- Arch-general - lists.archlinux.org

Race condition ? Issue with community/rtirq and probably core/systemd
by Ralf Mardorf 17 Oct '22

17 Oct '22

Hi, the machine was up for 19 days, hence in the meantime systemd was updated. Today I build 5.15.73.52.realtime1-1-rt-lts from AUR as provided by David, IOW without any changes done by me, but I didn't build in a clean chroot. Bevor that I run my own package of linux-rt 4.19.246_rt110-0.1000 without doing anything special to it, just CONFIG_AUDIT was disabled. Since last startup rtirq was updated, too, but the only changes were migrations from egrep to grep -E. In the past it happened very, very, very ... very seldom that after startup the PCI or PCIe card didn't get a high priority, which always was fixed by a single reboot. Today, when booting 5.15.73.52.realtime1-1-rt-lts 4 times, the first time snd_ice1 didn't get a high priority and the other times snd_ice1 as well as snd_hdsp didn't get a high priority. I doubt that it's related to the kernel. I suspect a systemd related race condition, but I might be completely mistaken. Any pointers are welcome. Regards, Ralf [rocketmouse@archlinux ~]$ grep RTIRQ_NAME_LIST..u /etc/rtirq.conf RTIRQ_NAME_LIST="usb snd_hdsp snd_ice1" [rocketmouse@archlinux ~]$ rtirq status PID CLS RTPRIO NI PRI %CPU STAT COMMAND 124 FF 90 - 130 0.0 S irq/16-ehci_hcd 184 FF 90 - 130 0.0 S irq/31-xhci_hcd 128 FF 89 - 129 0.0 S irq/23-ehci_hcd 88 FF 50 - 90 0.0 S irq/9-acpi 102 FF 50 - 90 0.0 S irq/29-ahci[000 116 FF 50 - 90 0.0 S irq/30-ahci[000 130 FF 50 - 90 0.0 S irq/8-rtc0 183 FF 50 - 90 0.0 S irq/1-i8042 324 FF 50 - 90 0.0 S irq/5-parport0 403 FF 50 - 90 0.0 S irq/32-mei_me 506 FF 50 - 90 0.0 S irq/18-i801_smb 572 FF 50 - 90 0.0 S irq/16-snd_hdsp 604 FF 50 - 90 0.0 S irq/34-i915 641 FF 50 - 90 0.0 S irq/35-snd_hda_ 644 FF 50 - 90 0.0 S irq/16-snd_ice1 16 TS - 0 19 0.0 S ksoftirqd/0 30 TS - 0 19 0.0 S ksoftirqd/1 [rocketmouse@archlinux ~]$ uname -rm 5.15.73.52.realtime1-1-rt-lts x86_64 [rocketmouse@archlinux ~]$ pacman -Q linux-rt linux-rt 4.19.246_rt110-0.1000 [rocketmouse@archlinux ~]$ pacman -Q rtirq rtirq 20220923-1 [rocketmouse@archlinux ~]$ pacman -Q systemd systemd 251.6-1 [root@archlinux rocketmouse]# rtirq stop Setting IRQ priorities: stop [ehci.hcd] irq=16 pid=124 pid=124 prio=50: OK. Setting IRQ priorities: stop [ehci.hcd] irq=23 pid=128 pid=128 prio=50: OK. Setting IRQ priorities: stop [xhci.hcd] irq=31 pid=184 pid=184 prio=50: OK. [root@archlinux rocketmouse]# rtirq start Setting IRQ priorities: start [ehci.hcd] irq=16 pid=124 prio=90: OK. Setting IRQ priorities: start [ehci.hcd] irq=23 pid=128 prio=89: OK. Setting IRQ priorities: start [xhci.hcd] irq=31 pid=184 prio=90: OK. Setting IRQ priorities: start [snd_hdsp] irq=16 pid=572 prio=85: OK. Setting IRQ priorities: start [snd_ice1] irq=16 pid=644 prio=80: OK. [root@archlinux rocketmouse]# rtirq status PID CLS RTPRIO NI PRI %CPU STAT COMMAND 124 FF 90 - 130 0.0 S irq/16-ehci_hcd 184 FF 90 - 130 0.0 S irq/31-xhci_hcd 128 FF 89 - 129 0.0 S irq/23-ehci_hcd 572 FF 85 - 125 0.0 S irq/16-snd_hdsp 644 FF 80 - 120 0.0 S irq/16-snd_ice1 88 FF 50 - 90 0.0 S irq/9-acpi 102 FF 50 - 90 0.0 S irq/29-ahci[000 116 FF 50 - 90 0.0 S irq/30-ahci[000 130 FF 50 - 90 0.0 S irq/8-rtc0 183 FF 50 - 90 0.0 S irq/1-i8042 324 FF 50 - 90 0.0 S irq/5-parport0 403 FF 50 - 90 0.0 S irq/32-mei_me 506 FF 50 - 90 0.0 S irq/18-i801_smb 604 FF 50 - 90 0.0 S irq/34-i915 641 FF 50 - 90 0.0 S irq/35-snd_hda_ 16 TS - 0 19 0.1 S ksoftirqd/0 30 TS - 0 19 0.1 S ksoftirqd/1

2 1

mkpkg - helping automate package rebuilds
by Genes Lists 14 Oct '22

14 Oct '22

Perhaps others might find this useful too. I'd like to give back a little to the arch community by sharing my mkpkg tool available on aur [1] and source is housed on github [2] where there is a detailed readme file. This tool helps to automate rebuilds when they're needed for reasons other than the package itself has updated. Rebuild triggers are defined as a list in the PKGBUILD. When a trigger is tripped then a rebuild is done. This helps when the actual package is unchanged but something the package uses changes in a way that makes a rebuild desirable or necessary: There are 3 kinds of triggers: 1 for files and 2 for packages. - a package has been installed since the last build - a package version has passed a specific threshold - a file is newer than the last build Some examples of build triggers. - 'python>minor' -> rebuild if major.minor vers > than at last build - 'xxx.service' -> rebuild if file has been updated since last build - 'gcc>major' -> rebuild if gcc major version newer than last build - 'some-lib' -> rebuild if more recent than last build - 'foo>last' -> rebuild if version greater than at last build - 'foo>=3.10' -> rebuild if version is >= 3.10 regards gene [1] https://aur.archlinux.org/packages/mkpkg [2] https://github.com/gene-git/Arch-mkpkg

1 0

Screen resolution while booting and on CLI
by bf 08 Oct '22

08 Oct '22

Hello, I have a problem with the screen resolution while booting the PC and afterwards. Part of the font is cut up at the left hand and upper side of the screen before booting into Arch Linux. And in Arch Linux itself (when the log in screen appears) the screen resolution is way too high, everything is located at the upper left corner of the screen. I don't use a graphical user interface so I can't use randr or anything else, I am on a Linux terminal. Is there a way to adjust screen resolution in BIOS? And is there a way to adjust screen resolution in a Linux terminal without xorg? I tried putting "vga=ask" into "GRUB_CMDLINE_LINUX_DEFAULT", but that didn't do anything. Regards Björn

2 4

[PATCH] broadcom-wl-dkms 6.30.223.271: Add a patch for Linux 6.0
by Masato TOYOSHIMA 05 Oct '22

05 Oct '22

[PATCH] broadcom-wl-dkms 6.30.223.271: Add a patch for Linux 6.0 https://gist.github.com/phoepsilonix/2edaf54aea36237e616b81b34b442a70

1 0

nvme drives
by Jude DaShiell 03 Oct '22

03 Oct '22

I have these on my new computer. I can partition the nvme scsi drive with parted but parted can't handle the other drive on the machine. Has archlinux got a separate tool to take control of nvme drives the way parted does? Jude <jdashiel at panix dot com> "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author, 1940) .

7 11

re: gave up on archinstall
by Jude DaShiell 03 Oct '22

03 Oct '22

It turns out archinstall works on highlighting the cursor so for espeak users who turn on highlighting inside of archinstall, their use of the keyboard puts them in the places they need to be to use archinstall. Highlighting isn't often needed but when it is needed it does work. Jude <jdashiel at panix dot com> "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author, 1940) .

2 3

gave up on archinstall
by Jude DaShiell 03 Oct '22

03 Oct '22

I use a screen reader and found archinstall locks the cursor on the top line of text on the display and scrolls the text on the screen whenever an action gets taken except for abort when it helpfully clears out of the way. I couldn't choose an audio server as a result even though choices got shown of pulseaudio and pipewire the cursor never made it into the options. All of this happened with latest release of archlinux. I think the network configuration was the only other one missing and I'm on ethernet so using dhcp usually allbeit managed by a network manager of some form. Jude <jdashiel at panix dot com> "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author, 1940) .

1 0

Enforced egrep warning by an if/then check
by Ralf Mardorf 02 Oct '22

02 Oct '22

Hi, I'm uncertain, if it makes sense to file a bug report against the xfce4-dev-tools or not. Actually upstream has got a reason to do this ;), but on Arch Linux it results in unwanted and annoying warnings :(. $ grep egrep /usr/bin/xdt-autogen -A3 if type egrep >/dev/null 2>&1; then EGREP=egrep else EGREP="grep -E" fi One time I reported an egrep related warning issue against the rtirq community package and upstream, but I didn't report it against e.g. AUR's ventoy-bin. However, the above if/then issue is something very special, since a workaround does cause the warnings in this particular case. The grep package does also provide a workaround. IOW two workarounds collide with each other :D. Regards, Ralf

1 0

Is the follwoing an arch packaging bug in the remmina package?
by u34＠net9.ga 02 Oct '22

02 Oct '22

$ ldd /usr/sbin/remmina | grep libsoup libsoup-2.4.so.1 => /usr/lib/libsoup-2.4.so.1 (0x00007fae45bde000) $ pacman -Qi remmina | grep libsoup $ I searched bugs.archlinux.org. Remmina has no bugs there. Is it an archlinux packaging bug? -- u34

5 7

Boot messages font issue
by u34＠net9.ga 26 Sep '22

26 Sep '22

There is a large, about 2M in size, screen shot at https://img.net9.ga/d9eYOM6/xbV+yBu/BootFontsIssue.jpg . Each text line is duplicated in a shifted manner because the picture was taken by a camera. The dupliacation of the lines is not the problem. It does not occour in reality. The problem are the rectangles enclosed in the red ellipse at the bottom of the picture. Have you encountred such a font issue? Any comments? This message is a partial duplication of the no replies Possibly only with UEFI or LVM, rectangles instead of letters at boot post at https://bbs.archlinux.org/viewtopic.php?id=279827 . -- u34

1 0