[D3D12] RTV pipeline creation: depth_bias_slope_scaled - memory leak/performance issue

Validation

[X] I've read the FAQ.
[X] The Xenia build used is from the master branch (not MLBS/AlexVS/Canary/pull requests, etc.)
[X] This issue isn't for tech support (help with Xenia).
[X] If this issue occurs in a specific game, I've done analysis to locate the faulty subsystem of the emulator and a potential reason in it.
[X] I've checked if this issue hasn't already been reported.
[X] My device meets the minimum requirements: https://github.com/xenia-project/xenia/wiki/Quickstart#system-requirements
[X] (If building) I have read the building doc: https://github.com/xenia-project/xenia/blob/master/docs/building.md

Describe what's going wrong

GetCurrentStateDescription in pipeline_cache.cc creates PipelineRuntimeDescription which in case of RTV uses GetPreferredFacePolygonOffset to generate depth_bias_slope_scaled. This is a problem because for some calls this is executed with different register values which causes to generate multiple pipelines for the same shader that will never be use again.

Test cases: https://github.com/xenia-project/game-compatibility/issues/52 https://github.com/xenia-project/game-compatibility/issues/21

In case of first test case values (from registers) are quite small: 0.0001 - 0.0000001 For second one they're quite huge: 10-150

Describe what should happen

No idea

If applicable, provide a callstack here, especially for crashes

No response

If applicable, upload a logfile and link it here

No response

xenia-project / xenia