aco: fix wrong IR in nir_intrinsic_load_barycentric_at_sample

Only GFX6 was affected, my mistake. The total number of SGPR operands
should be 4 when we want to create a vec4.

Fixes: dbdf3b3ef97 ("aco: implement nir_intrinsic_load_barycentric_at_sample on GFX6")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3477>

diff --git a/src/amd/compiler/aco_instruction_selection.cpp b/src/amd/compiler/aco_instruction_selection.cpp
index 22d6e3f341fdab78974d4582a83ae0e845329c39..47e1737b5d9172468de5444508a79922ecc65862 100644 (file)
--- a/src/amd/compiler/aco_instruction_selection.cpp
+++ b/src/amd/compiler/aco_instruction_selection.cpp
@@ -5579,7 +5579,7 @@ void visit_intrinsic(isel_context *ctx, nir_intrinsic_instr *instr)
 
          uint32_t rsrc_conf = S_008F0C_NUM_FORMAT(V_008F0C_BUF_NUM_FORMAT_FLOAT) |
                               S_008F0C_DATA_FORMAT(V_008F0C_BUF_DATA_FORMAT_32);
-         Temp rsrc = bld.pseudo(aco_opcode::p_create_vector, bld.def(s4), private_segment_buffer, Operand(rsrc_conf));
+         Temp rsrc = bld.pseudo(aco_opcode::p_create_vector, bld.def(s4), private_segment_buffer, Operand(0u), Operand(rsrc_conf));
 
          addr = bld.vop2(aco_opcode::v_lshlrev_b32, bld.def(v1), Operand(3u), addr);
          addr = bld.pseudo(aco_opcode::p_create_vector, bld.def(v2), addr, Operand(0u));