Show iterations grouped by GPU thread and block

[vasslitvinov]

Exends the test of @gpu.itersPerThread with a utility function that groups loop iterations by thread and block that execute them. For example:

numIndices 32  itersPerThread 5  blockSize 3  gridSize 3  block
bl th | iters
 0  0 |  1  2  3  4  5 .
 0  1 |  6  7  8  9 10 .
 0  2 | 11 12 13 14 15 .
 1  0 | 16 17 18 19 20 .
 1  1 | 21 22 23 24 25 .
 1  2 | 26 27 28 29 30 .
 2  0 | 31 32 .
 2  1 | .
 2  2 | .

numIndices 32  itersPerThread 5  blockSize 3  gridSize 3  cyclic
bl th | iters
 0  0 |  1  8 15 22 29 .
 0  1 |  2  9 16 23 30 .
 0  2 |  3 10 17 24 31 .
 1  0 |  4 11 18 25 32 .
 1  1 |  5 12 19 26 .
 1  2 |  6 13 20 27 .
 2  0 |  7 14 21 28 .
 2  1 | .
 2  2 | .

Also extends sanity checks.

Test enhancement. Not reviewed.