[Mesa-dev] [PATCH] r600g: Implement sm5 interpolation functions
Glenn Kennard
glenn.kennard at gmail.com
Mon Sep 1 12:37:50 PDT 2014
Requires evergreen/cayman
Signed-off-by: Glenn Kennard <glenn.kennard at gmail.com>
---
This patch depends on r600g: Implement GL_ARB_sample_shading
Implementation note: interpolateAtSample/Offset is doing affine
rather than perspective correct reverse projection, I think for
the small intra-pixel offsets used this should typically give
more accurate results due to not losing bits by the reciprocal
and additional multiplication. As a bonus slightly fewer ops,
and as far as I can tell it is within spec to skip the
perspective correction.
docs/GL3.txt | 2 +-
src/gallium/drivers/r600/evergreen_state.c | 3 +
src/gallium/drivers/r600/r600_shader.c | 211 ++++++++++++++++++++++++++++-
3 files changed, 211 insertions(+), 5 deletions(-)
diff --git a/docs/GL3.txt b/docs/GL3.txt
index 5baacc1..8e69068 100644
--- a/docs/GL3.txt
+++ b/docs/GL3.txt
@@ -107,7 +107,7 @@ GL 4.0, GLSL 4.00:
- Geometry shader instancing DONE (r600)
- Geometry shader multiple streams DONE ()
- Enhanced per-sample shading DONE (r600)
- - Interpolation functions DONE ()
+ - Interpolation functions DONE (r600)
- New overload resolution rules DONE
GL_ARB_gpu_shader_fp64 started (Dave)
GL_ARB_sample_shading DONE (i965, nv50, nvc0, r600, radeonsi)
diff --git a/src/gallium/drivers/r600/evergreen_state.c b/src/gallium/drivers/r600/evergreen_state.c
index 9531893..b857292 100644
--- a/src/gallium/drivers/r600/evergreen_state.c
+++ b/src/gallium/drivers/r600/evergreen_state.c
@@ -1427,6 +1427,9 @@ static void evergreen_set_framebuffer_state(struct pipe_context *ctx,
assert(rctx->framebuffer.nr_samples <= Elements(values)/4);
for (i = 0; i < rctx->framebuffer.nr_samples; i++) {
ctx->get_sample_position(ctx, rctx->framebuffer.nr_samples, i, &values[4*i]);
+ /* interpolateAtSample needs offset from center point */
+ values[4*i+2] = values[4*i+0] - 0.5f;
+ values[4*i+3] = values[4*i+1] - 0.5f;
}
constbuf.user_buffer = values;
constbuf.buffer_size = rctx->framebuffer.nr_samples * 4 * 4;
diff --git a/src/gallium/drivers/r600/r600_shader.c b/src/gallium/drivers/r600/r600_shader.c
index fd7d845..7e5a526 100644
--- a/src/gallium/drivers/r600/r600_shader.c
+++ b/src/gallium/drivers/r600/r600_shader.c
@@ -294,6 +294,8 @@ struct r600_shader_ctx {
int face_gpr;
/* sample id is .w component stored in fixed point position register */
int fixed_pt_position_gpr;
+ /* evergreen/cayman needs to add shader->input entries for SAMPLEMASK/SAMPLEID */
+ int num_system_value_inputs;
int colors_used;
boolean clip_vertex_write;
unsigned cv_output;
@@ -778,6 +780,8 @@ static int allocate_system_value_inputs(struct r600_shader_ctx *ctx, int gpr_off
}
}
+ ctx->num_system_value_inputs = num_regs;
+
return num_regs;
}
@@ -814,7 +818,7 @@ static int evergreen_gpr_count(struct r600_shader_ctx *ctx)
ctx->eg_interpolators[k].enabled = TRUE;
}
- /* Need to scan shader for system values */
+ /* Need to scan shader for interpolateAtSample/Offset/Centroid */
if (tgsi_parse_init(&parse, ctx->tokens) != TGSI_PARSE_OK) {
return 0;
}
@@ -824,7 +828,31 @@ static int evergreen_gpr_count(struct r600_shader_ctx *ctx)
while (!tgsi_parse_end_of_tokens(&parse)) {
tgsi_parse_token(&parse);
- if (parse.FullToken.Token.Type == TGSI_TOKEN_TYPE_DECLARATION) {
+ if (parse.FullToken.Token.Type == TGSI_TOKEN_TYPE_INSTRUCTION) {
+ const struct tgsi_full_instruction *inst
+ = &parse.FullToken.FullInstruction;
+ if (inst->Instruction.Opcode == TGSI_OPCODE_INTERP_SAMPLE ||
+ inst->Instruction.Opcode == TGSI_OPCODE_INTERP_OFFSET ||
+ inst->Instruction.Opcode == TGSI_OPCODE_INTERP_CENTROID)
+ {
+ int interpolate, location, k;
+
+ if (inst->Instruction.Opcode == TGSI_OPCODE_INTERP_SAMPLE) {
+ location = TGSI_INTERPOLATE_LOC_CENTER;
+ has_sampleid = true;
+ } else if (inst->Instruction.Opcode == TGSI_OPCODE_INTERP_OFFSET) {
+ location = TGSI_INTERPOLATE_LOC_CENTER;
+ /* Needs sample positions, currently those are always available */
+ } else {
+ location = TGSI_INTERPOLATE_LOC_CENTROID;
+ }
+
+ interpolate = ctx->info.input_interpolate[inst->Src[0].Register.Index];
+ k = eg_get_interpolator_index(interpolate, location);
+ ctx->eg_interpolators[k].enabled = true;
+ }
+ }
+ else if (parse.FullToken.Token.Type == TGSI_TOKEN_TYPE_DECLARATION) {
struct tgsi_full_declaration *d = &parse.FullToken.FullDeclaration;
if (d->Declaration.File == TGSI_FILE_SYSTEM_VALUE) {
if (d->Semantic.Name == TGSI_SEMANTIC_SAMPLEMASK) {
@@ -893,8 +921,8 @@ static int load_sample_position(struct r600_shader_ctx *ctx, struct r600_shader_
vtx.dst_gpr = t1;
vtx.dst_sel_x = 0;
vtx.dst_sel_y = 1;
- vtx.dst_sel_z = 7;
- vtx.dst_sel_w = 7;
+ vtx.dst_sel_z = 2;
+ vtx.dst_sel_w = 3;
vtx.data_format = FMT_32_32_32_32_FLOAT;
vtx.num_format_all = 2;
vtx.format_comp_all = 1;
@@ -1767,6 +1795,7 @@ static int r600_shader_from_tgsi(struct r600_context *rctx,
ctx.face_gpr = -1;
ctx.fixed_pt_position_gpr = -1;
+ ctx.num_system_value_inputs = 0;
ctx.fragcoord_input = -1;
ctx.colors_used = 0;
ctx.clip_vertex_write = 0;
@@ -4550,6 +4579,171 @@ static int tgsi_msb(struct r600_shader_ctx *ctx)
return 0;
}
+static int tgsi_interp_egcm(struct r600_shader_ctx *ctx)
+{
+ struct tgsi_full_instruction *inst = &ctx->parse.FullToken.FullInstruction;
+ struct r600_bytecode_alu alu;
+ int r, i = 0, k, interp_gpr, interp_base_chan, tmp, lasti;
+ unsigned location;
+ int input;
+
+ assert(inst->Src[0].Register.File == TGSI_FILE_INPUT);
+
+ input = inst->Src[0].Register.Index + ctx->num_system_value_inputs;
+
+ /* Interpolators have been marked for use already by allocate_system_value_inputs */
+ if (inst->Instruction.Opcode == TGSI_OPCODE_INTERP_OFFSET ||
+ inst->Instruction.Opcode == TGSI_OPCODE_INTERP_SAMPLE) {
+ location = TGSI_INTERPOLATE_LOC_CENTER; /* sample offset will be added explicitly */
+ }
+ else {
+ location = TGSI_INTERPOLATE_LOC_CENTROID;
+ }
+
+ k = eg_get_interpolator_index(ctx->shader->input[input].interpolate, location);
+ if (k < 0)
+ k = 0;
+ interp_gpr = ctx->eg_interpolators[k].ij_index / 2;
+ interp_base_chan = 2 * (ctx->eg_interpolators[k].ij_index % 2);
+
+ /* NOTE: currently offset is not perspective correct */
+ if (inst->Instruction.Opcode == TGSI_OPCODE_INTERP_OFFSET ||
+ inst->Instruction.Opcode == TGSI_OPCODE_INTERP_SAMPLE) {
+ int sample_gpr = -1;
+ int gradientsH, gradientsV;
+ struct r600_bytecode_tex tex;
+
+ if (inst->Instruction.Opcode == TGSI_OPCODE_INTERP_SAMPLE) {
+ sample_gpr = load_sample_position(ctx, &ctx->src[1], ctx->src[1].swizzle[0]);
+ }
+
+ gradientsH = r600_get_temp(ctx);
+ gradientsV = r600_get_temp(ctx);
+ for (i = 0; i < 2; i++) {
+ memset(&tex, 0, sizeof(struct r600_bytecode_tex));
+ tex.op = i == 0 ? FETCH_OP_GET_GRADIENTS_H : FETCH_OP_GET_GRADIENTS_V;
+ tex.src_gpr = interp_gpr;
+ tex.src_sel_x = interp_base_chan + 0;
+ tex.src_sel_y = interp_base_chan + 1;
+ tex.src_sel_z = 0;
+ tex.src_sel_w = 0;
+ tex.dst_gpr = i == 0 ? gradientsH : gradientsV;
+ tex.dst_sel_x = 0;
+ tex.dst_sel_y = 1;
+ tex.dst_sel_z = 7;
+ tex.dst_sel_w = 7;
+ tex.inst_mod = 1; // Use per pixel gradient calculation
+ tex.sampler_id = 0;
+ tex.resource_id = tex.sampler_id;
+ r = r600_bytecode_add_tex(ctx->bc, &tex);
+ if (r)
+ return r;
+ }
+
+ for (i = 0; i < 2; i++) {
+ memset(&alu, 0, sizeof(struct r600_bytecode_alu));
+ alu.op = ALU_OP3_MULADD;
+ alu.is_op3 = 1;
+ alu.src[0].sel = gradientsH;
+ alu.src[0].chan = i;
+ if (inst->Instruction.Opcode == TGSI_OPCODE_INTERP_SAMPLE) {
+ alu.src[1].sel = sample_gpr;
+ alu.src[1].chan = 2;
+ }
+ else {
+ r600_bytecode_src(&alu.src[1], &ctx->src[1], 0);
+ }
+ alu.src[2].sel = interp_gpr;
+ alu.src[2].chan = interp_base_chan + i;
+ alu.dst.sel = ctx->temp_reg;
+ alu.dst.chan = i;
+ alu.last = i == 1;
+
+ r = r600_bytecode_add_alu(ctx->bc, &alu);
+ if (r)
+ return r;
+ }
+
+ for (i = 0; i < 2; i++) {
+ memset(&alu, 0, sizeof(struct r600_bytecode_alu));
+ alu.op = ALU_OP3_MULADD;
+ alu.is_op3 = 1;
+ alu.src[0].sel = gradientsV;
+ alu.src[0].chan = i;
+ if (inst->Instruction.Opcode == TGSI_OPCODE_INTERP_SAMPLE) {
+ alu.src[1].sel = sample_gpr;
+ alu.src[1].chan = 3;
+ }
+ else {
+ r600_bytecode_src(&alu.src[1], &ctx->src[1], 1);
+ }
+ alu.src[2].sel = ctx->temp_reg;
+ alu.src[2].chan = i;
+ alu.dst.sel = ctx->temp_reg;
+ alu.dst.chan = i;
+ alu.last = i == 1;
+
+ r = r600_bytecode_add_alu(ctx->bc, &alu);
+ if (r)
+ return r;
+ }
+ }
+
+ tmp = r600_get_temp(ctx);
+ for (i = 0; i < 8; i++) {
+ memset(&alu, 0, sizeof(struct r600_bytecode_alu));
+ alu.op = i < 4 ? ALU_OP2_INTERP_ZW : ALU_OP2_INTERP_XY;
+
+ alu.dst.sel = tmp;
+ if ((i > 1 && i < 6)) {
+ alu.dst.write = 1;
+ }
+ else {
+ alu.dst.write = 0;
+ }
+ alu.dst.chan = i % 4;
+
+ if (inst->Instruction.Opcode == TGSI_OPCODE_INTERP_OFFSET ||
+ inst->Instruction.Opcode == TGSI_OPCODE_INTERP_SAMPLE) {
+ alu.src[0].sel = ctx->temp_reg;
+ alu.src[0].chan = 1 - (i % 2);
+ } else {
+ alu.src[0].sel = interp_gpr;
+ alu.src[0].chan = interp_base_chan + 1 - (i % 2);
+ }
+ alu.src[1].sel = V_SQ_ALU_SRC_PARAM_BASE + ctx->shader->input[input].lds_pos;
+ alu.src[1].chan = 0;
+
+ alu.last = i % 4 == 3;
+ alu.bank_swizzle_force = SQ_ALU_VEC_210;
+
+ r = r600_bytecode_add_alu(ctx->bc, &alu);
+ if (r)
+ return r;
+ }
+
+ // INTERP can't swizzle dst
+ lasti = tgsi_last_instruction(inst->Dst[0].Register.WriteMask);
+ for (i = 0; i <= lasti; i++) {
+ if (!(inst->Dst[0].Register.WriteMask & (1 << i)))
+ continue;
+
+ memset(&alu, 0, sizeof(struct r600_bytecode_alu));
+ alu.op = ALU_OP1_MOV;
+ alu.src[0].sel = tmp;
+ alu.src[0].chan = ctx->src[0].swizzle[i];
+ tgsi_dst(ctx, &inst->Dst[0], i, &alu.dst);
+ alu.dst.write = 1;
+ alu.last = i == lasti;
+ r = r600_bytecode_add_alu(ctx->bc, &alu);
+ if (r)
+ return r;
+ }
+
+ return 0;
+}
+
+
static int tgsi_helper_copy(struct r600_shader_ctx *ctx, struct tgsi_full_instruction *inst)
{
struct r600_bytecode_alu alu;
@@ -7076,6 +7270,9 @@ static struct r600_shader_tgsi_instruction r600_shader_tgsi_instruction[] = {
{TGSI_OPCODE_LSB, 0, ALU_OP1_FFBL_INT, tgsi_unsupported},
{TGSI_OPCODE_IMSB, 0, ALU_OP1_FFBH_INT, tgsi_unsupported},
{TGSI_OPCODE_UMSB, 0, ALU_OP1_FFBH_UINT, tgsi_unsupported},
+ {TGSI_OPCODE_INTERP_CENTROID, 0, ALU_OP0_NOP, tgsi_unsupported},
+ {TGSI_OPCODE_INTERP_SAMPLE, 0, ALU_OP0_NOP, tgsi_unsupported},
+ {TGSI_OPCODE_INTERP_OFFSET, 0, ALU_OP0_NOP, tgsi_unsupported},
{TGSI_OPCODE_LAST, 0, ALU_OP0_NOP, tgsi_unsupported},
};
@@ -7278,6 +7475,9 @@ static struct r600_shader_tgsi_instruction eg_shader_tgsi_instruction[] = {
{TGSI_OPCODE_LSB, 0, ALU_OP1_FFBL_INT, tgsi_op2},
{TGSI_OPCODE_IMSB, 0, ALU_OP1_FFBH_INT, tgsi_msb},
{TGSI_OPCODE_UMSB, 0, ALU_OP1_FFBH_UINT, tgsi_msb},
+ {TGSI_OPCODE_INTERP_CENTROID, 0, ALU_OP0_NOP, tgsi_interp_egcm},
+ {TGSI_OPCODE_INTERP_SAMPLE, 0, ALU_OP0_NOP, tgsi_interp_egcm},
+ {TGSI_OPCODE_INTERP_OFFSET, 0, ALU_OP0_NOP, tgsi_interp_egcm},
{TGSI_OPCODE_LAST, 0, ALU_OP0_NOP, tgsi_unsupported},
};
@@ -7481,5 +7681,8 @@ static struct r600_shader_tgsi_instruction cm_shader_tgsi_instruction[] = {
{TGSI_OPCODE_LSB, 0, ALU_OP1_FFBL_INT, tgsi_op2},
{TGSI_OPCODE_IMSB, 0, ALU_OP1_FFBH_INT, tgsi_msb},
{TGSI_OPCODE_UMSB, 0, ALU_OP1_FFBH_UINT, tgsi_msb},
+ {TGSI_OPCODE_INTERP_CENTROID, 0, ALU_OP0_NOP, tgsi_interp_egcm},
+ {TGSI_OPCODE_INTERP_SAMPLE, 0, ALU_OP0_NOP, tgsi_interp_egcm},
+ {TGSI_OPCODE_INTERP_OFFSET, 0, ALU_OP0_NOP, tgsi_interp_egcm},
{TGSI_OPCODE_LAST, 0, ALU_OP0_NOP, tgsi_unsupported},
};
--
1.9.1
More information about the mesa-dev
mailing list