Allow specifying allocator for benchmarks #1385

upsj · 2023-08-10T19:07:24Z

This adds an -allocator flag to specify how to allocate data in the benchmarks.

cuda/base/memory.cpp

upsj · 2023-08-10T19:15:50Z

format-rebase!

ginkgo-bot · 2023-08-10T19:16:37Z

Formatting rebase introduced changes, see Artifacts here to review them

pratikvn · 2023-08-10T20:10:35Z

benchmark/utils/general.hpp

+ return std::make_shared<gko::CudaAllocator>();
+ } else if (flag == "async") {


Is there any point in having an async allocator if you are not using a stream ?

The default stream is also a stream, the stream-ordered semantics of the allocator still apply

Does stream==nullptr mean that it is the default stream ? I cannot seem to find it in the docs anywhere.

It's listed as 0 in there https://docs.nvidia.com/cuda/cuda-runtime-api/stream-sync-behavior.html

but the default stream does not allow other streams created by cudaStreamNonBlocking execute at the same time, right?

yes, but the benchmarks only use a single stream, so we are fine for now. As soon as we move to multi-stream setups, we need to use a custom stream or CU_STREAM_PER_THREAD

benchmark/sparse_blas/sparse_blas.cpp

yhmtsai · 2023-08-11T07:45:58Z

benchmark/utils/general.hpp

+ return std::make_shared<gko::CudaAllocator>();
+ } else if (flag == "async") {


but the default stream does not allow other streams created by cudaStreamNonBlocking execute at the same time, right?

yhmtsai · 2023-08-11T07:49:30Z

core/device_hooks/cuda_hooks.cpp

@@ -75,6 +75,10 @@ bool CudaAsyncAllocator::check_environment(int device_id,
 GKO_NOT_COMPILED(cuda);


+CudaUnifiedAllocator::CudaUnifiedAllocator(int device_id)


Why did it not throw undefined reference before?

Because we did not call that constructor anywhere if CUDA is disabled

benchmark/sparse_blas/sparse_blas.cpp

sonarcloud · 2023-08-11T20:55:47Z

Kudos, SonarCloud Quality Gate passed!

0 Bugs
0 Vulnerabilities
0 Security Hotspots
6 Code Smells

8.3% Coverage
0.0% Duplication

The version of Java (11.0.3) you have used to run this analysis is deprecated and we will stop accepting it soon. Please update to at least Java 17.
Read more here

Release 1.7.0 to master The Ginkgo team is proud to announce the new Ginkgo minor release 1.7.0. This release brings new features such as: - Complete GPU-resident sparse direct solvers feature set and interfaces, - Improved Cholesky factorization performance, - A new MC64 reordering, - Batched iterative solver support with the BiCGSTAB solver with batched Dense and ELL matrix types, - MPI support for the SYCL backend, - Improved ParILU(T)/ParIC(T) preconditioner convergence, and more! If you face an issue, please first check our [known issues page](https://github.com/ginkgo-project/ginkgo/wiki/Known-Issues) and the [open issues list](https://github.com/ginkgo-project/ginkgo/issues) and if you do not find a solution, feel free to [open a new issue](https://github.com/ginkgo-project/ginkgo/issues/new/choose) or ask a question using the [github discussions](https://github.com/ginkgo-project/ginkgo/discussions). Supported systems and requirements: + For all platforms, CMake 3.16+ + C++14 compliant compiler + Linux and macOS + GCC: 5.5+ + clang: 3.9+ + Intel compiler: 2019+ + Apple Clang: 14.0 is tested. Earlier versions might also work. + NVHPC: 22.7+ + Cray Compiler: 14.0.1+ + CUDA module: CMake 3.18+, and CUDA 10.1+ or NVHPC 22.7+ + HIP module: ROCm 4.5+ + DPC++ module: Intel oneAPI 2022.1+ with oneMKL and oneDPL. Set the CXX compiler to `dpcpp` or `icpx`. + MPI: standard version 3.1+, ideally GPU Aware, for best performance + Windows + MinGW: GCC 5.5+ + Microsoft Visual Studio: VS 2019+ + CUDA module: CUDA 10.1+, Microsoft Visual Studio + OpenMP module: MinGW. ### Version support changes + CUDA 9.2 is no longer supported and 10.0 is untested [#1382](#1382) + Ginkgo now requires CMake version 3.16 (and 3.18 for CUDA) [#1368](#1368) ### Interface changes + `const` Factory parameters can no longer be modified through `with_*` functions, as this breaks const-correctness [#1336](#1336) [#1439](#1439) ### New Deprecations + The `device_reset` parameter of CUDA and HIP executors no longer has an effect, and its `allocation_mode` parameters have been deprecated in favor of the `Allocator` interface. [#1315](#1315) + The CMake parameter `GINKGO_BUILD_DPCPP` has been deprecated in favor of `GINKGO_BUILD_SYCL`. [#1350](#1350) + The `gko::reorder::Rcm` interface has been deprecated in favor of `gko::experimental::reorder::Rcm` based on `Permutation`. [#1418](#1418) + The Permutation class' `permute_mask` functionality. [#1415](#1415) + Multiple functions with typos (`set_complex_subpsace()`, range functions such as `conj_operaton` etc). [#1348](#1348) ### Summary of previous deprecations + `gko::lend()` is not necessary anymore. + The classes `RelativeResidualNorm` and `AbsoluteResidualNorm` are deprecated in favor of `ResidualNorm`. + The class `AmgxPgm` is deprecated in favor of `Pgm`. + Default constructors for the CSR `load_balance` and `automatical` strategies + The PolymorphicObject's move-semantic `copy_from` variant + The templated `SolverBase` class. + The class `MachineTopology` is deprecated in favor of `machine_topology`. + Logger constructors and create functions with the `executor` parameter. + The virtual, protected, Dense functions `compute_norm1_impl`, `add_scaled_impl`, etc. + Logger events for solvers and criterion without the additional `implicit_tau_sq` parameter. + The global `gko::solver::default_krylov_dim`, use instead `gko::solver::gmres_default_krylov_dim`. ### Added features + Adds a batch::BatchLinOp class that forms a base class for batched linear operators such as batched matrix formats, solver and preconditioners [#1379](#1379) + Adds a batch::MultiVector class that enables operations such as dot, norm, scale on batched vectors [#1371](#1371) + Adds a batch::Dense matrix format that stores batched dense matrices and provides gemv operations for these dense matrices. [#1413](#1413) + Adds a batch::Ell matrix format that stores batched Ell matrices and provides spmv operations for these batched Ell matrices. [#1416](#1416) [#1437](#1437) + Add a batch::Bicgstab solver (class, core, and reference kernels) that enables iterative solution of batched linear systems [#1438](#1438). + Add device kernels (CUDA, HIP, and DPCPP) for batch::Bicgstab solver. [#1443](#1443). + New MC64 reordering algorithm which optimizes the diagonal product or sum of a matrix by permuting the rows, and computes additional scaling factors for equilibriation [#1120](#1120) + New interface for (non-symmetric) permutation and scaled permutation of Dense and Csr matrices [#1415](#1415) + LU and Cholesky Factorizations can now be separated into their factors [#1432](#1432) + New symbolic LU factorization algorithm that is optimized for matrices with an almost-symmetric sparsity pattern [#1445](#1445) + Sorting kernels for SparsityCsr on all backends [#1343](#1343) + Allow passing pre-generated local solver as factory parameter for the distributed Schwarz preconditioner [#1426](#1426) + Add DPCPP kernels for Partition [#1034](#1034), and CSR's `check_diagonal_entries` and `add_scaled_identity` functionality [#1436](#1436) + Adds a helper function to create a partition based on either local sizes, or local ranges [#1227](#1227) + Add function to compute arithmetic mean of dense and distributed vectors [#1275](#1275) + Adds `icpx` compiler supports [#1350](#1350) + All backends can be built simultaneously [#1333](#1333) + Emits a CMake warning in downstream projects that use different compilers than the installed Ginkgo [#1372](#1372) + Reordering algorithms in sparse_blas benchmark [#1354](#1354) + Benchmarks gained an `-allocator` parameter to specify device allocators [#1385](#1385) + Benchmarks gained an `-input_matrix` parameter that initializes the input JSON based on the filename [#1387](#1387) + Benchmark inputs can now be reordered as a preprocessing step [#1408](#1408) ### Improvements + Significantly improve Cholesky factorization performance [#1366](#1366) + Improve parallel build performance [#1378](#1378) + Allow constrained parallel test execution using CTest resources [#1373](#1373) + Use arithmetic type more inside mixed precision ELL [#1414](#1414) + Most factory parameters of factory type no longer need to be constructed explicitly via `.on(exec)` [#1336](#1336) [#1439](#1439) + Improve ParILU(T)/ParIC(T) convergence by using more appropriate atomic operations [#1434](#1434) ### Fixes + Fix an over-allocation for OpenMP reductions [#1369](#1369) + Fix DPCPP's common-kernel reduction for empty input sizes [#1362](#1362) + Fix several typos in the API and documentation [#1348](#1348) + Fix inconsistent `Threads` between generations [#1388](#1388) + Fix benchmark median condition [#1398](#1398) + Fix HIP 5.6.0 compilation [#1411](#1411) + Fix missing destruction of rand_generator from cuda/hip [#1417](#1417) + Fix PAPI logger destruction order [#1419](#1419) + Fix TAU logger compilation [#1422](#1422) + Fix relative criterion to not iterate if the residual is already zero [#1079](#1079) + Fix memory_order invocations with C++20 changes [#1402](#1402) + Fix `check_diagonal_entries_exist` report correctly when only missing diagonal value in the last rows. [#1440](#1440) + Fix checking OpenMPI version in cross-compilation settings [#1446](#1446) + Fix false-positive deprecation warnings in Ginkgo, especially for the old Rcm (it doesn't emit deprecation warnings anymore as a result but is still considered deprecated) [#1444](#1444) ### Related PR: #1451

Release 1.7.0 to develop The Ginkgo team is proud to announce the new Ginkgo minor release 1.7.0. This release brings new features such as: - Complete GPU-resident sparse direct solvers feature set and interfaces, - Improved Cholesky factorization performance, - A new MC64 reordering, - Batched iterative solver support with the BiCGSTAB solver with batched Dense and ELL matrix types, - MPI support for the SYCL backend, - Improved ParILU(T)/ParIC(T) preconditioner convergence, and more! If you face an issue, please first check our [known issues page](https://github.com/ginkgo-project/ginkgo/wiki/Known-Issues) and the [open issues list](https://github.com/ginkgo-project/ginkgo/issues) and if you do not find a solution, feel free to [open a new issue](https://github.com/ginkgo-project/ginkgo/issues/new/choose) or ask a question using the [github discussions](https://github.com/ginkgo-project/ginkgo/discussions). Supported systems and requirements: + For all platforms, CMake 3.16+ + C++14 compliant compiler + Linux and macOS + GCC: 5.5+ + clang: 3.9+ + Intel compiler: 2019+ + Apple Clang: 14.0 is tested. Earlier versions might also work. + NVHPC: 22.7+ + Cray Compiler: 14.0.1+ + CUDA module: CMake 3.18+, and CUDA 10.1+ or NVHPC 22.7+ + HIP module: ROCm 4.5+ + DPC++ module: Intel oneAPI 2022.1+ with oneMKL and oneDPL. Set the CXX compiler to `dpcpp` or `icpx`. + MPI: standard version 3.1+, ideally GPU Aware, for best performance + Windows + MinGW: GCC 5.5+ + Microsoft Visual Studio: VS 2019+ + CUDA module: CUDA 10.1+, Microsoft Visual Studio + OpenMP module: MinGW. ### Version support changes + CUDA 9.2 is no longer supported and 10.0 is untested [#1382](#1382) + Ginkgo now requires CMake version 3.16 (and 3.18 for CUDA) [#1368](#1368) ### Interface changes + `const` Factory parameters can no longer be modified through `with_*` functions, as this breaks const-correctness [#1336](#1336) [#1439](#1439) ### New Deprecations + The `device_reset` parameter of CUDA and HIP executors no longer has an effect, and its `allocation_mode` parameters have been deprecated in favor of the `Allocator` interface. [#1315](#1315) + The CMake parameter `GINKGO_BUILD_DPCPP` has been deprecated in favor of `GINKGO_BUILD_SYCL`. [#1350](#1350) + The `gko::reorder::Rcm` interface has been deprecated in favor of `gko::experimental::reorder::Rcm` based on `Permutation`. [#1418](#1418) + The Permutation class' `permute_mask` functionality. [#1415](#1415) + Multiple functions with typos (`set_complex_subpsace()`, range functions such as `conj_operaton` etc). [#1348](#1348) ### Summary of previous deprecations + `gko::lend()` is not necessary anymore. + The classes `RelativeResidualNorm` and `AbsoluteResidualNorm` are deprecated in favor of `ResidualNorm`. + The class `AmgxPgm` is deprecated in favor of `Pgm`. + Default constructors for the CSR `load_balance` and `automatical` strategies + The PolymorphicObject's move-semantic `copy_from` variant + The templated `SolverBase` class. + The class `MachineTopology` is deprecated in favor of `machine_topology`. + Logger constructors and create functions with the `executor` parameter. + The virtual, protected, Dense functions `compute_norm1_impl`, `add_scaled_impl`, etc. + Logger events for solvers and criterion without the additional `implicit_tau_sq` parameter. + The global `gko::solver::default_krylov_dim`, use instead `gko::solver::gmres_default_krylov_dim`. ### Added features + Adds a batch::BatchLinOp class that forms a base class for batched linear operators such as batched matrix formats, solver and preconditioners [#1379](#1379) + Adds a batch::MultiVector class that enables operations such as dot, norm, scale on batched vectors [#1371](#1371) + Adds a batch::Dense matrix format that stores batched dense matrices and provides gemv operations for these dense matrices. [#1413](#1413) + Adds a batch::Ell matrix format that stores batched Ell matrices and provides spmv operations for these batched Ell matrices. [#1416](#1416) [#1437](#1437) + Add a batch::Bicgstab solver (class, core, and reference kernels) that enables iterative solution of batched linear systems [#1438](#1438). + Add device kernels (CUDA, HIP, and DPCPP) for batch::Bicgstab solver. [#1443](#1443). + New MC64 reordering algorithm which optimizes the diagonal product or sum of a matrix by permuting the rows, and computes additional scaling factors for equilibriation [#1120](#1120) + New interface for (non-symmetric) permutation and scaled permutation of Dense and Csr matrices [#1415](#1415) + LU and Cholesky Factorizations can now be separated into their factors [#1432](#1432) + New symbolic LU factorization algorithm that is optimized for matrices with an almost-symmetric sparsity pattern [#1445](#1445) + Sorting kernels for SparsityCsr on all backends [#1343](#1343) + Allow passing pre-generated local solver as factory parameter for the distributed Schwarz preconditioner [#1426](#1426) + Add DPCPP kernels for Partition [#1034](#1034), and CSR's `check_diagonal_entries` and `add_scaled_identity` functionality [#1436](#1436) + Adds a helper function to create a partition based on either local sizes, or local ranges [#1227](#1227) + Add function to compute arithmetic mean of dense and distributed vectors [#1275](#1275) + Adds `icpx` compiler supports [#1350](#1350) + All backends can be built simultaneously [#1333](#1333) + Emits a CMake warning in downstream projects that use different compilers than the installed Ginkgo [#1372](#1372) + Reordering algorithms in sparse_blas benchmark [#1354](#1354) + Benchmarks gained an `-allocator` parameter to specify device allocators [#1385](#1385) + Benchmarks gained an `-input_matrix` parameter that initializes the input JSON based on the filename [#1387](#1387) + Benchmark inputs can now be reordered as a preprocessing step [#1408](#1408) ### Improvements + Significantly improve Cholesky factorization performance [#1366](#1366) + Improve parallel build performance [#1378](#1378) + Allow constrained parallel test execution using CTest resources [#1373](#1373) + Use arithmetic type more inside mixed precision ELL [#1414](#1414) + Most factory parameters of factory type no longer need to be constructed explicitly via `.on(exec)` [#1336](#1336) [#1439](#1439) + Improve ParILU(T)/ParIC(T) convergence by using more appropriate atomic operations [#1434](#1434) ### Fixes + Fix an over-allocation for OpenMP reductions [#1369](#1369) + Fix DPCPP's common-kernel reduction for empty input sizes [#1362](#1362) + Fix several typos in the API and documentation [#1348](#1348) + Fix inconsistent `Threads` between generations [#1388](#1388) + Fix benchmark median condition [#1398](#1398) + Fix HIP 5.6.0 compilation [#1411](#1411) + Fix missing destruction of rand_generator from cuda/hip [#1417](#1417) + Fix PAPI logger destruction order [#1419](#1419) + Fix TAU logger compilation [#1422](#1422) + Fix relative criterion to not iterate if the residual is already zero [#1079](#1079) + Fix memory_order invocations with C++20 changes [#1402](#1402) + Fix `check_diagonal_entries_exist` report correctly when only missing diagonal value in the last rows. [#1440](#1440) + Fix checking OpenMPI version in cross-compilation settings [#1446](#1446) + Fix false-positive deprecation warnings in Ginkgo, especially for the old Rcm (it doesn't emit deprecation warnings anymore as a result but is still considered deprecated) [#1444](#1444) ### Related PR: #1454

ginkgo-bot added mod:core This is related to the core module. mod:cuda This is related to the CUDA module. reg:benchmarking This is related to benchmarking. labels Aug 10, 2023

upsj commented Aug 10, 2023

View reviewed changes

cuda/base/memory.cpp Show resolved Hide resolved

upsj requested a review from a team August 10, 2023 19:15

upsj added the 1:ST:ready-for-review This PR is ready for review label Aug 10, 2023

ginkgo-bot force-pushed the benchmark_allocator branch from b2ba4f0 to da42d8b Compare August 10, 2023 19:16

pratikvn reviewed Aug 10, 2023

View reviewed changes

pratikvn approved these changes Aug 11, 2023

View reviewed changes

yhmtsai reviewed Aug 11, 2023

View reviewed changes

yhmtsai approved these changes Aug 11, 2023

View reviewed changes

benchmark/sparse_blas/sparse_blas.cpp Show resolved Hide resolved

upsj added 3 commits August 11, 2023 13:39

allow specifying allocator for benchmarks

39125dd

fix CUDA_VERSION availability

90d88e3

increase repetitions for sparse_blas

21b4cd6

upsj force-pushed the benchmark_allocator branch from e8583e7 to 21b4cd6 Compare August 11, 2023 11:39

upsj 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 Aug 11, 2023

upsj merged commit ee648e2 into develop Aug 11, 2023
12 of 14 checks passed

upsj deleted the benchmark_allocator branch August 11, 2023 20:34

tcojean mentioned this pull request Nov 6, 2023

Release 1.7.0 to master #1451

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Allow specifying allocator for benchmarks #1385

Allow specifying allocator for benchmarks #1385

upsj commented Aug 10, 2023

upsj commented Aug 10, 2023

ginkgo-bot commented Aug 10, 2023

pratikvn Aug 10, 2023

upsj Aug 11, 2023

pratikvn Aug 11, 2023

upsj Aug 11, 2023 •

edited

Loading

yhmtsai Aug 11, 2023

upsj Aug 11, 2023

yhmtsai Aug 11, 2023

yhmtsai Aug 11, 2023

upsj Aug 11, 2023 •

edited

Loading

sonarcloud bot commented Aug 11, 2023

		return std::make_shared<gko::CudaAllocator>();
		} else if (flag == "async") {

		@@ -75,6 +75,10 @@ bool CudaAsyncAllocator::check_environment(int device_id,
		GKO_NOT_COMPILED(cuda);


		CudaUnifiedAllocator::CudaUnifiedAllocator(int device_id)

Allow specifying allocator for benchmarks #1385

Allow specifying allocator for benchmarks #1385

Conversation

upsj commented Aug 10, 2023

upsj commented Aug 10, 2023

ginkgo-bot commented Aug 10, 2023

pratikvn Aug 10, 2023

Choose a reason for hiding this comment

upsj Aug 11, 2023

Choose a reason for hiding this comment

pratikvn Aug 11, 2023

Choose a reason for hiding this comment

upsj Aug 11, 2023 • edited Loading

Choose a reason for hiding this comment

yhmtsai Aug 11, 2023

Choose a reason for hiding this comment

upsj Aug 11, 2023

Choose a reason for hiding this comment

yhmtsai Aug 11, 2023

Choose a reason for hiding this comment

yhmtsai Aug 11, 2023

Choose a reason for hiding this comment

upsj Aug 11, 2023 • edited Loading

Choose a reason for hiding this comment

sonarcloud bot commented Aug 11, 2023

upsj Aug 11, 2023 •

edited

Loading

upsj Aug 11, 2023 •

edited

Loading