Re: [linux-next:master] [lockref] d042dae6ad: unixbench.throughput -33.7% regression

[Linus Torvalds]

On Tue, 2 Jul 2024 at 13:33, Mateusz Guzik <mjguzik@xxxxxxxxx> wrote:
>
> If you are politely by lkml standards suggesting I should probably drop
> the idea due to unforseen complexities

Oh, absolutely not. I'd love to see how nasty - or not nasty - the
patch would end up being. I think it would be very interesting.

I'm just explaining why _I_ never got around to it.

I do think that the 'stat()' family of syscalls are some of the most
critical system calls out there. And while I suspect the benchmarks
that stat the same file over and over in parallel are worthless, if we
can do 'stat()' without ever even dirtying the dentry at all, that
would be absolutely lovely.

           Linus