@ptesarik@fosstodon.org avatar

ptesarik

@ptesarik@fosstodon.org

Open-source enthusiast since Slackware 2.3 (Linux kernel 1.2.8).

This profile is from a federated server and may be incomplete. Browse more on the original instance.

kernellogger, to linux
@kernellogger@fosstodon.org avatar

Looks like 6.10 will drop support for a few old machines:

https://lore.kernel.org/all/20240503081125.67990-1-arnd@kernel.org/

"'[…]
alpha: remove DECpc AXP150 (Jensen) support
alpha: sable: remove early machine support
alpha: remove LCA and APECS based machines
alpha: cabriolet: remove EV5 CPU support
alpha: drop pre-EV56 support
[…]
72 files changed, 166 insertions(+), 4545 deletions(-)'"

These changes from @arnd since today can be found in linux-next, too.

ptesarik,
@ptesarik@fosstodon.org avatar

@kernellogger @arnd Make Kernel Small Again!

jacob, to random
@jacob@jacobian.org avatar

There's something profoundly amazing about the fact that I've been able to work in computers for 25 years and still basically know almost nothing about how networking actually works.

ptesarik,
@ptesarik@fosstodon.org avatar

@MichaelTBacon @lukasa @jacob @glyph If you consider only TCP and UDP, you haven't gone deep enough in the networking stack. What about FCoE for example?

ptesarik,
@ptesarik@fosstodon.org avatar

@MichaelTBacon @lukasa @jacob @glyph Ack. Getting people to stop suggesting to tunnel raw USB frames as hex dump wrapped in JSON over HTTPS would be a lot…

kernellogger, to linux
@kernellogger@fosstodon.org avatar

🥳 My text on verifying bugs and bisecting regressions made it to the mainline for 6.9: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/plain/Documentation/admin-guide/verify-bugs-and-bisect-regressions.rst

You can still submit improvements by editing this document: https://docs.google.com/document/d/1Im7SPK0j6PUGQTSGZyCTSQv8h3S51EYsZuRRdyhfzto/edit?usp=sharing

Note, the rendered version mentioned in the text is not yet in place, but will be in a week or two.

ptesarik,
@ptesarik@fosstodon.org avatar

@kernellogger done

ptesarik, to random
@ptesarik@fosstodon.org avatar

Oh, Linux Kernel Programming 2nd edition is out!

https://github.com/PacktPublishing/Linux-Kernel-Programming_2E?tab=readme-ov-file

ptesarik, to random
@ptesarik@fosstodon.org avatar

Have you noticed that all public CNAs assign only negative numbers (like CVE-2017-5754)? So, what's with the positive CVE numbers (CVE+2024+xxxx)? They are assigned by secret CNAs! To hide even the fact there are any such CNAs, the positive-number CVEs must never be published.

Um, this conspiracy theory is even more stupid than flat Earth. So it may catch up…

craftyguy, to linuxphones
@craftyguy@freeradical.zone avatar

Do you know a lot about networking on Linux and want to help fix a multi-homed DNS routing issue experienced by all distros?

https://gitlab.com/postmarketOS/pmaports/-/issues/1430

Solving this directly benefits users of , , and others!

ptesarik,
@ptesarik@fosstodon.org avatar

@erikk @craftyguy Oh, are they using glibc? Yes, then this should work.

ptesarik,
@ptesarik@fosstodon.org avatar

@craftyguy Sigh…

The bug is apparently related to default address selection, which is described in RFC 6724. Section 10.3 of this document talks about configuring preference for IPv6 or IPv4, and section 10.5 talks about configuring a multi-homed site. Both assume that the selection is configurable. However, the configuration mechanism itself is implementation-specific. All I wanted to say is that the /etc/gai.conf file suggested by @erikk is specific to glibc.

https://datatracker.ietf.org/doc/html/rfc6724

ptesarik,
@ptesarik@fosstodon.org avatar

@craftyguy @erikk Sure. Unless you remove the IPv6 route completely, the BSD socket layer will always allow applications to use it.

So, what is it that you want? Prevent any use of the mobile connection except for IPv6-only MMS? In other words, should access to https://www.doesnotwork.eu/ fail from the browser if your phone is connected to an IPv6-capable mobile and an IPv4-only 802.11 network?

ptesarik,
@ptesarik@fosstodon.org avatar

@craftyguy @erikk To make it clear I'm trying to be constructive here. If the answer to my question is: “Yes,” then you want to run your MMS application in its own network namespace which has the IPv6 route. The system-wide namespace does not have it.

kernellogger, to linux
@kernellogger@fosstodon.org avatar

Submitted my text "How to verify [ ] bugs and bisect regressions"[1] for review and inclusion into the 's documentation[2].

Any feedback on it would be much appreciated!

[1] https://www.leemhuis.info/files/misc/How%20to%20verify%20bugs%20and%20bisect%20regressions%20%e2%80%94%20The%20Linux%20Kernel%20documentation.html

[2] https://lore.kernel.org/linux-doc/bf1d2eba0d291ff583e01b5985a0dec248eaf27a.1708072870.git.linux@leemhuis.info/

ptesarik,
@ptesarik@fosstodon.org avatar

@kernellogger “overall on average it will often just take about 10 to 15 minutes to compile each kernel on commodity x86 machines”

Indeed, and then it will take 5+ minutes to boot an enterprise-grade server. 😆

ptesarik, to random
@ptesarik@fosstodon.org avatar

bomb dropped^W^Wpatches submitted

Finally, my sandbox mode patches have been submitted.

Generic part:
https://lore.kernel.org/lkml/20240214113035.2117-1-petrtesarik@huaweicloud.com/

x86_64 implementation:
https://lore.kernel.org/lkml/20240214113516.2307-1-petrtesarik@huaweicloud.com/

kernellogger, (edited ) to linux
@kernellogger@fosstodon.org avatar

An engineer's guide to upgrades – slides and recording of this @fosdem talk from Cloudflare employee Ignat Korchagin are now online:

-Page with abstract: https://fosdem.org/2024/schedule/event/fosdem-2024-3370-an-engineer-s-guide-to-linux-kernel-upgrades/

Slides: https://fosdem.org/2024/events/attachments/fosdem-2024-3370-an-engineer-s-guide-to-linux-kernel-upgrades/slides/21307/Linux_Kernel_Upgrades_budIIgJ.pdf

Video: https://video.fosdem.org/2024/k1105/fosdem-2024-3370-an-engineer-s-guide-to-linux-kernel-upgrades.av1.webm (AV1/opus) and https://video.fosdem.org/2024/k1105/fosdem-2024-3370-an-engineer-s-guide-to-linux-kernel-upgrades.mp4 (mp4)

ptesarik,
@ptesarik@fosstodon.org avatar

@kernellogger @fosdem One (uncommon) way to make a system with >= 30 day uptime not vulnerable is live patching.

Ah, yes, sure, that was not the point here.

hrw, to random
@hrw@society.oftrolls.com avatar

What to watch in IC 6502 train? Some 6502 related videos of course!

ptesarik,
@ptesarik@fosstodon.org avatar

@hrw For my part, I believe a MOS 6502 can deliver more reliable locomotive control than a Newag PLC…

ptesarik, to random
@ptesarik@fosstodon.org avatar

@vbabka Did you know there's an O'RLY book for you?

b0rk, (edited ) to random
@b0rk@jvns.ca avatar

if you're an infrequent command line user -- what text editor do you use if you need to occasionally edit a file on the command line (other than vim/emacs)?

curious about what people use to edit a git commit message etc

if you picked 'other', I'd love to hear what you do in the replies!

ptesarik,
@ptesarik@fosstodon.org avatar

@pamelafox @b0rk What level is enough? Some people around me have only grasped that vi has two modes:

  1. flashing and beeping mode
  2. text-mangling mode
ptesarik, to random
@ptesarik@fosstodon.org avatar

If anyone (@kernellogger?) is curious, I probably “broke” resume on my laptop when I replaced the secondary SATA disk. The timing changed slightly, and I'm hitting a deadlock that shouldn't happen.

https://marc.info/?l=linux-pm&m=169634614718495&w=2

ptesarik,
@ptesarik@fosstodon.org avatar

@kernellogger It does hit this codepath, and no, it's still broken in rc4 (which contains the patch series).

kernellogger, to linux
@kernellogger@fosstodon.org avatar

: I have to perform a bisection on my Thinkpad T14s G1 (AMD) with Fedora's full-blown distro config, as the problem does not occur with a config trimmed by localmodconfig. 😟

[1] s2idle resume broke with 6.6-rc

ptesarik,
@ptesarik@fosstodon.org avatar

@kernellogger Oooh. I'm performing a post-mortem crash dump analysis using drgn on my Thinkpad E595 running . Broken resume, too. But mine broke with 6.5 already…

ptesarik,
@ptesarik@fosstodon.org avatar

@kernellogger Yeah, I realized I was able to trigger a crash dump with SysRq+c, even though my screen was blank or scrambled.

Congrats on your finding. My case is different. I know that it is somehow related to my NVMe disk. I can see blk_queue_enter() waiting for an event which never arrives.…

It only happens on battery power, so I'm going to check if it also fails after turning off autosuspend.

ptesarik,
@ptesarik@fosstodon.org avatar

@kernellogger That's it! My laptop has just survived 10 suspend-resume cycles on battery power with autosuspend off, but it hanged on the first attempt on AC power with autosuspend enabled.

Is it too cynical to expect that (some) distributions would fix the issue by adding the following command to their suspend hook?

echo on > /sys/bus/scsi/devices/0:0:0:0/power/control

ljs, to random
@ljs@social.kernel.org avatar

All I want is somebody to invent a kind of time compression machine where I can sleep for 12 hours and the clocks don't change.

IS THAT TOO MUCH TO ASK??

ptesarik,
@ptesarik@fosstodon.org avatar

@imrehg @ljs Or travel west at Earth's rotational speed at the given latitude.

karolherbst, to random
@karolherbst@chaos.social avatar

😓 😢

In case you don't hear anything from me this week, I'm going to be a total flop this entire week

ptesarik,
@ptesarik@fosstodon.org avatar

@kernellogger @wagi @karolherbst Yep. A friend of mine says it's not even worth taking out his bike if the ride is under 50 km.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • megavids
  • cubers
  • magazineikmin
  • GTA5RPClips
  • khanakhh
  • InstantRegret
  • Youngstown
  • mdbf
  • slotface
  • thenastyranch
  • everett
  • osvaldo12
  • kavyap
  • rosin
  • anitta
  • DreamBathrooms
  • Durango
  • modclub
  • ngwrru68w68
  • vwfavf
  • ethstaker
  • tester
  • cisconetworking
  • tacticalgear
  • Leos
  • provamag3
  • normalnudes
  • JUstTest
  • All magazines
    •