Message-ID: <5d1b5880-9bdc-4b04-81dc-341df7b02177@web.de>
Date: Wed, 18 Sep 2024 13:01:35 +0200
From: Frank Scheiner <frank.scheiner@....de>
To: torvalds@...ux-foundation.org
Cc: Tomáš Glozar <tglozar@...il.com>,
 linux-kernel@...r.kernel.org, Linux-Arch <linux-arch@...r.kernel.org>,
 linux-ia64@...r.kernel.org, t2@...de.org, gentoo-dev@...ts.gentoo.org
Subject: Re: Linux 6.11

Dear all,

here comes the usual update on Linux/ia64 for v6.11:

The v6.11 cycle was cut short by vacation, which thankfully didn't
create problems and the test results looked good for all tested machines
(up to and including v6.11-rc7 - more on that later). This cycle also
saw the switch from Binutils 2.42 to 2.43.1. And like for the v6.10
cycle (see last update on [1] for details) each release candidate and
release in this cycle was always built with the latest gcc-15 snapshot
(w/LRA enabled) available.

[1]:
https://lore.kernel.org/lkml/8dcb5d36-fc38-46f3-bb97-0e4e335c313b@online.de/

Unfortunately I was a little too optimistic at the end and changed my
used kernel config at the last minute (i.e. for the v6.11 release) to
help with testing on Tomas' rx2620 and OS installation with btrfs
support and 64 KB page size. But that change must have somehow broken
successful operation inside the initrd on my rx4640. Well, the rx4640 is
not technically broken with v6.11, because it continues to work when
using a kernel based on Tomas' tree ([2]) on top of v6.11 with the very
same config. And my per-release (candidate) branches also showed no
regressions for the other machines (rx2620. rx2660, rx6600, rx2800 i2,
not to forget hp-sim, but uses a different config) with the exact same
initrd and v6.11. So maybe I should retest all real machines with a
kernel made with my older kernel config to be sure.

[2]: https://github.com/linux-ia64/linux-ia64/

[3]: https://github.com/johnny-mnemonic/linux-ia64/

****

Again no new additions to the selection of Linux distributions for ia64
this time - or at least I'm not aware of any. In the meantime Gentoo
dropped support for ia64, too, which honestly came as a surprise. I used
Gentoo some years ago for some time starting on a rx2660. I later also
used it on a rx2800 i2, because their ia64 kernel was the only one that
could run on the latter back then. I never found out exactly what was
responsible for this oddity, but fortunately it was somehow solved in
mainline some time later. So, farewell Gentoo/ia64. You were very useful
in times, thanks for that and to all people involved to make that happen
in the past.

This leaves T2/SDE and EPIC Slack as possibly the last remaining active
Linux distributions supporting ia64 machines for the time being. But
that can change easily, like it happened for MIPS and Alpha, if we look
at the latest headline over at gentoo.org:

https://www.gentoo.org/news/2024/09/11/Improved-MIPS-and-Alpha-support.html

****

One more thing:

So far we missed a public place were people can find current information
about Linux/ia64. This gap is now closed by epic-linux.org ([4]). There
you'll find a collection of resources around ia64, including news,
current code, our CIs, our testing effort, Linux distributions
supporting ia64, supported machines and links to other sites related to
ia64. And possibly even more in the future.

[4]: http://epic-linux.org/

****

Thank you all for your hard work on Linux!

Cheers,
Frank et al

Powered by blists - more mailing lists