Merge pull request #1896 from CliMA/ck/bm_script

charleskawczynski · web-flow · commit 578b992e15a4 · 2024-07-23T09:15:51.000-04:00
Small fix to benchmark script
diff --git a/benchmarks/scripts/indexing_and_static_ndranges.jl b/benchmarks/scripts/indexing_and_static_ndranges.jl
@@ -345,6 +345,7 @@ function custom_kernel_bc!(X, Y, us::AbstractUniversalSizes; printtb=true, use_p
                 bc,
                 us,
             )
+        N = get_N(us)
         config = CUDA.launch_configuration(kernel.fun)
         threads = min(N, config.threads)
         blocks = cld(N, threads)

Original file line number	Diff line number	Diff line change
`@@ -345,6 +345,7 @@ function custom_kernel_bc!(X, Y, us::AbstractUniversalSizes; printtb=true, use_p`
`345`	`345`	`bc,`
`346`	`346`	`us,`
`347`	`347`	`)`
	`348`	`+ N = get_N(us)`
`348`	`349`	`config = CUDA.launch_configuration(kernel.fun)`
`349`	`350`	`threads = min(N, config.threads)`
`350`	`351`	`blocks = cld(N, threads)`