Re: LLVM "continuous" coverage profiling hack^Wpatch

From: Poul-Henning Kamp <phk_at_phk.freebsd.dk>
Date: Mon, 19 Oct 2026 06:06:42 UTC 
--------
Enji Cooper (yaneurabeya) writes:

> I'm a bit puzzled how this is different from the already 
> documented/implemented --coverage option in compiler_rt, 

It's the exact same code, but the counters are mmap(2)'ed, so they
survive core-dumps, setuid(), chroot(2) and other "weird cases"
where atexit(3) does not do what you expect.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.