367 lines
16 KiB
LLVM
367 lines
16 KiB
LLVM
; ModuleID = 'streamcluster_cuda_cpu-cuda-nvptx64-nvidia-cuda-sm_61.bc'
|
|
source_filename = "streamcluster_cuda_cpu.cu"
|
|
target datalayout = "e-i64:64-i128:128-v16:16-v32:32-n16:32:64"
|
|
target triple = "nvptx64-nvidia-cuda"
|
|
|
|
%struct.__cuda_builtin_blockIdx_t = type { i8 }
|
|
%struct.__cuda_builtin_gridDim_t = type { i8 }
|
|
%struct.__cuda_builtin_blockDim_t = type { i8 }
|
|
%struct.__cuda_builtin_threadIdx_t = type { i8 }
|
|
%struct.cudaFuncAttributes = type { i64, i64, i64, i32, i32, i32, i32, i32, i32, i32 }
|
|
%struct.Point = type { float, float*, i64, float }
|
|
|
|
$_ZN25__cuda_builtin_blockIdx_t17__fetch_builtin_xEv = comdat any
|
|
|
|
$_ZN24__cuda_builtin_gridDim_t17__fetch_builtin_xEv = comdat any
|
|
|
|
$_ZN25__cuda_builtin_blockIdx_t17__fetch_builtin_yEv = comdat any
|
|
|
|
$_ZN25__cuda_builtin_blockDim_t17__fetch_builtin_xEv = comdat any
|
|
|
|
$_ZN26__cuda_builtin_threadIdx_t17__fetch_builtin_xEv = comdat any
|
|
|
|
@blockIdx = extern_weak dso_local addrspace(1) global %struct.__cuda_builtin_blockIdx_t, align 1
|
|
@gridDim = extern_weak dso_local addrspace(1) global %struct.__cuda_builtin_gridDim_t, align 1
|
|
@blockDim = extern_weak dso_local addrspace(1) global %struct.__cuda_builtin_blockDim_t, align 1
|
|
@threadIdx = extern_weak dso_local addrspace(1) global %struct.__cuda_builtin_threadIdx_t, align 1
|
|
|
|
; Function Attrs: convergent noinline nounwind optnone
|
|
define weak dso_local i32 @cudaMalloc(i8** %p, i64 %s) #0 {
|
|
entry:
|
|
%p.addr = alloca i8**, align 8
|
|
%s.addr = alloca i64, align 8
|
|
store i8** %p, i8*** %p.addr, align 8
|
|
store i64 %s, i64* %s.addr, align 8
|
|
ret i32 999
|
|
}
|
|
|
|
; Function Attrs: convergent noinline nounwind optnone
|
|
define weak dso_local i32 @cudaFuncGetAttributes(%struct.cudaFuncAttributes* %p, i8* %c) #0 {
|
|
entry:
|
|
%p.addr = alloca %struct.cudaFuncAttributes*, align 8
|
|
%c.addr = alloca i8*, align 8
|
|
store %struct.cudaFuncAttributes* %p, %struct.cudaFuncAttributes** %p.addr, align 8
|
|
store i8* %c, i8** %c.addr, align 8
|
|
ret i32 999
|
|
}
|
|
|
|
; Function Attrs: convergent noinline nounwind optnone
|
|
define weak dso_local i32 @cudaDeviceGetAttribute(i32* %value, i32 %attr, i32 %device) #0 {
|
|
entry:
|
|
%value.addr = alloca i32*, align 8
|
|
%attr.addr = alloca i32, align 4
|
|
%device.addr = alloca i32, align 4
|
|
store i32* %value, i32** %value.addr, align 8
|
|
store i32 %attr, i32* %attr.addr, align 4
|
|
store i32 %device, i32* %device.addr, align 4
|
|
ret i32 999
|
|
}
|
|
|
|
; Function Attrs: convergent noinline nounwind optnone
|
|
define weak dso_local i32 @cudaGetDevice(i32* %device) #0 {
|
|
entry:
|
|
%device.addr = alloca i32*, align 8
|
|
store i32* %device, i32** %device.addr, align 8
|
|
ret i32 999
|
|
}
|
|
|
|
; Function Attrs: convergent noinline nounwind optnone
|
|
define weak dso_local i32 @cudaOccupancyMaxActiveBlocksPerMultiprocessor(i32* %numBlocks, i8* %func, i32 %blockSize, i64 %dynamicSmemSize) #0 {
|
|
entry:
|
|
%numBlocks.addr = alloca i32*, align 8
|
|
%func.addr = alloca i8*, align 8
|
|
%blockSize.addr = alloca i32, align 4
|
|
%dynamicSmemSize.addr = alloca i64, align 8
|
|
store i32* %numBlocks, i32** %numBlocks.addr, align 8
|
|
store i8* %func, i8** %func.addr, align 8
|
|
store i32 %blockSize, i32* %blockSize.addr, align 4
|
|
store i64 %dynamicSmemSize, i64* %dynamicSmemSize.addr, align 8
|
|
ret i32 999
|
|
}
|
|
|
|
; Function Attrs: convergent noinline nounwind optnone
|
|
define weak dso_local i32 @cudaOccupancyMaxActiveBlocksPerMultiprocessorWithFlags(i32* %numBlocks, i8* %func, i32 %blockSize, i64 %dynamicSmemSize, i32 %flags) #0 {
|
|
entry:
|
|
%numBlocks.addr = alloca i32*, align 8
|
|
%func.addr = alloca i8*, align 8
|
|
%blockSize.addr = alloca i32, align 4
|
|
%dynamicSmemSize.addr = alloca i64, align 8
|
|
%flags.addr = alloca i32, align 4
|
|
store i32* %numBlocks, i32** %numBlocks.addr, align 8
|
|
store i8* %func, i8** %func.addr, align 8
|
|
store i32 %blockSize, i32* %blockSize.addr, align 4
|
|
store i64 %dynamicSmemSize, i64* %dynamicSmemSize.addr, align 8
|
|
store i32 %flags, i32* %flags.addr, align 4
|
|
ret i32 999
|
|
}
|
|
|
|
; Function Attrs: convergent noinline nounwind optnone
|
|
define dso_local float @_Z6d_distiiiiPf(i32 %p1, i32 %p2, i32 %num, i32 %dim, float* %coord_d) #0 {
|
|
entry:
|
|
%p1.addr = alloca i32, align 4
|
|
%p2.addr = alloca i32, align 4
|
|
%num.addr = alloca i32, align 4
|
|
%dim.addr = alloca i32, align 4
|
|
%coord_d.addr = alloca float*, align 8
|
|
%retval1 = alloca float, align 4
|
|
%i = alloca i32, align 4
|
|
%tmp = alloca float, align 4
|
|
store i32 %p1, i32* %p1.addr, align 4
|
|
store i32 %p2, i32* %p2.addr, align 4
|
|
store i32 %num, i32* %num.addr, align 4
|
|
store i32 %dim, i32* %dim.addr, align 4
|
|
store float* %coord_d, float** %coord_d.addr, align 8
|
|
store float 0.000000e+00, float* %retval1, align 4
|
|
store i32 0, i32* %i, align 4
|
|
br label %for.cond
|
|
|
|
for.cond: ; preds = %for.inc, %entry
|
|
%0 = load i32, i32* %i, align 4
|
|
%1 = load i32, i32* %dim.addr, align 4
|
|
%cmp = icmp slt i32 %0, %1
|
|
br i1 %cmp, label %for.body, label %for.end
|
|
|
|
for.body: ; preds = %for.cond
|
|
%2 = load float*, float** %coord_d.addr, align 8
|
|
%3 = load i32, i32* %i, align 4
|
|
%4 = load i32, i32* %num.addr, align 4
|
|
%mul = mul nsw i32 %3, %4
|
|
%5 = load i32, i32* %p1.addr, align 4
|
|
%add = add nsw i32 %mul, %5
|
|
%idxprom = sext i32 %add to i64
|
|
%arrayidx = getelementptr inbounds float, float* %2, i64 %idxprom
|
|
%6 = load float, float* %arrayidx, align 4
|
|
%7 = load float*, float** %coord_d.addr, align 8
|
|
%8 = load i32, i32* %i, align 4
|
|
%9 = load i32, i32* %num.addr, align 4
|
|
%mul2 = mul nsw i32 %8, %9
|
|
%10 = load i32, i32* %p2.addr, align 4
|
|
%add3 = add nsw i32 %mul2, %10
|
|
%idxprom4 = sext i32 %add3 to i64
|
|
%arrayidx5 = getelementptr inbounds float, float* %7, i64 %idxprom4
|
|
%11 = load float, float* %arrayidx5, align 4
|
|
%sub = fsub contract float %6, %11
|
|
store float %sub, float* %tmp, align 4
|
|
%12 = load float, float* %tmp, align 4
|
|
%13 = load float, float* %tmp, align 4
|
|
%mul6 = fmul contract float %12, %13
|
|
%14 = load float, float* %retval1, align 4
|
|
%add7 = fadd contract float %14, %mul6
|
|
store float %add7, float* %retval1, align 4
|
|
br label %for.inc
|
|
|
|
for.inc: ; preds = %for.body
|
|
%15 = load i32, i32* %i, align 4
|
|
%inc = add nsw i32 %15, 1
|
|
store i32 %inc, i32* %i, align 4
|
|
br label %for.cond
|
|
|
|
for.end: ; preds = %for.cond
|
|
%16 = load float, float* %retval1, align 4
|
|
ret float %16
|
|
}
|
|
|
|
; Function Attrs: convergent noinline nounwind optnone
|
|
define dso_local void @_Z19kernel_compute_costiilP5PointiiPfS1_PiPb(i32 %num, i32 %dim, i64 %x, %struct.Point* %p, i32 %K, i32 %stride, float* %coord_d, float* %work_mem_d, i32* %center_table_d, i8* %switch_membership_d) #0 {
|
|
entry:
|
|
%num.addr = alloca i32, align 4
|
|
%dim.addr = alloca i32, align 4
|
|
%x.addr = alloca i64, align 8
|
|
%p.addr = alloca %struct.Point*, align 8
|
|
%K.addr = alloca i32, align 4
|
|
%stride.addr = alloca i32, align 4
|
|
%coord_d.addr = alloca float*, align 8
|
|
%work_mem_d.addr = alloca float*, align 8
|
|
%center_table_d.addr = alloca i32*, align 8
|
|
%switch_membership_d.addr = alloca i8*, align 8
|
|
%bid = alloca i32, align 4
|
|
%tid = alloca i32, align 4
|
|
%lower = alloca float*, align 8
|
|
%x_cost = alloca float, align 4
|
|
store i32 %num, i32* %num.addr, align 4
|
|
store i32 %dim, i32* %dim.addr, align 4
|
|
store i64 %x, i64* %x.addr, align 8
|
|
store %struct.Point* %p, %struct.Point** %p.addr, align 8
|
|
store i32 %K, i32* %K.addr, align 4
|
|
store i32 %stride, i32* %stride.addr, align 4
|
|
store float* %coord_d, float** %coord_d.addr, align 8
|
|
store float* %work_mem_d, float** %work_mem_d.addr, align 8
|
|
store i32* %center_table_d, i32** %center_table_d.addr, align 8
|
|
store i8* %switch_membership_d, i8** %switch_membership_d.addr, align 8
|
|
%call = call i32 @_ZN25__cuda_builtin_blockIdx_t17__fetch_builtin_xEv() #3
|
|
%call1 = call i32 @_ZN24__cuda_builtin_gridDim_t17__fetch_builtin_xEv() #3
|
|
%call2 = call i32 @_ZN25__cuda_builtin_blockIdx_t17__fetch_builtin_yEv() #3
|
|
%mul = mul i32 %call1, %call2
|
|
%add = add i32 %call, %mul
|
|
store i32 %add, i32* %bid, align 4
|
|
%call3 = call i32 @_ZN25__cuda_builtin_blockDim_t17__fetch_builtin_xEv() #3
|
|
%0 = load i32, i32* %bid, align 4
|
|
%mul4 = mul i32 %call3, %0
|
|
%call5 = call i32 @_ZN26__cuda_builtin_threadIdx_t17__fetch_builtin_xEv() #3
|
|
%add6 = add i32 %mul4, %call5
|
|
store i32 %add6, i32* %tid, align 4
|
|
%1 = load i32, i32* %tid, align 4
|
|
%2 = load i32, i32* %num.addr, align 4
|
|
%cmp = icmp slt i32 %1, %2
|
|
br i1 %cmp, label %if.then, label %if.end34
|
|
|
|
if.then: ; preds = %entry
|
|
%3 = load float*, float** %work_mem_d.addr, align 8
|
|
%4 = load i32, i32* %tid, align 4
|
|
%5 = load i32, i32* %stride.addr, align 4
|
|
%mul7 = mul nsw i32 %4, %5
|
|
%idxprom = sext i32 %mul7 to i64
|
|
%arrayidx = getelementptr inbounds float, float* %3, i64 %idxprom
|
|
store float* %arrayidx, float** %lower, align 8
|
|
%6 = load i32, i32* %tid, align 4
|
|
%7 = load i64, i64* %x.addr, align 8
|
|
%conv = trunc i64 %7 to i32
|
|
%8 = load i32, i32* %num.addr, align 4
|
|
%9 = load i32, i32* %dim.addr, align 4
|
|
%10 = load float*, float** %coord_d.addr, align 8
|
|
%call8 = call float @_Z6d_distiiiiPf(i32 %6, i32 %conv, i32 %8, i32 %9, float* %10) #3
|
|
%11 = load %struct.Point*, %struct.Point** %p.addr, align 8
|
|
%12 = load i32, i32* %tid, align 4
|
|
%idxprom9 = sext i32 %12 to i64
|
|
%arrayidx10 = getelementptr inbounds %struct.Point, %struct.Point* %11, i64 %idxprom9
|
|
%weight = getelementptr inbounds %struct.Point, %struct.Point* %arrayidx10, i32 0, i32 0
|
|
%13 = load float, float* %weight, align 8
|
|
%mul11 = fmul contract float %call8, %13
|
|
store float %mul11, float* %x_cost, align 4
|
|
%14 = load float, float* %x_cost, align 4
|
|
%15 = load %struct.Point*, %struct.Point** %p.addr, align 8
|
|
%16 = load i32, i32* %tid, align 4
|
|
%idxprom12 = sext i32 %16 to i64
|
|
%arrayidx13 = getelementptr inbounds %struct.Point, %struct.Point* %15, i64 %idxprom12
|
|
%cost = getelementptr inbounds %struct.Point, %struct.Point* %arrayidx13, i32 0, i32 3
|
|
%17 = load float, float* %cost, align 8
|
|
%cmp14 = fcmp olt float %14, %17
|
|
br i1 %cmp14, label %if.then15, label %if.else
|
|
|
|
if.then15: ; preds = %if.then
|
|
%18 = load i8*, i8** %switch_membership_d.addr, align 8
|
|
%19 = load i32, i32* %tid, align 4
|
|
%idxprom16 = sext i32 %19 to i64
|
|
%arrayidx17 = getelementptr inbounds i8, i8* %18, i64 %idxprom16
|
|
store i8 1, i8* %arrayidx17, align 1
|
|
%20 = load float, float* %x_cost, align 4
|
|
%21 = load %struct.Point*, %struct.Point** %p.addr, align 8
|
|
%22 = load i32, i32* %tid, align 4
|
|
%idxprom18 = sext i32 %22 to i64
|
|
%arrayidx19 = getelementptr inbounds %struct.Point, %struct.Point* %21, i64 %idxprom18
|
|
%cost20 = getelementptr inbounds %struct.Point, %struct.Point* %arrayidx19, i32 0, i32 3
|
|
%23 = load float, float* %cost20, align 8
|
|
%sub = fsub contract float %20, %23
|
|
%24 = load float*, float** %lower, align 8
|
|
%25 = load i32, i32* %K.addr, align 4
|
|
%idxprom21 = sext i32 %25 to i64
|
|
%arrayidx22 = getelementptr inbounds float, float* %24, i64 %idxprom21
|
|
%26 = load float, float* %arrayidx22, align 4
|
|
%add23 = fadd contract float %26, %sub
|
|
store float %add23, float* %arrayidx22, align 4
|
|
br label %if.end
|
|
|
|
if.else: ; preds = %if.then
|
|
%27 = load %struct.Point*, %struct.Point** %p.addr, align 8
|
|
%28 = load i32, i32* %tid, align 4
|
|
%idxprom24 = sext i32 %28 to i64
|
|
%arrayidx25 = getelementptr inbounds %struct.Point, %struct.Point* %27, i64 %idxprom24
|
|
%cost26 = getelementptr inbounds %struct.Point, %struct.Point* %arrayidx25, i32 0, i32 3
|
|
%29 = load float, float* %cost26, align 8
|
|
%30 = load float, float* %x_cost, align 4
|
|
%sub27 = fsub contract float %29, %30
|
|
%31 = load float*, float** %lower, align 8
|
|
%32 = load i32*, i32** %center_table_d.addr, align 8
|
|
%33 = load %struct.Point*, %struct.Point** %p.addr, align 8
|
|
%34 = load i32, i32* %tid, align 4
|
|
%idxprom28 = sext i32 %34 to i64
|
|
%arrayidx29 = getelementptr inbounds %struct.Point, %struct.Point* %33, i64 %idxprom28
|
|
%assign = getelementptr inbounds %struct.Point, %struct.Point* %arrayidx29, i32 0, i32 2
|
|
%35 = load i64, i64* %assign, align 8
|
|
%arrayidx30 = getelementptr inbounds i32, i32* %32, i64 %35
|
|
%36 = load i32, i32* %arrayidx30, align 4
|
|
%idxprom31 = sext i32 %36 to i64
|
|
%arrayidx32 = getelementptr inbounds float, float* %31, i64 %idxprom31
|
|
%37 = load float, float* %arrayidx32, align 4
|
|
%add33 = fadd contract float %37, %sub27
|
|
store float %add33, float* %arrayidx32, align 4
|
|
br label %if.end
|
|
|
|
if.end: ; preds = %if.else, %if.then15
|
|
br label %if.end34
|
|
|
|
if.end34: ; preds = %if.end, %entry
|
|
ret void
|
|
}
|
|
|
|
; Function Attrs: alwaysinline convergent nounwind
|
|
define linkonce_odr dso_local i32 @_ZN25__cuda_builtin_blockIdx_t17__fetch_builtin_xEv() #1 comdat align 2 {
|
|
entry:
|
|
%0 = call i32 @llvm.nvvm.read.ptx.sreg.ctaid.x()
|
|
ret i32 %0
|
|
}
|
|
|
|
; Function Attrs: alwaysinline convergent nounwind
|
|
define linkonce_odr dso_local i32 @_ZN24__cuda_builtin_gridDim_t17__fetch_builtin_xEv() #1 comdat align 2 {
|
|
entry:
|
|
%0 = call i32 @llvm.nvvm.read.ptx.sreg.nctaid.x()
|
|
ret i32 %0
|
|
}
|
|
|
|
; Function Attrs: alwaysinline convergent nounwind
|
|
define linkonce_odr dso_local i32 @_ZN25__cuda_builtin_blockIdx_t17__fetch_builtin_yEv() #1 comdat align 2 {
|
|
entry:
|
|
%0 = call i32 @llvm.nvvm.read.ptx.sreg.ctaid.y()
|
|
ret i32 %0
|
|
}
|
|
|
|
; Function Attrs: alwaysinline convergent nounwind
|
|
define linkonce_odr dso_local i32 @_ZN25__cuda_builtin_blockDim_t17__fetch_builtin_xEv() #1 comdat align 2 {
|
|
entry:
|
|
%0 = call i32 @llvm.nvvm.read.ptx.sreg.ntid.x()
|
|
ret i32 %0
|
|
}
|
|
|
|
; Function Attrs: alwaysinline convergent nounwind
|
|
define linkonce_odr dso_local i32 @_ZN26__cuda_builtin_threadIdx_t17__fetch_builtin_xEv() #1 comdat align 2 {
|
|
entry:
|
|
%0 = call i32 @llvm.nvvm.read.ptx.sreg.tid.x()
|
|
ret i32 %0
|
|
}
|
|
|
|
; Function Attrs: nounwind readnone
|
|
declare i32 @llvm.nvvm.read.ptx.sreg.ctaid.x() #2
|
|
|
|
; Function Attrs: nounwind readnone
|
|
declare i32 @llvm.nvvm.read.ptx.sreg.nctaid.x() #2
|
|
|
|
; Function Attrs: nounwind readnone
|
|
declare i32 @llvm.nvvm.read.ptx.sreg.ctaid.y() #2
|
|
|
|
; Function Attrs: nounwind readnone
|
|
declare i32 @llvm.nvvm.read.ptx.sreg.ntid.x() #2
|
|
|
|
; Function Attrs: nounwind readnone
|
|
declare i32 @llvm.nvvm.read.ptx.sreg.tid.x() #2
|
|
|
|
attributes #0 = { convergent noinline nounwind optnone "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "frame-pointer"="all" "less-precise-fpmad"="false" "min-legal-vector-width"="0" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="sm_61" "target-features"="+ptx64,+sm_61" "unsafe-fp-math"="false" "use-soft-float"="false" }
|
|
attributes #1 = { alwaysinline convergent nounwind "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "frame-pointer"="all" "less-precise-fpmad"="false" "min-legal-vector-width"="0" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="sm_61" "target-features"="+ptx64,+sm_61" "unsafe-fp-math"="false" "use-soft-float"="false" }
|
|
attributes #2 = { nounwind readnone }
|
|
attributes #3 = { convergent nounwind }
|
|
|
|
!llvm.module.flags = !{!0, !1, !2}
|
|
!nvvm.annotations = !{!3, !4, !5, !4, !6, !6, !6, !6, !7, !7, !6}
|
|
!llvm.ident = !{!8}
|
|
!nvvmir.version = !{!9}
|
|
|
|
!0 = !{i32 2, !"SDK Version", [2 x i32] [i32 10, i32 1]}
|
|
!1 = !{i32 1, !"wchar_size", i32 4}
|
|
!2 = !{i32 4, !"nvvm-reflect-ftz", i32 0}
|
|
!3 = !{void (i32, i32, i64, %struct.Point*, i32, i32, float*, float*, i32*, i8*)* @_Z19kernel_compute_costiilP5PointiiPfS1_PiPb, !"kernel", i32 1}
|
|
!4 = !{null, !"align", i32 8}
|
|
!5 = !{null, !"align", i32 8, !"align", i32 65544, !"align", i32 131080}
|
|
!6 = !{null, !"align", i32 16}
|
|
!7 = !{null, !"align", i32 16, !"align", i32 65552, !"align", i32 131088}
|
|
!8 = !{!"clang version 10.0.1 (https://github.com/llvm/llvm-project.git ef32c611aa214dea855364efd7ba451ec5ec3f74)"}
|
|
!9 = !{i32 1, i32 4}
|