MATSEQAIJCUSPARSE: optimize write access when possible

In MatMult and MatMultAdd_ with compressed rows, we only write access to the GPU array output

MatSEQAIJ{CUSPARSE|VIENNACL}: do not copy to the GPU if not at the final stage of assembly

CUDA: Removed check for ancient CUDA version >= 5000

CUDA: Remove ancient checks for CUDA >= 4.2

Merge branch 'hannah/gpu-gpu-logging' [PR #1881]

* hannah/gpu-gpu-logging:
This branch adds timers to include GPU data movement (device to device data movement) in GPU logging.

removing comments

adding GPU-GPU data movement to logging

Merge branch 'gropp/doc-fixes-1' [PR #1838]

Merge branch 'hannah/gpu-computation-logging' [PR #1843]

* hannah/gpu-computation-logging:
Adding GPU flop rate and GPU time.

PetscLogGpuTimeStart -> Begin

Adding Gpu flop rate and GPU time

Merge from doc-fixes branch

Merged in hannah/gpu-communication-logging (pull request #1814)

Hannah/gpu communication logging

Approved-by: BarryFSmith <bsmith@mcs.anl.gov>
Approved-by: Richard Mills <rtm@eecs.utk.edu>

Adding vector logging, started matrix logging

Fixes for doctext update

Merged in barry/cuda-multigrid-test (pull request #1763)

Various improvements for GPUs (mostly for performance and CUDA)

Various improvements for GPUs (mostly for performance and CUDA)

1) Add VecPinToCPU() for CUDA vector and matrices
2) Move initialization of cuBLAS to PetscInitialize() since it takes 1/2 second and

Various improvements for GPUs (mostly for performance and CUDA)

1) Add VecPinToCPU() for CUDA vector and matrices
2) Move initialization of cuBLAS to PetscInitialize() since it takes 1/2 second and distorts timing with -log_view
3) Add logging for DMCreateMatrix (for large meshes this is very large)
4) Add VecGet/RestoreArrayWrite() to prevent unneeded copies from GPU (only implemented so far for CUDA);
added a small number of usages in the source so that snes tutorials ex19 does not do unneeded communication from the GPU
5) Automatically convert MAIJ matrices to AIJ for CUDA since they are not yet supported natively in PETSc's CUDA matrix implementation
6) Pinned objects should still use the CUDA/ViennaCL versions of Destroy to clean up the GPU stuff

Commit-type: feature

show more ...

Merge remote-tracking branch 'origin/barry/add-petscarrayxxx-ops' [PR #1771]

Remove dead Matlab code, fix typo, fix caste needed for const free

Commit-type: portability-fix
Reported-by: Satish Balay <balay@mcs.anl.gov>

Add PetscArraycmp(), PetscArraycpy(), PetscArrayzero(), PetscArraymove(), update code to use new forms

Replace a couple multi-mallocs with a single malloc
Replace backwards loops in MatSetValues_*()

Add PetscArraycmp(), PetscArraycpy(), PetscArrayzero(), PetscArraymove(), update code to use new forms

Replace a couple multi-mallocs with a single malloc
Replace backwards loops in MatSetValues_*() with PetscArraymove()

Commit-type: style-fix, feature

show more ...

Merged in barry/update-collective-on (pull request #1744)

Update the use of Collective on in the manual pages to reflect the new style

Update the use of Collective on in the manual pages to reflect the new style

Commit-type: style-fix, documentation
Thanks-to: Patrick Sanan <patrick.sanan@gmail.com>

Merge branch 'barry/rm-readwrite' [PR #1710]

* barry/rm-readwrite:
Remove the ReadWrite notation from Vec array access for GPU vectors to match usage on the CPU.

Remove the ReadWrite notation from Vec array access for GPU vectors to match usage on the CPU

Commit-type: style-fix

Development Tools: Vim, Emacs, Eclipse

Merge branch 'master' into jpusztay/feature-swarm-symplectic-example