dialects: (snitch_stream) reverse bounds and strides in snitch_stream

[superlopuh]

This PR brings the documentation and interpreter implementation of snitch_stream stride patterns in line with the snitch_stream lowering to snitch.

This is a PR with only a functional change in the interpreter, but with consequences on upcoming lowerings from memref_stream to snitch_stream. When I first added snitch_stream and the corresponding interpreter functions, I was under the impression that the snrt functions took bounds and strides from the outermost loop inwards, but it's the reverse. The distinction is now somewhat arbitrary, since we are planning to link to snrt any more, and could actually implement the lowering from snitch_stream to snitch with the inverse order of bounds. I think that we should stick to the current order of bounds, implement the complete lowering first, and then see whether switching the order around would be an improvement.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (c8f19b9) 90.13% compared to head (1eb945f) 90.17%.
Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2139      +/-   ##
==========================================
+ Coverage   90.13%   90.17%   +0.03%     
==========================================
  Files         303      305       +2     
  Lines       36886    37041     +155     
  Branches     5484     5515      +31     
==========================================
+ Hits        33249    33402     +153     
- Misses       2859     2860       +1     
- Partials      778      779       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.