Kenneth Graunke <kenn...@whitecape.org> writes:
> On Monday, June 01, 2015 01:03:10 PM Matt Turner wrote:
>> OPCODE_MOV is in the op_trans[] array.
>> ---
>> src/mesa/program/prog_to_nir.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/src/mesa/program/prog_to_nir.c b/src/mesa/program/prog_to_nir.c
>> index 5e104e7..d6f165e 100644
>> --- a/src/mesa/program/prog_to_nir.c
>> +++ b/src/mesa/program/prog_to_nir.c
>> @@ -891,7 +891,7 @@ ptn_emit_instruction(struct ptn_compile *c, struct
>> prog_instruction *prog_inst)
>> break;
>>
>> default:
>> - if (op_trans[op] != 0 || op == OPCODE_MOV) {
>> + if (op_trans[op] != 0) {
>> ptn_alu(b, op_trans[op], dest, src);
>> } else {
>> fprintf(stderr, "unknown opcode: %s\n", _mesa_opcode_string(op));
>>
>
> This was copy and pasted from Eric's TGSI->NIR pass, which special cases
> TGSI_OPCODE_MOV. Apparently, that's pointless too.nir_op_fmov was 0, so the op_trans[op] test for != 0 would error out on TGSI_OPCODE_MOVs. It looks like nir_op_fmov has changed value since then, so this doesn't work any more.
signature.asc
Description: PGP signature
_______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
