Fixup lock handling in limit enforcer and compile cache.

[mar-cf]

No description provided.

[mar-cf]

@jasnell did we land on a compile time catch for this?

[jasnell]

did we land on a compile time catch for this?

Not that I'm aware of. @danlapid was looking into it

[codspeed-hq]

Merging this PR will not alter performance

✅ 70 untouched benchmarks
⏩ 129 skipped benchmarks¹

---

_{Comparing mar/lockfix (70da2f6) with main (edacbb5)}

Footnotes

1. 129 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[danlapid]

@mar-cf @jasnell haven't had time to make it merge-ready
Feel free to pick it up:
https://github.com/capnproto/capnproto/pull/new/dlapid/lockSafety
https://github.com/cloudflare/workerd/pull/new/dlapid/lockSafety
https://gitlab.cfdata.org/cloudflare/ew/edgeworker/-/merge_requests/new?merge_request%5Bsource_branch%5D=dlapid%2FlockSafety

[jasnell]

@mar-cf ... you'll likely need to rebase this on main to deal with spurious test failure on widows.