# Can't compile a GPU kernel function for =\>

**URL:** <https://discourse.julialang.org/t/cant-compile-a-gpu-kernel-function-for/53100>\
**Category:** GPU\
**Tags:** question\
**Created:** [January 10, 2021, 1:42am UTC](https://discourse.julialang.org/t/cant-compile-a-gpu-kernel-function-for/53100 "2021-01-10T01:42:15Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Stephan\_Meyn](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/stephan_meyn/32/13928_2.png) [@Stephan\_Meyn](https://discourse.julialang.org/u/Stephan_Meyn)\
**Post date:** [January 10, 2021, 1:42am UTC](https://discourse.julialang.org/t/cant-compile-a-gpu-kernel-function-for/53100/1 "2021-01-10T01:42:15Z")

</div>

I’m trying to persuade CUDA.jl to compile a kernel function. After reducing the issue I have come to this sample code

```julia
using CUDA

x=Float32(2.0)
knots =CuArray{Float32,1}([1.2, 2.0, 3.0])
mapOp(v::Float32)::Int = v >= x ? 1 : 0
mapreduce(mapOp, + , knots)

```

When I run this I get this

```julia
ERROR: InvalidIRError: compiling kernel partial_mapreduce_grid(typeof(identity), typeof(+), Int64, CartesianIndices{1,Tuple{Base.OneTo{Int64}}}, CartesianIndices{1,Tuple{Base.OneTo{Int64}}}, Val{true}, CuDeviceArray{Int64,2,1}, Base.Broadcast.Broadcasted{CUDA.CuArrayStyle{1},Tuple{Base.OneTo{Int64}},typeof(mapOp),Tuple{CuDeviceArray{Float32,1,1}}}) resulted in invalid LLVM IR        
Reason: unsupported dynamic function invocation (call to >=)
Stacktrace:
 [1] mapOp at D:\steph\Google Drive\src\julia\bspline\broke2.jl:4

```

so it seems it can’t compile the \>= function as it can’t determine the types that go into it. Hw do I address this?

---

<div class="post-metadata">

**Author:** ![stillyslalom](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/stillyslalom/32/45687_2.png) [@stillyslalom](https://discourse.julialang.org/u/stillyslalom)\
**Post date:** [January 10, 2021, 1:55am UTC](https://discourse.julialang.org/t/cant-compile-a-gpu-kernel-function-for/53100/2 "2021-01-10T01:55:44Z")

</div>

`x` is in the global scope, which prevents its type from being concretely inferred. You need to pass it as a function argument _a la_

```julia
julia> count_geq(arr, val) = count(>=(val), arr)
count_geq (generic function with 1 method)

julia> count_geq(cu([1.2, 2.0, 3.0]), 2.0f0)
2

```

edit: it’s almost as fast to just use `count(knots .>= 2)`.

```julia
julia> using BenchmarkTools

julia> knots = 3*CUDA.rand(2^14);

julia> @btime CUDA.@sync count_geq($knots, 2.0f0)
  53.600 μs (77 allocations: 2.06 KiB)
5437

julia> @btime CUDA.@sync count($knots .>= 2)
  55.299 μs (100 allocations: 2.69 KiB)
5437

```

---

<div class="post-metadata">

**Author:** ![maleadt](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/maleadt/32/10097_2.png) [@maleadt](https://discourse.julialang.org/u/maleadt)\
**Post date:** [January 11, 2021, 10:36am UTC](https://discourse.julialang.org/t/cant-compile-a-gpu-kernel-function-for/53100/3 "2021-01-11T10:36:55Z")

</div>

If `x` were const, it would have worked too.