DPC++ executor #648

tcojean · 2020-10-14T16:16:26Z

This is a first version of a DPC++ executor.

Everything is working under Intel, and tests are passing. Here are some
major changes:

Upgrade GTEST to a more recent version which allows using GTEST_SKIP()
to skip selected tests.
Add a new Dpcpp executor. Use it in all gko::Operation and
wherever appropriate.
Add the general architecture with all kernels set as GKO_NOT_IMPLEMENTED
Cope with unsupported copies between devices by going through the
master executors. This (should) allow to copy from an AMD GPU to a
NVIDIA GPU, or from Intel to AMD/NVIDIA (untested).
Implement a few simple kernels in the components and stop/criterion
to ensure kernel execution works properly.
Implement DPC++ specific executor tests.
Circumvent some issues with abs and sqrt by using the sycl
specific functions when appropriate. Note that it's still needed to
namespace abs (and probably sqrt as well) calls inside DPC++
kernels, otherwise the default (std) version is still used.

There are still a few things I want to do:

Add the DPC++ executor to all examples/benchmarks/documentation.
Generate documentation and check that it is complete.
Check that copying through master works (e.g. for CUDA and HIP).
Make Windows MSVC work with the new gtest.
Try to add an Intel CPU (hopefully GPU) CI job.

CMakeLists.txt

codecov · 2020-10-15T15:20:59Z

Codecov Report

Merging #648 into develop will decrease coverage by 0.08%.
The diff coverage is 64.28%.

@@             Coverage Diff             @@
##           develop     #648      +/-   ##
===========================================
- Coverage    92.92%   92.83%   -0.09%     
===========================================
  Files          320      322       +2     
  Lines        22833    22904      +71     
===========================================
+ Hits         21217    21263      +46     
- Misses        1616     1641      +25

Impacted Files	Coverage Δ
core/base/executor.cpp	`15.38% <0.00%> (-2.80%)`	⬇️
core/device_hooks/cuda_hooks.cpp	`51.72% <0.00%> (-3.84%)`	⬇️
core/device_hooks/hip_hooks.cpp	`51.72% <0.00%> (-3.84%)`	⬇️
core/test/base/array.cpp	`100.00% <ø> (ø)`
core/test/base/combination.cpp	`86.36% <ø> (ø)`
core/test/base/composition.cpp	`77.77% <ø> (ø)`
core/test/base/iterator_factory.cpp	`100.00% <ø> (ø)`
core/test/base/mtx_io.cpp	`97.45% <ø> (ø)`
core/test/factorization/par_ilu.cpp	`92.85% <ø> (ø)`
core/test/log/convergence.cpp	`100.00% <ø> (ø)`
... and 86 more

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 9caff82...a5cc5d2. Read the comment docs.

tcojean · 2020-10-15T16:24:30Z

A few comments:

The documentation was generated and is here: https://ginkgo-project.github.io/ginkgo/doc/dpcpp_executor I think it's complete, tell me if there is some issue
I tested the master copies by going setting HIP->HIP copies and CUDA->CUDA copies as NOT_SUPPORTED (66ca1bf). That means all related CUDA/HIP tests should have failed in the related pipeline, but they succeeded since the copies then went through the master executors: https://gitlab.com/ginkgo-project/ginkgo-public-ci/-/pipelines/202934371

yhmtsai

an issue is found when compiling it

dpcpp/factorization/factorization_kernels.dp.cpp

pratikvn

LGTM!

cmake/create_test.cmake

cuda/CMakeLists.txt

doc/examples/examples.hpp.in

include/ginkgo/core/base/executor.hpp

thoasm

LGTM!

CMakeLists.txt

core/device_hooks/dpcpp_hooks.cpp

core/test/base/executor.cpp

yhmtsai

LGTM in general.
minor format and something would like to give a try

yhmtsai · 2020-10-26T09:49:17Z

dpcpp/base/executor.dp.cpp

+        {"accelerator", sycl::info::device_type::accelerator},
+        {"all", sycl::info::device_type::all},
+        {"cpu", sycl::info::device_type::cpu},
+        {"gpu", sycl::info::device_type::gpu}};


maybe use enum to declare them?

I would rather keep using strings particularly since I can also be case insensitive (which I now added, actually).

dpcpp/base/executor.dp.cpp

yhmtsai · 2020-10-26T09:56:24Z

dpcpp/base/executor.dp.cpp

+    }
+    max_workgroup_size_ =
+        device.get_info<sycl::info::device::max_work_group_size>();
+    auto *queue = new sycl::queue{device, sycl::property::queue::in_order{}};


Could you write some comments about in_order here?

dpcpp/base/executor.dp.cpp

dpcpp/components/prefix_sum.dp.cpp

yhmtsai · 2020-10-26T10:18:34Z

include/ginkgo/core/base/executor.hpp

@@ -151,15 +165,16 @@ class ExecutorBase;
 * OMP
 * CUDA(0)
 * HIP(0)
+ * DPC++(0)


should we use DPC++ or DPCPP?

Well that's a good question, so far I use in documentation always DPC++ I believe, and only during the code I use DPCPP since we cannot use the + symbol where we want. With the code of this example, it will produce DPC++(0) indeed.

yhmtsai · 2020-10-26T10:24:18Z

include/ginkgo/core/base/math.hpp

@@ -908,7 +913,11 @@ GKO_INLINE GKO_ATTRIBUTES constexpr xstd::enable_if_t<is_complex_s<T>::value,
                                                      remove_complex<T>>
 abs(const T &x)
 {
+#ifdef CL_SYCL_LANGUAGE_VERSION


Could you add some todo here? like DPCPP currently needs to use OPENCL to support complex multiplication.

yhmtsai · 2020-10-26T10:32:30Z

cmake/autodetect_executors.cmake

 set(GINKGO_HAS_HIP OFF)
 find_package(OpenMP)
 include(CheckLanguage)
 check_language(CUDA)
+try_compile(GKO_CAN_COMPILE_DPCPP ${PROJECT_BINARY_DIR}/dpcpp
+    SOURCES ${PROJECT_SOURCE_DIR}/dpcpp/test_dpcpp.dp.cpp
+    CXX_STANDARD 17)


it does not use dpct header, so it might be compiled with std14.

I think this should also be the case for DPC++ itself. See https://spec.oneapi.com/versions/latest/elements/dpcpp/source/index.html

yhmtsai · 2020-10-26T10:42:16Z

cmake/create_test.cmake

+        ${PROJECT_BINARY_DIR} ${CMAKE_CURRENT_BINARY_DIR})
+    string(REPLACE "/" "_" TEST_TARGET_NAME "${REL_BINARY_DIR}/${test_name}")
+    add_executable(${TEST_TARGET_NAME} ${test_name}.dp.cpp)
+    target_compile_features("${TEST_TARGET_NAME}" PUBLIC cxx_std_17)


In the dpct beta update 10, they describe something related to std17 and std::max_element.
Maybe dpct does not require std17 anymore.
Because we do not use dpct header currently, could you try std14?

Note. I saw the DPCT beta update 10 release note yesterday, but I can not find it anymore.

upsj

LGTM!

core/devices/dpcpp/CMakeLists.txt

dpcpp/components/csr_spgeam.dp.hpp

dpcpp/components/fill_array.dp.cpp

dpcpp/components/format_conversion.dp.hpp

dpcpp/components/precision_conversion.dp.cpp

dpcpp/matrix/csr_kernels.dp.cpp

dpcpp/stop/criterion_kernels.dp.cpp

dpcpp/components/absolute_array.dp.cpp

dpcpp/base/executor.dp.cpp

+ `TYPED_TEST_CASE` is deprecated for the newer `TYPED_TEST_SUITE` version. + We can now use `GTEST_SKIP()` to skip specific tests.

+ Add a new `Dpcpp` executor. Use it in all `gko::Operation` and wherever appropriate. + Add the general architecture with all kernels set as `GKO_NOT_IMPLEMENTED` + Cope with unsupported copies between devices by going through the master executors. This (should) allow to copy from an AMD GPU to a NVIDIA GPU, or from Intel to AMD/NVIDIA (untested). + Implement a few simple kernels in the components and `stop/criterion` to ensure kernel execution works properly. + Implement DPC++ specific executor tests. + Circumvent some issues with `abs` and `sqrt` my using the sycl specific functions when appropriate. Note that it's still needed to namespace `abs` (and probably `sqrt` as well) calls inside DPC++ kernels, otherwise the default (std) version is still used.

`using T = x;` is banned inside any gtest class used in a `TYPED_TEST` fashion because gtest also defines a `T` of its own and somehow `MSVC` confuses the two.

+ Fix some documentation issues. + Fix some spacing issues. + Do not duplicate `GKO_DECLARE_EXECUTOR_FRIEND` macro. Co-authored-by: Pratik Nayak <pratikvn@protonmail.com> Co-authored-by: Thomas Grützmacher <thomas.gruetzmacher@kit.edu>

Co-authored-by: Yuhsiang M. Tsai <yhmtsai@gmail.com>

+ Add some environment variable output to `get_info.cmake`. + Fix formatting of `GKO_NOT_IMPLEMENTED` declarations. + Allow lowercase in the `DPC++` strings passed to executors. + Fix kernels storing SYCL `id` into `int` types. + Add some extra comments to tricky parts of the code. + Improve some other code formatting issues. Co-authored-by: Tobias Ribizel <ribizel@kit.edu> Co-authored-by: Yuhsiang M. Tsai <yhmtsai@gmail.com>

sonarqubecloud · 2020-11-04T15:13:05Z

Kudos, SonarCloud Quality Gate passed!

0 Bugs
0 Vulnerabilities (and 0 Security Hotspots to review)
67 Code Smells

17.5% Coverage
1.4% Duplication

The version of Java (1.8.0_121) you have used to run this analysis is deprecated and we will stop accepting it accepting it soon.Please update to at least Java 11.
Read more here

Ginkgo release 1.4.0 The Ginkgo team is proud to announce the new Ginkgo minor release 1.4.0. This release brings most of the Ginkgo functionality to the Intel DPC++ ecosystem which enables Intel-GPU and CPU execution. The only Ginkgo features which have not been ported yet are some preconditioners. Ginkgo's mixed-precision support is greatly enhanced thanks to: 1. The new Accessor concept, which allows writing kernels featuring on-the-fly memory compression, among other features. The accessor can be used as header-only, see the [accessor BLAS benchmarks repository](https://github.com/ginkgo-project/accessor-BLAS/tree/develop) as a usage example. 2. All LinOps now transparently support mixed-precision execution. By default, this is done through a temporary copy which may have a performance impact but already allows mixed-precision research. Native mixed-precision ELL kernels are implemented which do not see this cost. The accessor is also leveraged in a new CB-GMRES solver which allows for performance improvements by compressing the Krylov basis vectors. Many other features have been added to Ginkgo, such as reordering support, a new IDR solver, Incomplete Cholesky preconditioner, matrix assembly support (only CPU for now), machine topology information, and more! Supported systems and requirements: + For all platforms, cmake 3.13+ + C++14 compliant compiler + Linux and MacOS + gcc: 5.3+, 6.3+, 7.3+, all versions after 8.1+ + clang: 3.9+ + Intel compiler: 2018+ + Apple LLVM: 8.0+ + CUDA module: CUDA 9.0+ + HIP module: ROCm 3.5+ + DPC++ module: Intel OneAPI 2021.3. Set the CXX compiler to `dpcpp`. + Windows + MinGW and Cygwin: gcc 5.3+, 6.3+, 7.3+, all versions after 8.1+ + Microsoft Visual Studio: VS 2019 + CUDA module: CUDA 9.0+, Microsoft Visual Studio + OpenMP module: MinGW or Cygwin. Algorithm and important feature additions: + Add a new DPC++ Executor for SYCL execution and other base utilities [#648](#648), [#661](#661), [#757](#757), [#832](#832) + Port matrix formats, solvers and related kernels to DPC++. For some kernels, also make use of a shared kernel implementation for all executors (except Reference). [#710](#710), [#799](#799), [#779](#779), [#733](#733), [#844](#844), [#843](#843), [#789](#789), [#845](#845), [#849](#849), [#855](#855), [#856](#856) + Add accessors which allow multi-precision kernels, among other things. [#643](#643), [#708](#708) + Add support for mixed precision operations through apply in all LinOps. [#677](#677) + Add incomplete Cholesky factorizations and preconditioners as well as some improvements to ILU. [#672](#672), [#837](#837), [#846](#846) + Add an AMGX implementation and kernels on all devices but DPC++. [#528](#528), [#695](#695), [#860](#860) + Add a new mixed-precision capability solver, Compressed Basis GMRES (CB-GMRES). [#693](#693), [#763](#763) + Add the IDR(s) solver. [#620](#620) + Add a new fixed-size block CSR matrix format (for the Reference executor). [#671](#671), [#730](#730) + Add native mixed-precision support to the ELL format. [#717](#717), [#780](#780) + Add Reverse Cuthill-McKee reordering [#500](#500), [#649](#649) + Add matrix assembly support on CPUs. [#644](#644) + Extends ISAI from triangular to general and spd matrices. [#690](#690) Other additions: + Add the possibility to apply real matrices to complex vectors. [#655](#655), [#658](#658) + Add functions to compute the absolute of a matrix format. [#636](#636) + Add symmetric permutation and improve existing permutations. [#684](#684), [#657](#657), [#663](#663) + Add a MachineTopology class with HWLOC support [#554](#554), [#697](#697) + Add an implicit residual norm criterion. [#702](#702), [#818](#818), [#850](#850) + Row-major accessor is generalized to more than 2 dimensions and a new "block column-major" accessor has been added. [#707](#707) + Add an heat equation example. [#698](#698), [#706](#706) + Add ccache support in CMake and CI. [#725](#725), [#739](#739) + Allow tuning and benchmarking variables non intrusively. [#692](#692) + Add triangular solver benchmark [#664](#664) + Add benchmarks for BLAS operations [#772](#772), [#829](#829) + Add support for different precisions and consistent index types in benchmarks. [#675](#675), [#828](#828) + Add a Github bot system to facilitate development and PR management. [#667](#667), [#674](#674), [#689](#689), [#853](#853) + Add Intel (DPC++) CI support and enable CI on HPC systems. [#736](#736), [#751](#751), [#781](#781) + Add ssh debugging for Github Actions CI. [#749](#749) + Add pipeline segmentation for better CI speed. [#737](#737) Changes: + Add a Scalar Jacobi specialization and kernels. [#808](#808), [#834](#834), [#854](#854) + Add implicit residual log for solvers and benchmarks. [#714](#714) + Change handling of the conjugate in the dense dot product. [#755](#755) + Improved Dense stride handling. [#774](#774) + Multiple improvements to the OpenMP kernels performance, including COO, an exclusive prefix sum, and more. [#703](#703), [#765](#765), [#740](#740) + Allow specialization of submatrix and other dense creation functions in solvers. [#718](#718) + Improved Identity constructor and treatment of rectangular matrices. [#646](#646) + Allow CUDA/HIP executors to select allocation mode. [#758](#758) + Check if executors share the same memory. [#670](#670) + Improve test install and smoke testing support. [#721](#721) + Update the JOSS paper citation and add publications in the documentation. [#629](#629), [#724](#724) + Improve the version output. [#806](#806) + Add some utilities for dim and span. [#821](#821) + Improved solver and preconditioner benchmarks. [#660](#660) + Improve benchmark timing and output. [#669](#669), [#791](#791), [#801](#801), [#812](#812) Fixes: + Sorting fix for the Jacobi preconditioner. [#659](#659) + Also log the first residual norm in CGS [#735](#735) + Fix BiCG and HIP CSR to work with complex matrices. [#651](#651) + Fix Coo SpMV on strided vectors. [#807](#807) + Fix segfault of extract_diagonal, add short-and-fat test. [#769](#769) + Fix device_reset issue by moving counter/mutex to device. [#810](#810) + Fix `EnableLogging` superclass. [#841](#841) + Support ROCm 4.1.x and breaking HIP_PLATFORM changes. [#726](#726) + Decreased test size for a few device tests. [#742](#742) + Fix multiple issues with our CMake HIP and RPATH setup. [#712](#712), [#745](#745), [#709](#709) + Cleanup our CMake installation step. [#713](#713) + Various simplification and fixes to the Windows CMake setup. [#720](#720), [#785](#785) + Simplify third-party integration. [#786](#786) + Improve Ginkgo device arch flags management. [#696](#696) + Other fixes and improvements to the CMake setup. [#685](#685), [#792](#792), [#705](#705), [#836](#836) + Clarification of dense norm documentation [#784](#784) + Various development tools fixes and improvements [#738](#738), [#830](#830), [#840](#840) + Make multiple operators/constructors explicit. [#650](#650), [#761](#761) + Fix some issues, memory leaks and warnings found by MSVC. [#666](#666), [#731](#731) + Improved solver memory estimates and consistent iteration counts [#691](#691) + Various logger improvements and fixes [#728](#728), [#743](#743), [#754](#754) + Fix for ForwardIterator requirements in iterator_factory. [#665](#665) + Various benchmark fixes. [#647](#647), [#673](#673), [#722](#722) + Various CI fixes and improvements. [#642](#642), [#641](#641), [#795](#795), [#783](#783), [#793](#793), [#852](#852) Related PR: #857

Release 1.4.0 to master The Ginkgo team is proud to announce the new Ginkgo minor release 1.4.0. This release brings most of the Ginkgo functionality to the Intel DPC++ ecosystem which enables Intel-GPU and CPU execution. The only Ginkgo features which have not been ported yet are some preconditioners. Ginkgo's mixed-precision support is greatly enhanced thanks to: 1. The new Accessor concept, which allows writing kernels featuring on-the-fly memory compression, among other features. The accessor can be used as header-only, see the [accessor BLAS benchmarks repository](https://github.com/ginkgo-project/accessor-BLAS/tree/develop) as a usage example. 2. All LinOps now transparently support mixed-precision execution. By default, this is done through a temporary copy which may have a performance impact but already allows mixed-precision research. Native mixed-precision ELL kernels are implemented which do not see this cost. The accessor is also leveraged in a new CB-GMRES solver which allows for performance improvements by compressing the Krylov basis vectors. Many other features have been added to Ginkgo, such as reordering support, a new IDR solver, Incomplete Cholesky preconditioner, matrix assembly support (only CPU for now), machine topology information, and more! Supported systems and requirements: + For all platforms, cmake 3.13+ + C++14 compliant compiler + Linux and MacOS + gcc: 5.3+, 6.3+, 7.3+, all versions after 8.1+ + clang: 3.9+ + Intel compiler: 2018+ + Apple LLVM: 8.0+ + CUDA module: CUDA 9.0+ + HIP module: ROCm 3.5+ + DPC++ module: Intel OneAPI 2021.3. Set the CXX compiler to `dpcpp`. + Windows + MinGW and Cygwin: gcc 5.3+, 6.3+, 7.3+, all versions after 8.1+ + Microsoft Visual Studio: VS 2019 + CUDA module: CUDA 9.0+, Microsoft Visual Studio + OpenMP module: MinGW or Cygwin. Algorithm and important feature additions: + Add a new DPC++ Executor for SYCL execution and other base utilities [#648](#648), [#661](#661), [#757](#757), [#832](#832) + Port matrix formats, solvers and related kernels to DPC++. For some kernels, also make use of a shared kernel implementation for all executors (except Reference). [#710](#710), [#799](#799), [#779](#779), [#733](#733), [#844](#844), [#843](#843), [#789](#789), [#845](#845), [#849](#849), [#855](#855), [#856](#856) + Add accessors which allow multi-precision kernels, among other things. [#643](#643), [#708](#708) + Add support for mixed precision operations through apply in all LinOps. [#677](#677) + Add incomplete Cholesky factorizations and preconditioners as well as some improvements to ILU. [#672](#672), [#837](#837), [#846](#846) + Add an AMGX implementation and kernels on all devices but DPC++. [#528](#528), [#695](#695), [#860](#860) + Add a new mixed-precision capability solver, Compressed Basis GMRES (CB-GMRES). [#693](#693), [#763](#763) + Add the IDR(s) solver. [#620](#620) + Add a new fixed-size block CSR matrix format (for the Reference executor). [#671](#671), [#730](#730) + Add native mixed-precision support to the ELL format. [#717](#717), [#780](#780) + Add Reverse Cuthill-McKee reordering [#500](#500), [#649](#649) + Add matrix assembly support on CPUs. [#644](#644) + Extends ISAI from triangular to general and spd matrices. [#690](#690) Other additions: + Add the possibility to apply real matrices to complex vectors. [#655](#655), [#658](#658) + Add functions to compute the absolute of a matrix format. [#636](#636) + Add symmetric permutation and improve existing permutations. [#684](#684), [#657](#657), [#663](#663) + Add a MachineTopology class with HWLOC support [#554](#554), [#697](#697) + Add an implicit residual norm criterion. [#702](#702), [#818](#818), [#850](#850) + Row-major accessor is generalized to more than 2 dimensions and a new "block column-major" accessor has been added. [#707](#707) + Add an heat equation example. [#698](#698), [#706](#706) + Add ccache support in CMake and CI. [#725](#725), [#739](#739) + Allow tuning and benchmarking variables non intrusively. [#692](#692) + Add triangular solver benchmark [#664](#664) + Add benchmarks for BLAS operations [#772](#772), [#829](#829) + Add support for different precisions and consistent index types in benchmarks. [#675](#675), [#828](#828) + Add a Github bot system to facilitate development and PR management. [#667](#667), [#674](#674), [#689](#689), [#853](#853) + Add Intel (DPC++) CI support and enable CI on HPC systems. [#736](#736), [#751](#751), [#781](#781) + Add ssh debugging for Github Actions CI. [#749](#749) + Add pipeline segmentation for better CI speed. [#737](#737) Changes: + Add a Scalar Jacobi specialization and kernels. [#808](#808), [#834](#834), [#854](#854) + Add implicit residual log for solvers and benchmarks. [#714](#714) + Change handling of the conjugate in the dense dot product. [#755](#755) + Improved Dense stride handling. [#774](#774) + Multiple improvements to the OpenMP kernels performance, including COO, an exclusive prefix sum, and more. [#703](#703), [#765](#765), [#740](#740) + Allow specialization of submatrix and other dense creation functions in solvers. [#718](#718) + Improved Identity constructor and treatment of rectangular matrices. [#646](#646) + Allow CUDA/HIP executors to select allocation mode. [#758](#758) + Check if executors share the same memory. [#670](#670) + Improve test install and smoke testing support. [#721](#721) + Update the JOSS paper citation and add publications in the documentation. [#629](#629), [#724](#724) + Improve the version output. [#806](#806) + Add some utilities for dim and span. [#821](#821) + Improved solver and preconditioner benchmarks. [#660](#660) + Improve benchmark timing and output. [#669](#669), [#791](#791), [#801](#801), [#812](#812) Fixes: + Sorting fix for the Jacobi preconditioner. [#659](#659) + Also log the first residual norm in CGS [#735](#735) + Fix BiCG and HIP CSR to work with complex matrices. [#651](#651) + Fix Coo SpMV on strided vectors. [#807](#807) + Fix segfault of extract_diagonal, add short-and-fat test. [#769](#769) + Fix device_reset issue by moving counter/mutex to device. [#810](#810) + Fix `EnableLogging` superclass. [#841](#841) + Support ROCm 4.1.x and breaking HIP_PLATFORM changes. [#726](#726) + Decreased test size for a few device tests. [#742](#742) + Fix multiple issues with our CMake HIP and RPATH setup. [#712](#712), [#745](#745), [#709](#709) + Cleanup our CMake installation step. [#713](#713) + Various simplification and fixes to the Windows CMake setup. [#720](#720), [#785](#785) + Simplify third-party integration. [#786](#786) + Improve Ginkgo device arch flags management. [#696](#696) + Other fixes and improvements to the CMake setup. [#685](#685), [#792](#792), [#705](#705), [#836](#836) + Clarification of dense norm documentation [#784](#784) + Various development tools fixes and improvements [#738](#738), [#830](#830), [#840](#840) + Make multiple operators/constructors explicit. [#650](#650), [#761](#761) + Fix some issues, memory leaks and warnings found by MSVC. [#666](#666), [#731](#731) + Improved solver memory estimates and consistent iteration counts [#691](#691) + Various logger improvements and fixes [#728](#728), [#743](#743), [#754](#754) + Fix for ForwardIterator requirements in iterator_factory. [#665](#665) + Various benchmark fixes. [#647](#647), [#673](#673), [#722](#722) + Various CI fixes and improvements. [#642](#642), [#641](#641), [#795](#795), [#783](#783), [#793](#793), [#852](#852) Related PR: #866

tcojean self-assigned this Oct 14, 2020

tcojean force-pushed the dpcpp_executor branch 2 times, most recently from 321b069 to db9fbf1 Compare October 14, 2020 16:25

tcojean added the plat:intel This is related to the Intel compilers. label Oct 14, 2020

tcojean force-pushed the dpcpp_executor branch 2 times, most recently from df03dd9 to cc532f2 Compare October 14, 2020 16:48

yhmtsai reviewed Oct 14, 2020

View reviewed changes

CMakeLists.txt Outdated Show resolved Hide resolved

tcojean force-pushed the dpcpp_executor branch 2 times, most recently from 6fd8391 to 66ca1bf Compare October 15, 2020 09:27

tcojean force-pushed the dpcpp_executor branch 2 times, most recently from e8873b9 to 2325ee3 Compare October 15, 2020 16:18

tcojean force-pushed the dpcpp_executor branch 6 times, most recently from 541ae8b to 66caf9e Compare October 17, 2020 14:14

tcojean added 1:ST:ready-for-review This PR is ready for review and removed 1:ST:WIP This PR is a work in progress. Not ready for review. labels Oct 19, 2020

tcojean requested review from pratikvn, yhmtsai, upsj and fritzgoebel October 19, 2020 09:27

yhmtsai reviewed Oct 19, 2020

View reviewed changes

dpcpp/factorization/factorization_kernels.dp.cpp Outdated Show resolved Hide resolved

dpcpp/factorization/factorization_kernels.dp.cpp Show resolved Hide resolved

tcojean force-pushed the dpcpp_executor branch 3 times, most recently from 66eeab2 to 9a522b1 Compare October 20, 2020 18:09

pratikvn approved these changes Oct 23, 2020

View reviewed changes

cmake/create_test.cmake Show resolved Hide resolved

cuda/CMakeLists.txt Show resolved Hide resolved

doc/examples/examples.hpp.in Outdated Show resolved Hide resolved

include/ginkgo/core/base/executor.hpp Outdated Show resolved Hide resolved

thoasm approved these changes Oct 26, 2020

View reviewed changes

tcojean force-pushed the dpcpp_executor branch from f847396 to 6e6f178 Compare October 26, 2020 09:34

yhmtsai approved these changes Oct 26, 2020

View reviewed changes

upsj approved these changes Oct 26, 2020

View reviewed changes

tcojean force-pushed the dpcpp_executor branch 5 times, most recently from 74a7a31 to ea97fb7 Compare October 28, 2020 15:47

tcojean added 1:ST:ready-to-merge This PR is ready to merge. and removed 1:ST:ready-for-review This PR is ready for review labels Oct 29, 2020

tcojean and others added 8 commits November 4, 2020 10:55

Use a newer GTEST version.

42fb0fa

+ `TYPED_TEST_CASE` is deprecated for the newer `TYPED_TEST_SUITE` version. + We can now use `GTEST_SKIP()` to skip specific tests.

Update the JOSS paper.

eadf596

Fix the gtest issue with MSVC.

a343809

`using T = x;` is banned inside any gtest class used in a `TYPED_TEST` fashion because gtest also defines a `T` of its own and somehow `MSVC` confuses the two.

Add a CPU DPCPP job.

1f21042

Review updates.

55ab135

+ Fix some documentation issues. + Fix some spacing issues. + Do not duplicate `GKO_DECLARE_EXECUTOR_FRIEND` macro. Co-authored-by: Pratik Nayak <pratikvn@protonmail.com> Co-authored-by: Thomas Grützmacher <thomas.gruetzmacher@kit.edu>

Fix defines and header scripts.

c0d417a

Co-authored-by: Yuhsiang M. Tsai <yhmtsai@gmail.com>

tcojean force-pushed the dpcpp_executor branch from ea97fb7 to a5cc5d2 Compare November 4, 2020 10:39

tcojean merged commit 99c2004 into develop Nov 4, 2020

tcojean deleted the dpcpp_executor branch November 4, 2020 16:46

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

DPC++ executor #648

DPC++ executor #648

tcojean commented Oct 14, 2020 •

edited

Loading

codecov bot commented Oct 15, 2020 •

edited

Loading

tcojean commented Oct 15, 2020

yhmtsai left a comment

pratikvn left a comment

thoasm left a comment

yhmtsai left a comment

yhmtsai Oct 26, 2020

tcojean Oct 26, 2020

yhmtsai Oct 26, 2020

yhmtsai Oct 26, 2020

tcojean Oct 26, 2020

yhmtsai Oct 26, 2020

yhmtsai Oct 26, 2020

tcojean Oct 26, 2020 •

edited

Loading

yhmtsai Oct 26, 2020

upsj left a comment

sonarqubecloud bot commented Nov 4, 2020

DPC++ executor #648

DPC++ executor #648

Conversation

tcojean commented Oct 14, 2020 • edited Loading

codecov bot commented Oct 15, 2020 • edited Loading

Codecov Report

tcojean commented Oct 15, 2020

yhmtsai left a comment

Choose a reason for hiding this comment

pratikvn left a comment

Choose a reason for hiding this comment

thoasm left a comment

Choose a reason for hiding this comment

yhmtsai left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

tcojean Oct 26, 2020 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

upsj left a comment

Choose a reason for hiding this comment

sonarqubecloud bot commented Nov 4, 2020

tcojean commented Oct 14, 2020 •

edited

Loading

codecov bot commented Oct 15, 2020 •

edited

Loading

tcojean Oct 26, 2020 •

edited

Loading