feat: enable println via trace

[mooori]

• Adds support for println-style output via a new TRACE_PRINT_LN trace event, with decoding of [address, length] from the VM stack and reading the corresponding UTF-8 bytes from memory.
• Refactors trace handling so callbacks receive full ProcessorState instead of just the cycle, which lets the host capture trace events and read string-bytes from memory.
• Persists printed output through the executor and ExecutionTrace, exposing it as printed_lines() keyed by cycle.
• Extracts reusable memory-byte helpers and updates existing memory reads to use them.

Companion PR in the compiler: 1077

Panics vs propagating errors

In the first pass I went with panicking early instead of propagating errors because:

• It's simpler and the compiler is setting up println
• Glancing at surrounding code I saw panics instead of errors too

Though this PR might take panicking too far - let me know what you think.

[mooori]

Temporary fix to make this branch work in the compiler repo.

This is needed as the version on next is behind the released version (0.6.1). Probably main still needs to be merged into next after the latest release.

[mooori]

This does not yet handle displaying printed lines. So far it only makes them available programmatically (example).

Marking it as ready for review because I would like to:

• Get the internals (this PR) done first.
• Get aligned on how to present printed lines in the UI before working on it. Especially since I'm not yet familiar with the internals of the different UI modes.
  • I've opened Display printed lines in the UI #55 as an initial proposal. Any Thoughts?