From Newsgroup: comp.os.linux.advocacy

The new Linux kernel-6.18 has a new option: PERSISTENT_HUGE_ZERO_FOLIO

The help text states:

==================================================================
Enable this option to reduce the runtime refcounting overhead
of the huge zero folio and expand the places in the kernel
that can use huge zero folios. For instance, block I/O benefits
from access to large folios for zeroing memory.

With this option enabled, the huge zero folio is allocated
once and never freed. One full huge page's worth of memory shall
be used.

Say Y if your system has lots of memory. Say N if you are
memory constrained. =================================================================

What does this mean? I have never encountered the term "folio."

Since I have 32G memory I will enable it, but any comments will
be most appreciated.

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

L Thorpe wrote:

huge zero folio

I don't stay as up to date with kernel development as I once did, but ...

<https://lwn.net/Articles/1033058>
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

On Sun, 7 Dec 2025 17:40:18 +0000, Andy Burns wrote:

L Thorpe wrote:

huge zero folio

I don't stay as up to date with kernel development as I once did, but ...

<https://lwn.net/Articles/1033058>

So mapping a huge page full of zeros during context switching is more
efficient than just writing zeros to the page. That makes sense.

But I wonder how the distros will handle this. There is no way to determine memory size prior to installation. As usual, the distros will default to
an inefficient common ground.

It's ALWAYS better to build from scratch.


--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

L Thorpe <lt666@sixsixsix.net> wrote:

What does this mean? I have never encountered the term "folio."

Since I have 32G memory I will enable it, but any comments will
be most appreciated.

Why are you rolling your own kernel if you don't have the knowledge to
research such things yourself? The people doing Kernels in the
distributions are expertes. Why do you distrust them? And why do you
not distrust the people building your libc, your toolchain, your perl
and python?

Greetings
Marc
-- ---------------------------------------------------------------------------- Marc Haber | " Questions are the | Mailadresse im Header Rhein-Neckar, DE | Beginning of Wisdom " |
Nordisch by Nature | Lt. Worf, TNG "Rightful Heir" | Fon: *49 6224 1600402
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

On Mon, 08 Dec 2025 19:37:42 +0100, Marc Haber wrote:

Why are you rolling your own kernel if you don't have the knowledge to research such things yourself?

Because "rolling" is a learning process and that's something which
you have obviously abandoned.

The people doing Kernels in the
distributions are expertes.

Says who? Says you?

But I don't use a distribution and thus I must build my own
kernel.

Why do you distrust them?

It is not a question of trust. It is the fact that distros build
only for the "lowest common denominator." Thus, all distros are
un-optimized, inefficient, crippled

And why do you
not distrust the people building your libc, your toolchain, your perl
and python?

I don't. I build glibc, the toolchain, Perl, and Python all by
myself.

You should try it. But you can't, and that's why you defend the disrtos
in an attempt to defend only yourself.
--
Gentoo: the only road to GNU/Linux freedom and perfection.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.os.linux.advocacy

On 12/8/25 3:26 PM, Lester Thorpe wrote:

Gentoo: the only road to GNU/Linux freedom and perfection.

Where perfection aspires to be so stripped of essential material that
even limbs are too detailed, give me a stump for a hand, says Gentoo
compared to Mint or Debian, give me powerlessness in the name of idiotic vanity!
--
Joel W. Crump
--- Synchronet 3.21a-Linux NewsLink 1.2