search

rtic-scope / cargo-rtic-scope

Non-intrusive ITM tracing/replay toolset for RTIC programs with nanosecond timestamp accuracy.
16 stars 4 forks source link

Stabilize STLink probe source #18

Open tmplt opened 3 years ago

tmplt commented 3 years ago

As mentioned in https://github.com/rtic-scope/cargo-rtic-scope/commit/f2a73dabf540baa20642e0ed9345c79f6397a06e, the new CMSIS-DAP implementation is buggy. Investigate the cause and fix.

tmplt commented 3 years ago

Tangent: an STLink is not a CMSIS-DAP. DAPSource should be further abstracted to ProbeSource. We do not care what kind of probe is ultimately used, only that probe-rs supports it.

tmplt commented 3 years ago

Mostly we just get an invalid sync size. This could just be a variance from the ITM standard.

tmplt commented 3 years ago

Perhaps a feature-set of quirks should be maintained upstream?

tmplt commented 3 years ago

This would be an upstream issue in probe-rs. Presumably the amount of work is non-trivial. Postponing.

tmplt commented 2 years ago

I theorize that the root cause of this issue is transient configuration noise seen on the SWO pin. So far I've only been able to reproduce this issue on a Nucleo, where the SWO pin feeds into an on-chip debugger that is interfaced with via USB. Verifying if this indeed is the cause should be trivial, but I do not have a Nucleo at hand at the moment.