radeon/llvm: add support for some ALU instructions

Add support for IABS, NOT, AND, XOR, OR, UADD, UDIV, IDIV, MOD, UMOD, INEG,
I2F, U2F, F2U, F2I, USEQ, USGE, USLT, USNE, ISGE, ISLT, ROUND, MIN, MAX,
IMIN, IMAX, UMIN, UMAX

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>

diff --git a/src/gallium/drivers/radeon/AMDGPUGenInstrEnums.pl b/src/gallium/drivers/radeon/AMDGPUGenInstrEnums.pl
index 1fd4fb04b3e14814f99722964e5ece741626c5a7..c6d43873c0878ee19affb140ef5aa72d3ea1ea85 100644 (file)
--- a/src/gallium/drivers/radeon/AMDGPUGenInstrEnums.pl
+++ b/src/gallium/drivers/radeon/AMDGPUGenInstrEnums.pl
@@ -41,7 +41,7 @@ my $FILE_TYPE = $ARGV[0];
 
 open AMDIL, '<', 'AMDILInstructions.td';
 
-my @INST_ENUMS = ('NONE', 'FEQ', 'FGE', 'FLT', 'FNE', 'MOVE_f32', 'MOVE_i32', 'FTOI', 'ITOF', 'CMOVLOG_f32', 'UGT', 'IGE', 'INE', 'UGE', 'IEQ');
+my @INST_ENUMS = ('NONE', 'FEQ', 'FGE', 'FLT', 'FNE', 'MOVE_f32', 'MOVE_i32', 'FTOI', 'ITOF', 'CMOVLOG_f32', 'UGT', 'IGE', 'INE', 'UGE', 'IEQ', 'BINARY_OR_i32', 'BINARY_NOT_i32');
 
 while (<AMDIL>) {
   if ($_ =~ /defm\s+([A-Z_]+)\s+:\s+([A-Za-z0-9]+)</) {

diff --git a/src/gallium/drivers/radeon/AMDGPUIntrinsics.td b/src/gallium/drivers/radeon/AMDGPUIntrinsics.td
index 41f9ca21bcd2471ac043e5c6de55c7f842b549b0..d8ea45287a47bc07cc561f1ba6bab3bce8203b32 100644 (file)
--- a/src/gallium/drivers/radeon/AMDGPUIntrinsics.td
+++ b/src/gallium/drivers/radeon/AMDGPUIntrinsics.td
@@ -35,7 +35,7 @@ let TargetPrefix = "AMDGPU", isTarget = 1 in {
   def int_AMDGPU_rsq : Intrinsic<[llvm_float_ty], [llvm_float_ty], []>;
   def int_AMDGPU_seq : Intrinsic<[llvm_float_ty], [llvm_float_ty, llvm_float_ty], []>;
   def int_AMDGPU_sgt : Intrinsic<[llvm_float_ty], [llvm_float_ty, llvm_float_ty], []>;
-  def int_AMDGPU_sge : BinaryIntFloat;
+  def int_AMDGPU_sge : Intrinsic<[llvm_float_ty], [llvm_float_ty, llvm_float_ty], []>;
   def int_AMDGPU_sin : Intrinsic<[llvm_float_ty], [llvm_float_ty], []>;
   def int_AMDGPU_sle : Intrinsic<[llvm_float_ty], [llvm_float_ty, llvm_float_ty], []>;
   def int_AMDGPU_sne : Intrinsic<[llvm_float_ty], [llvm_float_ty, llvm_float_ty], []>;
@@ -50,6 +50,10 @@ let TargetPrefix = "AMDGPU", isTarget = 1 in {
   def int_AMDGPU_trunc : Intrinsic<[llvm_float_ty], [llvm_float_ty], []>;
   def int_AMDGPU_ddx : Intrinsic<[llvm_v4f32_ty], [llvm_v4f32_ty, llvm_i32_ty, llvm_i32_ty], []>;
   def int_AMDGPU_ddy : Intrinsic<[llvm_v4f32_ty], [llvm_v4f32_ty, llvm_i32_ty, llvm_i32_ty], []>;
+  def int_AMDGPU_imax : Intrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty], []>;
+  def int_AMDGPU_imin : Intrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty], []>;
+  def int_AMDGPU_umax : Intrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty], []>;
+  def int_AMDGPU_umin : Intrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty], []>;
 }
 
 let TargetPrefix = "TGSI", isTarget = 1 in {

diff --git a/src/gallium/drivers/radeon/R600Instructions.td b/src/gallium/drivers/radeon/R600Instructions.td
index 9df057025f2da88398e3e798947f608f72229242..edbade706270eeeade7de43426faaf1305674ed6 100644 (file)
--- a/src/gallium/drivers/radeon/R600Instructions.td
+++ b/src/gallium/drivers/radeon/R600Instructions.td
@@ -84,7 +84,7 @@ class R600_3OP <bits<32> inst, string opName, list<dag> pattern,
   InstR600 <inst,
           (outs R600_Reg32:$dst),
           (ins R600_Reg32:$src0, R600_Reg32:$src1, R600_Reg32:$src2, variable_ops),
-          !strconcat(opName, "$dst $src0, $src1, $src2"),
+          !strconcat(opName, " $dst, $src0, $src1, $src2"),
           pattern,
           itin>{
 
@@ -311,6 +311,18 @@ def TRUNC : R600_1OP <
   [(set R600_Reg32:$dst, (int_AMDGPU_trunc R600_Reg32:$src))]
 >;
 
+def CEIL : R600_1OP <
+  0x12, "CEIL",
+  [(set R600_Reg32:$dst, (int_AMDIL_round_neginf R600_Reg32:$src))]> {
+  let AMDILOp = AMDILInst.ROUND_NEGINF_f32;
+}
+
+def RNDNE : R600_1OP <
+  0x13, "RNDNE",
+  [(set R600_Reg32:$dst, (int_AMDIL_round_nearest R600_Reg32:$src))]> {
+  let AMDILOp = AMDILInst.ROUND_NEAREST_f32;
+}
+
 def FLOOR : R600_1OP <
   0x14, "FLOOR",
   [(set R600_Reg32:$dst, (int_AMDGPU_floor R600_Reg32:$src))]
@@ -329,59 +341,88 @@ def AND_INT : R600_2OP <
   let AMDILOp = AMDILInst.AND_i32;
 }
 
+def OR_INT : R600_2OP <
+  0x31, "OR_INT",
+  []>{
+  let AMDILOp = AMDILInst.BINARY_OR_i32;
+}
+
 def XOR_INT : R600_2OP <
   0x32, "XOR_INT",
   []
 >;
 
+def NOT_INT : R600_1OP <
+  0x33, "NOT_INT",
+  []>{
+  let AMDILOp = AMDILInst.BINARY_NOT_i32;
+}
+
 def ADD_INT : R600_2OP <
-  0x34, "ADD_INT $dst, $src0, $src1",
+  0x34, "ADD_INT",
   []>{
   let AMDILOp = AMDILInst.ADD_i32;
 }
 
 def SUB_INT : R600_2OP <
-       0x35, "SUB_INT $dst, $src0, $src1",
+       0x35, "SUB_INT",
        []
 >;
 
+def MAX_INT : R600_2OP <
+  0x36, "MAX_INT",
+  [(set R600_Reg32:$dst, (int_AMDGPU_imax R600_Reg32:$src0, R600_Reg32:$src1))]>;
+
+def MIN_INT : R600_2OP <
+  0x37, "MIN_INT",
+  [(set R600_Reg32:$dst, (int_AMDGPU_imin R600_Reg32:$src0, R600_Reg32:$src1))]>;
+
+def MAX_UINT : R600_2OP <
+  0x38, "MAX_UINT",
+  [(set R600_Reg32:$dst, (int_AMDGPU_umax R600_Reg32:$src0, R600_Reg32:$src1))]>;
+
+def MIN_UINT : R600_2OP <
+  0x39, "MIN_UINT",
+  [(set R600_Reg32:$dst, (int_AMDGPU_umin R600_Reg32:$src0, R600_Reg32:$src1))]>;
+
+
 def SETE_INT : R600_2OP <
-  0x3A, "SETE_INT $dst, $src0, $src1",
+  0x3A, "SETE_INT",
   []>{
   let AMDILOp = AMDILInst.IEQ;
 }
 
 def SETGT_INT : R600_2OP <
-  0x3B, "SGT_INT $dst, $src0, $src1",
+  0x3B, "SGT_INT",
   []
 >;
 
 def SETGE_INT : R600_2OP <
-       0x3C, "SETGE_INT $dst, $src0, $src1",
+       0x3C, "SETGE_INT",
        []>{
   let AMDILOp = AMDILInst.IGE;
 }
 
 def SETNE_INT : R600_2OP <
-  0x3D, "SETNE_INT $dst, $src0, $src1",
+  0x3D, "SETNE_INT",
   []>{
   let AMDILOp = AMDILInst.INE;
 }
 
 def SETGT_UINT : R600_2OP <
-  0x3E, "SETGT_UINT $dst, $src0, $src1",
+  0x3E, "SETGT_UINT",
   []>{
   let AMDILOp = AMDILInst.UGT;
 }
 
 def SETGE_UINT : R600_2OP <
-  0x3F, "SETGE_UINT $dst, $src0, $src1",
+  0x3F, "SETGE_UINT",
   []>{
   let AMDILOp = AMDILInst.UGE;
 }
 
 def CNDE_INT : R600_3OP <
-       0x1C, "CNDE_INT $dst, $src0, $src1, $src2",
+       0x1C, "CNDE_INT",
        []
 >;
 

diff --git a/src/gallium/drivers/radeon/radeon_setup_tgsi_llvm.c b/src/gallium/drivers/radeon/radeon_setup_tgsi_llvm.c
index 2932bdd9490c86fb236c5f33c6b496c706ea4cb2..502d551dce1481a01bf0a6e06884b39eb0bcfeef 100644 (file)
--- a/src/gallium/drivers/radeon/radeon_setup_tgsi_llvm.c
+++ b/src/gallium/drivers/radeon/radeon_setup_tgsi_llvm.c
@@ -533,6 +533,116 @@ static void tex_fetch_args(
        emit_data->dst_type = LLVMVectorType(bld_base->base.elem_type, 4);
 }
 
+static void emit_icmp(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       unsigned pred;
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       LLVMContextRef context = bld_base->base.gallivm->context;
+
+       switch (emit_data->inst->Instruction.Opcode) {
+       case TGSI_OPCODE_USEQ: pred = LLVMIntEQ; break;
+       case TGSI_OPCODE_USNE: pred = LLVMIntNE; break;
+       case TGSI_OPCODE_USGE: pred = LLVMIntUGE; break;
+       case TGSI_OPCODE_USLT: pred = LLVMIntULT; break;
+       case TGSI_OPCODE_ISGE: pred = LLVMIntSGE; break;
+       case TGSI_OPCODE_ISLT: pred = LLVMIntSLT; break;
+       default:
+               assert(!"unknown instruction");
+       }
+
+       LLVMValueRef v = LLVMBuildICmp(builder, pred,
+                       emit_data->args[0], emit_data->args[1],"");
+
+       v = LLVMBuildSExtOrBitCast(builder, v,
+                       LLVMInt32TypeInContext(context), "");
+
+       emit_data->output[emit_data->chan] = v;
+}
+
+static void emit_not(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       LLVMValueRef v = bitcast(bld_base, TGSI_TYPE_UNSIGNED,
+                       emit_data->args[0]);
+       emit_data->output[emit_data->chan] = LLVMBuildNot(builder, v, "");
+}
+
+static void emit_and(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildAnd(builder,
+                       emit_data->args[0], emit_data->args[1], "");
+}
+
+static void emit_or(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildOr(builder,
+                       emit_data->args[0], emit_data->args[1], "");
+}
+
+static void emit_uadd(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildAdd(builder,
+                       emit_data->args[0], emit_data->args[1], "");
+}
+
+static void emit_udiv(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildUDiv(builder,
+                       emit_data->args[0], emit_data->args[1], "");
+}
+
+static void emit_idiv(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildSDiv(builder,
+                       emit_data->args[0], emit_data->args[1], "");
+}
+
+static void emit_mod(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildSRem(builder,
+                       emit_data->args[0], emit_data->args[1], "");
+}
+
+static void emit_umod(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildURem(builder,
+                       emit_data->args[0], emit_data->args[1], "");
+}
+
 static void emit_shl(
                const struct lp_build_tgsi_action * action,
                struct lp_build_tgsi_context * bld_base,
@@ -562,6 +672,90 @@ static void emit_ishr(
                        emit_data->args[0], emit_data->args[1], "");
 }
 
+static void emit_xor(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildXor(builder,
+                       emit_data->args[0], emit_data->args[1], "");
+}
+
+static void emit_ssg(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+
+       LLVMValueRef cmp, val;
+
+       if (emit_data->inst->Instruction.Opcode == TGSI_OPCODE_ISSG) {
+               cmp = LLVMBuildICmp(builder, LLVMIntSGT, emit_data->args[0], bld_base->int_bld.zero, "");
+               val = LLVMBuildSelect(builder, cmp, bld_base->int_bld.one, emit_data->args[0], "");
+               cmp = LLVMBuildICmp(builder, LLVMIntSGE, val, bld_base->int_bld.zero, "");
+               val = LLVMBuildSelect(builder, cmp, val, LLVMConstInt(bld_base->int_bld.elem_type, -1, true), "");
+       } else { // float SSG
+               cmp = LLVMBuildFCmp(builder, LLVMRealUGT, emit_data->args[0], bld_base->int_bld.zero, "");
+               val = LLVMBuildSelect(builder, cmp, bld_base->base.one, emit_data->args[0], "");
+               cmp = LLVMBuildFCmp(builder, LLVMRealUGE, val, bld_base->base.zero, "");
+               val = LLVMBuildSelect(builder, cmp, val, LLVMConstReal(bld_base->base.elem_type, -1), "");
+       }
+
+       emit_data->output[emit_data->chan] = val;
+}
+
+static void emit_ineg(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildNeg(builder,
+                       emit_data->args[0], "");
+}
+
+static void emit_f2i(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildFPToSI(builder,
+                       emit_data->args[0], bld_base->int_bld.elem_type, "");
+}
+
+static void emit_f2u(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildFPToUI(builder,
+                       emit_data->args[0], bld_base->uint_bld.elem_type, "");
+}
+
+static void emit_i2f(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildSIToFP(builder,
+                       emit_data->args[0], bld_base->base.elem_type, "");
+}
+
+static void emit_u2f(
+               const struct lp_build_tgsi_action * action,
+               struct lp_build_tgsi_context * bld_base,
+               struct lp_build_emit_data * emit_data)
+{
+       LLVMBuilderRef builder = bld_base->base.gallivm->builder;
+       emit_data->output[emit_data->chan] = LLVMBuildUIToFP(builder,
+                       emit_data->args[0], bld_base->base.elem_type, "");
+}
+
 static void emit_immediate(struct lp_build_tgsi_context * bld_base,
                const struct tgsi_full_immediate *imm)
 {
@@ -635,17 +829,58 @@ void radeon_llvm_context_init(struct radeon_llvm_context * ctx)
 
        lp_set_default_actions(bld_base);
 
+       bld_base->op_actions[TGSI_OPCODE_IABS].emit = lp_build_tgsi_intrinsic;
+       bld_base->op_actions[TGSI_OPCODE_IABS].intr_name = "llvm.AMDIL.abs.";
+       bld_base->op_actions[TGSI_OPCODE_NOT].emit = emit_not;
+       bld_base->op_actions[TGSI_OPCODE_AND].emit = emit_and;
+       bld_base->op_actions[TGSI_OPCODE_XOR].emit = emit_xor;
+       bld_base->op_actions[TGSI_OPCODE_OR].emit = emit_or;
+       bld_base->op_actions[TGSI_OPCODE_UADD].emit = emit_uadd;
+       bld_base->op_actions[TGSI_OPCODE_UDIV].emit = emit_udiv;
+       bld_base->op_actions[TGSI_OPCODE_IDIV].emit = emit_idiv;
+       bld_base->op_actions[TGSI_OPCODE_MOD].emit = emit_mod;
+       bld_base->op_actions[TGSI_OPCODE_UMOD].emit = emit_umod;
+       bld_base->op_actions[TGSI_OPCODE_INEG].emit = emit_ineg;
        bld_base->op_actions[TGSI_OPCODE_SHL].emit = emit_shl;
        bld_base->op_actions[TGSI_OPCODE_ISHR].emit = emit_ishr;
        bld_base->op_actions[TGSI_OPCODE_USHR].emit = emit_ushr;
+       bld_base->op_actions[TGSI_OPCODE_SSG].emit = emit_ssg;
+       bld_base->op_actions[TGSI_OPCODE_ISSG].emit = emit_ssg;
+       bld_base->op_actions[TGSI_OPCODE_I2F].emit = emit_i2f;
+       bld_base->op_actions[TGSI_OPCODE_U2F].emit = emit_u2f;
+       bld_base->op_actions[TGSI_OPCODE_F2I].emit = emit_f2i;
+       bld_base->op_actions[TGSI_OPCODE_F2U].emit = emit_f2u;
        bld_base->op_actions[TGSI_OPCODE_DDX].intr_name = "llvm.AMDGPU.ddx";
        bld_base->op_actions[TGSI_OPCODE_DDX].fetch_args = tex_fetch_args;
        bld_base->op_actions[TGSI_OPCODE_DDY].intr_name = "llvm.AMDGPU.ddy";
        bld_base->op_actions[TGSI_OPCODE_DDY].fetch_args = tex_fetch_args;
+       bld_base->op_actions[TGSI_OPCODE_USEQ].emit = emit_icmp;
+       bld_base->op_actions[TGSI_OPCODE_USGE].emit = emit_icmp;
+       bld_base->op_actions[TGSI_OPCODE_USLT].emit = emit_icmp;
+       bld_base->op_actions[TGSI_OPCODE_USNE].emit = emit_icmp;
+       bld_base->op_actions[TGSI_OPCODE_ISGE].emit = emit_icmp;
+       bld_base->op_actions[TGSI_OPCODE_ISLT].emit = emit_icmp;
+       bld_base->op_actions[TGSI_OPCODE_ROUND].emit = lp_build_tgsi_intrinsic;
+       bld_base->op_actions[TGSI_OPCODE_ROUND].intr_name = "llvm.AMDIL.round.nearest.";
+       bld_base->op_actions[TGSI_OPCODE_MIN].emit = lp_build_tgsi_intrinsic;
+       bld_base->op_actions[TGSI_OPCODE_MIN].intr_name = "llvm.AMDIL.min.";
+       bld_base->op_actions[TGSI_OPCODE_MAX].emit = lp_build_tgsi_intrinsic;
+       bld_base->op_actions[TGSI_OPCODE_MAX].intr_name = "llvm.AMDIL.max.";
+       bld_base->op_actions[TGSI_OPCODE_IMIN].emit = lp_build_tgsi_intrinsic;
+       bld_base->op_actions[TGSI_OPCODE_IMIN].intr_name = "llvm.AMDGPU.imin";
+       bld_base->op_actions[TGSI_OPCODE_IMAX].emit = lp_build_tgsi_intrinsic;
+       bld_base->op_actions[TGSI_OPCODE_IMAX].intr_name = "llvm.AMDGPU.imax";
+       bld_base->op_actions[TGSI_OPCODE_UMIN].emit = lp_build_tgsi_intrinsic;
+       bld_base->op_actions[TGSI_OPCODE_UMIN].intr_name = "llvm.AMDGPU.umin";
+       bld_base->op_actions[TGSI_OPCODE_UMAX].emit = lp_build_tgsi_intrinsic;
+       bld_base->op_actions[TGSI_OPCODE_UMAX].intr_name = "llvm.AMDGPU.umax";
        bld_base->op_actions[TGSI_OPCODE_TXF].fetch_args = tex_fetch_args;
        bld_base->op_actions[TGSI_OPCODE_TXF].intr_name = "llvm.AMDGPU.txf";
        bld_base->op_actions[TGSI_OPCODE_TXQ].fetch_args = tex_fetch_args;
        bld_base->op_actions[TGSI_OPCODE_TXQ].intr_name = "llvm.AMDGPU.txq";
+       bld_base->op_actions[TGSI_OPCODE_CEIL].emit = lp_build_tgsi_intrinsic;
+       bld_base->op_actions[TGSI_OPCODE_CEIL].intr_name = "llvm.AMDIL.round.neginf.";
+
 
 
        bld_base->op_actions[TGSI_OPCODE_ABS].emit = lp_build_tgsi_intrinsic;
@@ -696,7 +931,7 @@ void radeon_llvm_context_init(struct radeon_llvm_context * ctx)
        bld_base->op_actions[TGSI_OPCODE_SSG].emit = lp_build_tgsi_intrinsic;
        bld_base->op_actions[TGSI_OPCODE_SSG].intr_name = "llvm.AMDGPU.ssg";
        bld_base->op_actions[TGSI_OPCODE_SGE].emit = lp_build_tgsi_intrinsic;
-       bld_base->op_actions[TGSI_OPCODE_SGE].intr_name = "llvm.AMDGPU.sge.";
+       bld_base->op_actions[TGSI_OPCODE_SGE].intr_name = "llvm.AMDGPU.sge";
        bld_base->op_actions[TGSI_OPCODE_SEQ].emit = lp_build_tgsi_intrinsic;
        bld_base->op_actions[TGSI_OPCODE_SEQ].intr_name = "llvm.AMDGPU.seq";
        bld_base->op_actions[TGSI_OPCODE_SLE].fetch_args = radeon_llvm_fetch_args_2_reverse_soa;