fix CUDA runtests

qutip · Sep 11, 2024 · f717463 · f717463
1 parent 0c7d58f
commit f717463
Show file tree

Hide file tree

Showing 3 changed files with 3 additions and 0 deletions.
diff --git a/docs/src/index.md b/docs/src/index.md
@@ -103,6 +103,7 @@ We can easily pass the computation to the GPU, by simply passing all the `Qobj`s
 ```julia
 using QuantumToolbox
 using CUDA
+CUDA.allowscalar(false) # Avoid unexpected scalar indexing
 
 a_gpu = cu(destroy(N)) # The only difference in the code is the cu() function
 

diff --git a/docs/src/users_guide/extensions/cuda.md b/docs/src/users_guide/extensions/cuda.md
@@ -13,6 +13,7 @@ This extension will be automatically loaded if user imports both `QuantumToolbox
 using QuantumToolbox
 using CUDA
 using CUDA.CUSPARSE
+CUDA.allowscalar(false) # Avoid unexpected scalar indexing
 ```
 
 We wrapped several functions in `CUDA` and `CUDA.CUSPARSE` in order to not only converting `QuantumObject.data` into GPU arrays, but also changing the element type and word size (`32` and `64`) since some of the GPUs perform better in `32`-bit. The functions are listed as follows (where input `A` is a [`QuantumObject`](@ref)):

diff --git a/test/cuda_ext.jl b/test/cuda_ext.jl
@@ -1,5 +1,6 @@
 using CUDA
 using CUDA.CUSPARSE
+CUDA.allowscalar(false) # Avoid unexpected scalar indexing
 
 QuantumToolbox.about()
 CUDA.versioninfo()