- Arch-general - lists.archlinux.org

Change BTRFS filesystem from R/O back to R/W
by Spencer Collyer 15 Nov '22

15 Nov '22

Hi, I've hit a problem with a BTRFS filesystem I have on my main machine. It's gone into read-only mode, I suspect because it ran out of disk space when doing a rebalance (it's got v little space left). Looking at the logs I see the following for BTRFS: Nov 15 09:09:35 selket kernel: BTRFS: Transaction aborted (error -28) Nov 15 09:09:35 selket kernel: BTRFS: error (device dm-1: state A) in btrfs_finish_ordered_io:3315: errno=-28 No space left Nov 15 09:09:35 selket kernel: BTRFS info (device dm-1: state EA): forced readonly Ironically I was going to shift a large chunk of data off this filesystem onto an external backup disk but hadn't gotten around to doing it. Digging around on the web I've seen suggestions involving unmounting the filesystem, remounting it R/W, removing any unwanted data, then continuing the rebalance. They generally seem to want to add another disk temporarily as well. I don't really want to go to the trouble of temporarily adding a secondary disk only to then remove it again. As I was already planning on moving a bunch of old data off to external backup, I'm wondering if the following would be safe to do: 1) Unmount the filesystem. 2) Remount it as R/W 3) Move data to the external disk 4) Resume the rebalance operation (is this required?) Does that look reasonable? Is there likely to be any problem with moving files off the filesystem when the rebalance operation failed? Thanks for you attention, Spencer

1 0

DKIM fail messages
by David Rosenstrauch 14 Nov '22

14 Nov '22

After posting a message to this list earlier today, I immediately received nearly a dozen DKIM fail messages, all being sent by the "OpenDMARC Filter" at various domains, and all saying that the DKIM fail reason was "signature verification failed". I raised this issue with support at my hosting company, who pointed out that a) my domain's DKIM records are set up correctly, and b) one of the headers on my email that got sent to the list indicated that Arch's mailman actually did successfully verify my domain's DKIM record when it received my email, before forwarding it on. Anyone know why these fail messages might be happening? Is this being caused by a misconfiguration in Arch's mailman installation? Or is this a misconfiguration of the individual OpenDMARC software installations at each of those domains? Thanks, DR

6 17

community/virtualbox 7.0.2-2 Windows 11 guest is unresponsive
by Ralf Mardorf 13 Nov '22

13 Nov '22

Hi, on my machine a Windows 11 guest still worked with [2022-11-01T17:48:40+0100] [ALPM] upgraded virtualbox-host-dkms (6.1.40-1 -> 7.0.2-1) [2022-11-01T17:48:40+0100] [ALPM] upgraded virtualbox (6.1.40-1 -> 7.0.2-1) [2022-11-01T17:48:40+0100] [ALPM] upgraded virtualbox-ext-oracle (6.1.40-1 -> 7.0.2-1) [2022-11-01T17:48:41+0100] [ALPM] upgraded virtualbox-guest-iso (6.1.40-1 -> 7.0.2-1) and a self build, rt patched linux 4.19.246_rt110-0.1000. A few settings are: 3D acceleration disable USB 3.0 enabled Network adapter attached to NAT The next time I run it (today) and it always does freeze, with [2022-11-05T11:36:20+0100] [ALPM] upgraded virtualbox-host-dkms (7.0.2-1 -> 7.0.2-2) [2022-11-05T11:36:20+0100] [ALPM] upgraded virtualbox (7.0.2-1 -> 7.0.2-2) and a self build, rt patched linux 1:4.19.255_rt113-0.1000 as well as with core/linux 6.0.8.arch1-1 threadirqs (not tested without threadirqs). FWIW I changed USB 3.0 enabled to USB 2.0 enabled, to no avail. [rocketmouse@archlinux ~]$ grep unresponsive /mnt/winos11/winos11/Logs/VBox.log 00:03:22.279545 VMMDev: vmmDevHeartbeatFlatlinedTimer: Guest seems to be unresponsive. Last heartbeat received 4 seconds ago [rocketmouse@archlinux ~]$ grep -i error /mnt/winos11/winos11/Logs/VBox.log 00:00:01.506420 TM: GIP - CPU: iCpuSet=0x0 idCpu=0x0 idApic=0x0 iGipCpu=0x1 i64TSCDelta=0 enmState=3 u64CpuHz=2793529547(*) cErrors=0 00:00:01.506423 TM: GIP - CPU: iCpuSet=0x1 idCpu=0x1 idApic=0x2 iGipCpu=0x0 i64TSCDelta=0 enmState=3 u64CpuHz=2793529898(*) cErrors=0 00:00:01.722439 HM: VM-entry Xcpt error-code optional = false 00:04:32.452728 /Devices/vga/VMSVGA/FifoErrors 0 times 00:04:32.452860 /Devices/vga/VMSVGA/Reg/GmrDescriptorWrite/Errors 0 times 00:04:32.454436 /PGM/PhysHandlerTree/ErrorsAllocatorR3 0 times 00:04:32.454438 /PGM/PhysHandlerTree/ErrorsTree 0 times 00:04:32.454712 /SELM/LoadHidSel/GstReadErrors 0 times The only differences are seemingly the same version of virtualbox build against another version of openssl and different kernels are used, while one of the tested kernels is an update of the 4.19 longterm kernel used when everything still was ok. I asked at https://bbs.archlinux.org/viewtopic.php?pid=2067415#p2067415 too and will report back to the mailing list, if it should be solved by the forum or to the forum, if it should be solved by the mailing list. Regards, Ralf

1 6

Python 2.7 depends on openssl-1.1
by Friedrich Romstedt 12 Nov '22

12 Nov '22

Hi, Today I encountered an error response in my browser when attempting to use my locally-installed Trac projects. Notice that Trac, as it is installed on my system, uses Python 2.x. Inspecting /var/log/httpd/error_log I tracked the error down to the hashlib-Module in Python 2.7.18-5. Attempting to import that module in Python2 reproduced the error logged by apache precisely. So I tried to install python2-2.7.18-8 from the AUR. This required gdb and xorg-server-xvfb. After having them installed, another dependency on openssl-1.1 was reported by makepkg. Notice that the package itself is called openssl-1.1, the -1.1 is *not* the Arch Package version, which is at the moment 1.1.1.s-2. While googling to resolve the issue, before attempting to make the AUR python2 package, in some way I spotted openssl as a probable reason for the issue. It got updated from 1.1.1.q-1 to 3.0.7-2 quite recently on my system (pacman.log). As a result of this guess, I tried to downgrade from 3.0.7-2 back to 1.1.1.q-1, which failed because it would break some dependencies. Now, I installed the openssl-1.1 package (which resolves to openssl-1.1-1.1.1.s-2), as it was required to build the AUR python2-2.7.18-8 package. After having installed openssl-1.1, 'import hashlib' worked in Python2 from Arch upstream (i.e., python2-2.7.18-5), and after a machine restart my Trac projects are now up and running again as well. So there was no need to build the AUR package in the end. I am proposing to add 'openssl-1.1' to the list of requirements for the 'python2' package. 'openssl-1.1' can peacefully coexist with the 'openssl' package, AFAICT from my system. At the moment, 'python2' depends on 'openssl'. I do not know if *both* openssl (3.0.7-2) *as well as* openssl-1.1 (1.1.1.s-2) are required, or if the openssl-dependency can be dropped. For all I know, keeping them both would not harm. Best, Friedrich

2 2

pacman gpg trust db corrupted by temporary failure in name-resolution
by David C. Rankin 12 Nov '22

12 Nov '22

All, Disregard the "how do I fix?" part of the prior thread on the pacman trust db getting corrupted. (I got it fixed) Backtracking, the problem occurred after an attempted update when there was a temporary failure in name-resolution due to the DNS server undergoing a reboot when the pacman update was attempted. Corruption of the trust store should not occur due to name resolution failure. Is this a "feature" of the latest pacman release? Does it need further investigation? -- David C. Rankin, J.D.,P.E.

1 0

/var/cache/pacman/pkg/archlinux-keyring-20221110-1-any.pkg.tar.zst is corrupted
by David C. Rankin 12 Nov '22

12 Nov '22

Full system update completed find earlier, but then attempting to install lightdm, the following error was received: :: Proceed with installation? [Y/n] :: Retrieving packages... libxklavier-5.4-3-x86_64 72.5 KiB 145 KiB/s 00:01 [######################################] 100% lightdm-1:1.32.0-4-x86_64 239.1 KiB 357 KiB/s 00:01 [######################################] 100% Total (2/2) 311.6 KiB 351 KiB/s 00:01 [######################################] 100% (2/2) checking keys in keyring [######################################] 100% warning: Public keyring not found; have you run 'pacman-key --init'? downloading required keys... error: keyring is not writable error: keyring is not writable error: required key missing from keyring error: failed to commit transaction (unexpected error) Errors occurred, no packages were upgraded. Huh? Attempting to update archlinux-keyring results in: :: Proceed with installation? [Y/n] (1/1) checking keys in keyring [######################################] 100% (1/1) checking package integrity [######################################] 100% error: archlinux-keyring: signature from "Christian Hesse <eworm(a)archlinux.org>" is unknown trust :: File /var/cache/pacman/pkg/archlinux-keyring-20221110-1-any.pkg.tar.zst is corrupted (invalid or corrupted package (PGP signature)). Do you want to delete it? [Y/n] error: failed to commit transaction (invalid or corrupted package (PGP signature)) Errors occurred, no packages were upgraded. Attempting lightdm again results in Allan's key not found and of unknown trust: :: Proceed with installation? [Y/n] (2/2) checking keys in keyring [######################################] 100% downloading required keys... :: Import PGP key F99FFE0FEAE999BD, "Allan McRae <allan(a)archlinux.org>"? [Y/n] :: Import PGP key 0F65C7D881506130, "Maxime Gauduin <alucryd(a)archlinux.org>"? [Y/n] (2/2) checking package integrity [######################################] 100% error: libxklavier: signature from "Allan McRae (Developer) <allan(a)archlinux.org>" is unknown trust :: File /var/cache/pacman/pkg/libxklavier-5.4-3-x86_64.pkg.tar.zst is corrupted (invalid or corrupted package (PGP signature)). Do you want to delete it? [Y/n] error: lightdm: signature from "Maxime Gauduin <alucryd(a)archlinux.org>" is unknown trust :: File /var/cache/pacman/pkg/lightdm-1:1.32.0-4-x86_64.pkg.tar.zst is corrupted (invalid or corrupted package (PGP signature)). Do you want to delete it? [Y/n] error: failed to commit transaction (invalid or corrupted package (PGP signature)) Errors occurred, no packages were upgraded. Is this a temporary issue or did something get fsck'ed up on my system? (and how to attempt to fix?) -- David C. Rankin, J.D.,P.E.

1 0

random junk appended to kernel command line
by Geert Hendrickx 10 Nov '22

10 Nov '22

Hi I'm UEFI-booting Arch Linux from a unified kernel image (UKI) generated by mkinitcpio: https://wiki.archlinux.org/title/Unified_kernel_image Recently, I'm seeing random junk being appended to the kernel commandline: $ cat /proc/cmdline "_VGA] +WQmapper/ssdvg-root rw quiet loglevel=3 udev.log_priority=3 ]GM?|-dQ2�H~A3y $ hexdump -C /proc/cmdline 00000000 72 6f 6f 74 3d 2f 64 65 76 2f 6d 61 70 70 65 72 |root=/dev/mapper| 00000010 2f 73 73 64 76 67 2d 72 6f 6f 74 20 72 77 20 71 |/ssdvg-root rw q| 00000020 75 69 65 74 20 6c 6f 67 6c 65 76 65 6c 3d 33 20 |uiet loglevel=3 | 00000030 75 64 65 76 2e 6c 6f 67 5f 70 72 69 6f 72 69 74 |udev.log_priorit| 00000040 79 3d 33 20 17 5d 47 4d 05 3f 7c 2d 64 51 32 1a |y=3 .]GM.?|-dQ2.| 00000050 1f 48 15 7e 41 33 79 0d 22 5f 56 47 41 5d 20 2b |.H.~A3y."_VGA] +| 00000060 57 51 0a |WQ.| 00000063 (the junk is only at the end, but with "cat" it garbles the terminal and overwrites the beginning of the line. hexdump shows it's exclusively at the end though.) Unrecognized junk is then passed on to init: $ journalctl -k -o cat | grep -a command\ line "_VGA] +WQmand line: root=/dev/mapper/ssdvg-root rw quiet loglevel=3 udev.log_priority=3 ]GM?|-dQ2�H~A3y Unknown kernel command line parameters "]GM?|-dQ2�H~A3y _VGA] +WQ", will be passed to user space. $ ps -ef | grep init root 1 0 1 20:30 ? 00:00:00 /sbin/init ?]GM??|-dQ2??H?~A3y _VGA] +WQ (journalctl without -o cat doesn't even print those lines.) The input file /etc/kernel/cmdline (used by mkinitcpio to generate the UKI) is clean though: $ cat /etc/kernel/cmdline root=/dev/mapper/ssdvg-root rw quiet loglevel=3 udev.log_priority=3 And the UKI itself doesn't seem to contain this random junk either: $ hexdump -C /boot/archlinux-linux.efi | grep -A5 root= 0000e200 72 6f 6f 74 3d 2f 64 65 76 2f 6d 61 70 70 65 72 |root=/dev/mapper| 0000e210 2f 73 73 64 76 67 2d 72 6f 6f 74 20 72 77 20 71 |/ssdvg-root rw q| 0000e220 75 69 65 74 20 6c 6f 67 6c 65 76 65 6c 3d 33 20 |uiet loglevel=3 | 0000e230 75 64 65 76 2e 6c 6f 67 5f 70 72 69 6f 72 69 74 |udev.log_priorit| 0000e240 79 3d 33 20 00 00 00 00 00 00 00 00 00 00 00 00 |y=3 ............| 0000e250 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| However when booting the kernel directly (not the UKI) from the UEFI shell, the same kernel command line is passed clean. No idea what is causing this, but it's a recent issue. I have regenerated the UKI several times, and also tried downgrading mkinitcpio back to v31, and the kernel to 5.19, but keep getting the same result. I'm not suspecting the hardware, since I'm seeing the same issue on two different machines. Any ideas? Geert

2 2

(Arch)Linux source routing help
by Pascal 09 Nov '22

09 Nov '22

*(sorry, the previous message is gone a bit fast)* hi, where can I get help for this ? my problem is : I'm on a corporate network that is filtered (proxy) and I need to run some tests on a virtual machine that needs to have direct access to the Internet. I set up this configuration and try to route everything that comes out of tap0 to wlan0 and then to Internet. - - \ / \ / \ / +--------\-------+ +-------/-------+ |free GW | |corporate GW | |192.168.144.254 | |10.10.10.254/24| <http://10.10.10.254/24%7C> +--------|-------+ +-------|-------+ \ / +---------------|-----------------|------+ |HOST |wlan0 | |eth0 | | |192.168.144.1/24| <http://192.168.144.1/24%7C> |10.10.10.1/24| <http://10.10.10.1/24%7C> | +----------------+ +-------------| | | | +-----------------+ | | |tap0 | | | |192.168.11.254/24| <http://192.168.11.254/24%7C> | | +--------|--------+ | | | | | +-------------|--------+ | | |VM |eth0 | | | | |192.168.11.1/24 | | | | +-----------------| | | +----------------------+ | +----------------------------------------+ here are the commands used: # ip tuntap add tap0 mode tap user me # ip addr add 192.168.11.254/24 dev tap0 # ip link set tap0 down # echo 11 tap0 >> /etc/iproute2/rt_tables # ip rule add from 192.168.11.254 lookup tap0 # ip route add default via 192.168.144.254 dev wlan0 proto dhcp src 192.168.144.1 # echo 1 > /proc/sys/net/ipv4/ip_forward # iptables -t nat -A POSTROUTING -o wlan0 -j MASQUERADE I can ping 8.8.8.8 from my virtual machine but all traffic (host/corporate) now seems to flow through wlan0 (tshark -i wlan0)... regards, lacsaP.

4 4

(Arch)Linux source routing help
by Pascal 09 Nov '22

09 Nov '22

hi, where can I get help with this ? my problem is this: I'm on a corporate network that is filtered (proxy) and I need to run some tests on a virtual machine that needs to have direct access to the Internet. I set up this configuration and try to route everything that comes out of tap0 to wlan0 and then to Internet. - - \ / \ / \ / +--------\-------+ +-------/-------+ |free GW | |corporate GW | |192.168.144.254 | |10.10.10.254/24| +--------|-------+ +-------|-------+ \ / +---------------|-----------------|------+ |HOST |wlan0 | |eth0 | | |192.168.144.1/24| |10.10.10.1/24| | +----------------+ +-------------| | | | +-----------------+ | | |tap0 | | | |192.168.11.254/24| | | +--------|--------+ | | | | | +-------------|--------+ | | |VM |eth0 | | | | |192.168.11.1/24 | | | | +-----------------| | | +----------------------+ | +----------------------------------------+

1 0

installing archlinux, get lynx accessible first
by Jude DaShiell 08 Nov '22

08 Nov '22

I had a read through the Installation_guide script and discovered it uses lynx to bring the guide up on the screen. I'm a screen reader user so whenever I use lynx and certain other console browsers I set certain options then save me a .lynxrc file before using lynx further. Two really critical options are show cursor on so my screen reader can track where I am on the screen and links and form fields are numbered so I can find links to run. Others are there too but for me these are most critical. Once the .lynxrc file is saved reading through the installation_guide for me is much better. elinks and links can show numbered links but their configuration is a bit different. So for screen reader users who find numbered links in lynx the cat useful, before running Installation_guide get lynx configured so it runs well for you. I did this on a flash drive so my .lynxrc got saved for future use. 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) .

3 4