[SYCL][NewOffload][E2E] add a single test for --offload-new-driver #14129
Conversation
jasonlizhengjian
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
jasonlizhengjian
had a problem deploying
to
WindowsCILock
GitHub Actions
Failure
| // RUN: %{build} --offload-new-driver -o %t.out | ||
| // RUN: %{run} %t.out | ||
|
|
||
| #include <sycl/sycl.hpp> |
There was a problem hiding this comment.
We try to use a subset here. Please do:
| #include <sycl/sycl.hpp> | |
| #include <sycl/detail/core.hpp> |
|
Currently, we support only JIT compilation flows in SYCL offloading using new offload model. This test will fail for AOT compilations. Thanks |
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
jasonlizhengjian
changed the title
| @@ -0,0 +1,48 @@ | |||
| // REQUIRES: level_zero | |||
| // RUN: %{build} --offload-new-driver -o %t.out | |||
There was a problem hiding this comment.
to force JIT i think we can do
%clangxx -fsycl --offload-new-driver ...
because if it's set up for AOT, %{build} will expand to AOT options
jasonlizhengjian
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
| @@ -0,0 +1,48 @@ | |||
| // REQUIRES: level_zero | |||
| // RUN: %clangxx --offload-new-driver -o %t.out | |||
There was a problem hiding this comment.
i think we need -fsycl here
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
|
LGTM, will approve after CI passes |
There was a problem hiding this comment.
Looks good to me. Couple of nits. (1) it is a good idea to add a comment at the top mentioning what the test does. I am ok to start doing this from the next PR. and (2) EOF seems to be missing a linebreak.
(2) can be addressed just before merging (after CI testing is complete).
Thanks again for adding this.
jasonlizhengjian
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
jasonlizhengjian
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
|
@jasonlizhengjian I think we just have to fix the formatting and then we are ready to merge! |
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
| // REQUIRES: level_zero | ||
| // RUN: %clangxx -fsycl --offload-new-driver %s -o %t.out | ||
| // RUN: %{run} %t.out |
There was a problem hiding this comment.
These should be below the copyright header
There was a problem hiding this comment.
Should we even have a copyright header in E2E tests? Currently, some of our tests have a copyright header, while others don't.
There was a problem hiding this comment.
imo this pr has been through a lot of iteration already, we should resolve this question in general for all tests outside of this pr and then make a changing implementing the rule (and ideally a CI check to enforce it in the future
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
|
@intel/llvm-reviewers-runtime Can someone please take a look at this? It's actually testing a dpcpp tools feature so should not be a difficult review! Thanks! |
jasonlizhengjian
had a problem deploying
to
WindowsCILock
GitHub Actions
Error
Co-authored-by: Marcos Maronas <maarquitos14@users.noreply.github.com>
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
jasonlizhengjian
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
…ntel#14129) A single basic file to compile and run to test functionality of --offload-new-driver --------- Co-authored-by: Marcos Maronas <maarquitos14@users.noreply.github.com>
A single basic file to compile and run to test functionality of --offload-new-driver