Cov - drop debug code from code coverage and run make tidy

CUDA - output generated operator code with debugging flag

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

Merge pull request #561 from CEED/jeremy/is-deterministic

Ceed - add IsDeterministic

spellcheck

Merge pull request #564 from CEED/jeremy/cuda-diag

CUDA - LinearAssemble*Diagonal

README - highlight reproducibility of backends

CUDA - add missing codecov exceptions

Ceed - add IsDeterministic

Merge pull request #569 from CEED/jeremy/cuda-shared-fix

CUDA - shared basis needs minimum of 1 elem per block

CUDA - shared basis needs minimum of 1 elem per block

CUDA - use d_ and h_ prefix for diagonal assembly data arrays

PETSc - add missing address-of operator in prolongation matshell operation

Ref - compute QF value bound once

CUDA - convert diagonal assembly into device kernel

CUDA - fix composite operator diagonal support

CUDA - inital version of LinearAssemble*

Merge pull request #568 from CEED/jeremy/cuda-destroy

Small CUDA Tidying

Tests - add t318 to test all AVX backend tensor cases

PETSc - adjust bpsraw test parameters to avoid too many threads for cuda backends

CUDA - fix small leak in cuda/gen QFunction load

CUDA - add some missing ierr =

CUDA - add small leak fix in QFunction JiT load

CUDA - clean up includes

CUDA - clean up minor memory leak