Re: performance regressions in 15.0

From: Mateusz Guzik <mjguzik_at_gmail.com>
Date: Mon, 08 Dec 2025 12:32:56 UTC 
On Sun, Dec 7, 2025 at 7:31 PM Rozhuk Ivan <rozhuk.im@gmail.com> wrote:
>
> On Sun, 7 Dec 2025 11:07:48 +0100
> Mateusz Guzik <mjguzik@gmail.com> wrote:
> > That is to say, if you have time, can you please benchmark a
> > statically linked clang vs clang which merely reverting the libprivate
> > change? Should be a small speed up on top.
>
> NO_SHARED - it was static link with libc and probably other system libs.
> Now clang splitted into set of libs, it is a bit different thing.
>

It is indeed a different thing and I am saying it also likely reduces
performance compared to what it can be without the recent libprivat
problem. It very much warrants benchmarking.