jaeger changed the topic of #crux to: CRUX 3.7 | Homepage: https://crux.nu/ | Ports: https://crux.nu/portdb/ https://crux.ninja/portdb/ | Logs: https://libera.irclog.whitequark.org/crux/
<cruxbot> [opt.git/3.7]: rust: fixed a build error when $CARGO/registry/src does not exist
tilman has quit [Ping timeout: 256 seconds]
tilman has joined #crux
<remiliascarlet> jaeger: The rtmpdump has a broken source link. I checked the link out directly, and it seems like the MPV guys took that repository down. So I asked a friend of mine to host the file on her servers, so could you replace it in the Pkgfile? New link: https://076.moe/repo/src/3rdparty/rtmpdump-2.4-f1b83c10-4.tar.gz
<cruxbot> [contrib.git/3.7]: suitesparse: 7.3.1 -> 7.4.0
<cruxbot> [contrib.git/3.7]: zjstatus: 0.11.1 -> 0.11.2
<remiliascarlet> Same goes for the compat-32 version as well.
<jaeger> noted, though I'm not the maintainer
<farkuhar> remiliascarlet: if we added to pkgmk the ability to recognize git urls in the source array, then it wouldn't matter when an upstream project switches to a different web frontend (silently dropping support for snapshot tarballs).
<farkuhar> it's not a common occurrence for an upstream host to make such a switch, so pkgmk can get away with only supporting the schemes http(s) and ftp. In the uncommon event like what you found with rtmpdump, it's easy enough to "git clone and host a tarball ourselves".
<remiliascarlet> farkuhar: The point is that they took the actual repo itself down.
<remiliascarlet> So no amount of git cloning is going to help you here.
<farkuhar> remiliascarlet: They just removed the link from the landing page. You can still load the actual repo, https://git.ffmpeg.org/gitweb/rtmpdump.git
<remiliascarlet> Oh, I see.
mnkydeth has joined #crux
<mnkydeth> I recently built a new system around sapphire rapids. I went to install Crux. I am using the updated iso, as the original just gives me a black screen. I got through the entire install until the last step. Installing my bootloader. I have a partition formatted with mkfs.fat /dev/nvme0n1p1 mounted at /boot/EFI. It tells me my partition is unrecognized and will not install. I have not tested the updated iso on another computer so I assume the install media i
<ukky> mnkydeth: what is the output for 'blkid /dev/nvme0n1*' ? What is your BIOS boot type, UEFI or Legacy? Are you installing Grub or Syslinux?
<mnkydeth> UEFI, the blkid is normal. I mean, I've been installing Crux for years on my computers. The partition shows up as vfat in blkid if I recall. I always use grub. I've never used syslinux.
<ukky> Is it grub-install that refuses to recognize your partition?
<mnkydeth> Correct
<ukky> MBR or GPT?
<mnkydeth> Gpt
<ukky> UEFI boot?
<mnkydeth> Yes, correct
<ukky> Did you set proper partition type, 'EFI System' if using fdisk?
<mnkydeth> I was thinking it might be something with secure boot. But from what I can tell I have disabled everything in the bios that I could relating to it.
<mnkydeth> I set it in cfdisk. Yes
<mnkydeth> Hmm, I wonder if I should use fdisk instead. ... I've seen weirder things happen over the years I suppose.
<ukky> I remember previous grub (pre 2.12) had issues with some file system properties. Forgot which one, searching...
<ukky> metadata checksum seed was not handled properly
<mnkydeth> I've tried mkfs.fat , mkfs.fat -F 32, and mkfs.vfat on that partition. Same error
<ukky> Which grub version do you have now?
<mnkydeth> It's what's on the updated iso from crux.ninja I assume it's 2.06 but I'm not sure.
<ukky> Yes, 2.06 is affected. Not sure if it is a patched version though.
<ukky> I am looking if there is a workaround available for your case
<mnkydeth> Well, I guess I could try and learn syslinux. I'll have to do some reading I guess. :)
<ukky> Do you have separate /boot partition?
<mnkydeth> My partition is mounted at /boot/EFI
<ukky> Grub installs its files at /boot/grub/. UEFI payload will be placed under /boot/EFI
<mnkydeth> Hah, if it's mounting it to /boot instead .. I never would have thought. Pretty sure I always mounted to /boot/EFI ... But it's been a good while since I last did a fresh install.
<ukky> What partioning scheme do you have? I mainly interested in /, /boot, and /boot/efi
<mnkydeth> I have /dev/md0 as / I have /dev/nvme0n1p1 as /boot/EFI
<mnkydeth> I run a Linux raid for / and /home ... /dev/md1 is /home
<ukky> What FS type is on /dev/md0?
<mnkydeth> xfs
<ukky> Never used xfs. But this might be an issue, i.e. grub might not recognize xfs. Do you have some spare space on your NVM device for separate /boot partition?
<mnkydeth> Yeah, I'll redo my /boot/EFI as /boot I stead.
<mnkydeth> Thanks for the help ukky.
<ukky> But you still need separate EFI partition for UEFI BIOS, that is where boot sequence starts from on UEFI platforms.
<mnkydeth> Oh I see yeah. Easy fix then. The way I do my drives I waste some space. I should revise the way I normally do it I suppose.
<mnkydeth> I have four NVMe drives. The first two I partition off some space for the EFI and swap. I can get rid of the swap space and move to using a file instead. Make that swap space the /boot
<ukky> It might help if you reserve two separate partitions for grub: /boot and /boot/efi.
<mnkydeth> Yep, /dev/nvme0n1p1 will be /boot/EFI and /dev/nvme1n1p1 will be /boot
<mnkydeth> I can try ext2 or 4 for /boot instead of xfs.
<ukky> According to wiki, grub supports xfs. But maybe 2.06 does not support some FS options.
<mnkydeth> I'll experiment a little. You pointed me in the right direction for something new to try. And it's much appreciated.
<ukky> Personally, I use ext4. But make sure to remove 'metadata_csum_seed' feature, as grub-2.06 doesn't support it.
<mnkydeth> Kk
<ukky> After you format /boot, test for 'metadata_csum_seed' feature: tune2fs -l /dev/nvme1n1p1 | grep metadata_csum_seed
<ukky> And to remove 'metadata_csum_seed': mkfs.ext4 -O ^metadata_csum_seed /dev/nvme1n1p1
<mnkydeth> Ok, much appreciated
<ukky> I have to leave, good luck
<mnkydeth> Thank you
<ukky> you are welcome
<jaeger> the grub2 port had a patch for that in version 2.06, removed in 2.12 as it seemed to be already applied but I can double check that tomorrow
<jaeger> And the updated ISO hasn't been updated with 2.12 yet (that's happening currently)
ppetrov^ has joined #crux
lavaball has joined #crux
deltahotel has joined #crux
deltahotel has quit [Client Quit]
zorz has joined #crux
lavaball has quit [Remote host closed the connection]
tarxvfz has joined #crux
<cruxbot> [opt.git/3.7]: gnuplot: update to 6.0.0
<cruxbot> [opt.git/3.7]: newsboat: update to 2.34, optional dependency asciidoctor
<cruxbot> [core.git/3.7]: inetutils: update to 2.5
remiliascarlet has quit [Quit: remiliascarlet]
remiliascarlet has joined #crux
<cruxbot> [contrib.git/3.7]: python3-rapidfuzz: 3.6.0 -> 3.6.1
<cruxbot> [contrib.git/3.7]: python3-tzdata: 2023.3 -> 2023.4
<cruxbot> [contrib.git/3.7]: sdl_gfx: update to 2.0.27
<cruxbot> [contrib.git/3.7]: sdl2_image: update to 2.8.1
<cruxbot> [contrib.git/3.7]: fltk: 1.3.8 -> 1.3.9
<cruxbot> [contrib.git/3.7]: python3-lxml: 4.9.4 -> 5.0.0
zorz has quit [Quit: WeeChat 4.1.2]
<cruxbot> [opt.git/3.7]: newsboat: update to 2.34
lavaball has joined #crux
<cruxbot> [opt.git/3.7]: neovim: 0.9.4 -> 0.9.5
zorz has joined #crux
joe9 has joined #crux
ivandi has quit [Quit: WeeChat 4.1.2]
ivandi has joined #crux
ivandi has quit [Client Quit]
ivandi has joined #crux
ivandi has quit [Quit: WeeChat 4.1.2]
ivandi has joined #crux
tarxvfz has quit [Ping timeout: 240 seconds]
ppetrov^ has quit [Quit: Leaving]
magnahelix has left #crux [The Lounge - https://thelounge.chat]
joe9 has quit [Quit: leaving]
<cruxbot> [contrib.git/3.7]: rav1e: 0.6.6 -> 0.7.0
<cruxbot> [contrib.git/3.7]: p5-image-exiftool: 12.71 -> 12.72
<cruxbot> [contrib.git/3.7]: p5-datetime-timezone: 2.60 -> 2.61
<cruxbot> [contrib.git/3.7]: protontricks: 1.10.5 -> 1.11.0
<cruxbot> [opt.git/3.7]: fontconfig: updated symlink
zorz has quit [Quit: WeeChat 4.1.2]
mnkydeth has quit [Ping timeout: 276 seconds]
mnkydeth has joined #crux
lavaball has quit [Remote host closed the connection]