On Fri, 1 Dec 2023 21:06:22 +0100
David Hildenbrand <david@xxxxxxxxxx> wrote:
> But
>
> > Cc: stable@xxxxxxxxxxxxxxx
>
> stable, really? Why?
>
> > Fixes:fb3d824d1a46c ("mm/rmap: split page_dup_rmap() into page_dup_file_rmap() and page_try_dup_anon_rmap()")
>
> and does it even fix a real bug?
As a performance person, who measures likely and unlikely results (the
ftrace ring buffer was sped up by over 50% with strategically placed
likely/unlikely annotation). I find this to be a real bug, and something I
would want backported to the kernels we maintain in ChromeOS (which uses
upstream stable kernels).
-- Steve
