CUDA: for each PR, show delta register usage for GPU kernels

[BenWibking]

Describe the proposal
List the GPU kernels with changed register usage as a comment in each PR.

This can done by using the --ptxas-options=v compiler flag, then parsing the compiler output with sed or another text parser, e.g.:

ptxas info    : Compiling entry function 'searchkernel(octree, int*, double, int, double*, double*, double*)' for 'sm_20'
ptxas info    : Function properties for searchkernel(octree, int*, double, int, double*, double*, double*)
    72 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info    : Used 46 registers, 176 bytes cmem[0], 16 bytes cmem[14]

We can parse the above into a *.csv file that consists of two columns: kernel name and register usage:

searchkernel 46

and then diff it against a reference file computed using the current development branch.

Describe alternatives you've considered
Profile GPU performance directly for each PR. This is tricky to do on a GPU kernel-by-kernel basis.

Additional context
GPU performance for our code is exquisitely sensitive to register usage. Conversely, register pressure is a good predictor of performance for our code. This metric should tell us whether there are major performance regressions on GPU.

See also: https://stackoverflow.com/questions/12388207/interpreting-the-verbose-output-of-ptxas-part-i

[BenWibking]

Perl one-liner from Weiqun:

$ perl -ne 'if (/Compiling entry function (.*?) for '\''sm_/) { print "$1 "; } elsif (/Used (\d+) registers/) { print " $1\n"; }' foo.txt
'searchkernelxyz(octree, int*, double, int, double*, double*, double*)'  256
'searchkernel(octree, int*, double, int, double*, double*, double*)'  46

$ cat foo.txt
111

222
ptxas info    : Compiling entry function 'searchkernelxyz(octree, int*, double, int, double*, double*, double*)' for 'sm_20'
ptxas info    : Function properties for searchkernel(octree, int*, double, int, double*, double*, double*)
    72 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info    : Used 256 registers, 176 bytes cmem[0], 16 bytes cmem[14]
sss
ddd
ptxas info    : Compiling entry function 'searchkernel(octree, int*, double, int, double*, double*, double*)' for 'sm_20'
ptxas info    : Function properties for searchkernel(octree, int*, double, int, double*, double*, double*)
    72 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info    : Used 46 registers, 176 bytes cmem[0], 16 bytes cmem[14]
xxxx
yyy