Re: [GIT PULL] MM updates for 5.20-rc1

[Linus Torvalds <torvalds@linux-foundation.org>]

On Wed, Aug 3, 2022 at 1:35 PM Andrew Morton <akpm@linux-foundation.org> wrote:
>
> - memcg debug/visibility work from Roman Gushchin

Well, not just memcg.

There's that new CONFIG_SHRINKER_DEBUG too.

Which looks like yet another "people add Kconfig options incorrectly".

We don't make new features "default y" unless there's some truly
pressing reason for it (ie "99.9% of all people will want this" or "we
used to always compile this in, now it's optional").

And shouldn't that thing be under the DEBUG_VM heading anyway?

I removed the completely bogus and inappropriate "default y".

People, please stop doing that 'default y' thing.

I realize that everybody always thinks their own code is *sooo*
important that it should be enabled everywhere, but if we've gone 30
years without it in the past, maybe it's not so critical after all,
and maybe people shouldn't randomly enable it unless they WANT it. Ok?

The main reasons for 'default y' tend to be literally

 - this used to be unconditional, now we have a config variable for
it, so let's make it 'default y' so that people don't suddenly lose
functionality

 - this feature truly does cure cancer

 - this isn't actually a feature, but is a gating question to other
features that you may want to just shut up

That last case is mainly used by the network driver subsystem, where
it asks 'Do you want to see drivers by vendor Xyzzy?', and it defaults
to indeed show those options. But admittedly that network driver case
is also _partly_ guided by that first case, ie it has often been
something where a group of drivers were moved to be under a "do you
care about this vendor" situation.

                     Linus