ON instead of WARN_ON, simplify the conditions. > > - Fix __iomem sparse errors. > > - Leave locking/syncing/pinning up to the caller, require only that the > > caller has pinned the object pages. > > - Check for iomem backing store before reading via an iomap. > > > > Cc: Chris Wilson <chris@chris-wilson.co.uk> > > Signed-off-by: Imre Deak <imre.deak@intel.com> > > --- > > drivers/gpu/drm/i915/gem/i915_gem_object.c | 64 ++++++++++++++++++++++ > > drivers/gpu/drm/i915/gem/i915_gem_object.h | 8 +++ > > 2 files changed, 72 insertions(+) > > > > diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.c b/drivers/gpu/drm/i915/gem/i915_gem_object.c > > index 00d24000b5e8..67956a5f5fe3 100644 > > --- a/drivers/gpu/drm/i915/gem/i915_gem_object.c > > +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.c > > @@ -32,6 +32,7 @@ > > #include "i915_gem_mman.h" > > #include "i915_gem_object.h" > > #include "i915_globals.h" > > +#include "i915_memcpy.h" > > #include "i915_trace.h" > > > > static struct i915_global_object { > > @@ -383,6 +384,69 @@ void __i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj, > > } > > } > > > > +static void > > +i915_gem_object_read_from_page_kmap(struct drm_i915_gem_object *obj, u64 offset, void *dst, int size) > > +{ > > + void *src_map; > > + void *src_ptr; > > + > > + src_map = kmap_atomic(i915_gem_object_get_page(obj, offset >> PAGE_SHIFT)); > > + > > + src_ptr = src_map + offset_in_page(offset); > > + if (!(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_READ)) > > + drm_clflush_virt_range(src_ptr, size); > > + memcpy(dst, src_ptr, size); > > + > > + kunmap_atomic(src_map); > > +} > > + > > +static void > > +i915_gem_object_read_from_page_iomap(struct drm_i915_gem_object *obj, u64 offset, void *dst, int size) > > +{ > > + void __iomem *src_map; > > + void __iomem *src_ptr; > > + > > + src_map = io_mapping_map_wc(&obj->mm.region->iomap, > > + i915_gem_object_get_dma_address(obj, offset >> PAGE_SHIFT), > > I've been corrected in that one needs to use > > dma_addr_t dma = > i915_gem_object_get_dma_address(obj, offset >> PAGE_SHIFT); > > src_map = io_mapping_map_wc(&obj->mm.region->iomap, > dma - obj->mm.region->region.start, Thanks, will fix this. It happened to work on DG1, where region.start is always 0. > > > + PAGE_SIZE); > > + > > + src_ptr = src_map + offset_in_page(offset); > > + if (!i915_memcpy_from_wc(dst, (void __force *)src_ptr, size)) > > + memcpy_fromio(dst, src_ptr, size); > > + > > + io_mapping_unmap(src_map); > > +} > > + > > +/** > > + * i915_gem_object_read_from_page - read data from the page of a GEM object > > + * @obj: GEM object to read from > > + * @offset: offset within the object > > + * @dst: buffer to store the read data > > + * @size: size to read > > + * > > + * Reads data from @obj at the specified offset. The requested region to read > > + * from can't cross a page boundary. The caller must ensure that @obj pages > > + * are pinned and that @obj is synced wrt. any related writes. > > + * > > + * Returns 0 on sucess or -ENODEV if the type of @obj's backing store is > > + * unsupported. > > + */ > > +int i915_gem_object_read_from_page(struct drm_i915_gem_object *obj, u64 offset, void *dst, int size) > > +{ > > + GEM_BUG_ON(offset >= obj->base.size); > > + GEM_BUG_ON(offset_in_page(offset) > PAGE_SIZE - size); > > + GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj)); > > + > > + if (i915_gem_object_has_struct_page(obj)) > > + i915_gem_object_read_from_page_kmap(obj, offset, dst, size); > > + else if (i915_gem_object_has_iomem(obj)) > > + i915_gem_object_read_from_page_iomap(obj, offset, dst, size); > > + else > > + return -ENODEV; > > Otherwise, that looks to be as simple as possible (offloading the setup > to the caller where it is already done), so > > Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> > > with the dma offset before Matthew corrects me, again. > -Chris
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.c b/drivers/gpu/drm/i915/gem/i915_gem_object.c index 00d24000b5e8..67956a5f5fe3 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.c @@ -32,6 +32,7 @@ #include "i915_gem_mman.h" #include "i915_gem_object.h" #include "i915_globals.h" +#include "i915_memcpy.h" #include "i915_trace.h" static struct i915_global_object { @@ -383,6 +384,69 @@ void __i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj, } } +static void +i915_gem_object_read_from_page_kmap(struct drm_i915_gem_object *obj, u64 offset, void *dst, int size) +{ + void *src_map; + void *src_ptr; + + src_map = kmap_atomic(i915_gem_object_get_page(obj, offset >> PAGE_SHIFT)); + + src_ptr = src_map + offset_in_page(offset); + if (!(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_READ)) + drm_clflush_virt_range(src_ptr, size); + memcpy(dst, src_ptr, size); + + kunmap_atomic(src_map); +} + +static void +i915_gem_object_read_from_page_iomap(struct drm_i915_gem_object *obj, u64 offset, void *dst, int size) +{ + void __iomem *src_map; + void __iomem *src_ptr; + + src_map = io_mapping_map_wc(&obj->mm.region->iomap, + i915_gem_object_get_dma_address(obj, offset >> PAGE_SHIFT), + PAGE_SIZE); + + src_ptr = src_map + offset_in_page(offset); + if (!i915_memcpy_from_wc(dst, (void __force *)src_ptr, size)) + memcpy_fromio(dst, src_ptr, size); + + io_mapping_unmap(src_map); +} + +/** + * i915_gem_object_read_from_page - read data from the page of a GEM object + * @obj: GEM object to read from + * @offset: offset within the object + * @dst: buffer to store the read data + * @size: size to read + * + * Reads data from @obj at the specified offset. The requested region to read + * from can't cross a page boundary. The caller must ensure that @obj pages + * are pinned and that @obj is synced wrt. any related writes. + * + * Returns 0 on sucess or -ENODEV if the type of @obj's backing store is + * unsupported. + */ +int i915_gem_object_read_from_page(struct drm_i915_gem_object *obj, u64 offset, void *dst, int size) +{ + GEM_BUG_ON(offset >= obj->base.size); + GEM_BUG_ON(offset_in_page(offset) > PAGE_SIZE - size); + GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj)); + + if (i915_gem_object_has_struct_page(obj)) + i915_gem_object_read_from_page_kmap(obj, offset, dst, size); + else if (i915_gem_object_has_iomem(obj)) + i915_gem_object_read_from_page_iomap(obj, offset, dst, size); + else + return -ENODEV; + + return 0; +} + void i915_gem_init__objects(struct drm_i915_private *i915) { INIT_WORK(&i915->mm.free_work, __i915_gem_free_work); diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.h b/drivers/gpu/drm/i915/gem/i915_gem_object.h index be14486f63a7..ae83737f1d48 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object.h +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.h @@ -200,6 +200,12 @@ i915_gem_object_has_struct_page(const struct drm_i915_gem_object *obj) return i915_gem_object_type_has(obj, I915_GEM_OBJECT_HAS_STRUCT_PAGE); } +static inline bool +i915_gem_object_has_iomem(const struct drm_i915_gem_object *obj) +{ + return i915_gem_object_type_has(obj, I915_GEM_OBJECT_HAS_IOMEM); +} + static inline bool i915_gem_object_is_shrinkable(const struct drm_i915_gem_object *obj) { @@ -540,4 +546,6 @@ i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj, __i915_gem_object_invalidate_frontbuffer(obj, origin); } +int i915_gem_object_read_from_page(struct drm_i915_gem_object *obj, u64 offset, void *dst, int size); + #endif
Add a simple helper to read data with the CPU from the page of a GEM object. Do the read either via a kmap if the object has struct pages or an iomap otherwise. This is needed by the next patch, reading a u64 value from the object (w/o requiring the obj to be mapped to the GPU). Suggested by Chris. v2 (Chris): - Sanitize the type and order of func params. - Avoid consts requiring too many casts. - Use BUG_ON instead of WARN_ON, simplify the conditions. - Fix __iomem sparse errors. - Leave locking/syncing/pinning up to the caller, require only that the caller has pinned the object pages. - Check for iomem backing store before reading via an iomap. Cc: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Imre Deak <imre.deak@intel.com> --- drivers/gpu/drm/i915/gem/i915_gem_object.c | 64 ++++++++++++++++++++++ drivers/gpu/drm/i915/gem/i915_gem_object.h | 8 +++ 2 files changed, 72 insertions(+)