nouveau: Add support for SV_WORK_DIM

[mesa.git] / src / gallium / drivers / nouveau / codegen / nv50_ir_from_tgsi.cpp

diff --git a/src/gallium/drivers/nouveau/codegen/nv50_ir_from_tgsi.cpp b/src/gallium/drivers/nouveau/codegen/nv50_ir_from_tgsi.cpp
index beb7b5310bf4545c97451cf7fa452117504c6d1c..c0816cbf0609b4d5379133303c5a091b2f9bbf5d 100644 (file)
--- a/src/gallium/drivers/nouveau/codegen/nv50_ir_from_tgsi.cpp
+++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_from_tgsi.cpp
@@ -408,6 +408,7 @@ static nv50_ir::SVSemantic translateSysVal(uint sysval)
    case TGSI_SEMANTIC_BASEVERTEX: return nv50_ir::SV_BASEVERTEX;
    case TGSI_SEMANTIC_BASEINSTANCE: return nv50_ir::SV_BASEINSTANCE;
    case TGSI_SEMANTIC_DRAWID:     return nv50_ir::SV_DRAWID;
+   case TGSI_SEMANTIC_WORK_DIM:   return nv50_ir::SV_WORK_DIM;
    default:
       assert(0);
       return nv50_ir::SV_CLOCK;
@@ -548,6 +549,9 @@ nv50_ir::DataType Instruction::inferSrcType() const
    case TGSI_OPCODE_UBFE:
    case TGSI_OPCODE_UMSB:
    case TGSI_OPCODE_UP2H:
+   case TGSI_OPCODE_VOTE_ALL:
+   case TGSI_OPCODE_VOTE_ANY:
+   case TGSI_OPCODE_VOTE_EQ:
       return nv50_ir::TYPE_U32;
    case TGSI_OPCODE_I2F:
    case TGSI_OPCODE_I2D:
@@ -835,6 +839,10 @@ static nv50_ir::operation translateOpcode(uint opcode)
    NV50_IR_OPCODE_CASE(IMSB, BFIND);
    NV50_IR_OPCODE_CASE(UMSB, BFIND);
 
+   NV50_IR_OPCODE_CASE(VOTE_ALL, VOTE);
+   NV50_IR_OPCODE_CASE(VOTE_ANY, VOTE);
+   NV50_IR_OPCODE_CASE(VOTE_EQ, VOTE);
+
    NV50_IR_OPCODE_CASE(END, EXIT);
 
    default:
@@ -861,6 +869,9 @@ static uint16_t opcodeToSubOp(uint opcode)
    case TGSI_OPCODE_IMUL_HI:
    case TGSI_OPCODE_UMUL_HI:
       return NV50_IR_SUBOP_MUL_HIGH;
+   case TGSI_OPCODE_VOTE_ALL: return NV50_IR_SUBOP_VOTE_ALL;
+   case TGSI_OPCODE_VOTE_ANY: return NV50_IR_SUBOP_VOTE_ANY;
+   case TGSI_OPCODE_VOTE_EQ: return NV50_IR_SUBOP_VOTE_UNI;
    default:
       return 0;
    }
@@ -2378,6 +2389,9 @@ Converter::getImageCoords(std::vector<Value *> &coords, int r, int s)
 
    for (int c = 0; c < arg; ++c)
       coords.push_back(fetchSrc(s, c));
+
+   if (t.isMS())
+      coords.push_back(fetchSrc(s, 3));
 }
 
 // For raw loads, granularity is 4 byte.
@@ -3187,6 +3201,17 @@ Converter::handleInstruction(const struct tgsi_full_instruction *insn)
          mkCmp(op, tgsi.getSetCond(), dstTy, dst0[c], srcTy, src0, src1);
       }
       break;
+   case TGSI_OPCODE_VOTE_ALL:
+   case TGSI_OPCODE_VOTE_ANY:
+   case TGSI_OPCODE_VOTE_EQ:
+      val0 = new_LValue(func, FILE_PREDICATE);
+      FOR_EACH_DST_ENABLED_CHANNEL(0, c, tgsi) {
+         mkCmp(OP_SET, CC_NE, TYPE_U32, val0, TYPE_U32, fetchSrc(0, c), zero);
+         mkOp1(op, dstTy, val0, val0)
+            ->subOp = tgsi::opcodeToSubOp(tgsi.getOpcode());
+         mkCvt(OP_CVT, TYPE_U32, dst0[c], TYPE_U8, val0);
+      }
+      break;
    case TGSI_OPCODE_KILL_IF:
       val0 = new_LValue(func, FILE_PREDICATE);
       mask = 0;