Converted a CUDA config to use clang host-only compilation

This is distinct from the clang_cuda configuration because that one
uses clang also for device compilation. This one doesn't use the
modified build system, and will be easier to update when NVIDIA
support newer clang versions on Linux.

Refs #2899

Change-Id: I82bdf700c5553fcf367809209481ab4e27b037a2

diff --git a/admin/builds/post-submit-matrix.txt b/admin/builds/post-submit-matrix.txt
index 69e892eaa83c3cee2baac3a318052d8c4bf1243a..617a79bd62ec284bac93e940553b2d15e96d7a7f 100644 (file)
--- a/admin/builds/post-submit-matrix.txt
+++ b/admin/builds/post-submit-matrix.txt
@@ -41,7 +41,7 @@ icc-19 simd=sse2 mkl release
 
 # Test SSE4.1 SIMD
 # Test single-rank GPU
-# Test clang + OpenMP + CUDA
+# Test clang host and device compilation, with OpenMP and CUDA
 # NOTE: using release-with-assert as debug build trigers some ptx warnings
 clang-8 simd=sse4.1 openmp nranks=1 gpuhw=nvidia cuda-10.0 clang_cuda release-with-assert
 
@@ -60,7 +60,8 @@ gcc-6 npme=1 nranks=2 opencl-1.2 gpuhw=nvidia simd=sse2 release
 # Test non-default use of mdrun -gpu_id
 # Test GPU-sharing among 4 PP ranks
 # Test no hwloc build, tests internal CPU topology detection (mainly for x86)
-gcc-5 gpuhw=nvidia nranks=4 gpu_id=1 cuda-10.0 no-hwloc release-with-assert
+# Test clang host compilation with CUDA on Linux
+clang-8 gpuhw=nvidia openmp nranks=4 gpu_id=1 cuda-10.1 no-hwloc release-with-assert
 
 # Test ARM HPC compier toolchain with gcc-7
 #gcc-7 armhpc-18.2 openmp simd=ARM_NEON_ASIMD release