2026-01-01 00:50:42 happy new year to everyone. can someone point me to the APKBUILD-file of 'vim' on gitlab.alpinelinux.org? it appears to have moved since my old link results in 404. :( 2026-01-01 00:50:54 https://gitlab.alpinelinux.org/alpine/aports/-/raw/master/main/vim/APKBUILD 2026-01-01 00:54:18 https://gitlab.alpinelinux.org/alpine/aports/-/blob/master/community/vim/APKBUILD 2026-01-01 00:54:45 thanks! 2026-01-01 00:54:59 use the "git repository link" fom https://pkgs.alpinelinux.org/package/edge/community/x86_64/vim 2026-01-01 00:55:00 so it moved from main to community, i see 2026-01-01 00:55:42 ah right, i think that's where i found it earlier, before it moved 2026-01-01 01:12:20 hmm getting 500 error from alpines gitlab instance 2026-01-01 14:04:38 Do we have documentation for the v3 APK format anywhere? Thanks! 2026-01-01 14:24:41 runxiyu: no complete one, but some basics are at https://gitlab.alpinelinux.org/alpine/apk-tools/-/blob/master/doc/apk-v3.5.scd 2026-01-01 14:28:05 ty! 2026-01-01 14:29:03 wait, uh 2026-01-01 14:29:12 is this actually used in apk v3? 2026-01-01 14:31:07 https://dl-cdn.alpinelinux.org/edge/main/riscv64/apk-tools-3.0.3-r1.apk's first few bytes just looks like gzip to me 2026-01-01 14:37:56 Yes, apkv3 can still parse v2 files and alpine only uses v2 files at the moment 2026-01-01 14:48:20 Yes, Alpine still does V2 format only. Migration to V3 formats happens later. Some other distributions (Chimera, OpenWRT) use the V3 formats already now. 2026-01-01 14:48:38 APK V3 supports both file formats at this time. 2026-01-01 16:02:14 ah, alright 2026-01-01 16:02:15 ty 2026-01-01 17:02:46 runxiyu: If you want some prebuild v3 files to play with look in https://downloads.openwrt.org/snapshots/packages/x86_64/packages/ 2026-01-01 17:03:02 all the '*.apk' and the packages.adb are all v3 2026-01-01 17:24:08 thanks! 2026-01-01 17:24:13 im familiar with cihmera so im using theirs 2026-01-01 19:26:17 fabled: afaik the newest alpine does v3 2026-01-01 19:26:45 https://alpinelinux.org/posts/Alpine-3.23.0-released.html <- yeah 2026-01-01 19:35:45 just the package manager, the index and the packages still use the v2 format 2026-01-01 19:36:54 oh oops, also i missed what channel i was in lol 2026-01-01 19:37:01 thanks for the clarification :) 2026-01-01 19:56:04 I've just upgraded from 3.22 to 3.23 and got the following errors from APK: https://termbin.com/9fvx 2026-01-01 19:56:11 What can I do about that? 2026-01-01 19:59:13 Maybe apk fix linux-firmware-nvidia 2026-01-01 19:59:23 Okay, let's see. 2026-01-01 20:00:21 That did it. Thanks. 2026-01-01 20:01:11 You're welcome :) 2026-01-01 20:54:17 oh they left already 2026-01-01 20:54:24 but tldr thats a apk bug 2026-01-01 20:54:47 we could probably work around it in aports 2026-01-02 04:16:48 I'm getting this error in a build in CI. This smells like a package bug; is it? 2026-01-02 04:16:52 lto1: fatal error: bytecode stream in file '/usr/lib/gcc/x86_64-alpine-linux-musl/15.2.0/../../../../lib/libgmp.a' generated with LTO version 14.0 instead of the expected 15.1 2026-01-02 09:02:15 The package containing libgmp.a needs to be rebuild 2026-01-02 14:18:41 !list 2026-01-02 14:20:01 oO 2026-01-02 14:29:00 I think they just called you a gano 2026-01-02 14:57:13 hello, i have problem with my alpine linux installation, for some reason when i try to boot it up, it shows that /dev/nvme0n1p2 mounting on /sysroot: Not found file or directory 2026-01-02 14:57:57 "mount /dev/nvme0n1p2 on /sysroot failed: No such file or directory" 2026-01-02 14:58:37 Maybe it's missing some modules from the initrd 2026-01-02 14:58:54 Did you customize /etc/mkinitfs/mkinitfs.conf? 2026-01-02 14:59:19 a little i think, but this is what i have in it: features="ata base cdrom ext4 keymap kms mmc nvme raid scsi usb virtio sd-mod" 2026-01-02 15:00:03 my root is ext4 btw 2026-01-02 15:00:13 Yeah, should be enough 2026-01-02 15:00:29 So, you're dropped to an emergency sh 2026-01-02 15:00:39 What do you have in /dev? 2026-01-02 15:01:25 yes 2026-01-02 15:01:40 this is my /dev 2026-01-02 15:01:41 autofs log tty tty45 ttyS24 vcs6 2026-01-02 15:01:41 block loop-control tty0 tty46 ttyS25 vcs7 2026-01-02 15:01:41 bus mcelog tty11 tty49 ttyS28 vcsa 2026-01-02 15:01:41 btrfs-control mapper tty10 tty48 ttyS27 vcs9 2026-01-02 15:01:41 bsg loop0 tty1 tty47 ttyS26 vcs8 2026-01-02 15:01:43 char media0 tty12 tty5 ttyS29 vcsa1 2026-01-02 15:01:45 console mem tty13 tty50 ttyS3 vcsa10 2026-01-02 15:01:47 core mqueue tty14 tty51 ttyS30 vcsa11 2026-01-02 15:01:49 cpu net tty15 tty52 ttyS31 vcsa12 2026-01-02 15:01:53 cpu_dma_latency ng0n1 tty16 tty53 ttyS4 vcsa2 2026-01-02 15:01:55 cuse null tty17 tty54 ttyS5 vcsa3 2026-01-02 15:01:57 dbc nvme0 tty18 tty55 ttyS6 vcsa4 2026-01-02 15:01:59 disk nvme0n1 tty19 tty56 ttyS7 vcsa5 2026-01-02 15:02:01 dm-0 nvme0n1p1 tty2 tty57 ttyS8 vcsa6 2026-01-02 15:02:03 dm-1 nvme0n1p2 tty20 tty58 ttyS9 vcsa7 2026-01-02 15:02:05 dma_heap nvme0n1p3 tty21 tty59 udmabuf vcsa8 2026-01-02 15:02:07 dri nvram tty22 tty6 uhid vcsa9 2026-01-02 15:02:09 drm_dp_aux0 port tty23 tty60 uinput vcsu 2026-01-02 15:02:11 drm_dp_aux1 ppp tty24 tty61 urandom vcsu1 2026-01-02 15:02:13 drm_dp_aux2 ptmx tty25 tty62 usb vcsu10 2026-01-02 15:02:15 drm_dp_aux3 pts tty26 tty63 usbmon0 vcsu11 2026-01-02 15:02:17 drm_dp_aux4 random tty27 tty7 usbmon1 vcsu12 2026-01-02 15:02:19 drm_dp_aux5 rfkill tty28 tty8 usbmon2 vcsu2 2026-01-02 15:02:23 drm_dp_aux6 rtc tty29 tty9 usbmon3 vcsu3 2026-01-02 15:02:24 But nvme0n1p2 is in there 2026-01-02 15:02:25 drm_dp_aux7 rtc0 tty3 ttyS0 usbmon4 vcsu4 2026-01-02 15:02:27 fd sda tty30 ttyS1 usbmon5 vcsu5 2026-01-02 15:02:29 full sda1 tty31 ttyS10 usbmon6 vcsu6 2026-01-02 15:02:31 fuse sda2 tty32 ttyS11 usbmon7 vcsu7 2026-01-02 15:02:33 gpiochip0 sg0 tty33 ttyS12 usbmon8 vcsu8 2026-01-02 15:02:35 gpmctl shm tty34 ttyS13 usbmon9 vcsu9 2026-01-02 15:02:37 hidraw0 snapshot tty35 ttyS14 userfaultfd vfio 2026-01-02 15:02:39 hidraw1 snd tty36 ttyS15 v4l vga_arbiter 2026-01-02 15:02:41 hidraw2 softsynth tty37 ttyS16 vcs vhci 2026-01-02 15:02:43 hidraw3 softsynthu tty38 ttyS17 vcs1 vhost-net 2026-01-02 15:02:45 hpet stderr tty39 ttyS18 vcs10 vhost-vsock 2026-01-02 15:02:46 (try not to paste such huge amount of data into IRC, better using a third-party pasting service) 2026-01-02 15:02:47 hwrng stdin tty4 ttyS19 vcs11 video0 2026-01-02 15:02:49 input stdout tty40 ttyS2 vcs12 video1 2026-01-02 15:02:53 kfd synth tty41 ttyS20 vcs2 watchdog 2026-01-02 15:02:55 kmsg synthu tty42 ttyS21 vcs3 watchdog0 2026-01-02 15:02:57 kvm tpm0 tty43 ttyS22 vcs4 zero 2026-01-02 15:02:59 live-base tpmrm0 tty44 ttyS23 vcs5 zfs 2026-01-02 15:03:01 woops 2026-01-02 15:03:03 im rn on gentoo live usb 2026-01-02 15:03:05 and doing it through chroot 2026-01-02 15:03:07 ah 2026-01-02 15:03:07 yeah thats the problem 2026-01-02 15:03:50 Are you on a second machine? Can you reboot into Alpine to see the actual state, while reporting on IRC? 2026-01-02 15:04:12 not really, i cant 2026-01-02 15:04:23 ok 2026-01-02 15:04:42 Check the output of lsmod 2026-01-02 15:05:14 Hummm, and/or rather the output of lspci -k 2026-01-02 15:05:21 here https://0x0.st/P-U3.txt 2026-01-02 15:06:03 here you go https://0x0.st/P-UY.txt 2026-01-02 15:06:33 also i got pcilib: Error reading /sys/bus/pci/devices/0000:00:08.3/label: Operation not permitted 2026-01-02 15:06:33 , but thats probably becuz im accessing alpine fro, chroot 2026-01-02 15:08:58 Hummm, I don't know, looks like it should work 2026-01-02 15:09:06 Maybe try regenerating your initrd 2026-01-02 15:09:19 how could i regenerate initrd 2026-01-02 15:09:45 that i also want to know, also how to set kernel cmdline arguments 2026-01-02 15:09:47 # mkinitfs $(basename /lib/modules/*) 2026-01-02 15:10:12 okay so i did initrd regeneration correctly 2026-01-02 15:10:48 The cmdline are part of your boot-loader, that depends on which 2026-01-02 15:11:14 fairly sure i am on grub for this machine, no idea about the pis i have now 2026-01-02 15:12:05 In /etc/default/grub, you have an entry GRUB_CMDLINE_LINUX_DEFAULT that you can use to customize your boot (will be used for every kernel entry) 2026-01-02 15:12:30 i will come back in sec, just gotta check if it works, and if not i will return 2026-01-02 15:12:38 i presume i have to do grub-mkconfig after or something like that 2026-01-02 15:12:40 good luck gentoo 2026-01-02 15:12:49 good luck also 2026-01-02 15:12:54 Yes penguinz_rule[m] :) 2026-01-02 15:13:09 # grub-mkconfig -o /boot/grub/grub.cfg 2026-01-02 15:16:33 okay so it doesnt work, the error is exactly the same 2026-01-02 15:17:05 Did you see anything suspicious in the kernel log while booting? 2026-01-02 15:17:44 "mount: mounting /dev/nvme0n1p2 on /sysroot: No such file or directory" 2026-01-02 15:17:50 also no, the boot was very fast 2026-01-02 15:18:00 maybe there is way to get the boot info back? 2026-01-02 15:18:02 Yeah but before that ;) 2026-01-02 15:18:28 I don't think so, it's ephemeral as it happens in memory, before being able to mount your permanent storage disk 2026-01-02 15:19:13 Could you share the kernel config of your gentoo one that correctly handles the nvme drive? 2026-01-02 15:20:32 okay im back 2026-01-02 15:22:21 okay i have no idea why it is not mounting 2026-01-02 15:22:47 maybe it doesnt see /dev/ ? 2026-01-02 15:22:52 at boot time 2026-01-02 15:24:19 but i have libevdev 2026-01-02 15:24:25 It should be populated by mdev/openrc 2026-01-02 15:24:41 mdev? 2026-01-02 15:25:17 maybe i need mdev 2026-01-02 15:25:33 Wait, no, just by the kernel when mounting devfs on /dev during init phase 2026-01-02 15:25:45 16:19:13 quinq$ Could you share the kernel config of your gentoo one that correctly handles the nvme drive? 2026-01-02 15:25:59 maybe it has no devfs 2026-01-02 15:26:09 It does 2026-01-02 15:26:21 so why it doesnt /dev/nvme0n1p2? 2026-01-02 15:26:34 That's what we are trying to find out 2026-01-02 15:26:52 wait if im in shell emergency can i see what is in /dev/? 2026-01-02 15:27:01 ls /dev/ 2026-01-02 15:27:01 Yes 2026-01-02 15:27:14 let me try it and then i will report 2026-01-02 15:27:29 Could you, first, share the gentoo kernel config? 2026-01-02 15:35:16 okay so i figured out that the kernel sees /dev/nvme0n1p2 2026-01-02 15:35:22 it simply refuses to mount it 2026-01-02 15:35:37 i tried just using mount command in emergency shell 2026-01-02 15:35:49 and it just refused to mount any /dev/nvme0 2026-01-02 15:36:08 it just says no such file or directory 2026-01-02 15:36:13 maybe it doesnt see /sysroot? 2026-01-02 15:36:50 but then it would different error 2026-01-02 15:38:22 so i have no idea what it would be 2026-01-02 15:42:57 if there is something to add, i use the mini root file system for chroot install 2026-01-02 15:44:17 so i have no idea whats happening with alpine 2026-01-02 15:55:04 Sorry, was away a bit 2026-01-02 15:55:08 ohhh, ok 2026-01-02 15:55:58 Well, that is weird indeed 2026-01-02 15:56:00 16:36:13 gentoo$ maybe it doesnt see /sysroot? 2026-01-02 15:56:03 Well, does it exist? 2026-01-02 15:57:07 yeah 2026-01-02 15:57:10 wait 2026-01-02 15:57:13 maybe it doesnt 2026-01-02 15:57:32 yeah there is /sysroot, there must be, i created directory for it 2026-01-02 15:58:33 Remember, everything's ephemeral at that point 2026-01-02 16:01:09 Well, it's created during init: /bin/busybox mkdir -p […] "$sysroot" 2026-01-02 16:02:19 Sorry but at this point I don't know, it's a bit hard without being able to poke at it directly 2026-01-02 16:02:46 how should i poke it idk 2026-01-02 16:03:05 sounds like a device enumeration issue 2026-01-02 16:03:06 it is weird it doesnt want to mount /dev/nvme0n1p2 even tho it sees it and has all the stuff 2026-01-02 16:03:21 gentoo: What is the filesystem type? 2026-01-02 16:03:54 maybe i will try different kernel 2026-01-02 16:04:06 ext4 2026-01-02 16:04:18 can you check /proc/cmdline? 2026-01-02 16:04:34 on chroot or boot? 2026-01-02 16:04:41 Boot 2026-01-02 16:04:52 well let me try it then 2026-01-02 16:09:29 so the /proc/cmdline is root=/dev/nvme0n1p2 rw 2026-01-02 16:10:21 just like in my limine conf 2026-01-02 16:10:31 What's a limine conf? 2026-01-02 16:10:55 a config for limine boot loader 2026-01-02 16:11:14 ftr, limine is not officially supported 2026-01-02 16:11:22 oh 2026-01-02 16:11:32 You seem to miss some cmdline options 2026-01-02 16:11:41 which are? 2026-01-02 16:11:48 Isn't it missing the rootfstype? 2026-01-02 16:12:04 yes 2026-01-02 16:12:08 modules and rootfstype 2026-01-02 16:12:30 modules=sd-mod,usb-storage,ext4 rootfstype=ext4 2026-01-02 16:12:36 gentoo, why not starting to state that you're using a custom system ;_; 2026-01-02 16:12:47 okay let me try 2026-01-02 16:14:06 yeah i did forget to mention that 2026-01-02 16:14:32 i usually use distros made for costumization 2026-01-02 16:19:35 anyways, sorry for not telling that and also thanks for the cmdline, so now to make wifi work i just need to add it as module? 2026-01-02 16:20:56 The module should already be loaded 2026-01-02 16:21:03 You need to connect to a wi-fi network 2026-01-02 16:21:39 gentoo: ftr, you can customize alpine, but 1: you should know what you do, 2: you should state that when asking for help :) 2026-01-02 16:22:44 okay so my issue is that, iwd (yeah im not using wpa_supplicant, it just didnt worked last time using alpine linux, and i prefer iwd) doesnt see wifi device 2026-01-02 16:23:14 Then it seems that neither wpa_supplicant nor iwd work 2026-01-02 16:23:39 It might just be missing the proprietary blobs for the chip 2026-01-02 16:24:25 gentoo: what NIC does it have? 2026-01-02 16:25:12 RTL8852BE PCIe 802.11ax Wireless Network Controller 2026-01-02 16:26:09 Verify that you have linux-firmware-rtlwifi installed 2026-01-02 16:26:52 yep i do have linux-firmware-rtlwif 2026-01-02 16:26:57 Humm, or maybe another 2026-01-02 16:27:20 Now that you're in Alpine, look at the dmesg 2026-01-02 16:27:26 Not sure if required, but I saw https://github.com/morrownr/rtw89 2026-01-02 16:27:52 well im not on alpine becuz i dont have wifi connection here, im still on chroot 2026-01-02 16:28:06 but i think dmesg still should work on chroot 2026-01-02 16:28:27 Then try linux-firmware-rtw89 2026-01-02 16:29:25 also i installed linux-firmware, so it should have all linux-firmware-rtw89 2026-01-02 16:31:01 i think i should just add wifi module 2026-01-02 16:31:01 $ grep -F 8852BE /boot/config-6.18.2-1-lts 2026-01-02 16:31:01 CONFIG_RTW89_8852BE=m 2026-01-02 16:31:04 Looks good 2026-01-02 16:31:27 gentoo, don't assume, check 2026-01-02 16:31:30 Also dmesg 2026-01-02 16:32:34 yep i have linux-firmware-rtw89 2026-01-02 16:33:32 okay let me send you dmesg 2026-01-02 16:33:36 ACTION . o O ( maybe I should replace my thinking noise “hmm” by “quinq” ) 2026-01-02 16:34:42 Am I making too much noise? ^_^ sorry 2026-01-02 16:35:20 don’t worry, I think it’s fun when that happens 2026-01-02 16:38:24 here dmesg https://0x0.st/P-GN.txt 2026-01-02 16:39:43 so thats about it 2026-01-02 16:41:17 Is rtw89_8852be loaded? 2026-01-02 16:41:45 i mean, if devices dont show up, then it is not 2026-01-02 16:42:36 Yes sorry, that was kind of a rhetorical question 2026-01-02 16:42:49 I'll reformulate, can you try loading the module and share again the dmesg? 2026-01-02 16:43:18 [ 6.793660] faux_driver regulatory: Direct firmware load for regulatory.db failed with error -2 2026-01-02 16:43:25 That shouldn't be fatal, would it? 2026-01-02 16:43:52 no it is not fatal 2026-01-02 16:44:05 i have no idea what is regulatory 2026-01-02 16:44:16 also i should load the driver using modprobe? 2026-01-02 16:44:27 That's to adapt your wireless device power according to your contry RF laws 2026-01-02 16:44:33 It makes sure you don't use frequencies that are not allowed in the specific region 2026-01-02 16:44:46 Yeah: modprobe rtw89_8852be 2026-01-02 16:44:57 *power and frequencies indeed 2026-01-02 16:45:07 anyways, let me try then this 2026-01-02 16:49:22 okay yeah, when i did modprobe, the device showed up in iwd 2026-01-02 16:49:34 now gotta make it load on startup 2026-01-02 16:49:46 It should load by itself, something else is going on 2026-01-02 16:50:16 You could add it to /etc/modules, but quinq is right that it should autoload 2026-01-02 16:50:45 ./etc/modules? 2026-01-02 16:51:24 Is there any other deviration from the standard install that you did? 2026-01-02 16:52:19 i mentioned that i did not used standard install, i have used mini root filesystem as i didnt found other way to do rootfs install 2026-01-02 16:52:43 That can explain a lot 2026-01-02 16:52:52 You're most likely missing services 2026-01-02 16:53:18 ok, thank you :) 2026-01-02 16:53:28 I missed the info indeed 2026-01-02 16:55:07 what services would i be missing? 2026-01-02 16:55:28 mdev I suppose 2026-01-02 16:55:45 hwdrivers 2026-01-02 16:55:52 okay let me install them then 2026-01-02 16:56:46 rc-status -fini boot sysinit 2026-01-02 16:56:50 what does that return? 2026-01-02 16:57:27 http://0x0.st/P-G4.bin 2026-01-02 16:59:10 Right, those 2 are most prominently missing 2026-01-02 16:59:56 okay which package i should add for mdev? busybox-mdev-openrc or mdevd-openrc 2026-01-02 17:00:27 You don't need to explicitly install it 2026-01-02 17:01:03 well i dont have them 2026-01-02 17:01:18 rc-update add mdev sysinit 2026-01-02 17:01:27 Maybe just reinstall with a base system 2026-01-02 17:01:30 rc-update add hwdrivers sysinit 2026-01-02 17:02:28 okay done 2026-01-02 17:03:19 Now if you reboot, I expect the interface to be available 2026-01-02 17:03:22 quinq, i mean, i could just use base system, but first i put whole day into this install and i wanted to try minimal system 2026-01-02 17:04:05 also how would i know how to do such things if i didnt not try doing them 2026-01-02 17:04:13 *did not 2026-01-02 17:04:27 gentoo: opposite order generally works better. First do a default install, figure out how things work, and then move to a minimal system 2026-01-02 17:04:45 well i was already on default alpine install 2026-01-02 17:05:06 well i had before default alpine install 2026-01-02 17:05:15 so i figured out that i will try doing more lower level stuff 2026-01-02 17:05:38 but thanks for your help 2026-01-02 17:07:10 let me reboot quick 2026-01-02 17:14:17 okay yep everything till now works 2026-01-02 17:14:33 now i need to figure out why seatd is giving me error about no permissions 2026-01-02 18:45:29 hi 2026-01-02 18:48:04 gentoo: Add your user to seat group 2026-01-02 18:54:32 visone[m], gone :/ 2026-01-02 18:56:40 quinq: Thanks!! 2026-01-02 20:56:07 hello, so I tried setting up gtkgreetd on my alpine install, and found out that mdev doesnt work with libinput so i needed eudev, but now i have very weird segfault with GTK, that it cannot draw anything for some reason? 2026-01-02 21:00:23 gentoo: libudev-zero 2026-01-02 21:03:50 i wont use libudev-zero as it doesnt support pipewire and such (according to their readme) 2026-01-02 21:05:46 gentoo: yeah, pipewire needs a patch for some specific scenarios (works on 2 out of 3 setups for me) 2026-01-02 21:06:01 If you don't want libudev-zero, then you need eudev and libudev, no mdev(d) 2026-01-03 05:44:47 greetings! i've got a machine running alpine linux, its wi-fi and ethernet worked at first in the install media and after install, but after a few hours the interfaces stopped showing in ip link show and i can no longer connect. happy to provide more information and follow troubleshooting steps on request. anyone online to advise? 2026-01-03 05:45:11 absynthaero: anything relevant in dmesg? 2026-01-03 05:45:23 will check 2026-01-03 05:45:25 ikke: hey 2026-01-03 05:46:52 not seeing anything in dmesg under wifi, wireless, or realtek (i know this to be a realtek chip). anything else i should check? 2026-01-03 05:48:21 absynthaero: yes 2026-01-03 05:51:08 Lesley: happy to keep looking. any keywords i should be looking for or more information i could give? 2026-01-03 05:51:42 absynthaero: okay 2026-01-03 14:18:10 Hello, i have issue with user services in openrc, for some reason when i try to enable pipewire/wireplumber as user service (according to the wiki) it shows that there is no 'gui' runlevel 2026-01-03 14:19:32 Did you add something like `openrc -U gui` to ~/.profile? 2026-01-03 14:19:49 https://wiki.alpinelinux.org/wiki/OpenRC#Runlevels_for_user_services 2026-01-03 14:20:09 well no, but also, when i try to run the command, it says that gui is not a valid runlevel 2026-01-03 14:21:36 i need to create gui runlevel? 2026-01-03 14:21:37 Create a custom gui user runlevel: 2026-01-03 14:21:39 $ mkdir -p ~/.config/rc/runlevels/gui 2026-01-03 14:21:58 okey yeah now it works 2026-01-03 23:50:04 Good evening everyone. I have a weird one that I've been trying to troubleshoot. I'm in the process of adding a software RAID to my fstab and every time I reboot, /dev/md0 does not exist. I have to manually start mdadm-raid to get /dev/md0 to appear. It is the last line in my /etc/fstab and rebuilding the initramfs doesn't seem to help. Of course, I have added the service to the boot runlevel, 2026-01-03 23:50:10 but it fails to start every time. I am not seeing anything in /var/log/messages outside of "mdadm-raid failed to start". rc-status doesn't show mdadm or mdadm-raid at all. Is there something I may have missed? Running a fresh install of 2.23.2. 2026-01-03 23:54:33 anything in dmesg? 2026-01-03 23:59:34 Not sure if it's related or not, but I found a message showing "/dev/md0: Can't lookup blockdev". 2026-01-04 00:00:17 It happens a few seconds after the SAS controller is initialized, so I'm not sure why it would be complaining. 2026-01-04 00:13:29 Alright, this is curious. Why does mdadm rely on the firewall? I'm looking at /etc/init.d/mdadm and it's told not to start until after the firewall is up. Even weirder, /etc/init.d/mdadm-raid doesn't seem to rely on the mdadm service. Why would mdadm need the firewall? 2026-01-04 00:14:08 very much guessing here: in case some devices are networked, like nbd? 2026-01-04 00:15:23 Maybe? My thought was syslog, but that's a possibility too. I've commented out the requirement for dns and networking in /etc/init.d/mdadm, and I put a dependency on mdadm into /etc/init.d/mdadm-raid. Let's see if that makes things a little better. 2026-01-04 00:16:21 No luck. Removing the mdadm dependency from mdadm-raid to see if that was a mistake. 2026-01-04 00:17:47 Still no luck. I've reverted /etc/init.d/mdadm and /etc/init.d/mdadm-raid back to their original state. 2026-01-04 00:19:25 The only other thing I can think of is that mdadm legitimately can't find the SCSI drives. According to dmesg, sda/sdb/sdc are all present before the message about md0. It's set up as a RAID 10, but the last drive is configured as missing because the drive is defective and I will be reaching out to Seagate once their support center opens. 2026-01-04 18:24:22 hey guys, i'm trying to create a virtual raspi4b libvirt machine with alpine raspberry pi aarch64 image. https://wiki.alpinelinux.org/wiki/Raspberry_Pi#Manual_method is the latest thing i've done through manual. Log https://0x0.st/PoQy.txt 2026-01-04 18:26:13 i try to get a rpi cluster but barely can afford one 2026-01-04 19:44:07 switch_root: can't execute '/sbin/init': No such file or directory 2026-01-04 19:44:24 should it come from initramfs? 2026-01-04 19:45:37 The message yes, the /sbin/init from the target root fs 2026-01-04 19:48:55 it's related to me attempting to have a virtual raspberry pi alpine edition. maybe there is a HOWTO 2026-01-04 19:49:58 There would be something on the wiki, yeah 2026-01-04 19:51:03 so i specify kernel,dtb,initrd to qemu-system-aarch64 and all seems to be working, but it turns out guest can't see /sbin/init 2026-01-04 19:57:38 passingbyuser, it could either not be present, or for the wrong architecture 2026-01-04 19:59:07 then the question is what can i do 2026-01-04 20:00:28 i'm kinda noob in emulating things, but i need a virtual rpi with alpine distro that badly 2026-01-04 20:04:06 for the record, i figured it out, my mistake was so stupid 2026-01-04 20:04:31 it was altering cmdline with root= 2026-01-04 22:41:16 Is this a good place to ask about securing services within an Alpine system, or can someone recommend a better place to ask? 2026-01-05 10:34:50 Aangularity[m]: This is as good of a place as any 2026-01-05 12:15:54 Hummm, not alpine-specific, but would you know how I can modify the library search path for alsa-lib? 2026-01-05 12:16:26 It has hardcoded /usr/lib/alsa-lib/ by default, I'd like to use something from my $HOME 2026-01-05 12:18:30 LD_LIBRARY_PATH? 2026-01-05 12:18:31 Bah, nevermind, went to the source directly, src/dlmisc.c:94: const char *env = getenv("ALSA_PLUGIN_DIR"); 2026-01-05 12:18:46 lopid, that's for libc 2026-01-05 12:18:54 i suspected as much 2026-01-05 12:19:02 But thanks :) 2026-01-05 13:28:34 is there some problem with colors in a bash prompt? I've copied my PS1 prompt from Fedora to Alpine (.bashrc), and while it works on Fedora, on Alpine the cursor is in the middle of the prompt, or it has linebreaks in it so it goes over 3 lines. but just on the first prompt of a new tab 2026-01-05 13:29:26 you've put weird escape codes in it? 2026-01-05 13:29:47 lopid: PS1="\[\e[1;33m\][\t] \[\e[1;32m\]\u@\h \[\033[1;36m\]\w: \[\e[0m\]" 2026-01-05 13:33:27 what shells? 2026-01-05 13:33:59 both bash? 2026-01-05 13:34:35 on both (Fedora/Alpine) I use bash via qterminal, accessed via xrdp. on Alpine I also testet vt and sakura, same problem 2026-01-05 13:35:00 it looks like this: https://ibb.co/9HV6T4t7 2026-01-05 13:35:11 Are those codes compatible with the terminal you're using? 2026-01-05 13:35:40 Are you using a shell that interpets those PS1 extensions? 2026-01-05 13:36:20 well it is bash, so I guess so. it does work on Fedora, so I thought it would work on Alpine too 2026-01-05 13:36:31 (changed it for the screenshot to PS1="\[\e[1;33m\][time] \[\e[1;32m\]user@host \[\033[1;36m\]dir: \[\e[0m\]" 2026-01-05 13:37:13 is the rest of the terminal limited to 10 chars width or just that prompt? 2026-01-05 13:38:14 just the first prompt of the tab. when I hit return, it seems normal starting with the second line 2026-01-05 13:38:54 it also works fine if I remove the color codes 2026-01-05 13:39:54 worksforme™ 2026-01-05 13:39:58 it might be something output before the prompt. are the rest of your shell start up files ok? 2026-01-05 13:40:02 (bash/qterminal) 2026-01-05 13:40:19 everything else seems just fine 2026-01-05 13:40:21 On edge 2026-01-05 13:40:54 installed from the 3.23.2 iso 2026-01-05 13:42:18 does tput output the same sequence the the colours, on both hosts? 2026-01-05 13:42:29 for the 2026-01-05 13:43:31 hm, no tput... looking to install it 2026-01-05 13:47:19 what exactly should I do with tput? 2026-01-05 13:47:27 (never used it) 2026-01-05 13:48:03 eg., to see yellow: v=$( tput setaf 3 ); declare -p v 2026-01-05 13:48:36 or tput setaf 3 | od -ta 2026-01-05 13:48:43 ah, you mean change the PS1 to it uses tput instead of \e? 2026-01-05 13:48:48 oh 2026-01-05 13:49:12 using tput would be more portable that hard coding a terminal's escape codes 2026-01-05 13:49:31 scratch the od example 2026-01-05 13:50:01 actually, it's fine 2026-01-05 13:50:55 for yellow, I get the same output on Fedora and Alpine 2026-01-05 13:51:18 does it happen in other terminals? 2026-01-05 13:53:25 hm, seems ok in st and sakura 2026-01-05 13:55:20 ahh... removed the qterminal.conf and it did work...changed the font to liberation mono 10 and it acts up 2026-01-05 13:56:58 it seems to be the 10pt. at 12pt it works 2026-01-05 14:05:38 it's not font related, only size related. even the default font at 10pt causes it (for me) 2026-01-05 14:17:27 hm 2026-01-05 15:47:06 btw abcde doesnt work because MusicBrainz/DiscID.pm doesnt exist and isnt pacakged 2026-01-05 15:47:31 i tried using perl-webservice-musicbrainz, libmusicbrainz{,-dev} and libdiscid{,-dev} and none had it 2026-01-05 15:48:21 https://pkgs.alpinelinux.org/package/edge/testing/x86_64/perl-musicbrainz-discid of course its an edge package 2026-01-05 15:48:45 using cdparanoia myself instead 2026-01-05 19:39:56 lopid: I seem to have fixed it by adding "resize > /dev/null" into .bashrc. without it, $COLUMNS was 10 and $LINES was 2 for the first prompt. but for a font size >10pt, those two variables are empty 2026-01-05 20:52:00 weird 2026-01-05 21:46:55 ikke: Thank you. I've got the alpine/git docker image which I'm downloading and extracting in a format **runc** can use. But I'm very new to containers so I'd like to get input on what makes for a good, hardened, reliable setup for an appliance like this? My hope is to use the things I learn in this small case for other more complex ones. Thanks again and sorry about being a noob but I'm excited to learn and grow. 2026-01-06 05:59:44 Aangularity[m]: Containers are not considered a good security boundary. There are some tools to harden them, similar to what docker does, but you should still be careful with untrusted workloads 2026-01-06 06:31:52 ikke: Can you go into a little more detail about what docker does versus runc? I'm not against adding a VM around the container if that's best practice 2026-01-06 06:45:16 mostly things like seccomp, which are in the end facilitated through runc 2026-01-06 08:11:54 ikke: Interesting. I would have expected the security model to be similar between Docker and runc. So Docker has additional hardening. That's so surprising, given the amount of cve's it and Moby come with. Is there any documentation or info where I can learn what's missing in runc? 2026-01-06 08:12:41 Runc is a low-level tool, so you have to setup everything yourself 2026-01-06 08:17:09 Docker is a much broader tool, so a lot more attack surface. I'm not saying docker is necessarily more secure, but it does come with some hardening by default. 2026-01-06 08:17:46 A large improvement is to use rootless containers 2026-01-06 13:11:58 or use podman who runs rootless by default? 2026-01-06 13:13:47 That's included in 'using rootless containers' 2026-01-06 13:36:53 afaik Docker supports rootless too nowadays 2026-01-06 13:36:59 *iirc 2026-01-06 14:37:26 Indeed it does, but at least for our application, podman rootless was easier to setup 2026-01-06 14:48:08