Ceed - add debugging mode for backends, lifted to library level from OCCA backend

Ceed - add IsDeterministic

Op - add AssembleAdd version of diagonal assembly functions, will be helpful for MFEM integration

Op - change LinearAssemble* to accept CeedVector instead of pointer to CeedVector, allows for less memory movement and interfaces with parent code better

style - rename AssembleLinear* to LinearAssemble*

style - change Get*Status to Is*

Composite Operator support for AssembleLinearDiag/PBDiag (#552)

* ref - add composite operator support for building diagonal/pb diagonal

* ref - add error for non-composite mult-field operator di

Composite Operator support for AssembleLinearDiag/PBDiag (#552)

* ref - add composite operator support for building diagonal/pb diagonal

* ref - add error for non-composite mult-field operator diagonal/pb diagonal assembly

* tap - add t538 exclusion because OCCA does not support galleries

* tests - adjust test cases for ceed examples for test coverage

* Op - fix documentation

show more ...

Convert CUDA ref/reg/shared E-Layout (#554)

* tests - update tests for multiple e-layouts

* CUDA - convert ref and reg restrictions to Q-layout

* CUDA - ref/reg/shared use gen/magma E-Layout f

Convert CUDA ref/reg/shared E-Layout (#554)

* tests - update tests for multiple e-layouts

* CUDA - convert ref and reg restrictions to Q-layout

* CUDA - ref/reg/shared use gen/magma E-Layout for multi elememnt basis apply and operator apply

* CUDA/MAGMA - drop eandqdiffer and separate MAGMA operator code

* CUDA - update operator comment

* reg - clarify read/write dofs/quads

* CUDA - drop dead code

show more ...

CPU Impl of AssemblePointBlockDiagonal (#503)

* Operator - add AssemblePointBlockDiagonal for CPU backends, with test

* CUDA - add point block diagonal not supported message

* make style

*

CPU Impl of AssemblePointBlockDiagonal (#503)

* Operator - add AssemblePointBlockDiagonal for CPU backends, with test

* CUDA - add point block diagonal not supported message

* make style

* Operator - improve point block description

* OCCA - explicitly remove OCCA fallback to CPU assembly functions, will update after new OCCA backend

* Op - remove gap removal in point block diagonal

* Op - update diagonal assembly documentation

* Update backends/ref/ceed-ref-operator.c

Co-authored-by: Jed Brown <jed@jedbrown.org>

* style - fix extra space in * with nopad

Co-authored-by: Jed Brown <jed@jedbrown.org>

show more ...

fortran: use standard z prefix instead of deprecated x prefix for hexadecimal

XSMM - fix q=1 index computation (#531)

* XSMM - use hash table for kernel index lookup, add khash to libCEED

* Hash - add CeedHashIJKLM to simplify xsmm tensor hash table code

* simplify use

XSMM - fix q=1 index computation (#531)

* XSMM - use hash table for kernel index lookup, add khash to libCEED

* Hash - add CeedHashIJKLM to simplify xsmm tensor hash table code

* simplify use of hash table, use kernels as values

* XSMM: more idiomatic use of khash

* make style

Co-authored-by: Jed Brown <jed@jedbrown.org>

show more ...

doc: remove obsolete references to CeedInterlaceMode

ElemRestriction - add check for strided status

ElemRestriction - add state counter for reading offsets array

ElemRestriction - add ElemRestrictionGetOffsets

Merge pull request #513 from CEED/jeremy/restriction-no-interlace

Convert restriction to offsets

Basis - better protect basis data from user meddling

Co-Authored-By: Jed Brown <jed@jedbrown.org>

Basis - move some functions to public API

Restriction - convert restiction to use offsets instead of interlace mode. CPU and Python impl only, CUDA and MAGMA left to do.

pragma vector for Intel; pragma simd is deprecated

Pragma simd is documented without deprecation notice, but the
icc-19.05.281 warns of its deprecation.
https://software.intel.com/en-us/cpp-compiler

pragma vector for Intel; pragma simd is deprecated

Pragma simd is documented without deprecation notice, but the
icc-19.05.281 warns of its deprecation.
https://software.intel.com/en-us/cpp-compiler-developer-guide-and-reference-simdvector

One would think Intel's pragma ivdep would be similar to GCC ivdep,
but icc evidently has different semantics/miscompiles, as demonstrated
in the example, which is left in the source to protect an enterprising
future committer from adding ivdep.

show more ...

Ceed - add CeedView()

Add VLA support for CUDA backends (#481)

* CUDA - add VLA support for CUDA backends

* Navier-Stokes - guard #include math.h for CUDA

* Navier-Stokes - fix struct definitions for code generatio

Add VLA support for CUDA backends (#481)

* CUDA - add VLA support for CUDA backends

* Navier-Stokes - guard #include math.h for CUDA

* Navier-Stokes - fix struct definitions for code generation

* Navier-Stokes - fix file location of mass QFunction

* Navier-Stokes - minor style consistency

* Navier-Stokes - more context sturct fixing

* Navier-Stokes - cast context to stop CUDA errors

* Navier-Stokes - change helper functions to inline to silence CUDA errors

* CUDA - clean up CEED_Q_VLA definition for cuda/gen

* Navier-Stokes - populate TODOs for undocumenced functions

* Doc - document libCEED macros

* CUDA - use -default-device for NVRTC

* Tests - add t507 to test VLA support

* CUDA - fix my bug in dim, Q1d for cuda/gen

* Tests - update tap/junit to reflect current CUDA capabilities

* make style

* CUDA - check output bases for dim, Q1d as well (this fixes multigrid example support)

* CUDA - fix readSliceQuads, not all arguments were defined

* Navier-Stokes - document all functions in Navier-Stokes QFunction files

* make style

show more ...

change restrict to __restrict__ for t

Interface - use __restrict__ instead of restrict for C++ compatability

doc: convert to Sphinx documentation at libceed.readthedocs.io

Some technical notes about the expanded docs/process:

* Doxygen is now restricted to API documentation. Sphinx docs are
incomplete

doc: convert to Sphinx documentation at libceed.readthedocs.io

Some technical notes about the expanded docs/process:

* Doxygen is now restricted to API documentation. Sphinx docs are
incomplete surrogates due to not hyperlinking code listings, though
we hope the Doxygen site can be removed eventually.

* PDF output uses XeTeX (unicode and OTF fonts) with TeX Gyre Pagella
and DejaVu Sans Mono. You may need to install these fonts in order
to build the PDF users manual.

* Obtaining the author list from Git has not worked on
ReadTheDocs (`git shortlog` is empty even though `git log` is not),
so we automatically generate it on developer machines with `make
doc`. If the author information is incorrect/mangled, edit .mailmap
to fix the mapping.

* Math semantics are based on the description here
https://github.com/sphinx-doc/sphinx/issues/3785

Sphinx chooses one of two setups for the maths.

* If there is a single equation, then we wrap the maths in
\begin{equation*} (or equation).

* If there are multiple equations (as determined by the presence of
an empty line), then we wrap the maths first in \begin{aligned}
and then in \begin{align*} (or align).

Co-authored-by: Jeremy L. Thompson <jeremy.thompson@colorado.edu>
Co-authored-by: Jed Brown <jed@jedbrown.org>

show more ...