Zalathar ab786d3b98 coverage: Eliminate more counters by giving them to unreachable nodes
When preparing a function's coverage counters and metadata during codegen, any
part of the original coverage graph that was removed by MIR optimizations can
be treated as having an execution count of zero.

Somewhat counter-intuitively, if we give those unreachable nodes a _higher_
priority for receiving physical counters (instead of counter expressions), that
ends up reducing the total number of physical counters needed.

This works because if a node is unreachable, we don't actually create a
physical counter for it. Instead that node gets a fixed zero counter, and any
other node that would have relied on that physical counter in its counter
expression can just ignore that term completely.
2025-02-13 13:45:53 +11:00
..
2024-12-13 00:04:56 +00:00
2025-01-09 13:55:50 +01:00

The tests in this directory are shared by two different test modes, and can be run in multiple different ways:

  • ./x.py test coverage-map (compiles to LLVM IR and checks coverage mappings)
  • ./x.py test coverage-run (runs a test binary and checks its coverage report)
  • ./x.py test coverage (runs both coverage-map and coverage-run)

Maintenance note

These tests can be sensitive to small changes in MIR spans or MIR control flow, especially in HIR-to-MIR lowering or MIR optimizations.

If you haven't touched the coverage code directly, and the tests still pass in coverage-run mode, then it should usually be OK to just re-bless the mappings as necessary with ./x.py test coverage-map --bless, without worrying too much about the exact changes.