-
Notifications
You must be signed in to change notification settings - Fork 98
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ROT* HIP: testing and improving rocBLAS support for ROT* kernels #1594
Conversation
The rotation functions now support stream execution and proper pointer mode setting in the rocBLAS TPL interface layer.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Does the function signature change from camelcase work with all versions of rocm? If not, we should throw a CMake error if a rocm version with the camelcase function names is detected.
Status Flag 'Pre-Test Inspection' - Auto Inspected - Inspection Is Not Necessary for this Pull Request. |
Status Flag 'Pull Request AutoTester' - Testing Jenkins Projects: Pull Request Auto Testing STARTING (click to expand)Build InformationTest Name: KokkosKernels_PullRequest_GCC930_Light_Tpls_GCC930
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_CUDA11_CUDA11_LayoutRight
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_GCC1020
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_GCC1020_Light_LayoutRight
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_Tpls_GCC1020
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_Tpls_INTEL19
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_CLANG1001
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_CLANG13CUDA10
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_Tpls_ARMPL2110_Tpls_ARMPL2030_GCC1020
Jenkins Parameters
Using Repos:
Pull Request Author: lucbv |
Thanks @lucbv for the PR! I tested on caraway with rocblas enabled and compilation succeeds. When enabling rocsparse as well I ran into other build errors
Looking at the rocsparse documentation it seems the "X" needs to be replaced by corresponding letter to the scalar type (s, d, c, or z)? Maybe this is resolved in #1589 |
@lucbv after letting the unit tests run, there are runtime failures with the tests
More detailed output:
|
Status Flag 'Pull Request AutoTester' - Jenkins Testing: all Jobs PASSED Pull Request Auto Testing has PASSED (click to expand)Build InformationTest Name: KokkosKernels_PullRequest_GCC930_Light_Tpls_GCC930
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_CUDA11_CUDA11_LayoutRight
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_GCC1020
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_GCC1020_Light_LayoutRight
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_Tpls_GCC1020
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_Tpls_INTEL19
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_CLANG1001
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_CLANG13CUDA10
Jenkins Parameters
Build InformationTest Name: KokkosKernels_PullRequest_Tpls_ARMPL2110_Tpls_ARMPL2030_GCC1020
Jenkins Parameters
|
Status Flag 'Pre-Merge Inspection' - SUCCESS: The last commit to this Pull Request has been INSPECTED AND APPROVED by [ e10harvey ]! |
Status Flag 'Pull Request AutoTester' - Pull Request MUST BE MERGED MANUALLY BY Project Team - This Repo does not support Automerge |
@lucbv @e10harvey should we merge this PR since it resolves the compilation errors reported in #1593 |
I think that is a good path forward. |
@e10harvey the issue with the Camel case is related to differences between cublas and rocblas, I have not verified what prior versions of rocblas are doing. But we should certainly settle on which minimal version of rocblas we support to get started, I do not think that we have that defined anywhere and we should check what compatibility requirements are with rocm/5.2.0 which is the lowest rocm Kokkos supports. |
@ndellingwood unfortunately I have not tested rocpsarse in some time, let us focus on cleaning up rocblas first in our nightly builds. |
The rotation functions now support stream execution and proper pointer mode setting in the rocBLAS TPL interface layer.
This PR should fix the issue reported in #1593 and hopefully lead to fully clean build and tests on caraway!