Andy Pan

I'm not sure I agree with your conclusion. Despite the similarity in inter-process communication, Pipes and FIFOs are fundamentally different: Pipes are created in memory, and their existence is transient...

> I think this is caused by an incomplete fix for #24164. We avoid adding FDs of type `S_IFIFO` to the poller if they are marked as `kindOpenFile`, but the...

Ping @azat

I think this PR is ready to land. @azat

@madolson Any updates?

> Since we don't automatically test on FreeBSD: [valkey-io/valkey/actions/runs/8994018097](https://github.com/valkey-io/valkey/actions/runs/8994018097) Thanks, I see it's succeeded. Just out of curiosity, why didn't we set up a macOS runner that would run by...

Hi @madolson, any new thoughts on this?

> @panjf2000 It makes sense to me, I just haven't had a sense to walk through it and convince myself the code is correct. Hopefully tomorrow? Sure, take your time....

> Do we expect this to make the code more performant? This PR is expected to conserve plenty of extra system calls to `kevent(2)`, which should have a significant benefit...

> > which should have a significant benefit to the performance. > > Would you mind using `valkey-benchmark -r 10000000 -d 100 -n 10000000` with and without to change to...