GitLab

* Add a workaround for infinite loop in computeNonbonded (HIP)

computeNonbonded hangs in some tests (without neighbor list).
Reproducible on ROCm 6.4 and 6.4.1 (maybe on older versions too) on various architectures (both CDNA and RDNA).
Affected tests: TestHipATMForce, TestHipMonteCarloBarostat, TestHipNonbondedForce, TestHipVirtualSites.

Disassembly shows that the compiler splits branches of `if (skipBase+tgx < NUM_TILES_WITH_EXCLUSIONS)` and does
`SHFL(skipTiles, TILE_SIZE-1) < pos` checks in them separately, even though `__builtin_amdgcn_ds_bpermute`
is a convergent function. Apparently in this case not all lanes participate in each call.

* Simplify includeTile check using ballot

Skip neighbor list for very small systems

    https://github.com/openmm/openmm/pull/4070

Store bounding box sizes in half precision

    https://github.com/openmm/openmm/commit/2ae50f9

Use large blocks to optimize building the neighbor list

    https://github.com/openmm/openmm/commit/3955033

Improved sorting of blocks when building neighbor list

    https://github.com/openmm/openmm/commit/796ffaa

Fixed bug in large blocks optimization with triclinic boxes

    https://github.com/openmm/openmm/commit/4c10732

Optimize sorting of non-uniformly distributed data

    https://github.com/openmm/openmm/commit/71d9bb1

Co-authored-by: bdenhollander <44237618+bdenhollander@users.noreply.github.com>

* All AMD GPUs support shuffle, double precision and 64-bit int atomics;
* Remove unused code: !ENABLE_SHUFFLE code paths in nonbonded.hip;
* Use intrinsics in single-precision;
* Use realToFixedPoint (faster float32-to-int64);
* Remove shared atomIndices, use shuffles;
* Check early if atoms are in the cutoff range, sometimes all lanes in
  a warp can skip computations, single pairs can also skip useless
  atomics with zero values;
* Remove volatile skipTiles access, use shuffles;
* Distribute work for warps in a strided order;
* Skip warps that may be still busy in the first loop;
* Unify conditions for excluded atoms with `includeInteraction`;
* Move multiprocessors to HipContext;
* Increase number of warps for computeNonbonded;
* Disable packed math for >=MI200 (it affects performance of some
  kernels like computeGKForces of amoebagk);
* Remove defaultOptimizationOptions and createModule's optimizationFlags
  as they are never used;
* Support -save-temps.

Port changes in CUDA backend to HIP

Fix a warning about arithmetic operations on void* in HipArray::uploadSubArray

Fix "Error Initializing context ROCm 5.3.0"

    https://github.com/StreamHPC/openmm-hip/issues/3


    hipDeviceSetCacheConfig returns hipErrorNotSupported on 5.3
Co-authored-by: Nick Curtis <nicholas.curtis@amd.com>

* Skip neighbor list for very small systems

* Fixed typos

* Don't skip box size check when not using neighbor list

* Made test larger to ensure it uses neighbor list

It allows to use a faster float-to-int64 in the HIP platform.

* Minor optimizations to computing single pairs

* Adjusted MAX_BITS_FOR_PAIRS on Ampere

* Fixed range overflow with very large numbers of atoms

* More fixes to overflow with large numbers of atoms

* Fix test failures

Converted the array containing atom block indices for the neighbor list from ushort2 to int.  This removes the hard limit of 2 million atoms.

Fixes a hard to catch bug when a boundingBoxSize increases in size in between a timestep, but does not trigger a rebuild of the neighbourlist. Affects the usage of singlePeriodicCopy in nonbonded force.  

When converting to fixed point, multiply by 0x100000000 instead of 0xFFFFFFFF.  This should be (very very slightly) more accurate, since its reciprocal can be exactly represented in floating point.