ibc: prevent invalid Height from being deserialised

[mina86]

Change revision_height from u64 to NonZeroU64 to make invalid heights
(ones where revision_height is zero) impossible to represent. This
makes borsh and serde deserialisation fail when given invalid height
(while in the past they would construct an invalid object).

---

PR author checklist:

• Added changelog entry, using unclog.
• Added tests.
• Linked to GitHub issue.
• Updated code comments and documentation (e.g., docs/).
• Tagged one reviewer who will be the one responsible for shepherding this PR.

Reviewer checklist:

• Reviewed Files changed in the GitHub PR explorer.
• Manually tested (in case integration/unit/mock tests are absent).

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (553bc79) 67.65% compared to head (75f64a9) 67.75%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #932      +/-   ##
==========================================
+ Coverage   67.65%   67.75%   +0.09%     
==========================================
  Files         130      130              
  Lines       16415    16401      -14     
==========================================
+ Hits        11106    11112       +6     
+ Misses       5309     5289      -20     

Files	Coverage Δ
crates/ibc/src/core/ics02_client/height.rs	92.68% <100.00%> (+1.43%)	⬆️

... and 7 files with indirect coverage changes

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

[mina86]

@Farhad-Shabani

[rnbguy]

Thanks for the PR @mina86.

We intend to change this behavior via #603. To summarize, ics-02 allows zero for heights. It is the ics-07 (tmclient) that doesn't allow zero for heights. Hence this validation should be client-specific.

But if this is blocking you, please open an issue and we will consider this PR.