Closed WangJialei-A closed 2 months ago
Ok, will do
What about moving the build to clang? It seems a bit faster.
Rebuilt runners with clang, testing it here https://github.com/intel/graph-compiler/actions/runs/10743821428 upd. testing passed @WangJialei-A please check if the following the fix work for you
ii clang 1:14.0-55~exp2 amd64 C, C++ and Objective-C compiler (LLVM based), clang binary
I have also unified pvc and cpu runner images, the only difference now is in /dev/dri/renderD128
access
@WangJialei-A please take a look if we can close this
for nightly test
Currently
clang/clang++
is not installed in ourself-hosted
runner. Please help to rebuild the image and redeploy the runner.https://github.com/intel/graph-compiler/blob/23269d7882419924ff1f4290e2668468ba1525db/.github/dockerfiles/runner/Dockerfile#L12-L17