Explicitly reject routes that double-back

[shaavan]

Resolves #2215

• If a path within a route passes through the same channelID twice, that shows the path is looped and will be rejected by nodes.
• Add a check to explicitly reject such payments before trying to send them.

[tnull]

Thanks for taking a stab at this!

[shaavan]

Updated from pr2741.01 -> pr2741.02 (diff)
Addressed @tnull suggestion

• Modified the code to use nested for loops instead of a HashSet. Since the number of routeHops will always be very small (<20 for a path), this approach optimized the space complexity over time.

[codecov-commenter]

Codecov Report

Attention: 2 lines in your changes are missing coverage. Please review.

Comparison is base (870a0f1) 88.55% compared to head (a06d158) 88.52%.
Report is 3 commits behind head on main.

Files	Patch %	Lines
lightning/src/ln/functional_tests.rs	91.30%	1 Missing and 1 partial ⚠️

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2741      +/-   ##
==========================================
- Coverage   88.55%   88.52%   -0.04%     
==========================================
  Files         113      113              
  Lines       89330    89351      +21     
  Branches    89330    89351      +21     
==========================================
- Hits        79110    79097      -13     
- Misses       7849     7877      +28     
- Partials     2371     2377       +6     

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

[tnull]

Pretty much looks good to me. Could we add some test coverage for it though?

You could add a tiny test case checking that we receive the correct APIError if we try to use a looping route. The setup should be similar to, e.g., fn test_update_add_htlc_bolt2_sender_cltv_expiry_too_high() in functional_tests.rs.

[shaavan]

Updated from pr2741.02 -> pr2741.03 (diff)
Addressed @tnull comments

• Added test coverage for the added code.
• Also modified the unwrap_send_error!() macro to handle PaymentPathFailure case.

Thanks, @tnull, for the suggestion and for providing a reference to take inspiration for writing up the test.

[tnull]

LGTM