X-Git-Url: https://git.libre-soc.org/?a=blobdiff_plain;f=src%2Fgallium%2Fdrivers%2Fnouveau%2Fcodegen%2Fnv50_ir_emit_gk110.cpp;h=2118c3153f7f0ceffbf3ab7ab59ec76ee27b27d5;hb=bab4f6c724d384cfee7e7f98ff3b52648850641d;hp=27d9b8e16895591ea3d4129b9933e1452ed47b0d;hpb=df2881381ac67c42aa8ec9e0ed28f21a1d253785;p=mesa.git

diff --git a/src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gk110.cpp b/src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gk110.cpp
index 27d9b8e1689..2118c3153f7 100644
--- a/src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gk110.cpp
+++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gk110.cpp
@@ -47,7 +47,7 @@ private:
 private:
    void emitForm_21(const Instruction *, uint32_t opc2, uint32_t opc1);
    void emitForm_C(const Instruction *, uint32_t opc, uint8_t ctg);
-   void emitForm_L(const Instruction *, uint32_t opc, uint8_t ctg, Modifier);
+   void emitForm_L(const Instruction *, uint32_t opc, uint8_t ctg, Modifier, int sCount = 3);
 
    void emitPredicate(const Instruction *);
 
@@ -96,6 +96,7 @@ private:
    void emitDMUL(const Instruction *);
    void emitIMAD(const Instruction *);
    void emitISAD(const Instruction *);
+   void emitSHLADD(const Instruction *);
    void emitFMAD(const Instruction *);
    void emitDMAD(const Instruction *);
    void emitMADSP(const Instruction *i);
@@ -108,6 +109,7 @@ private:
    void emitBFIND(const Instruction *);
    void emitPERMT(const Instruction *);
    void emitShift(const Instruction *);
+   void emitShift64(const Instruction *);
 
    void emitSFnOp(const Instruction *, uint8_t subOp);
 
@@ -133,6 +135,8 @@ private:
 
    void emitFlow(const Instruction *);
 
+   void emitSHFL(const Instruction *);
+
    void emitVOTE(const Instruction *);
 
    void emitSULDGB(const TexInstruction *);
@@ -196,14 +200,18 @@ void CodeEmitterGK110::srcAddr32(const ValueRef& src, const int pos)
 
 void CodeEmitterGK110::defId(const ValueDef& def, const int pos)
 {
-   code[pos / 32] |= (def.get() ? DDATA(def).id : GK110_GPR_ZERO) << (pos % 32);
+   code[pos / 32] |= (def.get() && def.getFile() != FILE_FLAGS ? DDATA(def).id : GK110_GPR_ZERO) << (pos % 32);
 }
 
 bool CodeEmitterGK110::isLIMM(const ValueRef& ref, DataType ty, bool mod)
 {
    const ImmediateValue *imm = ref.get()->asImm();
 
-   return imm && (imm->reg.data.u32 & ((ty == TYPE_F32) ? 0xfff : 0xfff00000));
+   if (ty == TYPE_F32)
+      return imm && imm->reg.data.u32 & 0xfff;
+   else
+      return imm && (imm->reg.data.s32 > 0x7ffff ||
+                     imm->reg.data.s32 < -0x80000);
 }
 
 void
@@ -338,7 +346,7 @@ CodeEmitterGK110::setShortImmediate(const Instruction *i, const int s)
       code[1] |= ((u64 & 0x7fe0000000000000ULL) >> 53);
       code[1] |= ((u64 & 0x8000000000000000ULL) >> 36);
    } else {
-      assert((u32 & 0xfff00000) == 0 || (u32 & 0xfff00000) == 0xfff00000);
+      assert((u32 & 0xfff80000) == 0 || (u32 & 0xfff80000) == 0xfff80000);
       code[0] |= (u32 & 0x001ff) << 23;
       code[1] |= (u32 & 0x7fe00) >> 9;
       code[1] |= (u32 & 0x80000) << 8;
@@ -363,7 +371,7 @@ CodeEmitterGK110::setImmediate32(const Instruction *i, const int s,
 
 void
 CodeEmitterGK110::emitForm_L(const Instruction *i, uint32_t opc, uint8_t ctg,
-                             Modifier mod)
+                             Modifier mod, int sCount)
 {
    code[0] = ctg;
    code[1] = opc << 20;
@@ -372,7 +380,7 @@ CodeEmitterGK110::emitForm_L(const Instruction *i, uint32_t opc, uint8_t ctg,
 
    defId(i->def(0), 2);
 
-   for (int s = 0; s < 3 && i->srcExists(s); ++s) {
+   for (int s = 0; s < sCount && i->srcExists(s); ++s) {
       switch (i->src(s).getFile()) {
       case FILE_GPR:
          srcId(i->src(s), s ? 42 : 10);
@@ -485,25 +493,41 @@ CodeEmitterGK110::emitNOP(const Instruction *i)
 void
 CodeEmitterGK110::emitFMAD(const Instruction *i)
 {
-   assert(!isLIMM(i->src(1), TYPE_F32));
+   bool neg1 = (i->src(0).mod ^ i->src(1).mod).neg();
 
-   emitForm_21(i, 0x0c0, 0x940);
+   if (isLIMM(i->src(1), TYPE_F32)) {
+      assert(i->getDef(0)->reg.data.id == i->getSrc(2)->reg.data.id);
 
-   NEG_(34, 2);
-   SAT_(35);
-   RND_(36, F);
-   FTZ_(38);
-   DNZ_(39);
+      // last source is dst, so force 2 sources
+      emitForm_L(i, 0x600, 0x0, 0, 2);
 
-   bool neg1 = (i->src(0).mod ^ i->src(1).mod).neg();
+      if (i->flagsDef >= 0)
+         code[1] |= 1 << 23;
 
-   if (code[0] & 0x1) {
-      if (neg1)
-         code[1] ^= 1 << 27;
-   } else
-   if (neg1) {
-      code[1] |= 1 << 19;
+      SAT_(3a);
+      NEG_(3c, 2);
+
+      if (neg1) {
+         code[1] |= 1 << 27;
+      }
+   } else {
+      emitForm_21(i, 0x0c0, 0x940);
+
+      NEG_(34, 2);
+      SAT_(35);
+      RND_(36, F);
+
+      if (code[0] & 0x1) {
+         if (neg1)
+            code[1] ^= 1 << 27;
+      } else
+      if (neg1) {
+         code[1] |= 1 << 19;
+      }
    }
+
+   FTZ_(38);
+   DNZ_(39);
 }
 
 void
@@ -613,7 +637,7 @@ CodeEmitterGK110::emitIMUL(const Instruction *i)
    assert(!i->src(0).mod.neg() && !i->src(1).mod.neg());
    assert(!i->src(0).mod.abs() && !i->src(1).mod.abs());
 
-   if (i->src(1).getFile() == FILE_IMMEDIATE) {
+   if (isLIMM(i->src(1), TYPE_S32)) {
       emitForm_L(i, 0x280, 2, Modifier(0));
 
       if (i->subOp == NV50_IR_SUBOP_MUL_HIGH)
@@ -701,7 +725,7 @@ CodeEmitterGK110::emitUADD(const Instruction *i)
       if (addOp & 2)
          code[1] |= 1 << 27;
 
-      assert(!i->defExists(1));
+      assert(i->flagsDef < 0);
       assert(i->flagsSrc < 0);
 
       SAT_(39);
@@ -712,7 +736,7 @@ CodeEmitterGK110::emitUADD(const Instruction *i)
 
       code[1] |= addOp << 19;
 
-      if (i->defExists(1))
+      if (i->flagsDef >= 0)
          code[1] |= 1 << 18; // write carry
       if (i->flagsSrc >= 0)
          code[1] |= 1 << 14; // add carry
@@ -721,12 +745,11 @@ CodeEmitterGK110::emitUADD(const Instruction *i)
    }
 }
 
-// TODO: shl-add
 void
 CodeEmitterGK110::emitIMAD(const Instruction *i)
 {
    uint8_t addOp =
-      (i->src(2).mod.neg() << 1) | (i->src(0).mod.neg() ^ i->src(1).mod.neg());
+      i->src(2).mod.neg() | ((i->src(0).mod.neg() ^ i->src(1).mod.neg()) << 1);
 
    emitForm_21(i, 0x100, 0xa00);
 
@@ -756,6 +779,54 @@ CodeEmitterGK110::emitISAD(const Instruction *i)
       code[1] |= 1 << 19;
 }
 
+void
+CodeEmitterGK110::emitSHLADD(const Instruction *i)
+{
+   uint8_t addOp = (i->src(0).mod.neg() << 1) | i->src(2).mod.neg();
+   const ImmediateValue *imm = i->src(1).get()->asImm();
+   assert(imm);
+
+   if (i->src(2).getFile() == FILE_IMMEDIATE) {
+      code[0] = 0x1;
+      code[1] = 0xc0c << 20;
+   } else {
+      code[0] = 0x2;
+      code[1] = 0x20c << 20;
+   }
+   code[1] |= addOp << 19;
+
+   emitPredicate(i);
+
+   defId(i->def(0), 2);
+   srcId(i->src(0), 10);
+
+   if (i->flagsDef >= 0)
+      code[1] |= 1 << 18;
+
+   assert(!(imm->reg.data.u32 & 0xffffffe0));
+   code[1] |= imm->reg.data.u32 << 10;
+
+   switch (i->src(2).getFile()) {
+   case FILE_GPR:
+      assert(code[0] & 0x2);
+      code[1] |= 0xc << 28;
+      srcId(i->src(2), 23);
+      break;
+   case FILE_MEMORY_CONST:
+      assert(code[0] & 0x2);
+      code[1] |= 0x4 << 28;
+      setCAddress14(i->src(2));
+      break;
+   case FILE_IMMEDIATE:
+      assert(code[0] & 0x1);
+      setShortImmediate(i, 2);
+      break;
+   default:
+      assert(!"bad src2 file");
+      break;
+   }
+}
+
 void
 CodeEmitterGK110::emitNOT(const Instruction *i)
 {
@@ -773,7 +844,7 @@ CodeEmitterGK110::emitNOT(const Instruction *i)
       break;
    case FILE_MEMORY_CONST:
       code[1] |= 0x4 << 28;
-      setCAddress14(i->src(1));
+      setCAddress14(i->src(0));
       break;
    default:
       assert(0);
@@ -887,6 +958,24 @@ CodeEmitterGK110::emitShift(const Instruction *i)
       code[1] |= 1 << 10;
 }
 
+void
+CodeEmitterGK110::emitShift64(const Instruction *i)
+{
+   if (i->op == OP_SHR) {
+      emitForm_21(i, 0x27c, 0xc7c);
+      if (isSignedType(i->sType))
+         code[1] |= 0x100;
+      if (i->subOp & NV50_IR_SUBOP_SHIFT_HIGH)
+         code[1] |= 1 << 19;
+   } else {
+      emitForm_21(i, 0xdfc, 0xf7c);
+   }
+   code[1] |= 0x200;
+
+   if (i->subOp & NV50_IR_SUBOP_SHIFT_WRAP)
+      code[1] |= 1 << 21;
+}
+
 void
 CodeEmitterGK110::emitPreOp(const Instruction *i)
 {
@@ -945,6 +1034,9 @@ CodeEmitterGK110::emitMINMAX(const Instruction *i)
    if (i->dType == TYPE_S32)
       code[1] |= 1 << 19;
    code[1] |= (i->op == OP_MIN) ? 0x1c00 : 0x3c00; // [!]pt
+   code[1] |= i->subOp << 14;
+   if (i->flagsDef >= 0)
+      code[1] |= i->subOp << 18;
 
    FTZ_(2f);
    ABS_(31, 0);
@@ -1091,6 +1183,8 @@ CodeEmitterGK110::emitSET(const CmpInstruction *i)
    } else {
       code[1] |= 0x7 << 10;
    }
+   if (i->flagsSrc >= 0)
+      code[1] |= 1 << 14;
    emitCondCode(i->setCond,
                 isFloatType(i->sType) ? 0x33 : 0x34,
                 isFloatType(i->sType) ? 0xf : 0x7);
@@ -1110,6 +1204,8 @@ CodeEmitterGK110::emitSLCT(const CmpInstruction *i)
    } else {
       emitForm_21(i, 0x1a0, 0xb20);
       emitCondCode(cc, 0x34, 0x7);
+      if (i->dType == TYPE_S32)
+         code[1] |= 1 << 19;
    }
 }
 
@@ -1321,15 +1417,12 @@ void
 CodeEmitterGK110::emitQUADOP(const Instruction *i, uint8_t qOp, uint8_t laneMask)
 {
    code[0] = 0x00000002 | ((qOp & 1) << 31);
-   code[1] = 0x7fc00000 | (qOp >> 1) | (laneMask << 12);
+   code[1] = 0x7fc00200 | (qOp >> 1) | (laneMask << 12); // dall
 
    defId(i->def(0), 2);
    srcId(i->src(0), 10);
    srcId((i->srcExists(1) && i->predSrc != 1) ? i->src(1) : i->src(0), 23);
 
-   if (i->op == OP_QUADOP && progType != Program::TYPE_FRAGMENT)
-      code[1] |= 1 << 9; // dall
-
    emitPredicate(i);
 }
 
@@ -1478,22 +1571,104 @@ CodeEmitterGK110::emitFlow(const Instruction *i)
    }
 }
 
+void
+CodeEmitterGK110::emitSHFL(const Instruction *i)
+{
+   const ImmediateValue *imm;
+
+   code[0] = 0x00000002;
+   code[1] = 0x78800000 | (i->subOp << 1);
+
+   emitPredicate(i);
+
+   defId(i->def(0), 2);
+   srcId(i->src(0), 10);
+
+   switch (i->src(1).getFile()) {
+   case FILE_GPR:
+      srcId(i->src(1), 23);
+      break;
+   case FILE_IMMEDIATE:
+      imm = i->getSrc(1)->asImm();
+      assert(imm && imm->reg.data.u32 < 0x20);
+      code[0] |= imm->reg.data.u32 << 23;
+      code[0] |= 1 << 31;
+      break;
+   default:
+      assert(!"invalid src1 file");
+      break;
+   }
+
+   switch (i->src(2).getFile()) {
+   case FILE_GPR:
+      srcId(i->src(2), 42);
+      break;
+   case FILE_IMMEDIATE:
+      imm = i->getSrc(2)->asImm();
+      assert(imm && imm->reg.data.u32 < 0x2000);
+      code[1] |= imm->reg.data.u32 << 5;
+      code[1] |= 1;
+      break;
+   default:
+      assert(!"invalid src2 file");
+      break;
+   }
+
+   if (!i->defExists(1))
+      code[1] |= 7 << 19;
+   else {
+      assert(i->def(1).getFile() == FILE_PREDICATE);
+      defId(i->def(1), 51);
+   }
+}
+
 void
 CodeEmitterGK110::emitVOTE(const Instruction *i)
 {
-   assert(i->src(0).getFile() == FILE_PREDICATE &&
-          i->def(1).getFile() == FILE_PREDICATE);
+   const ImmediateValue *imm;
+   uint32_t u32;
 
    code[0] = 0x00000002;
    code[1] = 0x86c00000 | (i->subOp << 19);
 
    emitPredicate(i);
 
-   defId(i->def(0), 2);
-   defId(i->def(1), 48);
-   if (i->src(0).mod == Modifier(NV50_IR_MOD_NOT))
-      code[1] |= 1 << 13;
-   srcId(i->src(0), 42);
+   unsigned rp = 0;
+   for (int d = 0; i->defExists(d); d++) {
+      if (i->def(d).getFile() == FILE_PREDICATE) {
+         assert(!(rp & 2));
+         rp |= 2;
+         defId(i->def(d), 48);
+      } else if (i->def(d).getFile() == FILE_GPR) {
+         assert(!(rp & 1));
+         rp |= 1;
+         defId(i->def(d), 2);
+      } else {
+         assert(!"Unhandled def");
+      }
+   }
+   if (!(rp & 1))
+      code[0] |= 255 << 2;
+   if (!(rp & 2))
+      code[1] |= 7 << 16;
+
+   switch (i->src(0).getFile()) {
+   case FILE_PREDICATE:
+      if (i->src(0).mod == Modifier(NV50_IR_MOD_NOT))
+         code[0] |= 1 << 13;
+      srcId(i->src(0), 42);
+      break;
+   case FILE_IMMEDIATE:
+      imm = i->getSrc(0)->asImm();
+      assert(imm);
+      u32 = imm->reg.data.u32;
+      assert(u32 == 0 || u32 == 1);
+      code[1] |= (u32 == 1 ? 0x7 : 0xf) << 10;
+      break;
+   default:
+      assert(!"Unhandled src");
+      break;
+   }
 }
 
 void
@@ -2122,6 +2297,7 @@ CodeEmitterGK110::getSRegEncoding(const ValueRef& ref)
    case SV_INVOCATION_ID: return 0x11;
    case SV_YDIR:          return 0x12;
    case SV_THREAD_KILL:   return 0x13;
+   case SV_COMBINED_TID:  return 0x20;
    case SV_TID:           return 0x21 + SDATA(ref).sv.index;
    case SV_CTAID:         return 0x25 + SDATA(ref).sv.index;
    case SV_NTID:          return 0x29 + SDATA(ref).sv.index;
@@ -2129,6 +2305,11 @@ CodeEmitterGK110::getSRegEncoding(const ValueRef& ref)
    case SV_NCTAID:        return 0x2d + SDATA(ref).sv.index;
    case SV_LBASE:         return 0x34;
    case SV_SBASE:         return 0x30;
+   case SV_LANEMASK_EQ:   return 0x38;
+   case SV_LANEMASK_LT:   return 0x39;
+   case SV_LANEMASK_LE:   return 0x3a;
+   case SV_LANEMASK_GT:   return 0x3b;
+   case SV_LANEMASK_GE:   return 0x3c;
    case SV_CLOCK:         return 0x50 + SDATA(ref).sv.index;
    default:
       assert(!"no sreg for system value");
@@ -2139,6 +2320,34 @@ CodeEmitterGK110::getSRegEncoding(const ValueRef& ref)
 void
 CodeEmitterGK110::emitMOV(const Instruction *i)
 {
+   if (i->def(0).getFile() == FILE_PREDICATE) {
+      if (i->src(0).getFile() == FILE_GPR) {
+         // Use ISETP.NE.AND dst, PT, src, RZ, PT
+         code[0] = 0x00000002;
+         code[1] = 0xdb500000;
+
+         code[0] |= 0x7 << 2;
+         code[0] |= 0xff << 23;
+         code[1] |= 0x7 << 10;
+         srcId(i->src(0), 10);
+      } else
+      if (i->src(0).getFile() == FILE_PREDICATE) {
+         // Use PSETP.AND.AND dst, PT, src, PT, PT
+         code[0] = 0x00000002;
+         code[1] = 0x84800000;
+
+         code[0] |= 0x7 << 2;
+         code[1] |= 0x7 << 0;
+         code[1] |= 0x7 << 10;
+
+         srcId(i->src(0), 14);
+      } else {
+         assert(!"Unexpected source for predicate destination");
+         emitNOP(i);
+      }
+      emitPredicate(i);
+      defId(i->def(0), 5);
+   } else
    if (i->src(0).getFile() == FILE_SYSTEM_VALUE) {
       code[0] = 0x00000002 | (getSRegEncoding(i->src(0)) << 23);
       code[1] = 0x86400000;
@@ -2363,6 +2572,9 @@ CodeEmitterGK110::emitInstruction(Instruction *insn)
    case OP_SAD:
       emitISAD(insn);
       break;
+   case OP_SHLADD:
+      emitSHLADD(insn);
+      break;
    case OP_NOT:
       emitNOT(insn);
       break;
@@ -2377,7 +2589,10 @@ CodeEmitterGK110::emitInstruction(Instruction *insn)
       break;
    case OP_SHL:
    case OP_SHR:
-      emitShift(insn);
+      if (typeSizeof(insn->sType) == 8)
+         emitShift64(insn);
+      else
+         emitShift(insn);
       break;
    case OP_SET:
    case OP_SET_AND:
@@ -2506,6 +2721,9 @@ CodeEmitterGK110::emitInstruction(Instruction *insn)
    case OP_CCTL:
       emitCCTL(insn);
       break;
+   case OP_SHFL:
+      emitSHFL(insn);
+      break;
    case OP_VOTE:
       emitVOTE(insn);
       break;