[igt-dev] [PATCH i-g-t v4 1/4] lib: Move common gpgpu/media fill functions to gpu_fill library
Katarzyna Dec
katarzyna.dec at intel.com
Tue Apr 10 08:12:29 UTC 2018
Gpgpu_fill and media_fill libraries are very similar and many
functions can be shared. I have created library gpu_fill with
all functions needed for implementing gpgpu_fill and media_fill
tests for all Gens. For reviewing and debugging purposes this patch
should be only moving functions from few libraries to one removing
functions identical for both media and gpgpu.
Places in the code that required more changes:
Removing gen7_fill_gpgpu_kernel function that is identical to
gen7_fill_media_kernel and introduces conflict with moving
genX_fill_interface_descriptor, which are the same for media and gpgpu.
Function gen8_fill_media_kernel is not removed in this patch
(although it is identical with gen7 version), because this patch
should be as much as possible functions movement.
gen8_fill_interface_descriptor was unified for media and gpgpu
by adding kernel and its size as a parameter (this parameters
were missing in media gen8, gen8lp and gen9 functions)
gen8_emit_state_base_address was unified, the one for gpgpu was
configured like it would be using indirect state (while we are
using CURBE). I have checked that media fill version
(OUT_BATCH(0 | BASE_ADDRESS_MODIFY)) works fine on gpgpu gen8 and newer.
v2: Changed code layout. GenX_fill_media_kernel was identical to
genX_fill_gpgpu_kernel so this function was unified to
gen7_fill_kernel. There were 2 very similar functions
gen8_emit_state_base_address for media and gpgpu, where the one
for gpgpu was configured like it would be using indirect state
(while we are using CURBE). I have checked if media fill version
works fine in gpgpu test on Gen8 and unified them.
v3: Made patch easier for reviewing moving changes unifying code for
various gens (that were included v1) to other patch, leaving only
the most critical code changes.
Signed-off-by: Katarzyna Dec <katarzyna.dec at intel.com>
Cc: Lukasz Kalamarz <lukasz.kalamarz at intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio at intel.com>
---
lib/Makefile.sources | 2 +
lib/gpgpu_fill.c | 571 +-----------------------------------
lib/gpu_fill.c | 758 ++++++++++++++++++++++++++++++++++++++++++++++++
lib/gpu_fill.h | 138 +++++++++
lib/media_fill_gen7.c | 271 +----------------
lib/media_fill_gen8.c | 290 +-----------------
lib/media_fill_gen8lp.c | 284 +-----------------
lib/media_fill_gen9.c | 298 +------------------
lib/meson.build | 1 +
9 files changed, 908 insertions(+), 1705 deletions(-)
create mode 100644 lib/gpu_fill.c
create mode 100644 lib/gpu_fill.h
diff --git a/lib/Makefile.sources b/lib/Makefile.sources
index 3d37ef1d..45e65dd7 100644
--- a/lib/Makefile.sources
+++ b/lib/Makefile.sources
@@ -64,6 +64,8 @@ lib_source_list = \
media_spin.c \
gpgpu_fill.h \
gpgpu_fill.c \
+ gpu_fill.h \
+ gpu_fill.c \
gen7_media.h \
gen8_media.h \
rendercopy_i915.c \
diff --git a/lib/gpgpu_fill.c b/lib/gpgpu_fill.c
index 4d98643d..f2765fd6 100644
--- a/lib/gpgpu_fill.c
+++ b/lib/gpgpu_fill.c
@@ -30,10 +30,9 @@
#include "intel_reg.h"
#include "drmtest.h"
-#include "intel_batchbuffer.h"
-#include "gen7_media.h"
-#include "gen8_media.h"
+
#include "gpgpu_fill.h"
+#include "gpu_fill.h"
/* shaders/gpgpu/gpgpu_fill.gxa */
static const uint32_t gen7_gpgpu_kernel[][4] = {
@@ -75,572 +74,6 @@ static const uint32_t gen9_gpgpu_kernel[][4] = {
{ 0x07800031, 0x20000a40, 0x06000e00, 0x82000010 },
};
-static uint32_t
-batch_used(struct intel_batchbuffer *batch)
-{
- return batch->ptr - batch->buffer;
-}
-
-static uint32_t
-batch_align(struct intel_batchbuffer *batch, uint32_t align)
-{
- uint32_t offset = batch_used(batch);
- offset = ALIGN(offset, align);
- batch->ptr = batch->buffer + offset;
- return offset;
-}
-
-static void *
-batch_alloc(struct intel_batchbuffer *batch, uint32_t size, uint32_t align)
-{
- uint32_t offset = batch_align(batch, align);
- batch->ptr += size;
- return memset(batch->buffer + offset, 0, size);
-}
-
-static uint32_t
-batch_offset(struct intel_batchbuffer *batch, void *ptr)
-{
- return (uint8_t *)ptr - batch->buffer;
-}
-
-static uint32_t
-batch_copy(struct intel_batchbuffer *batch, const void *ptr, uint32_t size,
- uint32_t align)
-{
- return batch_offset(batch, memcpy(batch_alloc(batch, size, align), ptr, size));
-}
-
-static void
-gen7_render_flush(struct intel_batchbuffer *batch, uint32_t batch_end)
-{
- int ret;
-
- ret = drm_intel_bo_subdata(batch->bo, 0, 4096, batch->buffer);
- if (ret == 0)
- ret = drm_intel_bo_mrb_exec(batch->bo, batch_end,
- NULL, 0, 0, 0);
- igt_assert(ret == 0);
-}
-
-static uint32_t
-gen7_fill_curbe_buffer_data(struct intel_batchbuffer *batch, uint8_t color)
-{
- uint8_t *curbe_buffer;
- uint32_t offset;
-
- curbe_buffer = batch_alloc(batch, sizeof(uint32_t) * 8, 64);
- offset = batch_offset(batch, curbe_buffer);
- *curbe_buffer = color;
-
- return offset;
-}
-
-static uint32_t
-gen7_fill_surface_state(struct intel_batchbuffer *batch,
- struct igt_buf *buf,
- uint32_t format,
- int is_dst)
-{
- struct gen7_surface_state *ss;
- uint32_t write_domain, read_domain, offset;
- int ret;
-
- if (is_dst) {
- write_domain = read_domain = I915_GEM_DOMAIN_RENDER;
- } else {
- write_domain = 0;
- read_domain = I915_GEM_DOMAIN_SAMPLER;
- }
-
- ss = batch_alloc(batch, sizeof(*ss), 64);
- offset = batch_offset(batch, ss);
-
- ss->ss0.surface_type = GEN7_SURFACE_2D;
- ss->ss0.surface_format = format;
- ss->ss0.render_cache_read_write = 1;
-
- if (buf->tiling == I915_TILING_X)
- ss->ss0.tiled_mode = 2;
- else if (buf->tiling == I915_TILING_Y)
- ss->ss0.tiled_mode = 3;
-
- ss->ss1.base_addr = buf->bo->offset;
- ret = drm_intel_bo_emit_reloc(batch->bo,
- batch_offset(batch, ss) + 4,
- buf->bo, 0,
- read_domain, write_domain);
- igt_assert(ret == 0);
-
- ss->ss2.height = igt_buf_height(buf) - 1;
- ss->ss2.width = igt_buf_width(buf) - 1;
-
- ss->ss3.pitch = buf->stride - 1;
-
- ss->ss7.shader_chanel_select_r = 4;
- ss->ss7.shader_chanel_select_g = 5;
- ss->ss7.shader_chanel_select_b = 6;
- ss->ss7.shader_chanel_select_a = 7;
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_surface_state(struct intel_batchbuffer *batch,
- struct igt_buf *buf,
- uint32_t format,
- int is_dst)
-{
- struct gen8_surface_state *ss;
- uint32_t write_domain, read_domain, offset;
- int ret;
-
- if (is_dst) {
- write_domain = read_domain = I915_GEM_DOMAIN_RENDER;
- } else {
- write_domain = 0;
- read_domain = I915_GEM_DOMAIN_SAMPLER;
- }
-
- ss = batch_alloc(batch, sizeof(*ss), 64);
- offset = batch_offset(batch, ss);
-
- ss->ss0.surface_type = GEN8_SURFACE_2D;
- ss->ss0.surface_format = format;
- ss->ss0.render_cache_read_write = 1;
- ss->ss0.vertical_alignment = 1; /* align 4 */
- ss->ss0.horizontal_alignment = 1; /* align 4 */
-
- if (buf->tiling == I915_TILING_X)
- ss->ss0.tiled_mode = 2;
- else if (buf->tiling == I915_TILING_Y)
- ss->ss0.tiled_mode = 3;
-
- ss->ss8.base_addr = buf->bo->offset;
-
- ret = drm_intel_bo_emit_reloc(batch->bo,
- batch_offset(batch, ss) + 8 * 4,
- buf->bo, 0,
- read_domain, write_domain);
- igt_assert_eq(ret, 0);
-
- ss->ss2.height = igt_buf_height(buf) - 1;
- ss->ss2.width = igt_buf_width(buf) - 1;
- ss->ss3.pitch = buf->stride - 1;
-
- ss->ss7.shader_chanel_select_r = 4;
- ss->ss7.shader_chanel_select_g = 5;
- ss->ss7.shader_chanel_select_b = 6;
- ss->ss7.shader_chanel_select_a = 7;
-
- return offset;
-
-}
-
-static uint32_t
-gen7_fill_binding_table(struct intel_batchbuffer *batch,
- struct igt_buf *dst)
-{
- uint32_t *binding_table, offset;
-
- binding_table = batch_alloc(batch, 32, 64);
- offset = batch_offset(batch, binding_table);
-
- binding_table[0] = gen7_fill_surface_state(batch, dst, GEN7_SURFACEFORMAT_R8_UNORM, 1);
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_binding_table(struct intel_batchbuffer *batch,
- struct igt_buf *dst)
-{
- uint32_t *binding_table, offset;
-
- binding_table = batch_alloc(batch, 32, 64);
- offset = batch_offset(batch, binding_table);
-
- binding_table[0] = gen8_fill_surface_state(batch, dst, GEN8_SURFACEFORMAT_R8_UNORM, 1);
-
- return offset;
-}
-
-static uint32_t
-gen7_fill_gpgpu_kernel(struct intel_batchbuffer *batch,
- const uint32_t kernel[][4],
- size_t size)
-{
- uint32_t offset;
-
- offset = batch_copy(batch, kernel, size, 64);
-
- return offset;
-}
-
-static uint32_t
-gen7_fill_interface_descriptor(struct intel_batchbuffer *batch, struct igt_buf *dst,
- const uint32_t kernel[][4], size_t size)
-{
- struct gen7_interface_descriptor_data *idd;
- uint32_t offset;
- uint32_t binding_table_offset, kernel_offset;
-
- binding_table_offset = gen7_fill_binding_table(batch, dst);
- kernel_offset = gen7_fill_gpgpu_kernel(batch, kernel, size);
-
- idd = batch_alloc(batch, sizeof(*idd), 64);
- offset = batch_offset(batch, idd);
-
- idd->desc0.kernel_start_pointer = (kernel_offset >> 6);
-
- idd->desc1.single_program_flow = 1;
- idd->desc1.floating_point_mode = GEN7_FLOATING_POINT_IEEE_754;
-
- idd->desc2.sampler_count = 0; /* 0 samplers used */
- idd->desc2.sampler_state_pointer = 0;
-
- idd->desc3.binding_table_entry_count = 0;
- idd->desc3.binding_table_pointer = (binding_table_offset >> 5);
-
- idd->desc4.constant_urb_entry_read_offset = 0;
- idd->desc4.constant_urb_entry_read_length = 1; /* grf 1 */
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_interface_descriptor(struct intel_batchbuffer *batch, struct igt_buf *dst,
- const uint32_t kernel[][4], size_t size)
-{
- struct gen8_interface_descriptor_data *idd;
- uint32_t offset;
- uint32_t binding_table_offset, kernel_offset;
-
- binding_table_offset = gen8_fill_binding_table(batch, dst);
- kernel_offset = gen7_fill_gpgpu_kernel(batch, kernel, size);
-
- idd = batch_alloc(batch, sizeof(*idd), 64);
- offset = batch_offset(batch, idd);
-
- idd->desc0.kernel_start_pointer = (kernel_offset >> 6);
-
- idd->desc2.single_program_flow = 1;
- idd->desc2.floating_point_mode = GEN8_FLOATING_POINT_IEEE_754;
-
- idd->desc3.sampler_count = 0; /* 0 samplers used */
- idd->desc3.sampler_state_pointer = 0;
-
- idd->desc4.binding_table_entry_count = 0;
- idd->desc4.binding_table_pointer = (binding_table_offset >> 5);
-
- idd->desc5.constant_urb_entry_read_offset = 0;
- idd->desc5.constant_urb_entry_read_length = 1; /* grf 1 */
-
- return offset;
-}
-
-static void
-gen7_emit_state_base_address(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN7_STATE_BASE_ADDRESS | (10 - 2));
-
- /* general */
- OUT_BATCH(0);
-
- /* surface */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* dynamic */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* indirect */
- OUT_BATCH(0);
-
- /* instruction */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* general/dynamic/indirect/instruction access Bound */
- OUT_BATCH(0);
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
- OUT_BATCH(0);
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
-}
-
-static void
-gen8_emit_state_base_address(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_STATE_BASE_ADDRESS | (16 - 2));
-
- /* general */
- OUT_BATCH(0 | (0x78 << 4) | (0 << 1) | BASE_ADDRESS_MODIFY);
- OUT_BATCH(0);
-
- /* stateless data port */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
-
- /* surface */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_SAMPLER, 0, BASE_ADDRESS_MODIFY);
-
- /* dynamic */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_RENDER | I915_GEM_DOMAIN_INSTRUCTION,
- 0, BASE_ADDRESS_MODIFY);
-
- /* indirect */
- OUT_BATCH(0);
- OUT_BATCH(0 );
-
- /* instruction */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* general state buffer size */
- OUT_BATCH(0xfffff000 | 1);
- /* dynamic state buffer size */
- OUT_BATCH(1 << 12 | 1);
- /* indirect object buffer size */
- OUT_BATCH(0xfffff000 | 1);
- /* intruction buffer size, must set modify enable bit, otherwise it may result in GPU hang */
- OUT_BATCH(1 << 12 | 1);
-}
-
-static void
-gen9_emit_state_base_address(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_STATE_BASE_ADDRESS | (19 - 2));
-
- /* general */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
- OUT_BATCH(0);
-
- /* stateless data port */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
-
- /* surface */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_SAMPLER, 0, BASE_ADDRESS_MODIFY);
-
- /* dynamic */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_RENDER | I915_GEM_DOMAIN_INSTRUCTION,
- 0, BASE_ADDRESS_MODIFY);
-
- /* indirect */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* instruction */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* general state buffer size */
- OUT_BATCH(0xfffff000 | 1);
- /* dynamic state buffer size */
- OUT_BATCH(1 << 12 | 1);
- /* indirect object buffer size */
- OUT_BATCH(0xfffff000 | 1);
- /* intruction buffer size, must set modify enable bit, otherwise it may result in GPU hang */
- OUT_BATCH(1 << 12 | 1);
-
- /* Bindless surface state base address */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
- OUT_BATCH(0);
- OUT_BATCH(0xfffff000);
-}
-
-static void
-gen7_emit_vfe_state_gpgpu(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN7_MEDIA_VFE_STATE | (8 - 2));
-
- /* scratch buffer */
- OUT_BATCH(0);
-
- /* number of threads & urb entries */
- OUT_BATCH(1 << 16 | /* max num of threads */
- 0 << 8 | /* num of URB entry */
- 1 << 2); /* GPGPU mode */
-
- OUT_BATCH(0);
-
- /* urb entry size & curbe size */
- OUT_BATCH(0 << 16 | /* URB entry size in 256 bits unit */
- 1); /* CURBE entry size in 256 bits unit */
-
- /* scoreboard */
- OUT_BATCH(0);
- OUT_BATCH(0);
- OUT_BATCH(0);
-}
-
-static void
-gen8_emit_vfe_state_gpgpu(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_MEDIA_VFE_STATE | (9 - 2));
-
- /* scratch buffer */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* number of threads & urb entries */
- OUT_BATCH(1 << 16 | 1 << 8);
-
- OUT_BATCH(0);
-
- /* urb entry size & curbe size */
- OUT_BATCH(0 << 16 | 1);
-
- /* scoreboard */
- OUT_BATCH(0);
- OUT_BATCH(0);
- OUT_BATCH(0);
-}
-
-static void
-gen7_emit_curbe_load(struct intel_batchbuffer *batch, uint32_t curbe_buffer)
-{
- OUT_BATCH(GEN7_MEDIA_CURBE_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* curbe total data length */
- OUT_BATCH(64);
- /* curbe data start address, is relative to the dynamics base address */
- OUT_BATCH(curbe_buffer);
-}
-
-static void
-gen7_emit_interface_descriptor_load(struct intel_batchbuffer *batch, uint32_t interface_descriptor)
-{
- OUT_BATCH(GEN7_MEDIA_INTERFACE_DESCRIPTOR_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* interface descriptor data length */
- OUT_BATCH(sizeof(struct gen7_interface_descriptor_data));
- /* interface descriptor address, is relative to the dynamics base address */
- OUT_BATCH(interface_descriptor);
-}
-
-static void
-gen8_emit_interface_descriptor_load(struct intel_batchbuffer *batch, uint32_t interface_descriptor)
-{
- OUT_BATCH(GEN8_MEDIA_INTERFACE_DESCRIPTOR_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* interface descriptor data length */
- OUT_BATCH(sizeof(struct gen8_interface_descriptor_data));
- /* interface descriptor address, is relative to the dynamics base address */
- OUT_BATCH(interface_descriptor);
-}
-
-static void
-gen7_emit_gpgpu_walk(struct intel_batchbuffer *batch,
- unsigned x, unsigned y,
- unsigned width, unsigned height)
-{
- uint32_t x_dim, y_dim, tmp, right_mask;
-
- /*
- * Simply do SIMD16 based dispatch, so every thread uses
- * SIMD16 channels.
- *
- * Define our own thread group size, e.g 16x1 for every group, then
- * will have 1 thread each group in SIMD16 dispatch. So thread
- * width/height/depth are all 1.
- *
- * Then thread group X = width / 16 (aligned to 16)
- * thread group Y = height;
- */
- x_dim = (width + 15) / 16;
- y_dim = height;
-
- tmp = width & 15;
- if (tmp == 0)
- right_mask = (1 << 16) - 1;
- else
- right_mask = (1 << tmp) - 1;
-
- OUT_BATCH(GEN7_GPGPU_WALKER | 9);
-
- /* interface descriptor offset */
- OUT_BATCH(0);
-
- /* SIMD size, thread w/h/d */
- OUT_BATCH(1 << 30 | /* SIMD16 */
- 0 << 16 | /* depth:1 */
- 0 << 8 | /* height:1 */
- 0); /* width:1 */
-
- /* thread group X */
- OUT_BATCH(0);
- OUT_BATCH(x_dim);
-
- /* thread group Y */
- OUT_BATCH(0);
- OUT_BATCH(y_dim);
-
- /* thread group Z */
- OUT_BATCH(0);
- OUT_BATCH(1);
-
- /* right mask */
- OUT_BATCH(right_mask);
-
- /* bottom mask, height 1, always 0xffffffff */
- OUT_BATCH(0xffffffff);
-}
-
-static void
-gen8_emit_gpgpu_walk(struct intel_batchbuffer *batch,
- unsigned x, unsigned y,
- unsigned width, unsigned height)
-{
- uint32_t x_dim, y_dim, tmp, right_mask;
-
- /*
- * Simply do SIMD16 based dispatch, so every thread uses
- * SIMD16 channels.
- *
- * Define our own thread group size, e.g 16x1 for every group, then
- * will have 1 thread each group in SIMD16 dispatch. So thread
- * width/height/depth are all 1.
- *
- * Then thread group X = width / 16 (aligned to 16)
- * thread group Y = height;
- */
- x_dim = (width + 15) / 16;
- y_dim = height;
-
- tmp = width & 15;
- if (tmp == 0)
- right_mask = (1 << 16) - 1;
- else
- right_mask = (1 << tmp) - 1;
-
- OUT_BATCH(GEN7_GPGPU_WALKER | 13);
-
- OUT_BATCH(0); /* kernel offset */
- OUT_BATCH(0); /* indirect data length */
- OUT_BATCH(0); /* indirect data offset */
-
- /* SIMD size, thread w/h/d */
- OUT_BATCH(1 << 30 | /* SIMD16 */
- 0 << 16 | /* depth:1 */
- 0 << 8 | /* height:1 */
- 0); /* width:1 */
-
- /* thread group X */
- OUT_BATCH(0);
- OUT_BATCH(0);
- OUT_BATCH(x_dim);
-
- /* thread group Y */
- OUT_BATCH(0);
- OUT_BATCH(0);
- OUT_BATCH(y_dim);
-
- /* thread group Z */
- OUT_BATCH(0);
- OUT_BATCH(1);
-
- /* right mask */
- OUT_BATCH(right_mask);
-
- /* bottom mask, height 1, always 0xffffffff */
- OUT_BATCH(0xffffffff);
-}
-
/*
* This sets up the gpgpu pipeline,
*
diff --git a/lib/gpu_fill.c b/lib/gpu_fill.c
new file mode 100644
index 00000000..172c6db6
--- /dev/null
+++ b/lib/gpu_fill.c
@@ -0,0 +1,758 @@
+#include "gpu_fill.h"
+
+uint32_t
+batch_used(struct intel_batchbuffer *batch)
+{
+ return batch->ptr - batch->buffer;
+}
+
+uint32_t
+batch_align(struct intel_batchbuffer *batch, uint32_t align)
+{
+ uint32_t offset = batch_used(batch);
+ offset = ALIGN(offset, align);
+ batch->ptr = batch->buffer + offset;
+ return offset;
+}
+
+void *
+batch_alloc(struct intel_batchbuffer *batch, uint32_t size, uint32_t align)
+{
+ uint32_t offset = batch_align(batch, align);
+ batch->ptr += size;
+ return memset(batch->buffer + offset, 0, size);
+}
+
+uint32_t
+batch_offset(struct intel_batchbuffer *batch, void *ptr)
+{
+ return (uint8_t *)ptr - batch->buffer;
+}
+
+uint32_t
+batch_copy(struct intel_batchbuffer *batch, const void *ptr, uint32_t size, uint32_t align)
+{
+ return batch_offset(batch, memcpy(batch_alloc(batch, size, align), ptr, size));
+}
+
+void
+gen7_render_flush(struct intel_batchbuffer *batch, uint32_t batch_end)
+{
+ int ret;
+
+ ret = drm_intel_bo_subdata(batch->bo, 0, 4096, batch->buffer);
+ if (ret == 0)
+ ret = drm_intel_bo_mrb_exec(batch->bo, batch_end,
+ NULL, 0, 0, 0);
+ igt_assert(ret == 0);
+}
+
+uint32_t
+gen7_fill_curbe_buffer_data(struct intel_batchbuffer *batch,
+ uint8_t color)
+{
+ uint8_t *curbe_buffer;
+ uint32_t offset;
+
+ curbe_buffer = batch_alloc(batch, sizeof(uint32_t) * 8, 64);
+ offset = batch_offset(batch, curbe_buffer);
+ *curbe_buffer = color;
+
+ return offset;
+}
+
+uint32_t
+gen7_fill_surface_state(struct intel_batchbuffer *batch,
+ struct igt_buf *buf,
+ uint32_t format,
+ int is_dst)
+{
+ struct gen7_surface_state *ss;
+ uint32_t write_domain, read_domain, offset;
+ int ret;
+
+ if (is_dst) {
+ write_domain = read_domain = I915_GEM_DOMAIN_RENDER;
+ } else {
+ write_domain = 0;
+ read_domain = I915_GEM_DOMAIN_SAMPLER;
+ }
+
+ ss = batch_alloc(batch, sizeof(*ss), 64);
+ offset = batch_offset(batch, ss);
+
+ ss->ss0.surface_type = GEN7_SURFACE_2D;
+ ss->ss0.surface_format = format;
+ ss->ss0.render_cache_read_write = 1;
+
+ if (buf->tiling == I915_TILING_X)
+ ss->ss0.tiled_mode = 2;
+ else if (buf->tiling == I915_TILING_Y)
+ ss->ss0.tiled_mode = 3;
+
+ ss->ss1.base_addr = buf->bo->offset;
+ ret = drm_intel_bo_emit_reloc(batch->bo,
+ batch_offset(batch, ss) + 4,
+ buf->bo, 0,
+ read_domain, write_domain);
+ igt_assert(ret == 0);
+
+ ss->ss2.height = igt_buf_height(buf) - 1;
+ ss->ss2.width = igt_buf_width(buf) - 1;
+
+ ss->ss3.pitch = buf->stride - 1;
+
+ ss->ss7.shader_chanel_select_r = 4;
+ ss->ss7.shader_chanel_select_g = 5;
+ ss->ss7.shader_chanel_select_b = 6;
+ ss->ss7.shader_chanel_select_a = 7;
+
+ return offset;
+}
+
+uint32_t
+gen7_fill_binding_table(struct intel_batchbuffer *batch,
+ struct igt_buf *dst)
+{
+ uint32_t *binding_table, offset;
+
+ binding_table = batch_alloc(batch, 32, 64);
+ offset = batch_offset(batch, binding_table);
+ binding_table[0] = gen7_fill_surface_state(batch, dst,
+ GEN7_SURFACEFORMAT_R8_UNORM, 1);
+
+ return offset;
+}
+
+uint32_t
+gen7_fill_media_kernel(struct intel_batchbuffer *batch,
+ const uint32_t kernel[][4],
+ size_t size)
+{
+ uint32_t offset;
+
+ offset = batch_copy(batch, kernel, size, 64);
+
+ return offset;
+}
+
+uint32_t
+gen8_fill_media_kernel(struct intel_batchbuffer *batch,
+ const uint32_t kernel[][4],
+ size_t size)
+{
+ uint32_t offset;
+
+ offset = batch_copy(batch, kernel, size, 64);
+
+ return offset;
+}
+
+uint32_t
+gen7_fill_interface_descriptor(struct intel_batchbuffer *batch, struct igt_buf *dst,
+ const uint32_t kernel[][4], size_t size)
+{
+ struct gen7_interface_descriptor_data *idd;
+ uint32_t offset;
+ uint32_t binding_table_offset, kernel_offset;
+
+ binding_table_offset = gen7_fill_binding_table(batch, dst);
+ kernel_offset = gen7_fill_media_kernel(batch, kernel, size);
+
+ idd = batch_alloc(batch, sizeof(*idd), 64);
+ offset = batch_offset(batch, idd);
+
+ idd->desc0.kernel_start_pointer = (kernel_offset >> 6);
+
+ idd->desc1.single_program_flow = 1;
+ idd->desc1.floating_point_mode = GEN7_FLOATING_POINT_IEEE_754;
+
+ idd->desc2.sampler_count = 0; /* 0 samplers used */
+ idd->desc2.sampler_state_pointer = 0;
+
+ idd->desc3.binding_table_entry_count = 0;
+ idd->desc3.binding_table_pointer = (binding_table_offset >> 5);
+
+ idd->desc4.constant_urb_entry_read_offset = 0;
+ idd->desc4.constant_urb_entry_read_length = 1; /* grf 1 */
+
+ return offset;
+}
+
+void
+gen7_emit_state_base_address(struct intel_batchbuffer *batch)
+{
+ OUT_BATCH(GEN7_STATE_BASE_ADDRESS | (10 - 2));
+
+ /* general */
+ OUT_BATCH(0);
+
+ /* surface */
+ OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
+
+ /* dynamic */
+ OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
+
+ /* indirect */
+ OUT_BATCH(0);
+
+ /* instruction */
+ OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
+
+ /* general/dynamic/indirect/instruction access Bound */
+ OUT_BATCH(0);
+ OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
+ OUT_BATCH(0);
+ OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
+}
+
+void
+gen7_emit_vfe_state(struct intel_batchbuffer *batch)
+{
+ OUT_BATCH(GEN7_MEDIA_VFE_STATE | (8 - 2));
+
+ /* scratch buffer */
+ OUT_BATCH(0);
+
+ /* number of threads & urb entries */
+ OUT_BATCH(1 << 16 |
+ 2 << 8);
+
+ OUT_BATCH(0);
+
+ /* urb entry size & curbe size */
+ OUT_BATCH(2 << 16 | /* in 256 bits unit */
+ 2); /* in 256 bits unit */
+
+ /* scoreboard */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+}
+
+void
+gen7_emit_vfe_state_gpgpu(struct intel_batchbuffer *batch)
+{
+ OUT_BATCH(GEN7_MEDIA_VFE_STATE | (8 - 2));
+
+ /* scratch buffer */
+ OUT_BATCH(0);
+
+ /* number of threads & urb entries */
+ OUT_BATCH(1 << 16 | /* max num of threads */
+ 0 << 8 | /* num of URB entry */
+ 1 << 2); /* GPGPU mode */
+
+ OUT_BATCH(0);
+
+ /* urb entry size & curbe size */
+ OUT_BATCH(0 << 16 | /* URB entry size in 256 bits unit */
+ 1); /* CURBE entry size in 256 bits unit */
+
+ /* scoreboard */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+}
+
+void
+gen7_emit_curbe_load(struct intel_batchbuffer *batch, uint32_t curbe_buffer)
+{
+ OUT_BATCH(GEN7_MEDIA_CURBE_LOAD | (4 - 2));
+ OUT_BATCH(0);
+ /* curbe total data length */
+ OUT_BATCH(64);
+ /* curbe data start address, is relative to the dynamics base address */
+ OUT_BATCH(curbe_buffer);
+}
+
+void
+gen7_emit_interface_descriptor_load(struct intel_batchbuffer *batch, uint32_t interface_descriptor)
+{
+ OUT_BATCH(GEN7_MEDIA_INTERFACE_DESCRIPTOR_LOAD | (4 - 2));
+ OUT_BATCH(0);
+ /* interface descriptor data length */
+ OUT_BATCH(sizeof(struct gen7_interface_descriptor_data));
+ /* interface descriptor address, is relative to the dynamics base address */
+ OUT_BATCH(interface_descriptor);
+}
+
+void
+gen7_emit_media_objects(struct intel_batchbuffer *batch,
+ unsigned x, unsigned y,
+ unsigned width, unsigned height)
+{
+ int i, j;
+
+ for (i = 0; i < width / 16; i++) {
+ for (j = 0; j < height / 16; j++) {
+ OUT_BATCH(GEN7_MEDIA_OBJECT | (8 - 2));
+
+ /* interface descriptor offset */
+ OUT_BATCH(0);
+
+ /* without indirect data */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+
+ /* scoreboard */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+
+ /* inline data (xoffset, yoffset) */
+ OUT_BATCH(x + i * 16);
+ OUT_BATCH(y + j * 16);
+ }
+ }
+}
+
+void
+gen7_emit_gpgpu_walk(struct intel_batchbuffer *batch,
+ unsigned x, unsigned y,
+ unsigned width, unsigned height)
+{
+ uint32_t x_dim, y_dim, tmp, right_mask;
+
+ /*
+ * Simply do SIMD16 based dispatch, so every thread uses
+ * SIMD16 channels.
+ *
+ * Define our own thread group size, e.g 16x1 for every group, then
+ * will have 1 thread each group in SIMD16 dispatch. So thread
+ * width/height/depth are all 1.
+ *
+ * Then thread group X = width / 16 (aligned to 16)
+ * thread group Y = height;
+ */
+ x_dim = (width + 15) / 16;
+ y_dim = height;
+
+ tmp = width & 15;
+ if (tmp == 0)
+ right_mask = (1 << 16) - 1;
+ else
+ right_mask = (1 << tmp) - 1;
+
+ OUT_BATCH(GEN7_GPGPU_WALKER | 9);
+
+ /* interface descriptor offset */
+ OUT_BATCH(0);
+
+ /* SIMD size, thread w/h/d */
+ OUT_BATCH(1 << 30 | /* SIMD16 */
+ 0 << 16 | /* depth:1 */
+ 0 << 8 | /* height:1 */
+ 0); /* width:1 */
+
+ /* thread group X */
+ OUT_BATCH(0);
+ OUT_BATCH(x_dim);
+
+ /* thread group Y */
+ OUT_BATCH(0);
+ OUT_BATCH(y_dim);
+
+ /* thread group Z */
+ OUT_BATCH(0);
+ OUT_BATCH(1);
+
+ /* right mask */
+ OUT_BATCH(right_mask);
+
+ /* bottom mask, height 1, always 0xffffffff */
+ OUT_BATCH(0xffffffff);
+}
+
+void
+gen8_render_flush(struct intel_batchbuffer *batch, uint32_t batch_end)
+{
+ int ret;
+
+ ret = drm_intel_bo_subdata(batch->bo, 0, 4096, batch->buffer);
+ if (ret == 0)
+ ret = drm_intel_bo_mrb_exec(batch->bo, batch_end,
+ NULL, 0, 0, 0);
+ igt_assert(ret == 0);
+}
+
+
+uint32_t
+gen8_fill_curbe_buffer_data(struct intel_batchbuffer *batch,
+ uint8_t color)
+{
+ uint8_t *curbe_buffer;
+ uint32_t offset;
+
+ curbe_buffer = batch_alloc(batch, sizeof(uint32_t) * 8, 64);
+ offset = batch_offset(batch, curbe_buffer);
+ *curbe_buffer = color;
+
+ return offset;
+}
+
+uint32_t
+gen8_fill_surface_state(struct intel_batchbuffer *batch,
+ struct igt_buf *buf,
+ uint32_t format,
+ int is_dst)
+{
+ struct gen8_surface_state *ss;
+ uint32_t write_domain, read_domain, offset;
+ int ret;
+
+ if (is_dst) {
+ write_domain = read_domain = I915_GEM_DOMAIN_RENDER;
+ } else {
+ write_domain = 0;
+ read_domain = I915_GEM_DOMAIN_SAMPLER;
+ }
+
+ ss = batch_alloc(batch, sizeof(*ss), 64);
+ offset = batch_offset(batch, ss);
+
+ ss->ss0.surface_type = GEN8_SURFACE_2D;
+ ss->ss0.surface_format = format;
+ ss->ss0.render_cache_read_write = 1;
+ ss->ss0.vertical_alignment = 1; /* align 4 */
+ ss->ss0.horizontal_alignment = 1; /* align 4 */
+
+ if (buf->tiling == I915_TILING_X)
+ ss->ss0.tiled_mode = 2;
+ else if (buf->tiling == I915_TILING_Y)
+ ss->ss0.tiled_mode = 3;
+
+ ss->ss8.base_addr = buf->bo->offset;
+
+ ret = drm_intel_bo_emit_reloc(batch->bo,
+ batch_offset(batch, ss) + 8 * 4,
+ buf->bo, 0,
+ read_domain, write_domain);
+ igt_assert(ret == 0);
+
+ ss->ss2.height = igt_buf_height(buf) - 1;
+ ss->ss2.width = igt_buf_width(buf) - 1;
+ ss->ss3.pitch = buf->stride - 1;
+
+ ss->ss7.shader_chanel_select_r = 4;
+ ss->ss7.shader_chanel_select_g = 5;
+ ss->ss7.shader_chanel_select_b = 6;
+ ss->ss7.shader_chanel_select_a = 7;
+
+ return offset;
+}
+
+uint32_t
+gen8_fill_binding_table(struct intel_batchbuffer *batch,
+ struct igt_buf *dst)
+{
+ uint32_t *binding_table, offset;
+
+ binding_table = batch_alloc(batch, 32, 64);
+ offset = batch_offset(batch, binding_table);
+
+ binding_table[0] = gen8_fill_surface_state(batch, dst,
+ GEN8_SURFACEFORMAT_R8_UNORM, 1);
+
+ return offset;
+}
+
+uint32_t
+gen8_fill_interface_descriptor(struct intel_batchbuffer *batch, struct igt_buf *dst, const uint32_t kernel[][4], size_t size)
+{
+ struct gen8_interface_descriptor_data *idd;
+ uint32_t offset;
+ uint32_t binding_table_offset, kernel_offset;
+
+ binding_table_offset = gen8_fill_binding_table(batch, dst);
+ kernel_offset = gen8_fill_media_kernel(batch, kernel, size);
+
+ idd = batch_alloc(batch, sizeof(*idd), 64);
+ offset = batch_offset(batch, idd);
+
+ idd->desc0.kernel_start_pointer = (kernel_offset >> 6);
+
+ idd->desc2.single_program_flow = 1;
+ idd->desc2.floating_point_mode = GEN8_FLOATING_POINT_IEEE_754;
+
+ idd->desc3.sampler_count = 0; /* 0 samplers used */
+ idd->desc3.sampler_state_pointer = 0;
+
+ idd->desc4.binding_table_entry_count = 0;
+ idd->desc4.binding_table_pointer = (binding_table_offset >> 5);
+
+ idd->desc5.constant_urb_entry_read_offset = 0;
+ idd->desc5.constant_urb_entry_read_length = 1; /* grf 1 */
+
+ return offset;
+}
+
+void
+gen8_emit_state_base_address(struct intel_batchbuffer *batch)
+{
+ OUT_BATCH(GEN8_STATE_BASE_ADDRESS | (16 - 2));
+
+ /* general */
+ OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
+ OUT_BATCH(0);
+
+ /* stateless data port */
+ OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
+
+ /* surface */
+ OUT_RELOC(batch->bo, I915_GEM_DOMAIN_SAMPLER, 0, BASE_ADDRESS_MODIFY);
+
+ /* dynamic */
+ OUT_RELOC(batch->bo, I915_GEM_DOMAIN_RENDER | I915_GEM_DOMAIN_INSTRUCTION,
+ 0, BASE_ADDRESS_MODIFY);
+
+ /* indirect */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+
+ /* instruction */
+ OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
+
+ /* general state buffer size */
+ OUT_BATCH(0xfffff000 | 1);
+ /* dynamic state buffer size */
+ OUT_BATCH(1 << 12 | 1);
+ /* indirect object buffer size */
+ OUT_BATCH(0xfffff000 | 1);
+ /* intruction buffer size, must set modify enable bit, otherwise it may result in GPU hang */
+ OUT_BATCH(1 << 12 | 1);
+}
+
+void
+gen8_emit_vfe_state(struct intel_batchbuffer *batch)
+{
+ OUT_BATCH(GEN8_MEDIA_VFE_STATE | (9 - 2));
+
+ /* scratch buffer */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+
+ /* number of threads & urb entries */
+ OUT_BATCH(1 << 16 |
+ 2 << 8);
+
+ OUT_BATCH(0);
+
+ /* urb entry size & curbe size */
+ OUT_BATCH(2 << 16 |
+ 2);
+
+ /* scoreboard */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+}
+
+void
+gen8_emit_vfe_state_gpgpu(struct intel_batchbuffer *batch)
+{
+ OUT_BATCH(GEN8_MEDIA_VFE_STATE | (9 - 2));
+
+ /* scratch buffer */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+
+ /* number of threads & urb entries */
+ OUT_BATCH(1 << 16 | 1 << 8);
+
+ OUT_BATCH(0);
+
+ /* urb entry size & curbe size */
+ OUT_BATCH(0 << 16 | 1);
+
+ /* scoreboard */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+}
+
+void
+gen8_emit_curbe_load(struct intel_batchbuffer *batch, uint32_t curbe_buffer)
+{
+ OUT_BATCH(GEN8_MEDIA_CURBE_LOAD | (4 - 2));
+ OUT_BATCH(0);
+ /* curbe total data length */
+ OUT_BATCH(64);
+ /* curbe data start address, is relative to the dynamics base address */
+ OUT_BATCH(curbe_buffer);
+}
+
+void
+gen8_emit_interface_descriptor_load(struct intel_batchbuffer *batch, uint32_t interface_descriptor)
+{
+ OUT_BATCH(GEN8_MEDIA_INTERFACE_DESCRIPTOR_LOAD | (4 - 2));
+ OUT_BATCH(0);
+ /* interface descriptor data length */
+ OUT_BATCH(sizeof(struct gen8_interface_descriptor_data));
+ /* interface descriptor address, is relative to the dynamics base address */
+ OUT_BATCH(interface_descriptor);
+}
+
+void
+gen8_emit_media_state_flush(struct intel_batchbuffer *batch)
+{
+ OUT_BATCH(GEN8_MEDIA_STATE_FLUSH | (2 - 2));
+ OUT_BATCH(0);
+}
+
+void
+gen8_emit_media_objects(struct intel_batchbuffer *batch,
+ unsigned x, unsigned y,
+ unsigned width, unsigned height)
+{
+ int i, j;
+
+ for (i = 0; i < width / 16; i++) {
+ for (j = 0; j < height / 16; j++) {
+ OUT_BATCH(GEN8_MEDIA_OBJECT | (8 - 2));
+
+ /* interface descriptor offset */
+ OUT_BATCH(0);
+
+ /* without indirect data */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+
+ /* scoreboard */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+
+ /* inline data (xoffset, yoffset) */
+ OUT_BATCH(x + i * 16);
+ OUT_BATCH(y + j * 16);
+ gen8_emit_media_state_flush(batch);
+ }
+ }
+}
+void
+gen8lp_emit_media_objects(struct intel_batchbuffer *batch,
+ unsigned x, unsigned y,
+ unsigned width, unsigned height)
+{
+ int i, j;
+
+ for (i = 0; i < width / 16; i++) {
+ for (j = 0; j < height / 16; j++) {
+ OUT_BATCH(GEN8_MEDIA_OBJECT | (8 - 2));
+
+ /* interface descriptor offset */
+ OUT_BATCH(0);
+
+ /* without indirect data */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+
+ /* scoreboard */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+
+ /* inline data (xoffset, yoffset) */
+ OUT_BATCH(x + i * 16);
+ OUT_BATCH(y + j * 16);
+ }
+ }
+}
+void
+gen8_emit_gpgpu_walk(struct intel_batchbuffer *batch,
+ unsigned x, unsigned y,
+ unsigned width, unsigned height)
+{
+ uint32_t x_dim, y_dim, tmp, right_mask;
+
+ /*
+ * Simply do SIMD16 based dispatch, so every thread uses
+ * SIMD16 channels.
+ *
+ * Define our own thread group size, e.g 16x1 for every group, then
+ * will have 1 thread each group in SIMD16 dispatch. So thread
+ * width/height/depth are all 1.
+ *
+ * Then thread group X = width / 16 (aligned to 16)
+ * thread group Y = height;
+ */
+ x_dim = (width + 15) / 16;
+ y_dim = height;
+
+ tmp = width & 15;
+ if (tmp == 0)
+ right_mask = (1 << 16) - 1;
+ else
+ right_mask = (1 << tmp) - 1;
+
+ OUT_BATCH(GEN7_GPGPU_WALKER | 13);
+
+ OUT_BATCH(0); /* kernel offset */
+ OUT_BATCH(0); /* indirect data length */
+ OUT_BATCH(0); /* indirect data offset */
+
+ /* SIMD size, thread w/h/d */
+ OUT_BATCH(1 << 30 | /* SIMD16 */
+ 0 << 16 | /* depth:1 */
+ 0 << 8 | /* height:1 */
+ 0); /* width:1 */
+
+ /* thread group X */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+ OUT_BATCH(x_dim);
+
+ /* thread group Y */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+ OUT_BATCH(y_dim);
+
+ /* thread group Z */
+ OUT_BATCH(0);
+ OUT_BATCH(1);
+
+ /* right mask */
+ OUT_BATCH(right_mask);
+
+ /* bottom mask, height 1, always 0xffffffff */
+ OUT_BATCH(0xffffffff);
+}
+
+void
+gen9_emit_state_base_address(struct intel_batchbuffer *batch)
+{
+ OUT_BATCH(GEN8_STATE_BASE_ADDRESS | (19 - 2));
+
+ /* general */
+ OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
+ OUT_BATCH(0);
+
+ /* stateless data port */
+ OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
+
+ /* surface */
+ OUT_RELOC(batch->bo, I915_GEM_DOMAIN_SAMPLER, 0, BASE_ADDRESS_MODIFY);
+
+ /* dynamic */
+ OUT_RELOC(batch->bo, I915_GEM_DOMAIN_RENDER | I915_GEM_DOMAIN_INSTRUCTION,
+ 0, BASE_ADDRESS_MODIFY);
+
+ /* indirect */
+ OUT_BATCH(0);
+ OUT_BATCH(0);
+
+ /* instruction */
+ OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
+
+ /* general state buffer size */
+ OUT_BATCH(0xfffff000 | 1);
+ /* dynamic state buffer size */
+ OUT_BATCH(1 << 12 | 1);
+ /* indirect object buffer size */
+ OUT_BATCH(0xfffff000 | 1);
+ /* intruction buffer size, must set modify enable bit, otherwise it may result in GPU hang */
+ OUT_BATCH(1 << 12 | 1);
+
+ /* Bindless surface state base address */
+ OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
+ OUT_BATCH(0);
+ OUT_BATCH(0xfffff000);
+}
diff --git a/lib/gpu_fill.h b/lib/gpu_fill.h
new file mode 100644
index 00000000..973513b0
--- /dev/null
+++ b/lib/gpu_fill.h
@@ -0,0 +1,138 @@
+#include <intel_bufmgr.h>
+#include <i915_drm.h>
+
+#include "media_fill.h"
+#include "gen7_media.h"
+#include "gen8_media.h"
+#include "intel_reg.h"
+#include "drmtest.h"
+#include "intel_batchbuffer.h"
+#include "intel_chipset.h"
+#include <assert.h>
+
+uint32_t
+batch_used(struct intel_batchbuffer *batch);
+
+uint32_t
+batch_align(struct intel_batchbuffer *batch, uint32_t align);
+
+void *
+batch_alloc(struct intel_batchbuffer *batch, uint32_t size, uint32_t align);
+
+uint32_t
+batch_offset(struct intel_batchbuffer *batch, void *ptr);
+
+uint32_t
+batch_copy(struct intel_batchbuffer *batch, const void *ptr, uint32_t size, uint32_t align);
+
+void
+gen7_render_flush(struct intel_batchbuffer *batch, uint32_t batch_end);
+
+uint32_t
+gen7_fill_curbe_buffer_data(struct intel_batchbuffer *batch,
+ uint8_t color);
+
+uint32_t
+gen7_fill_surface_state(struct intel_batchbuffer *batch,
+ struct igt_buf *buf,
+ uint32_t format,
+ int is_dst);
+
+uint32_t
+gen7_fill_binding_table(struct intel_batchbuffer *batch,
+ struct igt_buf *dst);
+
+uint32_t
+gen7_fill_media_kernel(struct intel_batchbuffer *batch,
+ const uint32_t kernel[][4],
+ size_t size);
+
+uint32_t
+gen8_fill_media_kernel(struct intel_batchbuffer *batch,
+ const uint32_t kernel[][4],
+ size_t size);
+
+uint32_t
+gen7_fill_interface_descriptor(struct intel_batchbuffer *batch, struct igt_buf *dst,
+ const uint32_t kernel[][4], size_t size);
+
+void
+gen7_emit_state_base_address(struct intel_batchbuffer *batch);
+
+void
+gen7_emit_vfe_state(struct intel_batchbuffer *batch);
+
+void
+gen7_emit_vfe_state_gpgpu(struct intel_batchbuffer *batch);
+
+void
+gen7_emit_curbe_load(struct intel_batchbuffer *batch, uint32_t curbe_buffer);
+
+void
+gen7_emit_interface_descriptor_load(struct intel_batchbuffer *batch, uint32_t interface_descriptor);
+
+void
+gen7_emit_media_objects(struct intel_batchbuffer *batch,
+ unsigned x, unsigned y,
+ unsigned width, unsigned height);
+
+void
+gen7_emit_gpgpu_walk(struct intel_batchbuffer *batch,
+ unsigned x, unsigned y,
+ unsigned width, unsigned height);
+
+void
+gen8_render_flush(struct intel_batchbuffer *batch, uint32_t batch_end);
+
+uint32_t
+gen8_fill_curbe_buffer_data(struct intel_batchbuffer *batch,
+ uint8_t color);
+
+uint32_t
+gen8_fill_surface_state(struct intel_batchbuffer *batch,
+ struct igt_buf *buf,
+ uint32_t format,
+ int is_dst);
+
+uint32_t
+gen8_fill_binding_table(struct intel_batchbuffer *batch,
+ struct igt_buf *dst);
+
+uint32_t
+gen8_fill_interface_descriptor(struct intel_batchbuffer *batch, struct igt_buf *dst, const uint32_t kernel[][4], size_t size);
+
+void
+gen8_emit_state_base_address(struct intel_batchbuffer *batch);
+
+void
+gen8_emit_vfe_state(struct intel_batchbuffer *batch);
+
+void
+gen8_emit_vfe_state_gpgpu(struct intel_batchbuffer *batch);
+
+void
+gen8_emit_curbe_load(struct intel_batchbuffer *batch, uint32_t curbe_buffer);
+
+void
+gen8_emit_interface_descriptor_load(struct intel_batchbuffer *batch, uint32_t interface_descriptor);
+
+void
+gen8_emit_media_state_flush(struct intel_batchbuffer *batch);
+
+void
+gen8_emit_media_objects(struct intel_batchbuffer *batch,
+ unsigned x, unsigned y,
+ unsigned width, unsigned height);
+
+void
+gen8lp_emit_media_objects(struct intel_batchbuffer *batch,
+ unsigned x, unsigned y,
+ unsigned width, unsigned height);
+
+void
+gen8_emit_gpgpu_walk(struct intel_batchbuffer *batch,
+ unsigned x, unsigned y,
+ unsigned width, unsigned height);
+
+void
+gen9_emit_state_base_address(struct intel_batchbuffer *batch);
diff --git a/lib/media_fill_gen7.c b/lib/media_fill_gen7.c
index 6fb44798..c97555a6 100644
--- a/lib/media_fill_gen7.c
+++ b/lib/media_fill_gen7.c
@@ -5,7 +5,7 @@
#include "gen7_media.h"
#include "intel_reg.h"
#include "drmtest.h"
-
+#include "gpu_fill.h"
#include <assert.h>
static const uint32_t media_kernel[][4] = {
@@ -22,275 +22,6 @@ static const uint32_t media_kernel[][4] = {
{ 0x07800031, 0x20001ca8, 0x00000e00, 0x82000010 },
};
-static uint32_t
-batch_used(struct intel_batchbuffer *batch)
-{
- return batch->ptr - batch->buffer;
-}
-
-static uint32_t
-batch_align(struct intel_batchbuffer *batch, uint32_t align)
-{
- uint32_t offset = batch_used(batch);
- offset = ALIGN(offset, align);
- batch->ptr = batch->buffer + offset;
- return offset;
-}
-
-static void *
-batch_alloc(struct intel_batchbuffer *batch, uint32_t size, uint32_t align)
-{
- uint32_t offset = batch_align(batch, align);
- batch->ptr += size;
- return memset(batch->buffer + offset, 0, size);
-}
-
-static uint32_t
-batch_offset(struct intel_batchbuffer *batch, void *ptr)
-{
- return (uint8_t *)ptr - batch->buffer;
-}
-
-static uint32_t
-batch_copy(struct intel_batchbuffer *batch, const void *ptr, uint32_t size, uint32_t align)
-{
- return batch_offset(batch, memcpy(batch_alloc(batch, size, align), ptr, size));
-}
-
-static void
-gen7_render_flush(struct intel_batchbuffer *batch, uint32_t batch_end)
-{
- int ret;
-
- ret = drm_intel_bo_subdata(batch->bo, 0, 4096, batch->buffer);
- if (ret == 0)
- ret = drm_intel_bo_mrb_exec(batch->bo, batch_end,
- NULL, 0, 0, 0);
- igt_assert(ret == 0);
-}
-
-static uint32_t
-gen7_fill_curbe_buffer_data(struct intel_batchbuffer *batch,
- uint8_t color)
-{
- uint8_t *curbe_buffer;
- uint32_t offset;
-
- curbe_buffer = batch_alloc(batch, sizeof(uint32_t) * 8, 64);
- offset = batch_offset(batch, curbe_buffer);
- *curbe_buffer = color;
-
- return offset;
-}
-
-static uint32_t
-gen7_fill_surface_state(struct intel_batchbuffer *batch,
- struct igt_buf *buf,
- uint32_t format,
- int is_dst)
-{
- struct gen7_surface_state *ss;
- uint32_t write_domain, read_domain, offset;
- int ret;
-
- if (is_dst) {
- write_domain = read_domain = I915_GEM_DOMAIN_RENDER;
- } else {
- write_domain = 0;
- read_domain = I915_GEM_DOMAIN_SAMPLER;
- }
-
- ss = batch_alloc(batch, sizeof(*ss), 64);
- offset = batch_offset(batch, ss);
-
- ss->ss0.surface_type = GEN7_SURFACE_2D;
- ss->ss0.surface_format = format;
- ss->ss0.render_cache_read_write = 1;
-
- if (buf->tiling == I915_TILING_X)
- ss->ss0.tiled_mode = 2;
- else if (buf->tiling == I915_TILING_Y)
- ss->ss0.tiled_mode = 3;
-
- ss->ss1.base_addr = buf->bo->offset;
- ret = drm_intel_bo_emit_reloc(batch->bo,
- batch_offset(batch, ss) + 4,
- buf->bo, 0,
- read_domain, write_domain);
- igt_assert(ret == 0);
-
- ss->ss2.height = igt_buf_height(buf) - 1;
- ss->ss2.width = igt_buf_width(buf) - 1;
-
- ss->ss3.pitch = buf->stride - 1;
-
- ss->ss7.shader_chanel_select_r = 4;
- ss->ss7.shader_chanel_select_g = 5;
- ss->ss7.shader_chanel_select_b = 6;
- ss->ss7.shader_chanel_select_a = 7;
-
- return offset;
-}
-
-static uint32_t
-gen7_fill_binding_table(struct intel_batchbuffer *batch,
- struct igt_buf *dst)
-{
- uint32_t *binding_table, offset;
-
- binding_table = batch_alloc(batch, 32, 64);
- offset = batch_offset(batch, binding_table);
-
- binding_table[0] = gen7_fill_surface_state(batch, dst, GEN7_SURFACEFORMAT_R8_UNORM, 1);
-
- return offset;
-}
-
-static uint32_t
-gen7_fill_media_kernel(struct intel_batchbuffer *batch,
- const uint32_t kernel[][4],
- size_t size)
-{
- uint32_t offset;
-
- offset = batch_copy(batch, kernel, size, 64);
-
- return offset;
-}
-
-static uint32_t
-gen7_fill_interface_descriptor(struct intel_batchbuffer *batch, struct igt_buf *dst,
- const uint32_t kernel[][4], size_t size)
-{
- struct gen7_interface_descriptor_data *idd;
- uint32_t offset;
- uint32_t binding_table_offset, kernel_offset;
-
- binding_table_offset = gen7_fill_binding_table(batch, dst);
- kernel_offset = gen7_fill_media_kernel(batch, kernel, size);
-
- idd = batch_alloc(batch, sizeof(*idd), 64);
- offset = batch_offset(batch, idd);
-
- idd->desc0.kernel_start_pointer = (kernel_offset >> 6);
-
- idd->desc1.single_program_flow = 1;
- idd->desc1.floating_point_mode = GEN7_FLOATING_POINT_IEEE_754;
-
- idd->desc2.sampler_count = 0; /* 0 samplers used */
- idd->desc2.sampler_state_pointer = 0;
-
- idd->desc3.binding_table_entry_count = 0;
- idd->desc3.binding_table_pointer = (binding_table_offset >> 5);
-
- idd->desc4.constant_urb_entry_read_offset = 0;
- idd->desc4.constant_urb_entry_read_length = 1; /* grf 1 */
-
- return offset;
-}
-
-static void
-gen7_emit_state_base_address(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN7_STATE_BASE_ADDRESS | (10 - 2));
-
- /* general */
- OUT_BATCH(0);
-
- /* surface */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* dynamic */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* indirect */
- OUT_BATCH(0);
-
- /* instruction */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* general/dynamic/indirect/instruction access Bound */
- OUT_BATCH(0);
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
- OUT_BATCH(0);
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
-}
-
-static void
-gen7_emit_vfe_state(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN7_MEDIA_VFE_STATE | (8 - 2));
-
- /* scratch buffer */
- OUT_BATCH(0);
-
- /* number of threads & urb entries */
- OUT_BATCH(1 << 16 |
- 2 << 8);
-
- OUT_BATCH(0);
-
- /* urb entry size & curbe size */
- OUT_BATCH(2 << 16 | /* in 256 bits unit */
- 2); /* in 256 bits unit */
-
- /* scoreboard */
- OUT_BATCH(0);
- OUT_BATCH(0);
- OUT_BATCH(0);
-}
-
-static void
-gen7_emit_curbe_load(struct intel_batchbuffer *batch, uint32_t curbe_buffer)
-{
- OUT_BATCH(GEN7_MEDIA_CURBE_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* curbe total data length */
- OUT_BATCH(64);
- /* curbe data start address, is relative to the dynamics base address */
- OUT_BATCH(curbe_buffer);
-}
-
-static void
-gen7_emit_interface_descriptor_load(struct intel_batchbuffer *batch, uint32_t interface_descriptor)
-{
- OUT_BATCH(GEN7_MEDIA_INTERFACE_DESCRIPTOR_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* interface descriptor data length */
- OUT_BATCH(sizeof(struct gen7_interface_descriptor_data));
- /* interface descriptor address, is relative to the dynamics base address */
- OUT_BATCH(interface_descriptor);
-}
-
-static void
-gen7_emit_media_objects(struct intel_batchbuffer *batch,
- unsigned x, unsigned y,
- unsigned width, unsigned height)
-{
- int i, j;
-
- for (i = 0; i < width / 16; i++) {
- for (j = 0; j < height / 16; j++) {
- OUT_BATCH(GEN7_MEDIA_OBJECT | (8 - 2));
-
- /* interface descriptor offset */
- OUT_BATCH(0);
-
- /* without indirect data */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* scoreboard */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* inline data (xoffset, yoffset) */
- OUT_BATCH(x + i * 16);
- OUT_BATCH(y + j * 16);
- }
- }
-}
-
/*
* This sets up the media pipeline,
*
diff --git a/lib/media_fill_gen8.c b/lib/media_fill_gen8.c
index 4a8fe5a2..4270997e 100644
--- a/lib/media_fill_gen8.c
+++ b/lib/media_fill_gen8.c
@@ -5,7 +5,7 @@
#include "gen8_media.h"
#include "intel_reg.h"
#include "drmtest.h"
-
+#include "gpu_fill.h"
#include <assert.h>
@@ -23,293 +23,7 @@ static const uint32_t media_kernel[][4] = {
{ 0x07800031, 0x20000a40, 0x0e000e00, 0x82000010 },
};
-static uint32_t
-batch_used(struct intel_batchbuffer *batch)
-{
- return batch->ptr - batch->buffer;
-}
-
-static uint32_t
-batch_align(struct intel_batchbuffer *batch, uint32_t align)
-{
- uint32_t offset = batch_used(batch);
- offset = ALIGN(offset, align);
- batch->ptr = batch->buffer + offset;
- return offset;
-}
-
-static void *
-batch_alloc(struct intel_batchbuffer *batch, uint32_t size, uint32_t align)
-{
- uint32_t offset = batch_align(batch, align);
- batch->ptr += size;
- return memset(batch->buffer + offset, 0, size);
-}
-
-static uint32_t
-batch_offset(struct intel_batchbuffer *batch, void *ptr)
-{
- return (uint8_t *)ptr - batch->buffer;
-}
-
-static uint32_t
-batch_copy(struct intel_batchbuffer *batch, const void *ptr, uint32_t size, uint32_t align)
-{
- return batch_offset(batch, memcpy(batch_alloc(batch, size, align), ptr, size));
-}
-
-static void
-gen8_render_flush(struct intel_batchbuffer *batch, uint32_t batch_end)
-{
- int ret;
-
- ret = drm_intel_bo_subdata(batch->bo, 0, 4096, batch->buffer);
- if (ret == 0)
- ret = drm_intel_bo_mrb_exec(batch->bo, batch_end,
- NULL, 0, 0, 0);
- igt_assert(ret == 0);
-}
-
-static uint32_t
-gen8_fill_curbe_buffer_data(struct intel_batchbuffer *batch,
- uint8_t color)
-{
- uint8_t *curbe_buffer;
- uint32_t offset;
-
- curbe_buffer = batch_alloc(batch, sizeof(uint32_t) * 8, 64);
- offset = batch_offset(batch, curbe_buffer);
- *curbe_buffer = color;
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_surface_state(struct intel_batchbuffer *batch,
- struct igt_buf *buf,
- uint32_t format,
- int is_dst)
-{
- struct gen8_surface_state *ss;
- uint32_t write_domain, read_domain, offset;
- int ret;
-
- if (is_dst) {
- write_domain = read_domain = I915_GEM_DOMAIN_RENDER;
- } else {
- write_domain = 0;
- read_domain = I915_GEM_DOMAIN_SAMPLER;
- }
-
- ss = batch_alloc(batch, sizeof(*ss), 64);
- offset = batch_offset(batch, ss);
-
- ss->ss0.surface_type = GEN8_SURFACE_2D;
- ss->ss0.surface_format = format;
- ss->ss0.render_cache_read_write = 1;
- ss->ss0.vertical_alignment = 1; /* align 4 */
- ss->ss0.horizontal_alignment = 1; /* align 4 */
-
- if (buf->tiling == I915_TILING_X)
- ss->ss0.tiled_mode = 2;
- else if (buf->tiling == I915_TILING_Y)
- ss->ss0.tiled_mode = 3;
-
- ss->ss8.base_addr = buf->bo->offset;
-
- ret = drm_intel_bo_emit_reloc(batch->bo,
- batch_offset(batch, ss) + 8 * 4,
- buf->bo, 0,
- read_domain, write_domain);
- igt_assert(ret == 0);
-
- ss->ss2.height = igt_buf_height(buf) - 1;
- ss->ss2.width = igt_buf_width(buf) - 1;
- ss->ss3.pitch = buf->stride - 1;
-
- ss->ss7.shader_chanel_select_r = 4;
- ss->ss7.shader_chanel_select_g = 5;
- ss->ss7.shader_chanel_select_b = 6;
- ss->ss7.shader_chanel_select_a = 7;
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_binding_table(struct intel_batchbuffer *batch,
- struct igt_buf *dst)
-{
- uint32_t *binding_table, offset;
-
- binding_table = batch_alloc(batch, 32, 64);
- offset = batch_offset(batch, binding_table);
-
- binding_table[0] = gen8_fill_surface_state(batch, dst, GEN8_SURFACEFORMAT_R8_UNORM, 1);
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_media_kernel(struct intel_batchbuffer *batch,
- const uint32_t kernel[][4],
- size_t size)
-{
- uint32_t offset;
-
- offset = batch_copy(batch, kernel, size, 64);
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_interface_descriptor(struct intel_batchbuffer *batch, struct igt_buf *dst)
-{
- struct gen8_interface_descriptor_data *idd;
- uint32_t offset;
- uint32_t binding_table_offset, kernel_offset;
-
- binding_table_offset = gen8_fill_binding_table(batch, dst);
- kernel_offset = gen8_fill_media_kernel(batch, media_kernel, sizeof(media_kernel));
-
- idd = batch_alloc(batch, sizeof(*idd), 64);
- offset = batch_offset(batch, idd);
-
- idd->desc0.kernel_start_pointer = (kernel_offset >> 6);
-
- idd->desc2.single_program_flow = 1;
- idd->desc2.floating_point_mode = GEN8_FLOATING_POINT_IEEE_754;
-
- idd->desc3.sampler_count = 0; /* 0 samplers used */
- idd->desc3.sampler_state_pointer = 0;
-
- idd->desc4.binding_table_entry_count = 0;
- idd->desc4.binding_table_pointer = (binding_table_offset >> 5);
-
- idd->desc5.constant_urb_entry_read_offset = 0;
- idd->desc5.constant_urb_entry_read_length = 1; /* grf 1 */
-
- return offset;
-}
-
-static void
-gen8_emit_state_base_address(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_STATE_BASE_ADDRESS | (16 - 2));
-
- /* general */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
- OUT_BATCH(0);
-
- /* stateless data port */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
-
- /* surface */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_SAMPLER, 0, BASE_ADDRESS_MODIFY);
-
- /* dynamic */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_RENDER | I915_GEM_DOMAIN_INSTRUCTION,
- 0, BASE_ADDRESS_MODIFY);
-
- /* indirect */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* instruction */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* general state buffer size */
- OUT_BATCH(0xfffff000 | 1);
- /* dynamic state buffer size */
- OUT_BATCH(1 << 12 | 1);
- /* indirect object buffer size */
- OUT_BATCH(0xfffff000 | 1);
- /* intruction buffer size, must set modify enable bit, otherwise it may result in GPU hang */
- OUT_BATCH(1 << 12 | 1);
-}
-
-static void
-gen8_emit_vfe_state(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_MEDIA_VFE_STATE | (9 - 2));
-
- /* scratch buffer */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* number of threads & urb entries */
- OUT_BATCH(1 << 16 |
- 2 << 8);
-
- OUT_BATCH(0);
-
- /* urb entry size & curbe size */
- OUT_BATCH(2 << 16 |
- 2);
-
- /* scoreboard */
- OUT_BATCH(0);
- OUT_BATCH(0);
- OUT_BATCH(0);
-}
-
-static void
-gen8_emit_curbe_load(struct intel_batchbuffer *batch, uint32_t curbe_buffer)
-{
- OUT_BATCH(GEN8_MEDIA_CURBE_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* curbe total data length */
- OUT_BATCH(64);
- /* curbe data start address, is relative to the dynamics base address */
- OUT_BATCH(curbe_buffer);
-}
-static void
-gen8_emit_interface_descriptor_load(struct intel_batchbuffer *batch, uint32_t interface_descriptor)
-{
- OUT_BATCH(GEN8_MEDIA_INTERFACE_DESCRIPTOR_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* interface descriptor data length */
- OUT_BATCH(sizeof(struct gen8_interface_descriptor_data));
- /* interface descriptor address, is relative to the dynamics base address */
- OUT_BATCH(interface_descriptor);
-}
-
-static void
-gen8_emit_media_state_flush(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_MEDIA_STATE_FLUSH | (2 - 2));
- OUT_BATCH(0);
-}
-
-static void
-gen8_emit_media_objects(struct intel_batchbuffer *batch,
- unsigned x, unsigned y,
- unsigned width, unsigned height)
-{
- int i, j;
-
- for (i = 0; i < width / 16; i++) {
- for (j = 0; j < height / 16; j++) {
- OUT_BATCH(GEN8_MEDIA_OBJECT | (8 - 2));
-
- /* interface descriptor offset */
- OUT_BATCH(0);
-
- /* without indirect data */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* scoreboard */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* inline data (xoffset, yoffset) */
- OUT_BATCH(x + i * 16);
- OUT_BATCH(y + j * 16);
- gen8_emit_media_state_flush(batch);
- }
- }
-}
/*
* This sets up the media pipeline,
@@ -349,7 +63,7 @@ gen8_media_fillfunc(struct intel_batchbuffer *batch,
batch->ptr = &batch->buffer[BATCH_STATE_SPLIT];
curbe_buffer = gen8_fill_curbe_buffer_data(batch, color);
- interface_descriptor = gen8_fill_interface_descriptor(batch, dst);
+ interface_descriptor = gen8_fill_interface_descriptor(batch, dst, media_kernel, sizeof(media_kernel));
igt_assert(batch->ptr < &batch->buffer[4095]);
/* media pipeline */
diff --git a/lib/media_fill_gen8lp.c b/lib/media_fill_gen8lp.c
index 1f8a4adc..dcc11982 100644
--- a/lib/media_fill_gen8lp.c
+++ b/lib/media_fill_gen8lp.c
@@ -5,7 +5,7 @@
#include "gen8_media.h"
#include "intel_reg.h"
#include "drmtest.h"
-
+#include "gpu_fill.h"
#include <assert.h>
@@ -23,286 +23,6 @@ static const uint32_t media_kernel[][4] = {
{ 0x07800031, 0x20000a40, 0x0e000e00, 0x82000010 },
};
-static uint32_t
-batch_used(struct intel_batchbuffer *batch)
-{
- return batch->ptr - batch->buffer;
-}
-
-static uint32_t
-batch_align(struct intel_batchbuffer *batch, uint32_t align)
-{
- uint32_t offset = batch_used(batch);
- offset = ALIGN(offset, align);
- batch->ptr = batch->buffer + offset;
- return offset;
-}
-
-static void *
-batch_alloc(struct intel_batchbuffer *batch, uint32_t size, uint32_t align)
-{
- uint32_t offset = batch_align(batch, align);
- batch->ptr += size;
- return memset(batch->buffer + offset, 0, size);
-}
-
-static uint32_t
-batch_offset(struct intel_batchbuffer *batch, void *ptr)
-{
- return (uint8_t *)ptr - batch->buffer;
-}
-
-static uint32_t
-batch_copy(struct intel_batchbuffer *batch, const void *ptr, uint32_t size, uint32_t align)
-{
- return batch_offset(batch, memcpy(batch_alloc(batch, size, align), ptr, size));
-}
-
-static void
-gen8_render_flush(struct intel_batchbuffer *batch, uint32_t batch_end)
-{
- int ret;
-
- ret = drm_intel_bo_subdata(batch->bo, 0, 4096, batch->buffer);
- if (ret == 0)
- ret = drm_intel_bo_mrb_exec(batch->bo, batch_end,
- NULL, 0, 0, 0);
- igt_assert(ret == 0);
-}
-
-static uint32_t
-gen8_fill_curbe_buffer_data(struct intel_batchbuffer *batch,
- uint8_t color)
-{
- uint8_t *curbe_buffer;
- uint32_t offset;
-
- curbe_buffer = batch_alloc(batch, sizeof(uint32_t) * 8, 64);
- offset = batch_offset(batch, curbe_buffer);
- *curbe_buffer = color;
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_surface_state(struct intel_batchbuffer *batch,
- struct igt_buf *buf,
- uint32_t format,
- int is_dst)
-{
- struct gen8_surface_state *ss;
- uint32_t write_domain, read_domain, offset;
- int ret;
-
- if (is_dst) {
- write_domain = read_domain = I915_GEM_DOMAIN_RENDER;
- } else {
- write_domain = 0;
- read_domain = I915_GEM_DOMAIN_SAMPLER;
- }
-
- ss = batch_alloc(batch, sizeof(*ss), 64);
- offset = batch_offset(batch, ss);
-
- ss->ss0.surface_type = GEN8_SURFACE_2D;
- ss->ss0.surface_format = format;
- ss->ss0.render_cache_read_write = 1;
- ss->ss0.vertical_alignment = 1; /* align 4 */
- ss->ss0.horizontal_alignment = 1; /* align 4 */
-
- if (buf->tiling == I915_TILING_X)
- ss->ss0.tiled_mode = 2;
- else if (buf->tiling == I915_TILING_Y)
- ss->ss0.tiled_mode = 3;
-
- ss->ss8.base_addr = buf->bo->offset;
-
- ret = drm_intel_bo_emit_reloc(batch->bo,
- batch_offset(batch, ss) + 8 * 4,
- buf->bo, 0,
- read_domain, write_domain);
- igt_assert(ret == 0);
-
- ss->ss2.height = igt_buf_height(buf) - 1;
- ss->ss2.width = igt_buf_width(buf) - 1;
- ss->ss3.pitch = buf->stride - 1;
-
- ss->ss7.shader_chanel_select_r = 4;
- ss->ss7.shader_chanel_select_g = 5;
- ss->ss7.shader_chanel_select_b = 6;
- ss->ss7.shader_chanel_select_a = 7;
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_binding_table(struct intel_batchbuffer *batch,
- struct igt_buf *dst)
-{
- uint32_t *binding_table, offset;
-
- binding_table = batch_alloc(batch, 32, 64);
- offset = batch_offset(batch, binding_table);
-
- binding_table[0] = gen8_fill_surface_state(batch, dst, GEN8_SURFACEFORMAT_R8_UNORM, 1);
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_media_kernel(struct intel_batchbuffer *batch,
- const uint32_t kernel[][4],
- size_t size)
-{
- uint32_t offset;
-
- offset = batch_copy(batch, kernel, size, 64);
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_interface_descriptor(struct intel_batchbuffer *batch, struct igt_buf *dst)
-{
- struct gen8_interface_descriptor_data *idd;
- uint32_t offset;
- uint32_t binding_table_offset, kernel_offset;
-
- binding_table_offset = gen8_fill_binding_table(batch, dst);
- kernel_offset = gen8_fill_media_kernel(batch, media_kernel, sizeof(media_kernel));
-
- idd = batch_alloc(batch, sizeof(*idd), 64);
- offset = batch_offset(batch, idd);
-
- idd->desc0.kernel_start_pointer = (kernel_offset >> 6);
-
- idd->desc2.single_program_flow = 1;
- idd->desc2.floating_point_mode = GEN8_FLOATING_POINT_IEEE_754;
-
- idd->desc3.sampler_count = 0; /* 0 samplers used */
- idd->desc3.sampler_state_pointer = 0;
-
- idd->desc4.binding_table_entry_count = 0;
- idd->desc4.binding_table_pointer = (binding_table_offset >> 5);
-
- idd->desc5.constant_urb_entry_read_offset = 0;
- idd->desc5.constant_urb_entry_read_length = 1; /* grf 1 */
-
- return offset;
-}
-
-static void
-gen8_emit_state_base_address(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_STATE_BASE_ADDRESS | (16 - 2));
-
- /* general */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
- OUT_BATCH(0);
-
- /* stateless data port */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
-
- /* surface */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_SAMPLER, 0, BASE_ADDRESS_MODIFY);
-
- /* dynamic */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_RENDER | I915_GEM_DOMAIN_INSTRUCTION,
- 0, BASE_ADDRESS_MODIFY);
-
- /* indirect */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* instruction */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* general state buffer size */
- OUT_BATCH(0xfffff000 | 1);
- /* dynamic state buffer size */
- OUT_BATCH(1 << 12 | 1);
- /* indirect object buffer size */
- OUT_BATCH(0xfffff000 | 1);
- /* intruction buffer size, must set modify enable bit, otherwise it may result in GPU hang */
- OUT_BATCH(1 << 12 | 1);
-}
-
-static void
-gen8_emit_vfe_state(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_MEDIA_VFE_STATE | (9 - 2));
-
- /* scratch buffer */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* number of threads & urb entries */
- OUT_BATCH(1 << 16 |
- 2 << 8);
-
- OUT_BATCH(0);
-
- /* urb entry size & curbe size */
- OUT_BATCH(2 << 16 |
- 2);
-
- /* scoreboard */
- OUT_BATCH(0);
- OUT_BATCH(0);
- OUT_BATCH(0);
-}
-
-static void
-gen8_emit_curbe_load(struct intel_batchbuffer *batch, uint32_t curbe_buffer)
-{
- OUT_BATCH(GEN8_MEDIA_CURBE_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* curbe total data length */
- OUT_BATCH(64);
- /* curbe data start address, is relative to the dynamics base address */
- OUT_BATCH(curbe_buffer);
-}
-
-static void
-gen8_emit_interface_descriptor_load(struct intel_batchbuffer *batch, uint32_t interface_descriptor)
-{
- OUT_BATCH(GEN8_MEDIA_INTERFACE_DESCRIPTOR_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* interface descriptor data length */
- OUT_BATCH(sizeof(struct gen8_interface_descriptor_data));
- /* interface descriptor address, is relative to the dynamics base address */
- OUT_BATCH(interface_descriptor);
-}
-
-static void
-gen8lp_emit_media_objects(struct intel_batchbuffer *batch,
- unsigned x, unsigned y,
- unsigned width, unsigned height)
-{
- int i, j;
-
- for (i = 0; i < width / 16; i++) {
- for (j = 0; j < height / 16; j++) {
- OUT_BATCH(GEN8_MEDIA_OBJECT | (8 - 2));
-
- /* interface descriptor offset */
- OUT_BATCH(0);
-
- /* without indirect data */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* scoreboard */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* inline data (xoffset, yoffset) */
- OUT_BATCH(x + i * 16);
- OUT_BATCH(y + j * 16);
- }
- }
-}
-
/*
* This sets up the media pipeline,
*
@@ -341,7 +61,7 @@ gen8lp_media_fillfunc(struct intel_batchbuffer *batch,
batch->ptr = &batch->buffer[BATCH_STATE_SPLIT];
curbe_buffer = gen8_fill_curbe_buffer_data(batch, color);
- interface_descriptor = gen8_fill_interface_descriptor(batch, dst);
+ interface_descriptor = gen8_fill_interface_descriptor(batch, dst, media_kernel, sizeof(media_kernel));
igt_assert(batch->ptr < &batch->buffer[4095]);
/* media pipeline */
diff --git a/lib/media_fill_gen9.c b/lib/media_fill_gen9.c
index 3fd21819..6accdbe4 100644
--- a/lib/media_fill_gen9.c
+++ b/lib/media_fill_gen9.c
@@ -4,11 +4,9 @@
#include "media_fill.h"
#include "gen8_media.h"
#include "intel_reg.h"
-
+#include "gpu_fill.h"
#include <assert.h>
-#define ALIGN(x, y) (((x) + (y)-1) & ~((y)-1))
-
static const uint32_t media_kernel[][4] = {
{ 0x00400001, 0x20202288, 0x00000020, 0x00000000 },
{ 0x00600001, 0x20800208, 0x008d0000, 0x00000000 },
@@ -23,298 +21,6 @@ static const uint32_t media_kernel[][4] = {
{ 0x07800031, 0x20000a40, 0x0e000e00, 0x82000010 },
};
-static uint32_t
-batch_used(struct intel_batchbuffer *batch)
-{
- return batch->ptr - batch->buffer;
-}
-
-static uint32_t
-batch_align(struct intel_batchbuffer *batch, uint32_t align)
-{
- uint32_t offset = batch_used(batch);
- offset = ALIGN(offset, align);
- batch->ptr = batch->buffer + offset;
- return offset;
-}
-
-static void *
-batch_alloc(struct intel_batchbuffer *batch, uint32_t size, uint32_t align)
-{
- uint32_t offset = batch_align(batch, align);
- batch->ptr += size;
- return memset(batch->buffer + offset, 0, size);
-}
-
-static uint32_t
-batch_offset(struct intel_batchbuffer *batch, void *ptr)
-{
- return (uint8_t *)ptr - batch->buffer;
-}
-
-static uint32_t
-batch_copy(struct intel_batchbuffer *batch, const void *ptr, uint32_t size, uint32_t align)
-{
- return batch_offset(batch, memcpy(batch_alloc(batch, size, align), ptr, size));
-}
-
-static void
-gen8_render_flush(struct intel_batchbuffer *batch, uint32_t batch_end)
-{
- int ret;
-
- ret = drm_intel_bo_subdata(batch->bo, 0, 4096, batch->buffer);
- if (ret == 0)
- ret = drm_intel_bo_mrb_exec(batch->bo, batch_end,
- NULL, 0, 0, 0);
- assert(ret == 0);
-}
-
-static uint32_t
-gen8_fill_curbe_buffer_data(struct intel_batchbuffer *batch,
- uint8_t color)
-{
- uint8_t *curbe_buffer;
- uint32_t offset;
-
- curbe_buffer = batch_alloc(batch, sizeof(uint32_t) * 8, 64);
- offset = batch_offset(batch, curbe_buffer);
- *curbe_buffer = color;
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_surface_state(struct intel_batchbuffer *batch,
- struct igt_buf *buf,
- uint32_t format,
- int is_dst)
-{
- struct gen8_surface_state *ss;
- uint32_t write_domain, read_domain, offset;
- int ret;
-
- if (is_dst) {
- write_domain = read_domain = I915_GEM_DOMAIN_RENDER;
- } else {
- write_domain = 0;
- read_domain = I915_GEM_DOMAIN_SAMPLER;
- }
-
- ss = batch_alloc(batch, sizeof(*ss), 64);
- offset = batch_offset(batch, ss);
-
- ss->ss0.surface_type = GEN8_SURFACE_2D;
- ss->ss0.surface_format = format;
- ss->ss0.render_cache_read_write = 1;
- ss->ss0.vertical_alignment = 1; /* align 4 */
- ss->ss0.horizontal_alignment = 1; /* align 4 */
-
- if (buf->tiling == I915_TILING_X)
- ss->ss0.tiled_mode = 2;
- else if (buf->tiling == I915_TILING_Y)
- ss->ss0.tiled_mode = 3;
-
- ss->ss8.base_addr = buf->bo->offset;
-
- ret = drm_intel_bo_emit_reloc(batch->bo,
- batch_offset(batch, ss) + 8 * 4,
- buf->bo, 0,
- read_domain, write_domain);
- assert(ret == 0);
-
- ss->ss2.height = igt_buf_height(buf) - 1;
- ss->ss2.width = igt_buf_width(buf) - 1;
- ss->ss3.pitch = buf->stride - 1;
-
- ss->ss7.shader_chanel_select_r = 4;
- ss->ss7.shader_chanel_select_g = 5;
- ss->ss7.shader_chanel_select_b = 6;
- ss->ss7.shader_chanel_select_a = 7;
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_binding_table(struct intel_batchbuffer *batch,
- struct igt_buf *dst)
-{
- uint32_t *binding_table, offset;
-
- binding_table = batch_alloc(batch, 32, 64);
- offset = batch_offset(batch, binding_table);
-
- binding_table[0] = gen8_fill_surface_state(batch, dst, GEN8_SURFACEFORMAT_R8_UNORM, 1);
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_media_kernel(struct intel_batchbuffer *batch,
- const uint32_t kernel[][4],
- size_t size)
-{
- uint32_t offset;
-
- offset = batch_copy(batch, kernel, size, 64);
-
- return offset;
-}
-
-static uint32_t
-gen8_fill_interface_descriptor(struct intel_batchbuffer *batch, struct igt_buf *dst)
-{
- struct gen8_interface_descriptor_data *idd;
- uint32_t offset;
- uint32_t binding_table_offset, kernel_offset;
-
- binding_table_offset = gen8_fill_binding_table(batch, dst);
- kernel_offset = gen8_fill_media_kernel(batch, media_kernel, sizeof(media_kernel));
-
- idd = batch_alloc(batch, sizeof(*idd), 64);
- offset = batch_offset(batch, idd);
-
- idd->desc0.kernel_start_pointer = (kernel_offset >> 6);
-
- idd->desc2.single_program_flow = 1;
- idd->desc2.floating_point_mode = GEN8_FLOATING_POINT_IEEE_754;
-
- idd->desc3.sampler_count = 0; /* 0 samplers used */
- idd->desc3.sampler_state_pointer = 0;
-
- idd->desc4.binding_table_entry_count = 0;
- idd->desc4.binding_table_pointer = (binding_table_offset >> 5);
-
- idd->desc5.constant_urb_entry_read_offset = 0;
- idd->desc5.constant_urb_entry_read_length = 1; /* grf 1 */
-
- return offset;
-}
-
-static void
-gen9_emit_state_base_address(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_STATE_BASE_ADDRESS | (19 - 2));
-
- /* general */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
- OUT_BATCH(0);
-
- /* stateless data port */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
-
- /* surface */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_SAMPLER, 0, BASE_ADDRESS_MODIFY);
-
- /* dynamic */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_RENDER | I915_GEM_DOMAIN_INSTRUCTION,
- 0, BASE_ADDRESS_MODIFY);
-
- /* indirect */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* instruction */
- OUT_RELOC(batch->bo, I915_GEM_DOMAIN_INSTRUCTION, 0, BASE_ADDRESS_MODIFY);
-
- /* general state buffer size */
- OUT_BATCH(0xfffff000 | 1);
- /* dynamic state buffer size */
- OUT_BATCH(1 << 12 | 1);
- /* indirect object buffer size */
- OUT_BATCH(0xfffff000 | 1);
- /* intruction buffer size, must set modify enable bit, otherwise it may result in GPU hang */
- OUT_BATCH(1 << 12 | 1);
-
- /* Bindless surface state base address */
- OUT_BATCH(0 | BASE_ADDRESS_MODIFY);
- OUT_BATCH(0);
- OUT_BATCH(0xfffff000);
-}
-
-static void
-gen8_emit_vfe_state(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_MEDIA_VFE_STATE | (9 - 2));
-
- /* scratch buffer */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* number of threads & urb entries */
- OUT_BATCH(1 << 16 |
- 2 << 8);
-
- OUT_BATCH(0);
-
- /* urb entry size & curbe size */
- OUT_BATCH(2 << 16 |
- 2);
-
- /* scoreboard */
- OUT_BATCH(0);
- OUT_BATCH(0);
- OUT_BATCH(0);
-}
-
-static void
-gen8_emit_curbe_load(struct intel_batchbuffer *batch, uint32_t curbe_buffer)
-{
- OUT_BATCH(GEN8_MEDIA_CURBE_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* curbe total data length */
- OUT_BATCH(64);
- /* curbe data start address, is relative to the dynamics base address */
- OUT_BATCH(curbe_buffer);
-}
-
-static void
-gen8_emit_interface_descriptor_load(struct intel_batchbuffer *batch, uint32_t interface_descriptor)
-{
- OUT_BATCH(GEN8_MEDIA_INTERFACE_DESCRIPTOR_LOAD | (4 - 2));
- OUT_BATCH(0);
- /* interface descriptor data length */
- OUT_BATCH(sizeof(struct gen8_interface_descriptor_data));
- /* interface descriptor address, is relative to the dynamics base address */
- OUT_BATCH(interface_descriptor);
-}
-
-static void
-gen8_emit_media_state_flush(struct intel_batchbuffer *batch)
-{
- OUT_BATCH(GEN8_MEDIA_STATE_FLUSH | (2 - 2));
- OUT_BATCH(0);
-}
-
-static void
-gen8_emit_media_objects(struct intel_batchbuffer *batch,
- unsigned x, unsigned y,
- unsigned width, unsigned height)
-{
- int i, j;
-
- for (i = 0; i < width / 16; i++) {
- for (j = 0; j < height / 16; j++) {
- OUT_BATCH(GEN8_MEDIA_OBJECT | (8 - 2));
-
- /* interface descriptor offset */
- OUT_BATCH(0);
-
- /* without indirect data */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* scoreboard */
- OUT_BATCH(0);
- OUT_BATCH(0);
-
- /* inline data (xoffset, yoffset) */
- OUT_BATCH(x + i * 16);
- OUT_BATCH(y + j * 16);
- gen8_emit_media_state_flush(batch);
- }
- }
-}
/*
* This sets up the media pipeline,
@@ -354,7 +60,7 @@ gen9_media_fillfunc(struct intel_batchbuffer *batch,
batch->ptr = &batch->buffer[BATCH_STATE_SPLIT];
curbe_buffer = gen8_fill_curbe_buffer_data(batch, color);
- interface_descriptor = gen8_fill_interface_descriptor(batch, dst);
+ interface_descriptor = gen8_fill_interface_descriptor(batch, dst, media_kernel, sizeof(media_kernel));
assert(batch->ptr < &batch->buffer[4095]);
/* media pipeline */
diff --git a/lib/meson.build b/lib/meson.build
index b3b8b14a..385e08b9 100644
--- a/lib/meson.build
+++ b/lib/meson.build
@@ -30,6 +30,7 @@ lib_sources = [
'media_fill_gen9.c',
'media_spin.c',
'gpgpu_fill.c',
+ 'gpu_fill.c',
'rendercopy_i915.c',
'rendercopy_i830.c',
'rendercopy_gen6.c',
--
2.14.3
More information about the igt-dev
mailing list