brkho
/
mesa
Watch 1
Star 0
Fork 0
Code Releases 562 Wiki Activity
Browse Source

anv: s/anv_batch_emit_blk/anv_batch_emit/ 

Acked-by: Kristian Høgsberg <krh@bitplanet.net>
tags/12.0-branchpoint
Jason Ekstrand 9 years ago
parent
0a45395902
commit
50018522d2
11 changed files with 133 additions and 138 deletions
Split View Show Diff Stats
  1. 3
    3
      src/intel/vulkan/anv_batch_chain.c
  2. 4
    4
      src/intel/vulkan/anv_device.c
  3. 1
    1
      src/intel/vulkan/anv_private.h
  4. 19
    20
      src/intel/vulkan/gen7_cmd_buffer.c
  5. 13
    14
      src/intel/vulkan/gen7_pipeline.c
  6. 18
    20
      src/intel/vulkan/gen8_cmd_buffer.c
  7. 15
    16
      src/intel/vulkan/gen8_pipeline.c
  8. 43
    43
      src/intel/vulkan/genX_cmd_buffer.c
  9. 1
    1
      src/intel/vulkan/genX_pipeline.c
  10. 6
    6
      src/intel/vulkan/genX_pipeline_util.h
  11. 10
    10
      src/intel/vulkan/genX_state.c

+ 3
- 3
src/intel/vulkan/anv_batch_chain.c View File

@@ -455,7 +455,7 @@ emit_batch_buffer_start(struct anv_cmd_buffer *cmd_buffer,
const uint32_t gen8_length =
GEN8_MI_BATCH_BUFFER_START_length - GEN8_MI_BATCH_BUFFER_START_length_bias;

anv_batch_emit_blk(&cmd_buffer->batch, GEN8_MI_BATCH_BUFFER_START, bbs) {
anv_batch_emit(&cmd_buffer->batch, GEN8_MI_BATCH_BUFFER_START, bbs) {
bbs.DWordLength = cmd_buffer->device->info.gen < 8 ?
gen7_length : gen8_length;
bbs._2ndLevelBatchBuffer = _1stlevelbatch;
@@ -712,11 +712,11 @@ anv_cmd_buffer_end_batch_buffer(struct anv_cmd_buffer *cmd_buffer)
cmd_buffer->batch.end += GEN8_MI_BATCH_BUFFER_START_length * 4;
assert(cmd_buffer->batch.end == batch_bo->bo.map + batch_bo->bo.size);

anv_batch_emit_blk(&cmd_buffer->batch, GEN7_MI_BATCH_BUFFER_END, bbe);
anv_batch_emit(&cmd_buffer->batch, GEN7_MI_BATCH_BUFFER_END, bbe);

/* Round batch up to an even number of dwords. */
if ((cmd_buffer->batch.next - cmd_buffer->batch.start) & 4)
anv_batch_emit_blk(&cmd_buffer->batch, GEN7_MI_NOOP, noop);
anv_batch_emit(&cmd_buffer->batch, GEN7_MI_NOOP, noop);

cmd_buffer->exec_mode = ANV_CMD_BUFFER_EXEC_MODE_PRIMARY;
}

+ 4
- 4
src/intel/vulkan/anv_device.c View File

@@ -1081,8 +1081,8 @@ VkResult anv_DeviceWaitIdle(
batch.start = batch.next = cmds;
batch.end = (void *) cmds + sizeof(cmds);

anv_batch_emit_blk(&batch, GEN7_MI_BATCH_BUFFER_END, bbe);
anv_batch_emit_blk(&batch, GEN7_MI_NOOP, noop);
anv_batch_emit(&batch, GEN7_MI_BATCH_BUFFER_END, bbe);
anv_batch_emit(&batch, GEN7_MI_NOOP, noop);

return anv_device_submit_simple_batch(device, &batch);
}
@@ -1423,8 +1423,8 @@ VkResult anv_CreateFence(
const uint32_t batch_offset = align_u32(sizeof(*fence), CACHELINE_SIZE);
batch.next = batch.start = fence->bo.map + batch_offset;
batch.end = fence->bo.map + fence->bo.size;
anv_batch_emit_blk(&batch, GEN7_MI_BATCH_BUFFER_END, bbe);
anv_batch_emit_blk(&batch, GEN7_MI_NOOP, noop);
anv_batch_emit(&batch, GEN7_MI_BATCH_BUFFER_END, bbe);
anv_batch_emit(&batch, GEN7_MI_NOOP, noop);

if (!device->info.has_llc) {
assert(((uintptr_t) batch.start & CACHELINE_MASK) == 0);

+ 1
- 1
src/intel/vulkan/anv_private.h View File

@@ -848,7 +848,7 @@ __gen_combine_address(struct anv_batch *batch, void *location,
VG(VALGRIND_CHECK_MEM_IS_DEFINED(dw, ARRAY_SIZE(dwords0) * 4));\
} while (0)

#define anv_batch_emit_blk(batch, cmd, name) \
#define anv_batch_emit(batch, cmd, name) \
for (struct cmd name = { __anv_cmd_header(cmd) }, \
*_dst = anv_batch_emit_dwords(batch, __anv_cmd_length(cmd)); \
__builtin_expect(_dst != NULL, 1); \

+ 19
- 20
src/intel/vulkan/gen7_cmd_buffer.c View File

@@ -57,8 +57,8 @@ gen7_cmd_buffer_emit_descriptor_pointers(struct anv_cmd_buffer *cmd_buffer,

anv_foreach_stage(s, stages) {
if (cmd_buffer->state.samplers[s].alloc_size > 0) {
anv_batch_emit_blk(&cmd_buffer->batch,
GENX(3DSTATE_SAMPLER_STATE_POINTERS_VS), ssp) {
anv_batch_emit(&cmd_buffer->batch,
GENX(3DSTATE_SAMPLER_STATE_POINTERS_VS), ssp) {
ssp._3DCommandSubOpcode = sampler_state_opcodes[s];
ssp.PointertoVSSamplerState = cmd_buffer->state.samplers[s].offset;
}
@@ -66,8 +66,8 @@ gen7_cmd_buffer_emit_descriptor_pointers(struct anv_cmd_buffer *cmd_buffer,

/* Always emit binding table pointers if we're asked to, since on SKL
* this is what flushes push constants. */
anv_batch_emit_blk(&cmd_buffer->batch,
GENX(3DSTATE_BINDING_TABLE_POINTERS_VS), btp) {
anv_batch_emit(&cmd_buffer->batch,
GENX(3DSTATE_BINDING_TABLE_POINTERS_VS), btp) {
btp._3DCommandSubOpcode = binding_table_opcodes[s];
btp.PointertoVSBindingTable = cmd_buffer->state.binding_tables[s].offset;
}
@@ -175,8 +175,8 @@ gen7_cmd_buffer_emit_scissor(struct anv_cmd_buffer *cmd_buffer)
}
}

anv_batch_emit_blk(&cmd_buffer->batch,
GEN7_3DSTATE_SCISSOR_STATE_POINTERS, ssp) {
anv_batch_emit(&cmd_buffer->batch,
GEN7_3DSTATE_SCISSOR_STATE_POINTERS, ssp) {
ssp.ScissorRectPointer = scissor_state.offset;
}

@@ -241,7 +241,7 @@ flush_compute_descriptor_set(struct anv_cmd_buffer *cmd_buffer)
unsigned push_constant_regs = reg_aligned_constant_size / 32;

if (push_state.alloc_size) {
anv_batch_emit_blk(&cmd_buffer->batch, GENX(MEDIA_CURBE_LOAD), curbe) {
anv_batch_emit(&cmd_buffer->batch, GENX(MEDIA_CURBE_LOAD), curbe) {
curbe.CURBETotalDataLength = push_state.alloc_size;
curbe.CURBEDataStartAddress = push_state.offset;
}
@@ -274,8 +274,8 @@ flush_compute_descriptor_set(struct anv_cmd_buffer *cmd_buffer)
pipeline->cs_thread_width_max);

const uint32_t size = GENX(INTERFACE_DESCRIPTOR_DATA_length) * sizeof(uint32_t);
anv_batch_emit_blk(&cmd_buffer->batch,
GENX(MEDIA_INTERFACE_DESCRIPTOR_LOAD), idl) {
anv_batch_emit(&cmd_buffer->batch,
GENX(MEDIA_INTERFACE_DESCRIPTOR_LOAD), idl) {
idl.InterfaceDescriptorTotalLength = size;
idl.InterfaceDescriptorDataStartAddress = state.offset;
}
@@ -312,7 +312,7 @@ genX(cmd_buffer_config_l3)(struct anv_cmd_buffer *cmd_buffer, bool enable_slm)
* flushed, which involves a first PIPE_CONTROL flush which stalls the
* pipeline...
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.DCFlushEnable = true;
pc.CommandStreamerStallEnable = true;
pc.PostSyncOperation = NoWrite;
@@ -332,7 +332,7 @@ genX(cmd_buffer_config_l3)(struct anv_cmd_buffer *cmd_buffer, bool enable_slm)
* previous and subsequent PIPE_CONTROLs already guarantee that there is
* no concurrent GPGPU kernel execution (see SKL HSD 2132585).
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.TextureCacheInvalidationEnable = true;
pc.ConstantCacheInvalidationEnable = true;
pc.InstructionCacheInvalidateEnable = true;
@@ -343,14 +343,14 @@ genX(cmd_buffer_config_l3)(struct anv_cmd_buffer *cmd_buffer, bool enable_slm)
/* Now send a third stalling flush to make sure that invalidation is
* complete when the L3 configuration registers are modified.
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.DCFlushEnable = true;
pc.CommandStreamerStallEnable = true;
pc.PostSyncOperation = NoWrite;
}

anv_finishme("write GEN7_L3SQCREG1");
anv_batch_emit_blk(&cmd_buffer->batch, GENX(MI_LOAD_REGISTER_IMM), lri) {
anv_batch_emit(&cmd_buffer->batch, GENX(MI_LOAD_REGISTER_IMM), lri) {
lri.RegisterOffset = GENX(L3CNTLREG2_num);
lri.DataDWord = l3cr2_val;
}
@@ -365,7 +365,7 @@ genX(cmd_buffer_config_l3)(struct anv_cmd_buffer *cmd_buffer, bool enable_slm)
.CAllocation = 8,
.TAllocation = 8);
const uint32_t l3cr3_val = enable_slm ? l3cr3_slm : l3cr3_noslm;
anv_batch_emit_blk(&cmd_buffer->batch, GENX(MI_LOAD_REGISTER_IMM), lri) {
anv_batch_emit(&cmd_buffer->batch, GENX(MI_LOAD_REGISTER_IMM), lri) {
lri.RegisterOffset = GENX(L3CNTLREG3_num);
lri.DataDWord = l3cr3_val;
}
@@ -455,8 +455,7 @@ genX(cmd_buffer_flush_dynamic_state)(struct anv_cmd_buffer *cmd_buffer)
if (!cmd_buffer->device->info.has_llc)
anv_state_clflush(cc_state);

anv_batch_emit_blk(&cmd_buffer->batch,
GENX(3DSTATE_CC_STATE_POINTERS), ccp) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_CC_STATE_POINTERS), ccp) {
ccp.ColorCalcStatePointer = cc_state.offset;
}
}
@@ -482,8 +481,8 @@ genX(cmd_buffer_flush_dynamic_state)(struct anv_cmd_buffer *cmd_buffer)
pipeline->gen7.depth_stencil_state,
GENX(DEPTH_STENCIL_STATE_length), 64);

anv_batch_emit_blk(&cmd_buffer->batch,
GENX(3DSTATE_DEPTH_STENCIL_STATE_POINTERS), dsp) {
anv_batch_emit(&cmd_buffer->batch,
GENX(3DSTATE_DEPTH_STENCIL_STATE_POINTERS), dsp) {
dsp.PointertoDEPTH_STENCIL_STATE = ds_state.offset;
}
}
@@ -495,13 +494,13 @@ genX(cmd_buffer_flush_dynamic_state)(struct anv_cmd_buffer *cmd_buffer)
uint32_t offset = cmd_buffer->state.gen7.index_offset;

#if GEN_IS_HASWELL
anv_batch_emit_blk(&cmd_buffer->batch, GEN75_3DSTATE_VF, vf) {
anv_batch_emit(&cmd_buffer->batch, GEN75_3DSTATE_VF, vf) {
vf.IndexedDrawCutIndexEnable = pipeline->primitive_restart;
vf.CutIndex = cmd_buffer->state.restart_index;
}
#endif

anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_INDEX_BUFFER), ib) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_INDEX_BUFFER), ib) {
#if !GEN_IS_HASWELL
ib.CutIndexEnable = pipeline->primitive_restart;
#endif

+ 13
- 14
src/intel/vulkan/gen7_pipeline.c View File

@@ -175,8 +175,7 @@ gen7_emit_cb_state(struct anv_pipeline *pipeline,
anv_state_clflush(pipeline->blend_state);
}

anv_batch_emit_blk(&pipeline->batch,
GENX(3DSTATE_BLEND_STATE_POINTERS), bsp) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_BLEND_STATE_POINTERS), bsp) {
bsp.BlendStatePointer = pipeline->blend_state.offset;
}
}
@@ -224,7 +223,7 @@ genX(graphics_pipeline_create)(
const VkPipelineRasterizationStateCreateInfo *rs_info =
pCreateInfo->pRasterizationState;

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_CLIP), clip) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_CLIP), clip) {
clip.FrontWinding = vk_to_gen_front_face[rs_info->frontFace],
clip.CullMode = vk_to_gen_cullmode[rs_info->cullMode],
clip.ClipEnable = !(extra && extra->use_rectlist),
@@ -248,12 +247,12 @@ genX(graphics_pipeline_create)(
uint32_t samples = 1;
uint32_t log2_samples = __builtin_ffs(samples) - 1;

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_MULTISAMPLE), ms) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_MULTISAMPLE), ms) {
ms.PixelLocation = PIXLOC_CENTER;
ms.NumberofMultisamples = log2_samples;
}

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_SAMPLE_MASK), sm) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_SAMPLE_MASK), sm) {
sm.SampleMask = 0xff;
}

@@ -279,9 +278,9 @@ genX(graphics_pipeline_create)(
#endif

if (pipeline->vs_vec4 == NO_KERNEL || (extra && extra->disable_vs))
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_VS), vs);
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_VS), vs);
else
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_VS), vs) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_VS), vs) {
vs.KernelStartPointer = pipeline->vs_vec4;
vs.ScratchSpaceBaseOffset = pipeline->scratch_start[MESA_SHADER_VERTEX];
vs.PerThreadScratchSpace = scratch_space(&vs_prog_data->base.base);
@@ -299,9 +298,9 @@ genX(graphics_pipeline_create)(
const struct brw_gs_prog_data *gs_prog_data = get_gs_prog_data(pipeline);

if (pipeline->gs_kernel == NO_KERNEL || (extra && extra->disable_vs)) {
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_GS), gs);
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_GS), gs);
} else {
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_GS), gs) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_GS), gs) {
gs.KernelStartPointer = pipeline->gs_kernel;
gs.ScratchSpaceBasePointer = pipeline->scratch_start[MESA_SHADER_GEOMETRY];
gs.PerThreadScratchSpace = scratch_space(&gs_prog_data->base.base);
@@ -332,9 +331,9 @@ genX(graphics_pipeline_create)(
}

if (pipeline->ps_ksp0 == NO_KERNEL) {
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_SBE), sbe);
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_SBE), sbe);

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_WM), wm) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_WM), wm) {
wm.StatisticsEnable = true;
wm.ThreadDispatchEnable = false;
wm.LineEndCapAntialiasingRegionWidth = 0; /* 0.5 pixels */
@@ -346,7 +345,7 @@ genX(graphics_pipeline_create)(
/* Even if no fragments are ever dispatched, the hardware hangs if we
* don't at least set the maximum number of threads.
*/
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_PS), ps) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_PS), ps) {
ps.MaximumNumberofThreads = device->info.max_wm_threads - 1;
}
} else {
@@ -359,7 +358,7 @@ genX(graphics_pipeline_create)(

emit_3dstate_sbe(pipeline);

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_PS), ps) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_PS), ps) {
ps.KernelStartPointer0 = pipeline->ps_ksp0;
ps.ScratchSpaceBasePointer = pipeline->scratch_start[MESA_SHADER_FRAGMENT];
ps.PerThreadScratchSpace = scratch_space(&wm_prog_data->base);
@@ -392,7 +391,7 @@ genX(graphics_pipeline_create)(
}

/* FIXME-GEN7: This needs a lot more work, cf gen7 upload_wm_state(). */
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_WM), wm) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_WM), wm) {
wm.StatisticsEnable = true;
wm.ThreadDispatchEnable = true;
wm.LineEndCapAntialiasingRegionWidth = 0; /* 0.5 pixels */

+ 18
- 20
src/intel/vulkan/gen8_cmd_buffer.c View File

@@ -80,12 +80,12 @@ gen8_cmd_buffer_emit_viewport(struct anv_cmd_buffer *cmd_buffer)
anv_state_clflush(cc_state);
}

anv_batch_emit_blk(&cmd_buffer->batch,
GENX(3DSTATE_VIEWPORT_STATE_POINTERS_CC), cc) {
anv_batch_emit(&cmd_buffer->batch,
GENX(3DSTATE_VIEWPORT_STATE_POINTERS_CC), cc) {
cc.CCViewportPointer = cc_state.offset;
}
anv_batch_emit_blk(&cmd_buffer->batch,
GENX(3DSTATE_VIEWPORT_STATE_POINTERS_SF_CLIP), clip) {
anv_batch_emit(&cmd_buffer->batch,
GENX(3DSTATE_VIEWPORT_STATE_POINTERS_SF_CLIP), clip) {
clip.SFClipViewportPointer = sf_clip_state.offset;
}
}
@@ -117,7 +117,7 @@ genX(cmd_buffer_config_l3)(struct anv_cmd_buffer *cmd_buffer, bool enable_slm)
* flushed, which involves a first PIPE_CONTROL flush which stalls the
* pipeline...
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.DCFlushEnable = true;
pc.PostSyncOperation = NoWrite;
pc.CommandStreamerStallEnable = true;
@@ -137,7 +137,7 @@ genX(cmd_buffer_config_l3)(struct anv_cmd_buffer *cmd_buffer, bool enable_slm)
* previous and subsequent PIPE_CONTROLs already guarantee that there is
* no concurrent GPGPU kernel execution (see SKL HSD 2132585).
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.TextureCacheInvalidationEnable = true,
pc.ConstantCacheInvalidationEnable = true,
pc.InstructionCacheInvalidateEnable = true,
@@ -148,13 +148,13 @@ genX(cmd_buffer_config_l3)(struct anv_cmd_buffer *cmd_buffer, bool enable_slm)
/* Now send a third stalling flush to make sure that invalidation is
* complete when the L3 configuration registers are modified.
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.DCFlushEnable = true;
pc.PostSyncOperation = NoWrite;
pc.CommandStreamerStallEnable = true;
}

anv_batch_emit_blk(&cmd_buffer->batch, GENX(MI_LOAD_REGISTER_IMM), lri) {
anv_batch_emit(&cmd_buffer->batch, GENX(MI_LOAD_REGISTER_IMM), lri) {
lri.RegisterOffset = GENX(L3CNTLREG_num);
lri.DataDWord = l3cr_val;
}
@@ -250,8 +250,7 @@ genX(cmd_buffer_flush_dynamic_state)(struct anv_cmd_buffer *cmd_buffer)
if (!cmd_buffer->device->info.has_llc)
anv_state_clflush(cc_state);

anv_batch_emit_blk(&cmd_buffer->batch,
GENX(3DSTATE_CC_STATE_POINTERS), ccp) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_CC_STATE_POINTERS), ccp) {
ccp.ColorCalcStatePointer = cc_state.offset;
ccp.ColorCalcStatePointerValid = true;
}
@@ -295,8 +294,7 @@ genX(cmd_buffer_flush_dynamic_state)(struct anv_cmd_buffer *cmd_buffer)
if (!cmd_buffer->device->info.has_llc)
anv_state_clflush(cc_state);

anv_batch_emit_blk(&cmd_buffer->batch,
GEN9_3DSTATE_CC_STATE_POINTERS, ccp) {
anv_batch_emit(&cmd_buffer->batch, GEN9_3DSTATE_CC_STATE_POINTERS, ccp) {
ccp.ColorCalcStatePointer = cc_state.offset;
ccp.ColorCalcStatePointerValid = true;
}
@@ -329,7 +327,7 @@ genX(cmd_buffer_flush_dynamic_state)(struct anv_cmd_buffer *cmd_buffer)

if (cmd_buffer->state.dirty & (ANV_CMD_DIRTY_PIPELINE |
ANV_CMD_DIRTY_INDEX_BUFFER)) {
anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_VF), vf) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_VF), vf) {
vf.IndexedDrawCutIndexEnable = pipeline->primitive_restart;
vf.CutIndex = cmd_buffer->state.restart_index;
}
@@ -359,7 +357,7 @@ void genX(CmdBindIndexBuffer)(

cmd_buffer->state.restart_index = restart_index_for_type[indexType];

anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_INDEX_BUFFER), ib) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_INDEX_BUFFER), ib) {
ib.IndexFormat = vk_to_gen_index_type[indexType];
ib.MemoryObjectControlState = GENX(MOCS);
ib.BufferStartingAddress =
@@ -399,7 +397,7 @@ flush_compute_descriptor_set(struct anv_cmd_buffer *cmd_buffer)
unsigned push_constant_regs = reg_aligned_constant_size / 32;

if (push_state.alloc_size) {
anv_batch_emit_blk(&cmd_buffer->batch, GENX(MEDIA_CURBE_LOAD), curbe) {
anv_batch_emit(&cmd_buffer->batch, GENX(MEDIA_CURBE_LOAD), curbe) {
curbe.CURBETotalDataLength = push_state.alloc_size;
curbe.CURBEDataStartAddress = push_state.offset;
}
@@ -432,8 +430,8 @@ flush_compute_descriptor_set(struct anv_cmd_buffer *cmd_buffer)
pipeline->cs_thread_width_max);

uint32_t size = GENX(INTERFACE_DESCRIPTOR_DATA_length) * sizeof(uint32_t);
anv_batch_emit_blk(&cmd_buffer->batch,
GENX(MEDIA_INTERFACE_DESCRIPTOR_LOAD), mid) {
anv_batch_emit(&cmd_buffer->batch,
GENX(MEDIA_INTERFACE_DESCRIPTOR_LOAD), mid) {
mid.InterfaceDescriptorTotalLength = size;
mid.InterfaceDescriptorDataStartAddress = state.offset;
}
@@ -476,7 +474,7 @@ void genX(CmdSetEvent)(
ANV_FROM_HANDLE(anv_cmd_buffer, cmd_buffer, commandBuffer);
ANV_FROM_HANDLE(anv_event, event, _event);

anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.DestinationAddressType = DAT_PPGTT,
pc.PostSyncOperation = WriteImmediateData,
pc.Address = (struct anv_address) {
@@ -495,7 +493,7 @@ void genX(CmdResetEvent)(
ANV_FROM_HANDLE(anv_cmd_buffer, cmd_buffer, commandBuffer);
ANV_FROM_HANDLE(anv_event, event, _event);

anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.DestinationAddressType = DAT_PPGTT;
pc.PostSyncOperation = WriteImmediateData;
pc.Address = (struct anv_address) {
@@ -523,7 +521,7 @@ void genX(CmdWaitEvents)(
for (uint32_t i = 0; i < eventCount; i++) {
ANV_FROM_HANDLE(anv_event, event, pEvents[i]);

anv_batch_emit_blk(&cmd_buffer->batch, GENX(MI_SEMAPHORE_WAIT), sem) {
anv_batch_emit(&cmd_buffer->batch, GENX(MI_SEMAPHORE_WAIT), sem) {
sem.WaitMode = PollingMode,
sem.CompareOperation = COMPARE_SAD_EQUAL_SDD,
sem.SemaphoreDataDword = VK_EVENT_SET,

+ 15
- 16
src/intel/vulkan/gen8_pipeline.c View File

@@ -39,7 +39,7 @@ emit_ia_state(struct anv_pipeline *pipeline,
const VkPipelineInputAssemblyStateCreateInfo *info,
const struct anv_graphics_pipeline_create_info *extra)
{
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_VF_TOPOLOGY), vft) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_VF_TOPOLOGY), vft) {
vft.PrimitiveTopologyType = pipeline->topology;
}
}
@@ -192,7 +192,7 @@ emit_cb_state(struct anv_pipeline *pipeline,

struct GENX(BLEND_STATE_ENTRY) *bs0 = &blend_state.Entry[0];

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_PS_BLEND), blend) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_PS_BLEND), blend) {
blend.AlphaToCoverageEnable = blend_state.AlphaToCoverageEnable;
blend.HasWriteableRT = has_writeable_rt;
blend.ColorBufferBlendEnable = bs0->ColorBufferBlendEnable;
@@ -209,8 +209,7 @@ emit_cb_state(struct anv_pipeline *pipeline,
if (!device->info.has_llc)
anv_state_clflush(pipeline->blend_state);

anv_batch_emit_blk(&pipeline->batch,
GENX(3DSTATE_BLEND_STATE_POINTERS), bsp) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_BLEND_STATE_POINTERS), bsp) {
bsp.BlendStatePointer = pipeline->blend_state.offset;
bsp.BlendStatePointerValid = true;
}
@@ -291,7 +290,7 @@ emit_ms_state(struct anv_pipeline *pipeline,
if (info && info->sampleShadingEnable)
anv_finishme("VkPipelineMultisampleStateCreateInfo::sampleShadingEnable");

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_MULTISAMPLE), ms) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_MULTISAMPLE), ms) {
/* The PRM says that this bit is valid only for DX9:
*
* SW can choose to set this bit only for DX9 API. DX10/OGL API's
@@ -303,7 +302,7 @@ emit_ms_state(struct anv_pipeline *pipeline,
ms.NumberofMultisamples = log2_samples;
}

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_SAMPLE_MASK), sm) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_SAMPLE_MASK), sm) {
sm.SampleMask = sample_mask;
}
}
@@ -351,7 +350,7 @@ genX(graphics_pipeline_create)(
emit_urb_setup(pipeline);

const struct brw_wm_prog_data *wm_prog_data = get_wm_prog_data(pipeline);
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_CLIP), clip) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_CLIP), clip) {
clip.ClipEnable = !(extra && extra->use_rectlist);
clip.EarlyCullEnable = true;
clip.APIMode = 1; /* D3D */
@@ -373,7 +372,7 @@ genX(graphics_pipeline_create)(
clip.MaximumVPIndex = pCreateInfo->pViewportState->viewportCount - 1;
}

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_WM), wm) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_WM), wm) {
wm.StatisticsEnable = true;
wm.LineEndCapAntialiasingRegionWidth = _05pixels;
wm.LineAntialiasingRegionWidth = _10pixels;
@@ -386,13 +385,13 @@ genX(graphics_pipeline_create)(
}

if (pipeline->gs_kernel == NO_KERNEL) {
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_GS), gs);
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_GS), gs);
} else {
const struct brw_gs_prog_data *gs_prog_data = get_gs_prog_data(pipeline);
offset = 1;
length = (gs_prog_data->base.vue_map.num_slots + 1) / 2 - offset;

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_GS), gs) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_GS), gs) {
gs.SingleProgramFlow = false;
gs.KernelStartPointer = pipeline->gs_kernel;
gs.VectorMaskEnable = false;
@@ -444,7 +443,7 @@ genX(graphics_pipeline_create)(
pipeline->vs_vec4;

if (vs_start == NO_KERNEL || (extra && extra->disable_vs)) {
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_VS), vs) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_VS), vs) {
vs.FunctionEnable = false;
/* Even if VS is disabled, SBE still gets the amount of
* vertex data to read from this field. */
@@ -452,7 +451,7 @@ genX(graphics_pipeline_create)(
vs.VertexURBEntryOutputLength = length;
}
} else {
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_VS), vs) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_VS), vs) {
vs.KernelStartPointer = vs_start;
vs.SingleVertexDispatch = false;
vs.VectorMaskEnable = false;
@@ -493,14 +492,14 @@ genX(graphics_pipeline_create)(

const int num_thread_bias = GEN_GEN == 8 ? 2 : 1;
if (pipeline->ps_ksp0 == NO_KERNEL) {
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_PS), ps);
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_PS_EXTRA), extra) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_PS), ps);
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_PS_EXTRA), extra) {
extra.PixelShaderValid = false;
}
} else {
emit_3dstate_sbe(pipeline);

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_PS), ps) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_PS), ps) {
ps.KernelStartPointer0 = pipeline->ps_ksp0;
ps.KernelStartPointer1 = 0;
ps.KernelStartPointer2 = pipeline->ps_ksp2;
@@ -527,7 +526,7 @@ genX(graphics_pipeline_create)(
bool per_sample_ps = pCreateInfo->pMultisampleState &&
pCreateInfo->pMultisampleState->sampleShadingEnable;

anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_PS_EXTRA), ps) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_PS_EXTRA), ps) {
ps.PixelShaderValid = true;
ps.PixelShaderKillsPixel = wm_prog_data->uses_kill;
ps.PixelShaderComputedDepthMode = wm_prog_data->computed_depth_mode;

+ 43
- 43
src/intel/vulkan/genX_cmd_buffer.c View File

@@ -49,12 +49,12 @@ genX(cmd_buffer_emit_state_base_address)(struct anv_cmd_buffer *cmd_buffer)
* this, we get GPU hangs when using multi-level command buffers which
* clear depth, reset state base address, and then go render stuff.
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.RenderTargetCacheFlushEnable = true;
}
#endif

anv_batch_emit_blk(&cmd_buffer->batch, GENX(STATE_BASE_ADDRESS), sba) {
anv_batch_emit(&cmd_buffer->batch, GENX(STATE_BASE_ADDRESS), sba) {
sba.GeneralStateBaseAddress = (struct anv_address) { scratch_bo, 0 };
sba.GeneralStateMemoryObjectControlState = GENX(MOCS);
sba.GeneralStateBaseAddressModifyEnable = true;
@@ -131,7 +131,7 @@ genX(cmd_buffer_emit_state_base_address)(struct anv_cmd_buffer *cmd_buffer)
* units cache the binding table in the texture cache. However, we have
* yet to be able to actually confirm this.
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.TextureCacheInvalidationEnable = true;
}
}
@@ -295,10 +295,10 @@ cmd_buffer_flush_push_constants(struct anv_cmd_buffer *cmd_buffer)
struct anv_state state = anv_cmd_buffer_push_constants(cmd_buffer, stage);

if (state.offset == 0) {
anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_CONSTANT_VS), c)
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_CONSTANT_VS), c)
c._3DCommandSubOpcode = push_constant_opcodes[stage];
} else {
anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_CONSTANT_VS), c) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_CONSTANT_VS), c) {
c._3DCommandSubOpcode = push_constant_opcodes[stage],
c.ConstantBody = (struct GENX(3DSTATE_CONSTANT_BODY)) {
#if GEN_GEN >= 9
@@ -420,7 +420,7 @@ genX(cmd_buffer_flush_state)(struct anv_cmd_buffer *cmd_buffer)
* PIPE_CONTROL needs to be sent before any combination of VS
* associated 3DSTATE."
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.DepthStallEnable = true;
pc.PostSyncOperation = WriteImmediateData;
pc.Address =
@@ -521,7 +521,7 @@ void genX(CmdDraw)(
if (vs_prog_data->uses_basevertex || vs_prog_data->uses_baseinstance)
emit_base_vertex_instance(cmd_buffer, firstVertex, firstInstance);

anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DPRIMITIVE), prim) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DPRIMITIVE), prim) {
prim.VertexAccessType = SEQUENTIAL;
prim.PrimitiveTopologyType = pipeline->topology;
prim.VertexCountPerInstance = vertexCount;
@@ -549,7 +549,7 @@ void genX(CmdDrawIndexed)(
if (vs_prog_data->uses_basevertex || vs_prog_data->uses_baseinstance)
emit_base_vertex_instance(cmd_buffer, vertexOffset, firstInstance);

anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DPRIMITIVE), prim) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DPRIMITIVE), prim) {
prim.VertexAccessType = RANDOM;
prim.PrimitiveTopologyType = pipeline->topology;
prim.VertexCountPerInstance = indexCount;
@@ -572,7 +572,7 @@ static void
emit_lrm(struct anv_batch *batch,
uint32_t reg, struct anv_bo *bo, uint32_t offset)
{
anv_batch_emit_blk(batch, GENX(MI_LOAD_REGISTER_MEM), lrm) {
anv_batch_emit(batch, GENX(MI_LOAD_REGISTER_MEM), lrm) {
lrm.RegisterAddress = reg;
lrm.MemoryAddress = (struct anv_address) { bo, offset };
}
@@ -581,7 +581,7 @@ emit_lrm(struct anv_batch *batch,
static void
emit_lri(struct anv_batch *batch, uint32_t reg, uint32_t imm)
{
anv_batch_emit_blk(batch, GENX(MI_LOAD_REGISTER_IMM), lri) {
anv_batch_emit(batch, GENX(MI_LOAD_REGISTER_IMM), lri) {
lri.RegisterOffset = reg;
lri.DataDWord = imm;
}
@@ -612,7 +612,7 @@ void genX(CmdDrawIndirect)(
emit_lrm(&cmd_buffer->batch, GEN7_3DPRIM_START_INSTANCE, bo, bo_offset + 12);
emit_lri(&cmd_buffer->batch, GEN7_3DPRIM_BASE_VERTEX, 0);

anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DPRIMITIVE), prim) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DPRIMITIVE), prim) {
prim.IndirectParameterEnable = true;
prim.VertexAccessType = SEQUENTIAL;
prim.PrimitiveTopologyType = pipeline->topology;
@@ -645,7 +645,7 @@ void genX(CmdDrawIndexedIndirect)(
emit_lrm(&cmd_buffer->batch, GEN7_3DPRIM_BASE_VERTEX, bo, bo_offset + 12);
emit_lrm(&cmd_buffer->batch, GEN7_3DPRIM_START_INSTANCE, bo, bo_offset + 16);

anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DPRIMITIVE), prim) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DPRIMITIVE), prim) {
prim.IndirectParameterEnable = true;
prim.VertexAccessType = RANDOM;
prim.PrimitiveTopologyType = pipeline->topology;
@@ -697,7 +697,7 @@ void genX(CmdDispatch)(

genX(cmd_buffer_flush_compute_state)(cmd_buffer);

anv_batch_emit_blk(&cmd_buffer->batch, GENX(GPGPU_WALKER), ggw) {
anv_batch_emit(&cmd_buffer->batch, GENX(GPGPU_WALKER), ggw) {
ggw.SIMDSize = prog_data->simd_size / 16;
ggw.ThreadDepthCounterMaximum = 0;
ggw.ThreadHeightCounterMaximum = 0;
@@ -709,7 +709,7 @@ void genX(CmdDispatch)(
ggw.BottomExecutionMask = 0xffffffff;
}

anv_batch_emit_blk(&cmd_buffer->batch, GENX(MEDIA_STATE_FLUSH), msf);
anv_batch_emit(&cmd_buffer->batch, GENX(MEDIA_STATE_FLUSH), msf);
}

#define GPGPU_DISPATCHDIMX 0x2500
@@ -761,7 +761,7 @@ void genX(CmdDispatchIndirect)(
emit_lrm(batch, MI_PREDICATE_SRC0, bo, bo_offset + 0);

/* predicate = (compute_dispatch_indirect_x_size == 0); */
anv_batch_emit_blk(batch, GENX(MI_PREDICATE), mip) {
anv_batch_emit(batch, GENX(MI_PREDICATE), mip) {
mip.LoadOperation = LOAD_LOAD;
mip.CombineOperation = COMBINE_SET;
mip.CompareOperation = COMPARE_SRCS_EQUAL;
@@ -771,7 +771,7 @@ void genX(CmdDispatchIndirect)(
emit_lrm(batch, MI_PREDICATE_SRC0, bo, bo_offset + 4);

/* predicate |= (compute_dispatch_indirect_y_size == 0); */
anv_batch_emit_blk(batch, GENX(MI_PREDICATE), mip) {
anv_batch_emit(batch, GENX(MI_PREDICATE), mip) {
mip.LoadOperation = LOAD_LOAD;
mip.CombineOperation = COMBINE_OR;
mip.CompareOperation = COMPARE_SRCS_EQUAL;
@@ -781,7 +781,7 @@ void genX(CmdDispatchIndirect)(
emit_lrm(batch, MI_PREDICATE_SRC0, bo, bo_offset + 8);

/* predicate |= (compute_dispatch_indirect_z_size == 0); */
anv_batch_emit_blk(batch, GENX(MI_PREDICATE), mip) {
anv_batch_emit(batch, GENX(MI_PREDICATE), mip) {
mip.LoadOperation = LOAD_LOAD;
mip.CombineOperation = COMBINE_OR;
mip.CompareOperation = COMPARE_SRCS_EQUAL;
@@ -789,14 +789,14 @@ void genX(CmdDispatchIndirect)(

/* predicate = !predicate; */
#define COMPARE_FALSE 1
anv_batch_emit_blk(batch, GENX(MI_PREDICATE), mip) {
anv_batch_emit(batch, GENX(MI_PREDICATE), mip) {
mip.LoadOperation = LOAD_LOADINV;
mip.CombineOperation = COMBINE_OR;
mip.CompareOperation = COMPARE_FALSE;
}
#endif

anv_batch_emit_blk(batch, GENX(GPGPU_WALKER), ggw) {
anv_batch_emit(batch, GENX(GPGPU_WALKER), ggw) {
ggw.IndirectParameterEnable = true;
ggw.PredicateEnable = GEN_GEN <= 7;
ggw.SIMDSize = prog_data->simd_size / 16;
@@ -807,7 +807,7 @@ void genX(CmdDispatchIndirect)(
ggw.BottomExecutionMask = 0xffffffff;
}

anv_batch_emit_blk(batch, GENX(MEDIA_STATE_FLUSH), msf);
anv_batch_emit(batch, GENX(MEDIA_STATE_FLUSH), msf);
}

static void
@@ -825,7 +825,7 @@ flush_pipeline_before_pipeline_select(struct anv_cmd_buffer *cmd_buffer,
* hardware too.
*/
if (pipeline == GPGPU)
anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_CC_STATE_POINTERS), t);
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_CC_STATE_POINTERS), t);
#elif GEN_GEN <= 7
/* From "BXML » GT » MI » vol1a GPU Overview » [Instruction]
* PIPELINE_SELECT [DevBWR+]":
@@ -837,7 +837,7 @@ flush_pipeline_before_pipeline_select(struct anv_cmd_buffer *cmd_buffer,
* command to invalidate read only caches prior to programming
* MI_PIPELINE_SELECT command to change the Pipeline Select Mode.
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.RenderTargetCacheFlushEnable = true;
pc.DepthCacheFlushEnable = true;
pc.DCFlushEnable = true;
@@ -845,7 +845,7 @@ flush_pipeline_before_pipeline_select(struct anv_cmd_buffer *cmd_buffer,
pc.CommandStreamerStallEnable = true;
}

anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.TextureCacheInvalidationEnable = true;
pc.ConstantCacheInvalidationEnable = true;
pc.StateCacheInvalidationEnable = true;
@@ -861,7 +861,7 @@ genX(flush_pipeline_select_3d)(struct anv_cmd_buffer *cmd_buffer)
if (cmd_buffer->state.current_pipeline != _3D) {
flush_pipeline_before_pipeline_select(cmd_buffer, _3D);

anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPELINE_SELECT), ps) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPELINE_SELECT), ps) {
#if GEN_GEN >= 9
ps.MaskBits = 3;
#endif
@@ -878,7 +878,7 @@ genX(flush_pipeline_select_gpgpu)(struct anv_cmd_buffer *cmd_buffer)
if (cmd_buffer->state.current_pipeline != GPGPU) {
flush_pipeline_before_pipeline_select(cmd_buffer, GPGPU);

anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPELINE_SELECT), ps) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPELINE_SELECT), ps) {
#if GEN_GEN >= 9
ps.MaskBits = 3;
#endif
@@ -937,7 +937,7 @@ cmd_buffer_emit_depth_stencil(struct anv_cmd_buffer *cmd_buffer)

/* Emit 3DSTATE_DEPTH_BUFFER */
if (has_depth) {
anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_DEPTH_BUFFER), db) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_DEPTH_BUFFER), db) {
db.SurfaceType = SURFTYPE_2D;
db.DepthWriteEnable = true;
db.StencilWriteEnable = has_stencil;
@@ -984,7 +984,7 @@ cmd_buffer_emit_depth_stencil(struct anv_cmd_buffer *cmd_buffer)
* nor stencil buffer is present. Also, D16_UNORM is not allowed to
* be combined with a stencil buffer so we use D32_FLOAT instead.
*/
anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_DEPTH_BUFFER), db) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_DEPTH_BUFFER), db) {
db.SurfaceType = SURFTYPE_2D;
db.SurfaceFormat = D32_FLOAT;
db.Width = fb->width - 1;
@@ -995,7 +995,7 @@ cmd_buffer_emit_depth_stencil(struct anv_cmd_buffer *cmd_buffer)

/* Emit 3DSTATE_STENCIL_BUFFER */
if (has_stencil) {
anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_STENCIL_BUFFER), sb) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_STENCIL_BUFFER), sb) {
#if GEN_GEN >= 8 || GEN_IS_HASWELL
sb.StencilBufferEnable = true,
#endif
@@ -1017,14 +1017,14 @@ cmd_buffer_emit_depth_stencil(struct anv_cmd_buffer *cmd_buffer)
};
}
} else {
anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_STENCIL_BUFFER), sb);
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_STENCIL_BUFFER), sb);
}

/* Disable hierarchial depth buffers. */
anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_HIER_DEPTH_BUFFER), hz);
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_HIER_DEPTH_BUFFER), hz);

/* Clear the clear params. */
anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_CLEAR_PARAMS), cp);
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_CLEAR_PARAMS), cp);
}

/**
@@ -1058,7 +1058,7 @@ void genX(CmdBeginRenderPass)(

const VkRect2D *render_area = &pRenderPassBegin->renderArea;

anv_batch_emit_blk(&cmd_buffer->batch, GENX(3DSTATE_DRAWING_RECTANGLE), r) {
anv_batch_emit(&cmd_buffer->batch, GENX(3DSTATE_DRAWING_RECTANGLE), r) {
r.ClippedDrawingRectangleYMin = MAX2(render_area->offset.y, 0);
r.ClippedDrawingRectangleXMin = MAX2(render_area->offset.x, 0);
r.ClippedDrawingRectangleYMax =
@@ -1098,7 +1098,7 @@ static void
emit_ps_depth_count(struct anv_batch *batch,
struct anv_bo *bo, uint32_t offset)
{
anv_batch_emit_blk(batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(batch, GENX(PIPE_CONTROL), pc) {
pc.DestinationAddressType = DAT_PPGTT;
pc.PostSyncOperation = WritePSDepthCount;
pc.DepthStallEnable = true;
@@ -1110,7 +1110,7 @@ static void
emit_query_availability(struct anv_batch *batch,
struct anv_bo *bo, uint32_t offset)
{
anv_batch_emit_blk(batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(batch, GENX(PIPE_CONTROL), pc) {
pc.DestinationAddressType = DAT_PPGTT;
pc.PostSyncOperation = WriteImmediateData;
pc.Address = (struct anv_address) { bo, offset };
@@ -1135,7 +1135,7 @@ void genX(CmdBeginQuery)(
*/
if (cmd_buffer->state.need_query_wa) {
cmd_buffer->state.need_query_wa = false;
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.DepthCacheFlushEnable = true;
pc.DepthStallEnable = true;
}
@@ -1192,11 +1192,11 @@ void genX(CmdWriteTimestamp)(

switch (pipelineStage) {
case VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT:
anv_batch_emit_blk(&cmd_buffer->batch, GENX(MI_STORE_REGISTER_MEM), srm) {
anv_batch_emit(&cmd_buffer->batch, GENX(MI_STORE_REGISTER_MEM), srm) {
srm.RegisterAddress = TIMESTAMP;
srm.MemoryAddress = (struct anv_address) { &pool->bo, offset };
}
anv_batch_emit_blk(&cmd_buffer->batch, GENX(MI_STORE_REGISTER_MEM), srm) {
anv_batch_emit(&cmd_buffer->batch, GENX(MI_STORE_REGISTER_MEM), srm) {
srm.RegisterAddress = TIMESTAMP + 4;
srm.MemoryAddress = (struct anv_address) { &pool->bo, offset + 4 };
}
@@ -1204,7 +1204,7 @@ void genX(CmdWriteTimestamp)(

default:
/* Everything else is bottom-of-pipe */
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.DestinationAddressType = DAT_PPGTT,
pc.PostSyncOperation = WriteTimestamp,
pc.Address = (struct anv_address) { &pool->bo, offset };
@@ -1253,11 +1253,11 @@ static void
emit_load_alu_reg_u64(struct anv_batch *batch, uint32_t reg,
struct anv_bo *bo, uint32_t offset)
{
anv_batch_emit_blk(batch, GENX(MI_LOAD_REGISTER_MEM), lrm) {
anv_batch_emit(batch, GENX(MI_LOAD_REGISTER_MEM), lrm) {
lrm.RegisterAddress = reg,
lrm.MemoryAddress = (struct anv_address) { bo, offset };
}
anv_batch_emit_blk(batch, GENX(MI_LOAD_REGISTER_MEM), lrm) {
anv_batch_emit(batch, GENX(MI_LOAD_REGISTER_MEM), lrm) {
lrm.RegisterAddress = reg + 4;
lrm.MemoryAddress = (struct anv_address) { bo, offset + 4 };
}
@@ -1267,13 +1267,13 @@ static void
store_query_result(struct anv_batch *batch, uint32_t reg,
struct anv_bo *bo, uint32_t offset, VkQueryResultFlags flags)
{
anv_batch_emit_blk(batch, GENX(MI_STORE_REGISTER_MEM), srm) {
anv_batch_emit(batch, GENX(MI_STORE_REGISTER_MEM), srm) {
srm.RegisterAddress = reg;
srm.MemoryAddress = (struct anv_address) { bo, offset };
}

if (flags & VK_QUERY_RESULT_64_BIT) {
anv_batch_emit_blk(batch, GENX(MI_STORE_REGISTER_MEM), srm) {
anv_batch_emit(batch, GENX(MI_STORE_REGISTER_MEM), srm) {
srm.RegisterAddress = reg + 4;
srm.MemoryAddress = (struct anv_address) { bo, offset + 4 };
}
@@ -1296,7 +1296,7 @@ void genX(CmdCopyQueryPoolResults)(
uint32_t slot_offset, dst_offset;

if (flags & VK_QUERY_RESULT_WAIT_BIT) {
anv_batch_emit_blk(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
anv_batch_emit(&cmd_buffer->batch, GENX(PIPE_CONTROL), pc) {
pc.CommandStreamerStallEnable = true;
pc.StallAtPixelScoreboard = true;
}

+ 1
- 1
src/intel/vulkan/genX_pipeline.c View File

@@ -105,7 +105,7 @@ genX(compute_pipeline_create)(
const uint32_t vfe_curbe_allocation =
push_constant_regs * pipeline->cs_thread_width_max;

anv_batch_emit_blk(&pipeline->batch, GENX(MEDIA_VFE_STATE), vfe) {
anv_batch_emit(&pipeline->batch, GENX(MEDIA_VFE_STATE), vfe) {
vfe.ScratchSpaceBasePointer = pipeline->scratch_start[MESA_SHADER_COMPUTE];
vfe.PerThreadScratchSpace = ffs(cs_prog_data->base.total_scratch / 2048);
#if GEN_GEN > 7

+ 6
- 6
src/intel/vulkan/genX_pipeline_util.h View File

@@ -130,7 +130,7 @@ emit_vertex_input(struct anv_pipeline *pipeline,
* that controls instancing. On Haswell and prior, that's part of
* VERTEX_BUFFER_STATE which we emit later.
*/
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_VF_INSTANCING), vfi) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_VF_INSTANCING), vfi) {
vfi.InstancingEnable = pipeline->instancing_enable[desc->binding],
vfi.VertexElementIndex = slot,
/* Vulkan so far doesn't have an instance divisor, so
@@ -173,7 +173,7 @@ emit_vertex_input(struct anv_pipeline *pipeline,
}

#if GEN_GEN >= 8
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_VF_SGVS), sgvs) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_VF_SGVS), sgvs) {
sgvs.VertexIDEnable = vs_prog_data->uses_vertexid;
sgvs.VertexIDComponentNumber = 2;
sgvs.VertexIDElementOffset = id_slot;
@@ -198,7 +198,7 @@ emit_urb_setup(struct anv_pipeline *pipeline)
* 3DSTATE_SAMPLER_STATE_POINTER_VS command. Only one PIPE_CONTROL
* needs to be sent before any combination of VS associated 3DSTATE."
*/
anv_batch_emit_blk(&pipeline->batch, GEN7_PIPE_CONTROL, pc) {
anv_batch_emit(&pipeline->batch, GEN7_PIPE_CONTROL, pc) {
pc.DepthStallEnable = true;
pc.PostSyncOperation = WriteImmediateData;
pc.Address = (struct anv_address) { &device->workaround_bo, 0 };
@@ -208,8 +208,8 @@ emit_urb_setup(struct anv_pipeline *pipeline)
unsigned push_start = 0;
for (int i = MESA_SHADER_VERTEX; i <= MESA_SHADER_FRAGMENT; i++) {
unsigned push_size = pipeline->urb.push_size[i];
anv_batch_emit_blk(&pipeline->batch,
GENX(3DSTATE_PUSH_CONSTANT_ALLOC_VS), alloc) {
anv_batch_emit(&pipeline->batch,
GENX(3DSTATE_PUSH_CONSTANT_ALLOC_VS), alloc) {
alloc._3DCommandSubOpcode = 18 + i;
alloc.ConstantBufferOffset = (push_size > 0) ? push_start : 0;
alloc.ConstantBufferSize = push_size;
@@ -218,7 +218,7 @@ emit_urb_setup(struct anv_pipeline *pipeline)
}

for (int i = MESA_SHADER_VERTEX; i <= MESA_SHADER_GEOMETRY; i++) {
anv_batch_emit_blk(&pipeline->batch, GENX(3DSTATE_URB_VS), urb) {
anv_batch_emit(&pipeline->batch, GENX(3DSTATE_URB_VS), urb) {
urb._3DCommandSubOpcode = 48 + i;
urb.VSURBStartingAddress = pipeline->urb.start[i];
urb.VSURBEntryAllocationSize = pipeline->urb.size[i] - 1;

+ 10
- 10
src/intel/vulkan/genX_state.c View File

@@ -44,30 +44,30 @@ genX(init_device_state)(struct anv_device *device)
batch.start = batch.next = cmds;
batch.end = (void *) cmds + sizeof(cmds);

anv_batch_emit_blk(&batch, GENX(PIPELINE_SELECT), ps) {
anv_batch_emit(&batch, GENX(PIPELINE_SELECT), ps) {
#if GEN_GEN >= 9
ps.MaskBits = 3;
#endif
ps.PipelineSelection = _3D;
}

anv_batch_emit_blk(&batch, GENX(3DSTATE_VF_STATISTICS), vfs)
anv_batch_emit(&batch, GENX(3DSTATE_VF_STATISTICS), vfs)
vfs.StatisticsEnable = true;

anv_batch_emit_blk(&batch, GENX(3DSTATE_HS), hs);
anv_batch_emit_blk(&batch, GENX(3DSTATE_TE), ts);
anv_batch_emit_blk(&batch, GENX(3DSTATE_DS), ds);
anv_batch_emit(&batch, GENX(3DSTATE_HS), hs);
anv_batch_emit(&batch, GENX(3DSTATE_TE), ts);
anv_batch_emit(&batch, GENX(3DSTATE_DS), ds);

anv_batch_emit_blk(&batch, GENX(3DSTATE_STREAMOUT), so);
anv_batch_emit_blk(&batch, GENX(3DSTATE_AA_LINE_PARAMETERS), aa);
anv_batch_emit(&batch, GENX(3DSTATE_STREAMOUT), so);
anv_batch_emit(&batch, GENX(3DSTATE_AA_LINE_PARAMETERS), aa);

#if GEN_GEN >= 8
anv_batch_emit_blk(&batch, GENX(3DSTATE_WM_CHROMAKEY), ck);
anv_batch_emit(&batch, GENX(3DSTATE_WM_CHROMAKEY), ck);

/* See the Vulkan 1.0 spec Table 24.1 "Standard sample locations" and
* VkPhysicalDeviceFeatures::standardSampleLocations.
*/
anv_batch_emit_blk(&batch, GENX(3DSTATE_SAMPLE_PATTERN), sp) {
anv_batch_emit(&batch, GENX(3DSTATE_SAMPLE_PATTERN), sp) {
sp._1xSample0XOffset = 0.5;
sp._1xSample0YOffset = 0.5;
sp._2xSample0XOffset = 0.25;
@@ -135,7 +135,7 @@ genX(init_device_state)(struct anv_device *device)
}
#endif

anv_batch_emit_blk(&batch, GENX(MI_BATCH_BUFFER_END), bbe);
anv_batch_emit(&batch, GENX(MI_BATCH_BUFFER_END), bbe);

assert(batch.next <= batch.end);


Write Preview
Loading…
Cancel
Save