nvk: Use MME for vkCmdDraw[Indexed]()

This lets us move the instance loop to the MME and stop emitting state
per-instance.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24326>
This commit is contained in:
Faith Ekstrand 2023-01-30 20:11:57 -06:00 committed by Marge Bot
parent 02783c822c
commit 9ceec19adb
3 changed files with 92 additions and 38 deletions

View file

@ -19,6 +19,7 @@
#include "nvk_clb197.h"
#include "nvk_clc397.h"
#include "nvk_clc597.h"
#include "drf.h"
static inline uint16_t
nvk_cmd_buffer_3d_cls(struct nvk_cmd_buffer *cmd)
@ -1243,6 +1244,33 @@ vk_to_nv9097_primitive_topology(VkPrimitiveTopology prim)
}
}
void
nvk_mme_draw(struct nvk_device *dev, struct mme_builder *b)
{
struct mme_value begin = mme_load(b);
struct mme_value vertex_count = mme_load(b);
struct mme_value instance_count = mme_load(b);
struct mme_value first_vertex = mme_load(b);
struct mme_value first_instance = mme_load(b);
mme_mthd(b, NV9097_SET_GLOBAL_BASE_INSTANCE_INDEX);
mme_emit(b, first_instance);
mme_loop(b, instance_count) {
mme_mthd(b, NV9097_BEGIN);
mme_emit(b, begin);
mme_mthd(b, NV9097_SET_VERTEX_ARRAY_START);
mme_emit(b, first_vertex);
mme_emit(b, vertex_count);
mme_mthd(b, NV9097_END);
mme_emit(b, mme_zero());
mme_set_field_enum(b, begin, NV9097_BEGIN_INSTANCE_ID, SUBSEQUENT);
}
}
VKAPI_ATTR void VKAPI_CALL
nvk_CmdDraw(VkCommandBuffer commandBuffer,
uint32_t vertexCount,
@ -1256,27 +1284,54 @@ nvk_CmdDraw(VkCommandBuffer commandBuffer,
nvk_flush_gfx_state(cmd);
const uint32_t begin_op =
vk_to_nv9097_primitive_topology(dyn->ia.primitive_topology);
uint32_t begin;
V_NV9097_BEGIN(begin, {
.op = vk_to_nv9097_primitive_topology(dyn->ia.primitive_topology),
.primitive_id = NV9097_BEGIN_PRIMITIVE_ID_FIRST,
.instance_id = NV9097_BEGIN_INSTANCE_ID_FIRST,
.split_mode = SPLIT_MODE_NORMAL_BEGIN_NORMAL_END,
});
struct nouveau_ws_push_buffer *p = P_SPACE(cmd->push, 2 + 7 * instanceCount);
struct nouveau_ws_push_buffer *p = P_SPACE(cmd->push, 6);
P_1INC(p, NV9097, CALL_MME_MACRO(NVK_MME_DRAW));
P_INLINE_DATA(p, begin);
P_INLINE_DATA(p, vertexCount);
P_INLINE_DATA(p, instanceCount);
P_INLINE_DATA(p, firstVertex);
P_INLINE_DATA(p, firstInstance);
}
P_IMMD(p, NV9097, SET_GLOBAL_BASE_INSTANCE_INDEX, firstInstance);
void
nvk_mme_draw_indexed(struct nvk_device *dev, struct mme_builder *b)
{
struct mme_value begin = mme_load(b);
struct mme_value index_count = mme_load(b);
struct mme_value instance_count = mme_load(b);
struct mme_value first_index = mme_load(b);
struct mme_value vertex_offset = mme_load(b);
struct mme_value first_instance = mme_load(b);
for (unsigned i = 0; i < instanceCount; i++) {
P_IMMD(p, NV9097, BEGIN, {
.op = begin_op,
.primitive_id = NV9097_BEGIN_PRIMITIVE_ID_FIRST,
.instance_id = (i == 0) ? NV9097_BEGIN_INSTANCE_ID_FIRST :
NV9097_BEGIN_INSTANCE_ID_SUBSEQUENT,
.split_mode = SPLIT_MODE_NORMAL_BEGIN_NORMAL_END,
});
mme_mthd(b, NV9097_SET_GLOBAL_BASE_VERTEX_INDEX);
mme_emit(b, vertex_offset);
P_MTHD(p, NV9097, SET_VERTEX_ARRAY_START);
P_NV9097_SET_VERTEX_ARRAY_START(p, firstVertex);
P_NV9097_DRAW_VERTEX_ARRAY(p, vertexCount);
mme_mthd(b, NV9097_SET_VERTEX_ID_BASE);
mme_emit(b, vertex_offset);
P_IMMD(p, NV9097, END, 0);
mme_mthd(b, NV9097_SET_GLOBAL_BASE_INSTANCE_INDEX);
mme_emit(b, first_instance);
mme_loop(b, instance_count) {
mme_mthd(b, NV9097_BEGIN);
mme_emit(b, begin);
mme_mthd(b, NV9097_SET_INDEX_BUFFER_F);
mme_emit(b, first_index);
mme_emit(b, index_count);
mme_mthd(b, NV9097_END);
mme_emit(b, mme_zero());
mme_set_field_enum(b, begin, NV9097_BEGIN_INSTANCE_ID, SUBSEQUENT);
}
}
@ -1294,27 +1349,20 @@ nvk_CmdDrawIndexed(VkCommandBuffer commandBuffer,
nvk_flush_gfx_state(cmd);
const uint32_t begin_op =
vk_to_nv9097_primitive_topology(dyn->ia.primitive_topology);
uint32_t begin;
V_NV9097_BEGIN(begin, {
.op = vk_to_nv9097_primitive_topology(dyn->ia.primitive_topology),
.primitive_id = NV9097_BEGIN_PRIMITIVE_ID_FIRST,
.instance_id = NV9097_BEGIN_INSTANCE_ID_FIRST,
.split_mode = SPLIT_MODE_NORMAL_BEGIN_NORMAL_END,
});
struct nouveau_ws_push_buffer *p = P_SPACE(cmd->push, 6 + 7 * instanceCount);
P_IMMD(p, NV9097, SET_GLOBAL_BASE_VERTEX_INDEX, vertexOffset);
P_IMMD(p, NV9097, SET_VERTEX_ID_BASE, vertexOffset);
P_IMMD(p, NV9097, SET_GLOBAL_BASE_INSTANCE_INDEX, firstInstance);
for (unsigned i = 0; i < instanceCount; i++) {
P_IMMD(p, NV9097, BEGIN, {
.op = begin_op,
.primitive_id = NV9097_BEGIN_PRIMITIVE_ID_FIRST,
.instance_id = (i == 0) ? NV9097_BEGIN_INSTANCE_ID_FIRST :
NV9097_BEGIN_INSTANCE_ID_SUBSEQUENT,
.split_mode = SPLIT_MODE_NORMAL_BEGIN_NORMAL_END,
});
P_MTHD(p, NV9097, SET_INDEX_BUFFER_F);
P_NV9097_SET_INDEX_BUFFER_F(p, firstIndex);
P_NV9097_DRAW_INDEX_BUFFER(p, indexCount);
P_IMMD(p, NV9097, END, 0);
}
struct nouveau_ws_push_buffer *p = P_SPACE(cmd->push, 7);
P_1INC(p, NV9097, CALL_MME_MACRO(NVK_MME_DRAW_INDEXED));
P_INLINE_DATA(p, begin);
P_INLINE_DATA(p, indexCount);
P_INLINE_DATA(p, instanceCount);
P_INLINE_DATA(p, firstIndex);
P_INLINE_DATA(p, vertexOffset);
P_INLINE_DATA(p, firstInstance);
}

View file

@ -5,6 +5,8 @@
static const nvk_mme_builder_func mme_builders[NVK_MME_COUNT] = {
[NVK_MME_CLEAR_VIEWS] = nvk_mme_clear_views,
[NVK_MME_CLEAR_LAYERS] = nvk_mme_clear_layers,
[NVK_MME_DRAW] = nvk_mme_draw,
[NVK_MME_DRAW_INDEXED] = nvk_mme_draw_indexed,
};
uint32_t *

View file

@ -8,6 +8,8 @@ struct nvk_device;
enum nvk_mme {
NVK_MME_CLEAR_VIEWS,
NVK_MME_CLEAR_LAYERS,
NVK_MME_DRAW,
NVK_MME_DRAW_INDEXED,
NVK_MME_COUNT,
};
@ -19,5 +21,7 @@ uint32_t *nvk_build_mme(struct nvk_device *dev, enum nvk_mme mme,
void nvk_mme_clear_views(struct nvk_device *dev, struct mme_builder *b);
void nvk_mme_clear_layers(struct nvk_device *dev, struct mme_builder *b);
void nvk_mme_draw(struct nvk_device *dev, struct mme_builder *b);
void nvk_mme_draw_indexed(struct nvk_device *dev, struct mme_builder *b);
#endif /* NVK_MME_H */