Change the default smoother to two-filters smoother

[beomki-yeo]

This will fix the issue on bad matching rate between CPU and GPU, reported by @stephenswat and @krasznaa

[beomki-yeo]

Yeah it's not working for the drift chamber where the navigation can get unstable.
Fitting with wire chamber won't be OK until we implement (1) direct navigator and (2) deterministic annealing filter.
The current tests with the wire chamber should not be taken seriously.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[krasznaa]

Okay... Let's treat this as a quick fix to the observed issue.

What I'll (or somebody else will) do is to introduce traccc::opts::track_finding, to make this (and the other track finding options) configurable from the command line. As it really should be...

[beomki-yeo]

This kills the performance a bit but it is expected as propagation runs twice..

[krasznaa]

This kills the performance a bit but it is expected as propagation runs twice..

"A bit" may be a bit of an understatement unfortunately. 😭 Lately our code has slowed down a lot... Something to be discussed in tomorrow's meeting.

[beomki-yeo]

It's a drop from correcting the mathematics so... take it as a necessary evil..

[beomki-yeo]

And KF will be trivial once we implement the ambinsolver after CKF.

If i am correct, KF is used in refitting only but it is not a part of ITk chain