Return-Path: Received: from mga14.intel.com (mga14.intel.com [192.55.52.115]) by gabe.freedesktop.org (Postfix) with ESMTPS id 9853C89ACD for ; Mon, 9 Aug 2021 06:09:23 +0000 (UTC) Date: Mon, 9 Aug 2021 08:09:18 +0200 From: Zbigniew =?utf-8?Q?Kempczy=C5=84ski?= Message-ID: <20210809060918.GA5318@zkempczy-mobl2> References: <20210806134145.24634-1-zbigniew.kempczynski@intel.com> <20210806134145.24634-6-zbigniew.kempczynski@intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <20210806134145.24634-6-zbigniew.kempczynski@intel.com> Subject: Re: [igt-dev] [PATCH i-g-t v4 05/56] lib/intel_batchbuffer: Add allocator support in blitter fast copy List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: igt-dev-bounces@lists.freedesktop.org Sender: "igt-dev" To: igt-dev@lists.freedesktop.org Cc: Bhanuprakash Modem , Petri Latvala , Ashutosh Dixit , Chris Wilson List-ID: On Fri, Aug 06, 2021 at 03:40:54PM +0200, Zbigniew Kempczyński wrote: > From: Bhanuprakash Modem > > For newer gens kernel will reject relocations by returning -EINVAL > so we should support allocator and acquire offsets for blit. > > Signed-off-by: Bhanuprakash Modem > Cc: Zbigniew Kempczyński > Cc: Petri Latvala > Cc: Ashutosh Dixit > Cc: Chris Wilson LGTM: Reviewed-by: Zbigniew Kempczyński -- Zbigniew > --- > lib/igt_fb.c | 17 ++++++++++++- > lib/intel_batchbuffer.c | 55 ++++++++++++++++++++++++++++++----------- > lib/intel_batchbuffer.h | 6 ++++- > 3 files changed, 62 insertions(+), 16 deletions(-) > > diff --git a/lib/igt_fb.c b/lib/igt_fb.c > index 75ab217b8..1bb32cd8a 100644 > --- a/lib/igt_fb.c > +++ b/lib/igt_fb.c > @@ -2417,6 +2417,8 @@ static void blitcopy(const struct igt_fb *dst_fb, > const struct igt_fb *src_fb) > { > uint32_t src_tiling, dst_tiling; > + uint32_t ctx = 0; > + uint64_t ahnd = 0; > > igt_assert_eq(dst_fb->fd, src_fb->fd); > igt_assert_eq(dst_fb->num_planes, src_fb->num_planes); > @@ -2424,6 +2426,12 @@ static void blitcopy(const struct igt_fb *dst_fb, > src_tiling = igt_fb_mod_to_tiling(src_fb->modifier); > dst_tiling = igt_fb_mod_to_tiling(dst_fb->modifier); > > + if (!gem_has_relocations(dst_fb->fd)) { > + igt_require(gem_has_contexts(dst_fb->fd)); > + ctx = gem_context_create(dst_fb->fd); > + ahnd = get_reloc_ahnd(dst_fb->fd, ctx); > + } > + > for (int i = 0; i < dst_fb->num_planes; i++) { > igt_assert_eq(dst_fb->plane_bpp[i], src_fb->plane_bpp[i]); > igt_assert_eq(dst_fb->plane_width[i], src_fb->plane_width[i]); > @@ -2435,11 +2443,13 @@ static void blitcopy(const struct igt_fb *dst_fb, > */ > if (fast_blit_ok(src_fb) && fast_blit_ok(dst_fb)) { > igt_blitter_fast_copy__raw(dst_fb->fd, > + ahnd, ctx, > src_fb->gem_handle, > src_fb->offsets[i], > src_fb->strides[i], > src_tiling, > 0, 0, /* src_x, src_y */ > + src_fb->size, > dst_fb->plane_width[i], > dst_fb->plane_height[i], > dst_fb->plane_bpp[i], > @@ -2447,7 +2457,8 @@ static void blitcopy(const struct igt_fb *dst_fb, > dst_fb->offsets[i], > dst_fb->strides[i], > dst_tiling, > - 0, 0 /* dst_x, dst_y */); > + 0, 0 /* dst_x, dst_y */, > + dst_fb->size); > } else { > igt_blitter_src_copy(dst_fb->fd, > src_fb->gem_handle, > @@ -2465,6 +2476,10 @@ static void blitcopy(const struct igt_fb *dst_fb, > 0, 0 /* dst_x, dst_y */); > } > } > + > + if (ctx) > + gem_context_destroy(dst_fb->fd, ctx); > + put_ahnd(ahnd); > } > > static void free_linear_mapping(struct fb_blit_upload *blit) > diff --git a/lib/intel_batchbuffer.c b/lib/intel_batchbuffer.c > index 3747019a5..d9cc4d89c 100644 > --- a/lib/intel_batchbuffer.c > +++ b/lib/intel_batchbuffer.c > @@ -706,12 +706,13 @@ fill_object(struct drm_i915_gem_exec_object2 *obj, > > static void exec_blit(int fd, > struct drm_i915_gem_exec_object2 *objs, uint32_t count, > - unsigned int gen) > + unsigned int gen, uint32_t ctx) > { > struct drm_i915_gem_execbuffer2 exec = { > .buffers_ptr = to_user_pointer(objs), > .buffer_count = count, > .flags = gen >= 6 ? I915_EXEC_BLT : 0, > + .rsvd1 = ctx, > }; > > gem_execbuf(fd, &exec); > @@ -896,7 +897,7 @@ void igt_blitter_src_copy(int fd, > objs[0].flags |= EXEC_OBJECT_NEEDS_FENCE; > objs[1].flags |= EXEC_OBJECT_NEEDS_FENCE; > > - exec_blit(fd, objs, 3, gen); > + exec_blit(fd, objs, 3, gen, 0); > > gem_close(fd, batch_handle); > } > @@ -904,12 +905,15 @@ void igt_blitter_src_copy(int fd, > /** > * igt_blitter_fast_copy__raw: > * @fd: file descriptor of the i915 driver > + * @ahnd: handle to an allocator > + * @ctx: context within which execute copy blit > * @src_handle: GEM handle of the source buffer > * @src_delta: offset into the source GEM bo, in bytes > * @src_stride: Stride (in bytes) of the source buffer > * @src_tiling: Tiling mode of the source buffer > * @src_x: X coordinate of the source region to copy > * @src_y: Y coordinate of the source region to copy > + * @src_size: size of the src bo required for allocator and softpin > * @width: Width of the region to copy > * @height: Height of the region to copy > * @bpp: source and destination bits per pixel > @@ -919,16 +923,20 @@ void igt_blitter_src_copy(int fd, > * @dst_tiling: Tiling mode of the destination buffer > * @dst_x: X coordinate of destination > * @dst_y: Y coordinate of destination > + * @dst_size: size of the dst bo required for allocator and softpin > * > * Like igt_blitter_fast_copy(), but talking to the kernel directly. > */ > void igt_blitter_fast_copy__raw(int fd, > + uint64_t ahnd, > + uint32_t ctx, > /* src */ > uint32_t src_handle, > unsigned int src_delta, > unsigned int src_stride, > unsigned int src_tiling, > unsigned int src_x, unsigned src_y, > + uint64_t src_size, > > /* size */ > unsigned int width, unsigned int height, > @@ -941,7 +949,8 @@ void igt_blitter_fast_copy__raw(int fd, > unsigned dst_delta, > unsigned int dst_stride, > unsigned int dst_tiling, > - unsigned int dst_x, unsigned dst_y) > + unsigned int dst_x, unsigned dst_y, > + uint64_t dst_size) > { > uint32_t batch[12]; > struct drm_i915_gem_exec_object2 objs[3]; > @@ -949,8 +958,20 @@ void igt_blitter_fast_copy__raw(int fd, > uint32_t batch_handle; > uint32_t dword0, dword1; > uint32_t src_pitch, dst_pitch; > + uint64_t batch_offset, src_offset, dst_offset; > int i = 0; > > + batch_handle = gem_create(fd, 4096); > + if (ahnd) { > + src_offset = get_offset(ahnd, src_handle, src_size, 0); > + dst_offset = get_offset(ahnd, dst_handle, dst_size, 0); > + batch_offset = get_offset(ahnd, batch_handle, 4096, 0); > + } else { > + src_offset = 16 << 20; > + dst_offset = ALIGN(src_offset + src_size, 1 << 20); > + batch_offset = ALIGN(dst_offset + dst_size, 1 << 20); > + } > + > src_pitch = fast_copy_pitch(src_stride, src_tiling); > dst_pitch = fast_copy_pitch(dst_stride, dst_tiling); > dword0 = fast_copy_dword0(src_tiling, dst_tiling); > @@ -967,30 +988,36 @@ void igt_blitter_fast_copy__raw(int fd, > batch[i++] = dword1 | dst_pitch; > batch[i++] = (dst_y << 16) | dst_x; /* dst x1,y1 */ > batch[i++] = ((dst_y + height) << 16) | (dst_x + width); /* dst x2,y2 */ > - batch[i++] = dst_delta; /* dst address lower bits */ > - batch[i++] = 0; /* dst address upper bits */ > + batch[i++] = dst_offset + dst_delta; /* dst address lower bits */ > + batch[i++] = (dst_offset + dst_delta) >> 32; /* dst address upper bits */ > batch[i++] = (src_y << 16) | src_x; /* src x1,y1 */ > batch[i++] = src_pitch; > - batch[i++] = src_delta; /* src address lower bits */ > - batch[i++] = 0; /* src address upper bits */ > + batch[i++] = src_offset + src_delta; /* src address lower bits */ > + batch[i++] = (src_offset + src_delta) >> 32; /* src address upper bits */ > batch[i++] = MI_BATCH_BUFFER_END; > batch[i++] = MI_NOOP; > > igt_assert(i == ARRAY_SIZE(batch)); > > - batch_handle = gem_create(fd, 4096); > gem_write(fd, batch_handle, 0, batch, sizeof(batch)); > > - fill_relocation(&relocs[0], dst_handle, -1, dst_delta, 4, > + fill_relocation(&relocs[0], dst_handle, dst_offset, dst_delta, 4, > I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER); > - fill_relocation(&relocs[1], src_handle, -1, src_delta, 8, > + fill_relocation(&relocs[1], src_handle, src_offset, src_delta, 8, > I915_GEM_DOMAIN_RENDER, 0); > > - fill_object(&objs[0], dst_handle, 0, NULL, 0); > - fill_object(&objs[1], src_handle, 0, NULL, 0); > - fill_object(&objs[2], batch_handle, 0, relocs, 2); > + fill_object(&objs[0], dst_handle, dst_offset, NULL, 0); > + objs[0].flags |= EXEC_OBJECT_WRITE; > + fill_object(&objs[1], src_handle, src_offset, NULL, 0); > + fill_object(&objs[2], batch_handle, batch_offset, relocs, !ahnd ? 2 : 0); > + > + if (ahnd) { > + objs[0].flags |= EXEC_OBJECT_PINNED; > + objs[1].flags |= EXEC_OBJECT_PINNED; > + objs[2].flags |= EXEC_OBJECT_PINNED; > + } > > - exec_blit(fd, objs, 3, intel_gen(intel_get_drm_devid(fd))); > + exec_blit(fd, objs, 3, intel_gen(intel_get_drm_devid(fd)), ctx); > > gem_close(fd, batch_handle); > } > diff --git a/lib/intel_batchbuffer.h b/lib/intel_batchbuffer.h > index bd417e998..74c21c40e 100644 > --- a/lib/intel_batchbuffer.h > +++ b/lib/intel_batchbuffer.h > @@ -300,12 +300,15 @@ void igt_blitter_fast_copy(struct intel_batchbuffer *batch, > unsigned dst_x, unsigned dst_y); > > void igt_blitter_fast_copy__raw(int fd, > + uint64_t ahnd, > + uint32_t ctx, > /* src */ > uint32_t src_handle, > unsigned int src_delta, > unsigned int src_stride, > unsigned int src_tiling, > unsigned int src_x, unsigned src_y, > + uint64_t src_size, > > /* size */ > unsigned int width, unsigned int height, > @@ -318,7 +321,8 @@ void igt_blitter_fast_copy__raw(int fd, > unsigned int dst_delta, > unsigned int dst_stride, > unsigned int dst_tiling, > - unsigned int dst_x, unsigned dst_y); > + unsigned int dst_x, unsigned dst_y, > + uint64_t dst_size); > > /** > * igt_render_copyfunc_t: > -- > 2.26.0 >