aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
-rw-r--r--src/mesa/drivers/dri/i965/brw_fs_nir.cpp42
1 files changed, 30 insertions, 12 deletions
diff --git a/src/mesa/drivers/dri/i965/brw_fs_nir.cpp b/src/mesa/drivers/dri/i965/brw_fs_nir.cpp
index 788315f6c52..ad347fcdbaf 100644
--- a/src/mesa/drivers/dri/i965/brw_fs_nir.cpp
+++ b/src/mesa/drivers/dri/i965/brw_fs_nir.cpp
@@ -1849,15 +1849,33 @@ fs_visitor::nir_emit_tes_intrinsic(const fs_builder &bld,
fs_inst *inst;
if (indirect_offset.file == BAD_FILE) {
- /* Replicate the patch handle to all enabled channels */
- const fs_reg srcs[] = {
- retype(brw_vec1_grf(0, 0), BRW_REGISTER_TYPE_UD)
- };
- fs_reg patch_handle = bld.vgrf(BRW_REGISTER_TYPE_UD, 1);
- bld.LOAD_PAYLOAD(patch_handle, srcs, ARRAY_SIZE(srcs), 0);
-
- inst = bld.emit(SHADER_OPCODE_URB_READ_SIMD8, dest, patch_handle);
- inst->mlen = 1;
+ /* Arbitrarily only push up to 32 vec4 slots worth of data,
+ * which is 16 registers (since each holds 2 vec4 slots).
+ */
+ const unsigned max_push_slots = 32;
+ if (imm_offset < max_push_slots) {
+ fs_reg src = fs_reg(ATTR, imm_offset / 2, dest.type);
+ for (int i = 0; i < instr->num_components; i++) {
+ bld.MOV(offset(dest, bld, i),
+ component(src, 4 * (imm_offset % 2) + i));
+ }
+ tes_prog_data->base.urb_read_length =
+ MAX2(tes_prog_data->base.urb_read_length,
+ DIV_ROUND_UP(imm_offset + 1, 2));
+ } else {
+ /* Replicate the patch handle to all enabled channels */
+ const fs_reg srcs[] = {
+ retype(brw_vec1_grf(0, 0), BRW_REGISTER_TYPE_UD)
+ };
+ fs_reg patch_handle = bld.vgrf(BRW_REGISTER_TYPE_UD, 1);
+ bld.LOAD_PAYLOAD(patch_handle, srcs, ARRAY_SIZE(srcs), 0);
+
+ inst = bld.emit(SHADER_OPCODE_URB_READ_SIMD8, dest, patch_handle);
+ inst->mlen = 1;
+ inst->offset = imm_offset;
+ inst->base_mrf = -1;
+ inst->regs_written = instr->num_components;
+ }
} else {
/* Indirect indexing - use per-slot offsets as well. */
const fs_reg srcs[] = {
@@ -1869,10 +1887,10 @@ fs_visitor::nir_emit_tes_intrinsic(const fs_builder &bld,
inst = bld.emit(SHADER_OPCODE_URB_READ_SIMD8_PER_SLOT, dest, payload);
inst->mlen = 2;
+ inst->offset = imm_offset;
+ inst->base_mrf = -1;
+ inst->regs_written = instr->num_components;
}
- inst->offset = imm_offset;
- inst->base_mrf = -1;
- inst->regs_written = instr->num_components;
break;
}
default: