Magma dev optimizations (#111)

* makefile changes

* update magma backend

* magma qfunctions updated to new interface

* in the magmabackend we manage where pointers are - if on CPU, on some

Magma dev optimizations (#111)

* makefile changes

* update magma backend

* magma qfunctions updated to new interface

* in the magmabackend we manage where pointers are - if on CPU, on some cases we still need and may call the CPU code

* update the reflect changes in the API

* update the reflect changes in the API

* add the q functions for ex1.c

* Switch to CeedIntPow

* Fix merge errors

* Clean up Magma operator loops

* Move zeroing lvec

* fix bug in the rebase and add some qfunctions. This passes the tests now

* adding new files, changing -O to -O3

* new faster way of checking CPU vs. GPU pointers

* core magma device functions for basis apply

* new kernels for basis apply

* use the new magma_isdevptr function

* minor cleanup

* new headers and defs

* calling the new magma functions for basis apply

* undo O3, and change default magma directory

* use static

* use static

* modify the generator to add before __global__

* remove unnecessary header

* silence some warnings

* Makefile: restore NVCC and NVCCFLAGS to match master

* first pass as updating new Magma work, untested

* Use CUDA backend to dispatch

* Device memory for MAGMA

* Add copyright messages and tidy

* WIP: starting fresh on magma-dev-rebae. Add magma_is_devptr

* WIP: starting fresh on magma-dev-rebae. Fix build issue

* WIP: starting fresh on magma-dev-rebae. Fix build issue

* WIP: starting fresh on magma-dev-rebae. Fix build issue

* WIP: starting fresh on magma-dev-rebae. Fix this include

* WIP: starting fresh on magma-dev-rebae. Fix build issue.

* WIP: starting fresh on magma-dev-rebase. Mostly fixing compilation errors

* WIP: starting fresh on magma-dev-rebase. Disbale magma-basis for now

* WIP: starting fresh on magma-dev-rebase. Edit the required magma src files

* move CeedVector_Magma functions from magma-dev branch

* add includes

* fix build errors

* disable magma vector logic for now

* Remove reference in CeedDelegate

* add RestoreArray calls to CeedBasisApply_Magma

* add basis for magma

* magma batched operator

* Change batch calls to match Q-vector ordering

* minor cleanup of unused variable

* update magma portion of Makefile

* remove magma vectors

* remove unused contract variable

* change extern to CEED_INTERN

* merge in updates from master branch

* MAGMA - add lcov markers

* remove stray example script

show more ...

Merge pull request #360 from CEED/jed/config-flags

Jed/config flags and string processing

Merge pull request #353 from CEED/jeremy/xsmm-pkgconfig

Makefile - Detect MKLROOT for LIBXSMM linking

Makefile: add flags to config.mk and add NVCC_CXX

CUDA usually does not support the latest gcc, but an older version C++
compiler is only needed internally by nvcc. NVCC_CXX allows the user
to spec

Makefile: add flags to config.mk and add NVCC_CXX

CUDA usually does not support the latest gcc, but an older version C++
compiler is only needed internally by nvcc. NVCC_CXX allows the user
to specify a compatible C++ compiler for CUDA while still using a
modern release for *.cpp sources.

show more ...

Makefile - Only add -L -rpath for MKL if MKLROOT defined

Makefile - add MKL -rpath flag

Makefile - add MKL -L flag

Makefile - improve MKL check

Makefile - Detect MKLROOT for LIBXSMM linking

Merge pull request #344 from CEED/jeremy/ifort-fix

Fortran - Use preprocessor directive to provide abs path to true qfuntion source
resolves Issue #71

Fortran - Use preprocessor directive to provide abs path to true qfunction source

QFunction Create by Name (#311)

This PR adds a QFunction gallery to libCEED with 1D, 2D, and 3D mass and Poisson operators.

Closes issue #37, issue #340

* Add QFunction gallery, rename focca

QFunction Create by Name (#311)

This PR adds a QFunction gallery to libCEED with 1D, 2D, and 3D mass and Poisson operators.

Closes issue #37, issue #340

* Add QFunction gallery, rename focca

* Gallery - add initial QFunctions

* Add a test for using the QF gallery

* Modify ex1 to use gallery

* Add multiple test configs to tap

* Move output to test directory

* Update junit

* Add OCCA galley exception

* Add ex2

* Update ex2 for dim->ncompx

* Gallery - modify to work for CUDA as is

* Update Documentation

* Gallery - typo fix

* Gallery - convention change, postappend qfunction family variant

* Gallery - update template with new name checking convention

* Gallery - condense diff3DBuild QFunction

* Gallery - rename diff -> poisson

* Gallery - clarify poisson3DBuild comment

* Gallery - use Pragma SIMD, store Qdata in Voigt convention

* Examples - Convert BP3-6 to Voigt convention

* Examples - add cl option to switch between header and gallery qfs in CEED examples

* Examples - clean up construction of QF name

* Gallery - Switch to PascalCase for gallery names

* Doc - fix function type page

* Interface - Make sure strncpy result is null terminated

* Gallery - Update Poisson 2/3D Apply to new QF body

* make style

* make style - fix worst style problems

* make style - add gallery to make style

* Doc - update documentation errors and inconsistencies

* Examples - test ex1 ex2 with and without gallary

* Examples - reduce testing of ex1/ex2 without gallery, clean up non-gallery qfunctions

* MFEM - revert another make style mistake

* Manual make style updates

* Doc - update function documentation page

* Style updates, document test numbering conventions

* doc: resolve ambiguous image location warning, allow more Dot nodes

* Tests - style and cast cleanup

* Tests - fix README indentation

show more ...

Single Source QFunction (#304)

Introduce a new macro CEED_QFUNCTION that allows to define qFunctions in a single source code independently of the targeted backend.

Thanks-to: Jeremy Thompson
Tha

Single Source QFunction (#304)

Introduce a new macro CEED_QFUNCTION that allows to define qFunctions in a single source code independently of the targeted backend.

Thanks-to: Jeremy Thompson
Thanks-to: Jed Brown
This work is the result of a fruitful discussion between Jed Brown, Jeremy Thompson and Yohann Dudouit. Jeremy Thompson also implemented important features in this commit and was very active and helpful all along the progress of this work.

[NEWS] Breaking change: QFunctionField parameter 'ncomp' changed to 'size'. This change requires setting the previous value of 'ncomp' to 'ncomp*dim' when adding a QFunctionField with eval mode 'CEED_EVAL_GRAD'.

* First steps toward cuda-gen backend!

* Closer to real code generation.

* Generated code should be ready for nvrtc.

* The code generation skeleton is ready.

* Hack with the qfunction to make the operator kernel compile.

* Some tweaks in the makefile + Input fields structure change.

* Remove using cout.

* 1d interp and grad device functions.

* 1d readDofs, readQuads, writeDofs, writeQuads.

* Remove dead code.

* readDofs, readQuads, writeDofs, writeQuads for 2d and 3d

* 2d interp and grad

* 3d interp and grad

* - weight functions for 1d,2d,3d
- link the indices to the kernel
- link the fields to the kernel
- link the basis to the kernel

* Add the qFunction reader + inlining

* Add qf files for the tests.

* Add qf file for ceed/ex1

* Add qf file for mfem/bp1

* All tests pass.

* Add qFunction for mfem/bp3, petsc/bp1, and petsc/bp3.

* mfem/bp1 passes + remove dead code

* Fix a bug in n_quads_out for writeQuads

* mfem/bp3 passes.

* All tests all examples pass.

* Temporary tweaks for mfem benchmarking

* Add Context management.

* Modify .qf files to take into account the context.

* Enable optimizations.

* First set of optimization for 2D and 3D.

* double pointer format for the qFunction.

* Change the .qf files to have the same code as the C functions.

* Make previous Cuda backends use .qf files.

* Add a return value to qFunctions.

* Make cpu backends use .qf files.

* Minor: clean commented code.

* Add guarded math.h for petsc examples.

* Remove previous nek qf files.

* Remove .cu files.

* Remove .qf files.

* Remove dead code in the tests.

* make style

* Make style fix.

* more make style fixes.

* CEED_QFUNCTION - improve macro for CPU filenames

* Add CEED_QFUNCTION macro to navierstokes.c

* Fix PETSc gitignore

* Change default NS problemtype to density_current (#307) in navierstokes.c

* Fix petsc bp1.h

* Real Fix for petsc bp1.h...

* fix

* README - Add /gpu/cuda/gen

* PETSc - Update dmplex example to use *_loc

* cuda/reg - fix typo

* Revert a couple of small changes

* Fix a bug in mfem bp3 similar to the previous bug in petsc bp3.

* Make PETSc qfunctions look closer to master, and minor style for debugging.

* More uniformity changes

* Fix a strange CUDA_OUT_OF_RESSOURCE bug.

* NS - fix fname variables

* Use a different convention for qFunction ncomp.

* update cuda-gen backend and bpsdmplex.

* PETSc - style update

* update mfem bp1 and bp3.

* Interface - Use size instead of ncomp for QFunction fields

* update ceed example and tests.

* Tests - Update ncomp to size

* CPU Backends - Update ncomp to size

* CPU Backends - style

* Nek - Update ncomp to size

* Opt - fix style

* CUDA - update ncomp to size

* Doc - Update API documentation for QFunction \ncomp->size

* OCCA - Patch QFunction ncomp -> size, work but revamp will be better

* OCCA - assert dim>0 for clang-tidy

* CUDA - Change GetNumComp to GetSize

* Basis - Shift check for dim > 0 to interface

* Doc update

* Update NS field size

* NS - Fix problem options

show more ...

Merge pull request #302 from CEED/jeremy/dmplex

Create DMPlex PETSc BPs

Travis: Use PETSc master, commit 73f49a0

PETSc DMPlex BP1-6 working in unified code

Merge pull request #303 from CEED/nvcc-host-compiler

In Makefile, use the user defined c++ compiler as the nvcc host compiler

In Makefile, use the user defined c++ compiler as the
nvcc host compiler.

Full jit compiled operator: cuda-gen backend (#275)

* First steps toward cuda-gen backend!

* Closer to real code generation.

* Generated code should be ready for nvrtc.

* The code generatio

Full jit compiled operator: cuda-gen backend (#275)

* First steps toward cuda-gen backend!

* Closer to real code generation.

* Generated code should be ready for nvrtc.

* The code generation skeleton is ready.

* Hack with the qfunction to make the operator kernel compile.

* Some tweaks in the makefile + Input fields structure change.

* Remove using cout.

* 1d interp and grad device functions.

* 1d readDofs, readQuads, writeDofs, writeQuads.

* Remove dead code.

* readDofs, readQuads, writeDofs, writeQuads for 2d and 3d

* 2d interp and grad

* 3d interp and grad

* - weight functions for 1d,2d,3d
- link the indices to the kernel
- link the fields to the kernel
- link the basis to the kernel

* Add the qFunction reader + inlining

* Add qf files for the tests.

* Add qf file for ceed/ex1

* Add qf file for mfem/bp1

* All tests pass.

* Add qFunction for mfem/bp3, petsc/bp1, and petsc/bp3.

* mfem/bp1 passes + remove dead code

* Fix a bug in n_quads_out for writeQuads

* mfem/bp3 passes.

* All tests all examples pass.

* Temporary tweaks for mfem benchmarking

* Add Context management.

* Modify .qf files to take into account the context.

* Enable optimizations.

* First set of optimization for 2D and 3D.

* Makefile tweaks and destructor code.

* make style.

* Add -MP flag.

* Fix linking issues with the tests.

* Update .qf files for the tests.

* Add .qf files for nek5000 examples.

* Use shared memory for B and G matrices.

* Fix bug introduced in previous commit.

show more ...

Update nek5000 example scripts (#241)

* Update nek5000 examples to use NEKTOOLS_DIR

* Remove +x

* Nek: Fix a bug in generate-boxes.sh script

* Use one script for nek exaamples

* Update m

Update nek5000 example scripts (#241)

* Update nek5000 examples to use NEKTOOLS_DIR

* Remove +x

* Nek: Fix a bug in generate-boxes.sh script

* Use one script for nek exaamples

* Update make-nek-tests.sh

* Update nek-examples.sh

* Change permissions of make-nek-tests.sh

* Update examples Makefile to use new nek script

* Update nek5000 examples build process

* Trying to make travis happy

* Polishing the script

* This should make Travis happy

* Polish nek-examples.sh

* Update examples/Makefile

* Update top-level makefile

* Delete make-nek-tests.sh & update nek-examples.sh

* Update Makefile and examples/Makefile

* Update nek-examples.sh

* Update TESTARGS in bp .usr files

* Update Makefile

* Update .travis.yml

* Update nek-examples.sh

* Add CEED_DIR to tap.sh

* Uncomment OCCA_DIR in Makefile

* Fix make prove and update tap.sh

* Pass NEK5K_DIR and CEED_DIR in prove-all

* Set MPI=0 default for Nek examples/tests

* Nek5K make updates

* Make prove update

* Update make-prove to make all ex except Nek5K in parallel

* Set Nek5K examples to use MPI by default

* Makefile: Improved Nek5k handling

* Makefile: build Nek bps in separate directories

* Nek: rename dir for f77 file length

* debugging

* Nek: use single source

* Makefile: fix Nek mpif.h issue

* Nek: Clean up build proces, improve documentation

* Makefile: enable building single Nek example in parallel with other examples

* Nek: Fix CC/FC and TESTARGS

show more ...

Merge branch 'master' into yohann/cuda-restr-opt

Conflicts:
backends/cuda-reg/ceed-cuda-reg-restriction.c
backends/cuda-shared/ceed-cuda-shared-basis.c

Merge pull request #284 from CEED/jed/blas-lib

Makefile: MKL link flags for non-Intel compilers

Makefile: MKL link flags for non-Intel compilers

One can always run

make BLAS_LIB=-mkl

when using the Intel compiler, but it's typically important for us to
link the sequential version and we'd

Makefile: MKL link flags for non-Intel compilers

One can always run

make BLAS_LIB=-mkl

when using the Intel compiler, but it's typically important for us to
link the sequential version and we'd like a system that works with
non-Intel compilers, where the incantation is so hard to remember that
Intel made a web app:

https://software.intel.com/en-us/articles/intel-mkl-link-line-advisor/

show more ...

Makefile: AVX test cannot match -mno-avx

Merge branch 'yohann/cuda-non-tensor' [PR #249]

* yohann/cuda-non-tensor:
ceed-cuda: resolve -Wsign-compare for CUresult (unsigned enum) in CeedError
make style.
namespace cuda backends functi

Merge branch 'yohann/cuda-non-tensor' [PR #249]

* yohann/cuda-non-tensor:
ceed-cuda: resolve -Wsign-compare for CUresult (unsigned enum) in CeedError
make style.
namespace cuda backends functions.
Minor: styling
Add CUDA_LIB_DIR_STUBS for systems that don't have CUDA drivers installed
make style
Remove useless function declaration.
Add a reference non-tensor BasisApply for cuda backends.

show more ...