[stdpar] Update stdpar to remove Product, ScopedContext, device abstraction

[esseivaju]

As discussed in #368 this PR contains an updated to cudart api utilization to get closer to standard c++ and no cuda dependencies.

Changes

• No longer use the cudaStream and cudaEvent APIs.
• Removed multiple classes in CUDACore which are no longer necessary
• Updated Plugins and tests to be compatible with the previously mentioned changes.
• Refactored multiple producers to inherit from EDProducer instead of EDProducerExternalWork

Current issues to be fixed

• Multiple tests still manage memory through cudaMalloc / cudaMemcpy, this should be updated to use std::unique_ptr
• There are still runtime errors. I am currently working on figuring out exactly what's the issue. I've noticed that some kernels seem to produce incorrect results (e.g. theSiPixelRawToClusterCUDA module). I updated the gpuClustering_t test to work with std::unique_ptr and it fails with the following output:

[esseivaj@zeus pixeltrack-standalone]$ test/stdpar/gpuClustering_t 
created 175031 digis in 8960 clusters
CUDA countModules kernel launch with 2000 blocks of 256 threads
CUDA findModules kernel launch with 2000 blocks of 256 threads
before charge cut found 0 clusters
ERROR!!!!! wrong number of cluster found
found 897 Modules active
gpuClustering_t: /bld4/home/esseivaj/devel/pixeltrack-standalone/src/stdpar/test/gpuClustering_t.h:284: int main(): Assertion `h_clus[i] < int(nclus[h_id[i]])' failed.
Aborted (core dumped)

[esseivaju]

Changes listed in the original comment have been implemented. The tests below are still failing:

• HistoContainer_t
• OneToManyAssoc_t
• prefixScan_t
• radixSort_t
• Tests depending on a plugin

Fixing those will require updating many Kernels so it should probably be done in a future PR as this one is getting fairly large. Some of these tests (e.g. prefixScan_t / prefixScan.h) are still pretty much identical to the original cudauvm version, the only change being update to #define directives to compile the correct code with nvc++ so it's suggesting some issues with nvc++

[makortel]

it should probably be done in a future PR as this one is getting fairly large.

I agree, this starts to be a good chunk to integrate. Visually looks good, could you just confirm that you agree with merging now?

[esseivaju]

Yes we can merge now I will fix the remaining issues with HistoContainer, prefixScan and radixSort in the next pull request.