On Tue, Nov 19, 2024 at 6:53 PM Theodore Ts'o <tytso@xxxxxxx> wrote: > > On Tue, Nov 19, 2024 at 10:45:52AM +0100, Mateusz Guzik wrote: > > > > On my v1 Jan remarked 1.5% is not a particularly high win questioning > > whether doing this makes sense. I noted the value is only this small > > because of other slowdowns. > > Do you have a workload in mind which calls readlink() at sufficiently > high numbers such that this would be noticeable in > non-micro-benchmarks? What motiviated you to do this work? > I'm just messing about here. Given the triviality of the patch I'm not sure where the objection is coming from. I can point a finger at changes made by other people for supposed perf gains which are virtually guaranteed to be invisible in isolation, just like this one. Anyhow, I landed here from strlen -- the sucker is operating one byte at a time which I was looking to sort out, but then I noticed that one of the more commonly executing consumers does not even need to call it. -- Mateusz Guzik <mjguzik gmail.com>
