ChanServ changed the topic of #kisslinux to: KISS Linux | https://k1sslinux.org | /msg zr for kisslinux/* cloaks | logs: https://k1sslinux.org/irc#2.0 | thing of the day: https://vid.puffyan.us/watch?v=HVzC6WZImGY
<Guest68> So if the base dir is /usr/lib/firmware, the amdgpu files would be: CONFIG_EXTRA_FIRMWARE="a.ucode amdgpu/amdgpu1 amdgpu/amdgpu2" and so on?
<claudia> yep
<Guest68> Ok
<claudia> This exactly the gist.
<claudia> So bad, that whietequark cant search through the entire history
<claudia> I would exactly send this snippet to the next person questioning this.
<claudia> logbot was pretty comfortable.
<Guest68> Oh wait, I guess I could just type amdgpu/polaris10_{ce,ce_2,k_smc,k2_smc,k_mc,mc,me,me_2,mec2,mec2_2,mec,mec_2,pfp,pfp_2,rlc,sdma1,sdma,smc,smc_sk,uvd,vce}.bin
<Guest68> Would the .config accept this syntax?
<claudia> it will error on you if so.
<Guest68> Dang so I have type them out :(
<claudia> dilyn, do you have your .config handy? I think I can add an example to our wiki/firmware
<claudia> *.config with amdgpu
<dilyn> i've fixed webengine, claudia
<dilyn> I hate it tho :)
<claudia> this is quite the example. ty
<dilyn> lol ofc
<claudia> which gpu is this?
<dilyn> wx2100
<claudia> Oh this is good new. very nice.
<claudia> you fixed it for now :v
<dilyn> i haven't looked closely at which firmware files are required for this specific card tho
<dilyn> yeah it's new enough... the price rose by $50 so if I could nab a 6800 for MSRP I can make a few bucks on this card xD
<dilyn> yeah it's a very unpleasant package. IDK why they have to make this so hard
<dilyn> hopefully the bug report that was filed gets some actual attention :|
<Guest68> Typing this out is painful I don't have copy paste :D
<dilyn> you can edit the .config manually to add them
<dilyn> it's safe in this case haha
<Guest68> Okay
<Guest68> Can I test for errors too or will they come up when I compile
<dilyn> there's no good way to identify errors outside of compiling :\
<dilyn> maybe we should start recommending that users use something like arch's kernel and tinyramfs lmfao
micro_O has joined #kisslinux
<Guest68> Okay gonna start compiling the kernel now lets see if it works
<Guest68> Wow 5 seconds in and it already failed...
<konimex> jedavies: unfortunately, nope
<Guest68> Getting errors in /usr/include/linux/byteorder/little_endian.h. No idea
<dilyn> five seconds is better than twenty minutes...
micro_O has quit [Ping timeout: 252 seconds]
<Guest68> Ok that seems to have worked. Im going to go eat while this compiles
<dilyn> why do distributions pull this bs https://repology.org/projects/?inrepo=siduction_experimental
<claudia> Maybe a smaller example would be better fitting but is this comprehensive?
<dilyn> throw off everyone's repology stats because you decided not to track upstream's official releases smh
<claudia> so bad that there is no other musl libressl latest qt5 distro to spy on patches..
<dilyn> yeah. it's just me out here (:
<Guest68> "bzimage is ready" means that it compiled with no issues?
<dilyn> yessir!
<acheam> could openbsd patches work? obvs not musl, but musl is closer to the BSD libc than to glibc
<claudia> obsd is on older qt5 version ;)
<dilyn> yeah obsd is where I initially go the libressl patches from
<dilyn> they're on an almost similar version now tho! keeping up with kde dev work from what I can tell
<dilyn> but they don't have any patches for this problem. nobody does, in fact
<dilyn> which is why I think i'm missing something.
<dilyn> harump. who was I looking at...
<claudia> There is some huge work to update to qt5 stack and some kde stuff https://undeadly.org/cgi?action=article;sid=20210124113220
<claudia> Frome the tone of this post, there is also somebody doing this with their notebook ^^
<claudia> dilyn, you should get in touch and tell them that a faster computer does not solve the problem :D
<dilyn> > I can build it 1-2 times a day which makes testing a little bit annoying and time intensive.
<dilyn> can relate! lowkey built this computer so i could build webengine in < 30 minutes
<dilyn> faster computers only fix some problems for sure :v
<dilyn> if bsd cannot support plasma I can never use bsd :'(
<claudia> freebsd does.
<claudia> they also do wayland
<dilyn> yeah but freebsd is not the way
<claudia> there is no bsd
<claudia> they are totaly different :D
<dilyn> lol
<dilyn> openbsd == bsd, change my mindn
dilyn has quit [Quit: Connection closed]
<claudia> netbsd == bsd
<claudia> netbsd != openbsd
dilyn has joined #kisslinux
<dilyn> i appear to have angered the gods. i take it all back.
<claudia> yyy
<dilyn> oh... hmmm... hrngg....
<dilyn> this better not...
<dilyn> my head might explode here. one sec
<illiliti> Unfortunatelly, the Wayland development is strongly focused on Linux.
<illiliti> the dark side of wayland
<claudia> netbsd prob can do plasma, because they can do wayland(protocols) so the apps should build maybe.
<dilyn> honestly I'd rather *BSDs solved this problem in their own way
<dilyn> they do good work. their devs are wicked smart (usually). could they come up with something better than wayland?
<claudia> I havent watched openbsd over a long time, but from what I can tell they just sit it out and dont care.
<dilyn> lol
<claudia> because xorg works. no reason to fix. we need our dev power on important things.
<dilyn> dammit. >I am the stupidest fucking trash in the multiverse.
<claudia> context
<dilyn> remember this whole fiasco with qt5? "# Without this folder, Qt can't find its header files."
<dilyn> guess WHAT
<dilyn> realized I did not apply this fix to webengine.
<dilyn> all my time, wasted.
<acheam> kernel 5.13 is out
<claudia> gee
<claudia> gj for realising this.
<claudia> But I thought you already fixed webengine?
<claudia> Or you just thought
<dilyn> I did, but that entire patch of work I did is exactly equivalent to making a .git directory
<dilyn> my blessed KDE devs made some super dumb breaking changes when they forked Qt 5 :'(
<acheam> includes landlock.io and clang CFI enforcement among other things
<acheam> I wonder if they fixed the broken buold
<acheam> I might upgrade from 5.10.44
<illiliti> landlock is cool. i'm planning to implement it into king
<dilyn> that's interesting...
<claudia> dilyn, you are an absolute madman and pull my head that you fight this mess, still.
<Guest68> I'm back
<dilyn> the maddest of lads
<dilyn> it gave me something to brag about on my cover letter. that's why I did it. lmfao
<dilyn> if anybody happens to use qt5-webengine and could test if falkon/viper-browser/something works with dtls disabled in qt5, that would be wonderful to hear about
<claudia> I will prob do this tomorrow.
<dilyn> tyty <3
<Guest68> Is there a genfstab-like command for KISS
<dilyn> cem has one somewhere
<dilyn> a very interesting bug in `kiss`. without rust installed, and the bin repo before the main repo in $KISS_PATH, if you do `cd repo/extra/firefox`; `kiss b`, `kiss` will build the rust in repo, not the bin repo
<dilyn> how cute
<illiliti> this is intentional and called CRUX-like usage
<dilyn> I know that
<dilyn> but I thought it was only for the package who's directory you're in
<dilyn> not for the whole repository
<illiliti> oh, i misread
<illiliti> can you provide reproducible example?
zenomat has quit [Ping timeout: 244 seconds]
zenomat has joined #kisslinux
zenomat has joined #kisslinux
zenomat has quit [Changing host]
<Guest68> Okay I think I just need to get encryption to work at boot now
<Guest68> illiliti Srry I can't remember if you said to use tinyramfs or kiss-encryption for encrypted root
<illiliti> both
<Guest68> Ok
<illiliti> tinyramfs to mount /
<illiliti> kiss-encryption to properly unmount /
<dilyn> sure can. clone $/kiss-community/{repo,bin}. export KISS_PATH=$PWD/bin/bin:$PWD/repo/extra. cd repo/extra/firefox. kiss r rust. kiss b firefox. watch as kiss rebuilds rust
<Guest68> Dang just remembered I forgot to add dm-crypt to the kernel......
<Guest68> Anybody here know what parameter that is?
<illiliti> run `make menuconfig`, press /; type dm_crypt
<Guest68> Im looking in Device Drivers -> Multiple devices driver support (RAID and LVM) -> Device mapper support just as the location says but it's not there
<Guest68> Oh nvm
<Guest68> It's Crypt target support i think
claudia has quit [Quit: Leaving]
<illiliti> dilyn: can reproduce!
<illiliti> test -d /home/illiliti/projects/kisslinux/repo-main/extra/firefox/bin/bin/rust
<illiliti> something weird happened lol
<dilyn> lol
<Guest68> Any special kernel parameters Id need to set to use DisplayPort audio from my graphics card?
<Guest68> The onboard audio on my motherboard is broken
necromansy has joined #kisslinux
<Guest68> The DP cable goes to my monitor to which I have headphones plugged in
<dilyn> you need to have HDMI audio enabled in the kernel, and then you can use something like alsamixer to unmute the channel
<Guest68> This one I take it CONFIG_SND_HDA_CODEC_HDMI HDMI/DisplayPort HD-audio codec support
<dilyn> yessir
<Guest68> Also all my amdgpu files got overwritten for some reason D:
<Guest68> In the config, that is
<illiliti> CRUX-like usage gives priority to current directory/repository regardless of dependencies
<dilyn> yeah exactly
<illiliti> we need to disallow CRUX-like usage for dependencies
<dilyn> that seems like the most straightforward way
<dilyn> firefox 89.0.2 built perfectly with no changes. I must be in heaven.
<dilyn> and now I play house sitter to water my girlfriend's plants :o gngn everyone
dilyn has quit [Quit: Connection closed]
<acheam> but wee need babysitting too!
<illiliti> i hope posix will standardize readlink
<illiliti> because parsing ls output and doing some magic is ridiculous
<Guest68> illiliti Hi, would you mind sharing your tinyramfs config file and grub config :D
<illiliti> here: https://termbin.com/lydd
<illiliti> i don't use grub
<Guest68> Well, just the command parameters for whatever you're using
<Guest68> kernel parameters
<illiliti> vmlinuz initrd=\initramfs quiet i915.fastboot=1
<Guest68> Okay ty
<illiliti> there's also tinyramfs man pages if you didn't know
<illiliti> man tinyramfs.config
<Guest68> Yes im reading them now
<illiliti> they poorly written though
<illiliti> i need to rewrite/revisit them at some point...
<Guest68> So root would be /dev/sdc2 and luks_root would be /dev/mapper/crypt
<Guest68> Or do I have those backwards
<illiliti> yes
<illiliti> backwards
<illiliti> /dev/mapper/crypt is real unlocked root
<Guest68> Huh, weird. The UUID for /dev/mapper/crypt isn't showing up when I type blkid
<illiliti> make sure that you correctly formatted /dev/mapper/crypt
<illiliti> also, run blkid as root
<Guest68> I see a weird device called /dev/dm-0
<Guest68> With file type ext4
<illiliti> this is exactly what you need
<Guest68> When I type nlkid /dev/mapper/crypt it shows up with the same UUID as the /dev/dm-0
<Guest68> Not sure why it's labelled as such tho
<illiliti> /dev/mapper/crypt symlink doesn't created for some reason
<illiliti> wait
<illiliti> /dev/mapper/crypt is a symlink to /dev/dm-0, right?
midfavila-laptop has joined #kisslinux
<Guest68> Apparently it is logical volume but I don't remember setting it up that way
<illiliti> answers are false
<illiliti> this is not lvm unless you did setup it
<illiliti> this is unlocked luks container
<Guest68> Yeah I didn't touch any physical volume, volume group, logical volume settings
<Guest68> I'll just use the UUID for dm-0 and see if it works I guess
<illiliti> yeah
<illiliti> dm-0 is exactly what you need
<Guest68> Also, what's your output for uname -r? Mine is showing as the Arch iso I am on
<illiliti> 5.12.5
<Guest68> oh okay the version num
<illiliti> 1
<illiliti> what are you trying to do?
<Guest68> tinyramfs -o "/boot/initramfs-$(uname -r)"
<Guest68> Am getting the error cp can't stat /lib/modules/5.11.16-artix1-1/modules.order
<Guest68> Because I am on an Artix Linux iso and that's supposed to be just 5.12.13
<Guest68> And /lib/modules/5.12.13 does exist btw
<Guest68> Not sure why it's coming up as artix
<illiliti> you need to explictly tell tinyramfs which kernel to use
<illiliti> tinyramfs -k 5.12.13
<Guest68> thanks
<Guest68> Btw, do I need to specify a key or will I be prompted at boot?
<Guest68> To type in the passphrase
<illiliti> you will be prompted you don't setup key
<illiliti> the key is optional
<illiliti> you can store your password in file and use it as key
<illiliti> or you can create key with random data, add it via cryptsetup and use it in tinyramfs
<Guest68> I prefer password
<Guest68> And I think I am done... gonna reboot and see what happens
<Guest68> Dang got a kernel panic :(
<Guest68> VFS: unable to mount root fs on unknown-block(0,0)
Uks3 has joined #kisslinux
Uks2 has quit [Ping timeout: 268 seconds]
<Guest68> Oh was I supposed to type in update-grub before rebooting
<illiliti> yeah
Uks3 has quit [Client Quit]
Uks2 has joined #kisslinux
schillingklaus has joined #kisslinux
Uks2 has quit [Ping timeout: 268 seconds]
<Guest68> So I did actually generate the grub.cfg
<Guest68> I am not sure why it's giving me a kernel panic
Uks2 has joined #kisslinux
<acheam> are your kernel parameters right? did you compile in support for the filesystem and stuff?
<Guest68> Is there an ext4 or rootfs kernel parameter
<Guest68> EXT4_FS=y
<illiliti> looks like kernel didn't use parameters because they aren't defined
<illiliti> are you sure that grub passed parameters to kernel?
<Guest68> I think so
<illiliti> send me your kernel and grub config
<Guest68> kernel config https://pastebin.com/pNd8W8Gy
<illiliti> grub.cfg doesn't contain kernel parameters
<Guest68> line 104
<Guest68> And here's fstab https://pastebin.com/tfqMADtb
<illiliti> UUID=e3e7c07e-f5b8-4113-af4b-8a96df707e51 / ext4 rw,relatime 0 1
<illiliti> remove this
<Guest68> Ok
<illiliti> the problem is that grub doesn't save kernel parameters
<illiliti> how did you configure grub?
<Guest68> I typed grub-mkconfig -o /boot/grub/grub.cfg
<Guest68> After editing /etc/default/grub
<illiliti> send me /etc/default/grub
dylanaraps has joined #kisslinux
* midfavila-laptop waves
<Guest68> Under GRUB_CMDLINE_LINUX_DEFAULT are the kernel parameters
<Guest68> Here they are talking about compiling support for ext4 which I've made sure is set to yes https://old.reddit.com/r/kisslinux/comments/mljopc/kiss_linux_kernel_panic_unable_to_mount_root_fs/
<Guest68> And I did not find a rootfs setting
<illiliti> no no, you're using wrong parameters
dylanaraps has quit [Ping timeout: 268 seconds]
<illiliti> you need to pass: initrd=\<location of generated initramfs on /boot>
<illiliti> that's the main parameter
<illiliti> cryptdevice is meaningless to tinyramfs
<Guest68> So in the grub I just say vmlinuz initrd=\boot\initramfs-5.12.13
<Guest68> vmlinuz initrd=\initramfs-5.12.13
<illiliti> yes
<illiliti> exactly
<Guest68> Since /boot/initramfs-5.12.13 is the name
<illiliti> yep
<Guest68> Ok I will try that and report back
<illiliti> vmlinuz initrd=\initramfs-5.12.13
dylanaraps has joined #kisslinux
<schillingklaus> is chainloading possible?
<illiliti> what do you mean?
<Guest68> It has the parameters /vmlinuz-5.12.13 root=/dev/mapper/cry[t ro vmlinuz initrd=\initramfs-5.12,13
<Guest68> Despite my typing just the vmlinuz initrd=\initramfs-5.12.13 part
<Guest68> After generating the grub config file again
<Guest68> Should I take out the root=/dev/mapper/crypt part manually?
<schillingklaus> something like set root=(hd0,1) followed by chainloader +1
<illiliti> Guest68: it's harmless but you can remote this
<illiliti> schillingklaus: idk
<Guest68> Ok ill try rebooting again
<Guest68> Same error but more output. I can see eight penguins at the top of the screen...
<illiliti> post output
<illiliti> this is important
dylanaraps has quit [Quit: dylanaraps]
<Guest68> Dang can't upload picture on my POS phone
<Guest68> i will brb
Guest68 has quit [Quit: Client closed]
Guest12 has joined #kisslinux
<Guest12> Here is the image https://ibb.co/DLZ8mjk
<Guest12> Sorry for the delay my phone is a POS and deleted my photos
<illiliti> ah
<illiliti> you need to remove root= parameter
<Guest12> When I did that it says vmlinuz file not found
<Guest12> error: invalid file name 'vmlinuz'
<Guest12> Because of the parameter vmlinuz initramfs blah blah
<Guest12> The one you told me to put in
<illiliti> vmlinuz initrd=\\initramfs-5.12.13
<illiliti> try this
<illiliti> or: vmlinuz initrd=/initramfs-5.12.13
<Guest12> Ok
<Guest12> I guess it was complaining about the vmlinuz part tho
<Guest12> Same error with two backslashes, will try a forward one
<illiliti> afaik grub should automatically detect initramfs
<Guest12> ohhhh
<Guest12> i think vmlinuz should be vmlinuz-5.12.13 , no?
<Guest12> What's your vmlinuz file named as?
<illiliti> i don't use grub
<illiliti> try to rename vmlinuz to vmlinuz-<kernel> and initramfs to initrd-<kernel>.img
<illiliti> and run grub-mkconfig
<testuser[m]1> Hi
<Guest12> Did /vmlinuz-5.12.13 initrd=/initramfs-5.12.13
<Guest12> It says cannot open root device "(null)" so I will leave the other stuff in there too
<illiliti> try to do what i say above
<Guest12> Ok so change to vmlinuz-<kernel> and initrd-<kernel>.img in both parameters and /boot?
<illiliti> yes
micro_O has joined #kisslinux
<Guest12> What do you want me to put as the parameters? Any slashes or anything for vmlinuz? Backwards/forwards etc
<illiliti> no no, i mean rename files
<Guest12> Okay the file names are vmlinuz-5.12.13 and initrd-5.12.13.img
<illiliti> good
<Guest12> When I run grub-mkconfig it adds a separate line with "initrd /initrd-5.12.13.img"
<Guest12> Should I take out the initrd=\ ... parameter then?
<illiliti> no
<Guest12> Ok so its initrd=\initramfs-5.12.13
<Guest12> vmlinuz initrd=\initramfs-5.12.13
<illiliti> drop the all parameters under GRUB_CMDLINE_LINUX_DEFAULT
<illiliti> and re-run grub-mkconfig
<Guest12> kk
<Guest12> There are two lines: first is "linux /vmlinuz-5.12.13 root=/dev/mapper/crypt ro" and the second is "initrd /initrd-5.12.13.img"
<Guest12> In grub.cfg
<Guest12> After running grub-mkconfig with no parameters under GRUB_CMDLINE_LINUX_DEFAULT
<illiliti> drop 'root=/dev/mapper/crypt ro'
<illiliti> and try to boot
<Guest12> kk
<schillingklaus> my GRUB_CMDLINE_LINUX_DEFAULT is "quiet nomodeset"
<Guest12> Ok we made some progress
<Guest12> There was a bunch of text but it eventually said Enter passphrase for /dev/disk/by-uuid/53d4...
<Guest12> But there was more initialization text and didnt let me type in password
<Guest12> Last line is !! failed to unlock LUKS
<Guest12> Will send you a picture
<illiliti> don't afraid to type password
<illiliti> the password will get to cryptsetup regardless of initialization text
<illiliti> just like sudo/doas
<illiliti> yeah
<Guest12> You can see it says Enter passphrase but it is quicky overridden by more text
<Guest12> I tried typing in my password but it threw me into a shell
<illiliti> you can still type password
<illiliti> try to type it again
<Guest12> I am in a shell and it says command not found
<illiliti> don't rush
<illiliti> reboot
<Guest12> kk
<Guest12> Blinking cursor after text is done initializing
<Guest12> Will type in my pass now
<Guest12> Yeah it still says failed to unlock LUKS
<Guest12> device-mapper: table: 253:0: crypt: Error allocating crypto tfm
<Guest12> device-mapper: ioctl: error adding target to table
<illiliti> send output
<schillingklaus> tells me to not even think about considering that LUKS thing
<illiliti> something wrong with kernel
<schillingklaus> why is LUKS so popular?
<illiliti> schillingklaus: because there's no good alternative
<illiliti> Guest12: your kernel missing some essential crypt modules
<Guest12> Okay
<Guest12> I made sure dm-crypt enabled, I will see what others there are
<illiliti> XTS is required too iirc
<Guest12> Am i looking at pkg_setup() here?
<illiliti> yes
<illiliti> > Enabling cryptographic API functions
<Guest12> CRYPTO_XTS only one disabled from that list
<Guest12> I will turn it on and recompile
<illiliti> good
<Guest12> There are a bunch of others such as cfb cts lrw ofb pcbc that are not on
<Guest12> Should i touch those or no?
<illiliti> no
gtms has joined #kisslinux
<Guest12> Trying again with xts
<Guest12> Error is !! failed to lookup partition
<Guest12> After typing in the passphrase
dylanaraps has joined #kisslinux
<illiliti> send your tinyramfs config and blkid output from that shell
<illiliti> dylanaraps: hi!
<Guest12> hi
<Guest12> https://ibb.co/DpSMP2F heres the blkid
<Guest12> will get tinyramfs in a sec
schillingklaus has quit [Quit: schillingklaus]
<dylanaraps> illiliti: o/
<Guest12> real root is /dev/dm-0 i take it and luks_root is /dev/sdc2
<illiliti> yes, config is correct
<illiliti> run 'echo /dev/disk/by-*' in shell
micro_O has quit [Ping timeout: 272 seconds]
<Guest12> by-id/ by-label/ by-partuuid/ by-path/ by-uuid/ is the output
<illiliti> no no, i mean in shell after 'failed to lookup' error
<illiliti> not livecd
<Guest12> Ok
<illiliti> and run this too: 'echo /dev/disk/by-uuid/*'
<Guest12> It says /dev/disk-by-label /dev/disk/by-uuid
<Guest12> I meant /dev/disk/by-label
<illiliti> and the second command?
<illiliti> echo /dev/disk/by-uuid/*
<Guest12> "/dev/disk/by-uuid/53d4beaf-225d-41cb-841d-72ad7b2d990c /dev/disk/by-uuid/7f1b314b-36f8-4af9-bfd9-318e7147f5c9 /dev/disk/by-uuid/B1BB-21B8 /dev/disk/by-uuid/D4E8-9CC9"
<Guest12> Not sure what the last one is, it doesn't show up in blkid
<illiliti> did you compile kernel with CONFIG*=y ?
<illiliti> i mean kernel shouldn't have CONFIG*=m
<illiliti> check this
<Guest12> Ok I will check
<Guest12> This turns on all with CONFIG_
<Guest12> ?
<illiliti> yes
<Guest12> Can I find this in menuconfig or .config manually?
<Guest12> Don't see CONFIG* anywhere
<illiliti> by CONFIG*, i mean all CONFIG
<Guest12> I don't see any with m'
<Guest12> m's
<Guest12> There are a bunch commented out, e.g. # CONFIG_CRC8 is not set"
<Guest12> But none with m's
<illiliti> grep '=m' .config
<Guest12> Yes there are 13 with m
<illiliti> find them in menuconfig and turn them into '=y'
<Guest12> Ok
<illiliti> recompile kernel, regenerate initramfs
<illiliti> before regenerating initramfs, add 'monolith=1' to tinyramfs config
<Guest12> I think it could be EFIVAR_FS=m causing an problem
<Guest12> It say EFI variable filesystem
<Guest12> I will turn all on to be safe
<illiliti> dylanaraps: cool! i don't like this *fullpath function too.
<illiliti> we could actually have multilib support without ldd. the problem is that there's no way to parse symlink without hacks
<Guest12> Ok I will try again
<Guest12> I turned them into y's and added monolith=1
<Guest12> Regenerated initramfs and all that
<Guest12> And same error...
<illiliti> that's odd
<illiliti> try to change root option to /dev/dm-0 in tinyramfs config
<Guest12> I am in :D
<Guest12> root=/dev/dm-0 did the trick
<illiliti> finally, great
<testuser[m]1> nice
<illiliti> /dev/dm-0 is a hack though
<Guest12> I know
<illiliti> but at least it's good to have bootable system lol
<illiliti> you can fix that later
<Guest12> Yes
dylanaraps has quit [Quit: dylanaraps]
Uks2 has quit [Quit: Byee]
Uks2 has joined #kisslinux
Guest12 has quit [Quit: Client closed]
Uks2 has quit [Quit: Byee]
Uks2 has joined #kisslinux
micro_O has joined #kisslinux
LordRatte has joined #kisslinux
micro_O has quit [Ping timeout: 252 seconds]
LordRatte has quit [Quit: Quit]
Uks3 has joined #kisslinux
Uks2 has quit [Ping timeout: 268 seconds]
jleightcap has quit [Ping timeout: 272 seconds]
jleightcap has joined #kisslinux
dilyn has joined #kisslinux
dylanaraps has joined #kisslinux
<dylanaraps> o/
<acheam> hey Dylan
<testuser[m]1> hi
<acheam> glad to see you back! been using kiss since the fall, its seriously great software!
<omanom> welcome back, dylan! good to see ya!
<acheam> I feel the most in control of my computer ever
Uks3 has quit [Ping timeout: 268 seconds]
Uks2 has joined #kisslinux
<dylanaraps> Glad you find it useful :)
smartin has joined #kisslinux
<dilyn> dylan: I'll run these outstandings PRs through some ringers today after work
<dilyn> I still have no clue how GitHub's interface works
<dylanaraps> dilyn: All good. Take your time. :)
<dilyn> well my interviewers are radio silent until next week so instead of going crazy with anxiety... ;)
<dylanaraps> heh
<ang> dylanaraps: welcome back
<testuser[m]1> $< is used to get the first dependency of a target, how do you list all the dependencies ?
<testuser[m]1> Makefile*
<dylanaraps> ang: Thanks
<testuser[m]1> ah its $?
<dilyn> of course this strictly faster way of doing things breaks stuff
<dilyn> fast, efficient, or safe. pick two, get none
<testuser[m]1> what broke ?
<dylanaraps> See my link
<dilyn> Only things in master :)
<dilyn> these haven't been tagged yet
<testuser[m]1> oh i just downloaded the .patch , didnt open on shithub
<dilyn> lol
<testuser[m]1> dylanaraps: btw did you get to know about https://github.com/illiliti/king ?
<dylanaraps> No, haven't seen it before
<dylanaraps> Go version of the kiss package manager? Sounds neat.
<acheam> testuser[m]: if you can avoid it, I wouldn't use those special variables
<acheam> because they are GNU specific
<acheam> or call the Makefile GNUmakefile
<acheam> it will still be found by gnu make
<testuser[m]1> it seems to work fine with bmake too
<testuser[m]1> the $?
<acheam> maybe $< doesnt work then? but iirc many of the special variables dont work on bmake
<testuser[m]1> both work , ive only had issues with `ifeq` statements
<acheam> hm weird
<acheam> I wonder if I'm using a different version of bmake than you
<acheam> version/implementations
<konimex> it might don't work with *older* versions of bmake (<2021 versions perhaps?), since I had to patch libffi with $< last year
<konimex> anyway, time to update bmake
<testuser[m]1> what should you use in place of $< ? Add it manually ?
<testuser[m]1> bmake seems pretty limited :/
<dilyn> what all do you want from a `make` :P
<dilyn> if my make can read the shit autoconf produces, I don't want it
<acheam> testuser[m]1: it is, and it leads to less flexible makefiles
<acheam> konimex: ah got it, I think I kiss forked it from you a few months ago, so maybe due for an upgrade
<acheam> dilyn: make? don't you mean ./make
<acheam> ./make is the only build system permitted in this channel
dylan02 has joined #kisslinux
<konimex> I think there's only one implementation of bmake (in that crufty.net site which I use), which uses NetBSD make as a base iirc
<konimex> what's your bmake version?
<testuser[m]1> in my case you need gmake to build a bundled dependency anyway so i guess its OK :p
dylanaraps has quit [Ping timeout: 272 seconds]
<illiliti> iirc $< is a posix thing
<acheam> konimex: will report back later, am on my mobile
<illiliti> i can't believe that bmake doesn't support it
<acheam> maybe im just misremembering, or there was something else wrong with the rule that was causing it not to build with bmake
<acheam> but if I am wrong, I have some make files to rewrite lol
dylanaraps has joined #kisslinux
micro_O has joined #kisslinux
dylan02 has quit [Ping timeout: 272 seconds]
micro_O has quit [Ping timeout: 272 seconds]
<konimex> dilyn: autoconf, for all its shit is actually quite resilient, if there's no special GNU make things it should be readable by bmake
<illiliti> ^
<nerditup> dylanaraps: welcome back
<dilyn> it's fine usually but man oh man are the makefiles it makes sometimes the most unreadable noise...
<dylanaraps> nerditup: Thanks
micro_O has joined #kisslinux
dylanaraps has quit [Ping timeout: 272 seconds]
midfavila has quit [Remote host closed the connection]
midfavila has joined #kisslinux
konimex[m] has joined #kisslinux
<konimex[m]> i'll take unreadable noise over not working makefiles all day
<clemens3> working makefiles are kewl, but there is a point where exponentially growing unreadable source is worse then some binary..
<clemens3> I wonder the average linux distro contains how many lines of source..
<clemens3> maybe not a billion yet..
<clemens3> 100 million I guess
claudia has joined #kisslinux
<testuser[m]1> wont be hard to do on kiss, just make a script to untar every tarball in kiss/sources and call tokei
<testuser[m]1> or whatever tool
mahmutov has joined #kisslinux
dylanaraps has joined #kisslinux
<dylanaraps> > just make a script to untar every tarball in kiss/sources and call tokei
<dylanaraps> That doesn't work when software is cross-platform and doesn't take into account ifdefs, etc.
<rio6> contained source as in patched and build files or just combined source line of all packages
mahmutov has quit [Ping timeout: 258 seconds]
<claudia> dilyn: I build qt5-webengine successfully with gcc.
<dilyn> tyty
<dilyn> we should just package gn :v
<claudia> But since my production machine is with clang only I cant test it :p
<claudia> Did you happen to build qt5 with clang?
<claudia> I have found "-platform linux-clang" for qt5. but this results in a linking error in the end.
<dilyn> yeah lemme dig up my build...
<dilyn> so i had to specify -platform linux-clang-libc++, along with a small patch
<dilyn> but I never encountered a linker error at the end. this fixes a problem with socklen for qt5-network
<claudia> you are right. this is the error.
<dilyn> I should work a little harder at getting it upstreamed. afaik it's just a 'woops we forgot to make this change' sort of thing
<claudia> thx
<acheam> konimex: ah okay yeah I was using an old bmake version
midfavila-laptop has quit [Ping timeout: 272 seconds]
<midfavila> rip me
<acheam> why
<acheam> can anyone share a GCC package?
<acheam> accidentally screwed up my toolchain
<testuser[m]1> Just untar kiss-chroot and kiss export
<dilyn> there's one provided by kiss-static :v
<acheam> Thank you thank you
micro_O has quit [Ping timeout: 272 seconds]
micro_O has joined #kisslinux
claudia has quit [Quit: zzz]
midfavila has quit [Quit: ircII EPIC5-2.1.2 -- Are we there yet?]
midfavila has joined #kisslinux
<midfavila> question:
<midfavila> was the ssu package updated recently to include a post-install file that automatically adjusts the permissions of /bin/ssu
* midfavila grumbles
<midfavila> suppose I should have checked before updating my system. still, that caught me by surprise
<midfavila> time to liveboot
midfavila has quit [Quit: ircII EPIC5-2.1.2 -- Are we there yet?]
midfavila-laptop has joined #kisslinux
<acheam> anybody know how to resolve this?
<midfavila-laptop> format c:\, install gentoo
<acheam> ooh thank
<midfavila-laptop> np bby
<dilyn> cc1plus: error: /root/.cache/kiss/proc/24463/build/llvm/../libunwind/include: Permission denied
<dilyn> bruh
<acheam> hmmmmmmmm
<acheam> I'm not building as root
<acheam> and none of my installed packages have /root/.cache in their manifests
<dilyn> KISS_TMPDIR?
<acheam> empty
<acheam> midfavila-laptop: why does sdf git not work half the time I try and pull
<dilyn> building lld in community?
<acheam> yes
<midfavila-laptop> acheam no idea. works on my machine
<midfavila-laptop> are you getting SSL errors?
<acheam> no it fails to resolve
<acheam> genuinely half the time I "kiss u"
<midfavila-laptop> sounds like a you problem. is your DNS running?
<acheam> I use 9.9.9.9, and this is the only host with that issue
<midfavila-laptop> never tried that resolver. i use an opennic resolver and it works fine for me
<acheam> weird
<midfavila-laptop> yeah, just cloned it
<dilyn> -- Check for working CXX compiler: /usr/lib/ccache/bin/c++ - skipped <- don't use ccache?
<acheam> trying now
<midfavila-laptop> gotta download fecal fossa because puppy's grub always breaks...
<midfavila-laptop> fml
<acheam> dilyn: help
<acheam> didn't help"
mahmutov has joined #kisslinux
<testuser[m]1> What if you put it on tmpfs
<dilyn> looks like something in LLVM is referring to its builddir?
<dilyn> see the things llvm-config's various options outputs
<testuser[m]1> Its the missing libunwind include
<testuser[m]1> So it's Tarball is included in the lld package
<dilyn> wdym?
<dilyn> the error is referring to llvm's builddir, not lld's
<dilyn> using cc=clang -v, I get some of this output: 17 ignoring nonexistent directory "/tmp/9941/build/llvm/llvm/../libunwind/include"
<dilyn> but a lot of this too -I /tmp/9941/build/llvm/llvm/../libunwind/include
Uks3 has joined #kisslinux
<phoebos[m]> acheam: i had a problem with lld 12 a while ago, have you updated llvm?
Uks2 has quit [Ping timeout: 258 seconds]
<acheam> llvm is up to date from the bin repo
<dilyn> can reproduce :v cc1plus: error: /root/.cache/kiss/proc/10904/build/llvm/../libunwind/include: Permission denied
claudia has joined #kisslinux
<acheam> :(
<acheam> omnitay isn't here, right?
<acheam> can someone open a GH issue
<claudia> qtwebengine failed with clang http://ix.io/3roz . Anyway I am going to switch back to gcc anyway.
<dilyn> yeah gn is dumb
<dilyn> LDFLAGS="$LDFLAGS -lc++ -lc++abi -lunwind" make sure you build with this
<claudia> ok.
<claudia> Ahh, I am to unskilled to properly use this. A pity.
<phoebos[m]> I'd try building llvm :p
<dilyn> you could do this instead sed -i 's/gn_gen_args = /gn_gen_args = --no-static-libstdc++ /' \
<dilyn> 7 - src/buildtools/gn.pro
<dilyn> lld build fines when not using the llvm bin package
strajder has joined #kisslinux
<claudia> ok nice. It got past that error and started :)
midfavila has joined #kisslinux
riteo has joined #kisslinux
<riteo> hiiii!
<testuser[m]1> Hi riteo
Guest28 has joined #kisslinux
<Guest28> Hi again
<testuser[m]1> Hi guest
<Guest28> Ok so it appears wifi works on my machine
<Guest28> Tried building xorg and stuff, it took a long time to build but ended up failing because a directory or something was missing
<Guest28> Err took a long time for xf86-video-amdgpu because of llvm
Guest3 has joined #kisslinux
<claudia> Guest3: we need the log from the failed package.
<cem> dylanaraps: it's great to see you :)
<riteo> oh dylan's back on IRC too? Nice!
<dylanaraps> :)
<riteo> welcome back dylan :)
<sh4> i wonder if dylan took his absence to escape the covid 1984 madness
<testuser[m]1> bruh
<riteo> according to https://k1sslinux.org/news/20210519a he was just burnt out
claudia has quit [Quit: zzz]
<Guest28> covid is not a big deal if you don't go outside
<Guest28> so for me it is no problem
<riteo> yeah I agree, I actually felt kinda better doing school from my computer and not having to go out once every week
<riteo> it's a very subjective thing
midfavila-laptop has quit [Remote host closed the connection]
<dilyn> acheam: solution appears to be me building llvm in /tmp :v testing rn
<dilyn> there's... almost certainly an actual way, but /shrug
Guest3 has quit [Quit: Client closed]
<Guest28> I liked going to school physically but doinf it online was so much easier lol
<riteo> the weird thing is that most people hated doing it online
<dilyn> built fine for me. lmk
<acheam> dilyn: still fails
<acheam> built with "KISS_TMPDIR=/tmp kiss b old"
<testuser[m]1> thw log is incomplete
sh4 has left #kisslinux [Leaving]
<acheam> oops fixed, testuser[m]1
<Guest28> Here's what happened when I tried to build xf86-video-amdgpu https://pastebin.com/A2pUrz3b
<Guest28> Not sure why xorg-server didnt build I didn't even get a log for that
<Guest28> It says xorg-server not found
<dilyn> how strange
<dilyn> wait so you do or don't have xorg-server installed Guest28: ?
<Guest28> It wasn't build for some reason but I tried again and it worked
<Guest28> So I will try amdgpu again
dylanaraps has quit [Ping timeout: 272 seconds]
strajder has quit [Quit: nn]
mahmutov has quit [Ping timeout: 268 seconds]
mahmutov has joined #kisslinux
<Guest28> Can start xorg as root but not regular user. amdgpu is built into the kernel. Here's the xorg log https://pastebin.com/m7JRfBiw
<Guest28> My user is in wheel, audio, input, video group
<Guest28> dmesg shows that amdgpu is loadaed
<dilyn> [ 826.742] (EE) open /dev/dri/card0: Permission denied
<dilyn> your card's permissions are wrong
<dilyn> are you using udev or mdev?
<Guest28> mdev
<dilyn> is amdgpu built into the kernel, or built as a module?
<Guest28> Kernel
<Guest28> I will send dmesg
<dilyn> is the mdev service running at all?
<Guest28> Yes "run: mdev: (pid 399) 1333s"
<Guest28> When I type sv status mdev as user I get "warning: mdev: can't open supervise/ok: Permission denied"
<dilyn> you gotta run sv status as root
<Guest28> I did that was the output "run: mdev: (pid 399) 1333s"
<Guest28> Just didn't know if that was normal to get perm denied
<omanom> may be a dumb question, but have you rebooted since rebuilding everything?
<dilyn> what happens if you manually run mdev -vsf ?
<dilyn> i.e. ls -l /dev/dri/* should report root:input
<dilyn> errr root:video
<Guest28> yes sir
<Guest28> Ok I got it sorry
<Guest28> Ok I think I should be good with configuring the rest of my system now. @dilyn Sorry for the questions just not familar with device managers like mdev so I will have to learn more.
Uks3 has quit [Ping timeout: 268 seconds]
<dilyn> It's one of the most common questions :)
Uks2 has joined #kisslinux
micro_O has quit [Quit: micro_O]
<midfavila> man, that reminds me
<midfavila> i need to set up UUID for my disks
<midfavila> my SAS card passes them through to Linux in a different order on every boot
<midfavila> it's so annoying
<acheam> oof
<midfavila> cue spending half an hour waiting for my machine to finally pass disks through in something resembling the right order
<acheam> woah from 4 second boot with eudev to 1 second with mdev
<midfavila> >he doesn't measure his boot times in miliseconds
<midfavila> W e a k
<acheam> I'm just basing it off of the timer that's shown after boot
<acheam> do you have a better way of measuring boot time?
<Guest28> Atomic clock
<midfavila> atomic clock hooked up via i2c
<midfavila> guest28 gets it
<midfavila> my man!
<acheam> Thanks
<midfavila> yw acheam
<midfavila> always glad to help you out
<acheam> damn I regret making my boot partition so small
<acheam> I can't fit a backup kernel into it :(
<midfavila> how large is your boot partition?
<acheam> 27m available for use
<acheam> after ext4 stuff
<midfavila> you should be able to fit two kernels in there... hm
<midfavila> i normally go with 50mb and xfs
<acheam> my kernel is 10.7m rn
<midfavila> mine's like thirteen
* midfavila dabs on acheam
<acheam> and system.map is 5.4
<dilyn> imagine having a SAS card and not using zfs smh
<midfavila> i don't even bother with system.map
<acheam> what does it do?
<midfavila> nothing useful for me.
jaafarrc has joined #kisslinux
<acheam> Thanks
<jaafarrc> Hi.
<dilyn> o/
<Guest28> So wait you can delete System.map
<midfavila> yup.
<dilyn> acheam iirc system.map just contains all the symbols for the kernel
<dilyn> but they're already in /sys
<midfavila> symbol table, looks like-
<midfavila> yeah
<acheam> ooh
<midfavila> it's mostly only useful for debugging
<acheam> nice now I can fit two kernels
<midfavila> see? *sometimes* my advice isn't a complete meme
<midfavila> it's only *mostly* a complete meme
<dilyn> lol
<dilyn> wait how tiny is your boot partition that you can only fit one kernel
<dilyn> :thinking:
<midfavila> 27mb
<dilyn> imagine...
<dilyn> my mbp won't even boot with an EFI partition < 250MB
<acheam> oi I followed mid's advice!!!!
<acheam> check the freenode logs
<acheam> he said anything above 32m was excessive or somthng
<midfavila> >muh ebbel
<midfavila> that's why
<jaafarrc> Does KISS Linux use binutils by default?
<acheam> yes
<midfavila> if it's in repo/core it's standard
<midfavila> everything else can't be assumed
<jaafarrc> Hmmm... either binutils is big or my internet is slow... I'm assuming the latter.
<midfavila> it's just under 22mb
<jaafarrc> What the fuck is wrong with my internet?
<jaafarrc> This is shameful.
<midfavila> try running a traceroute
<midfavila> you might have high latency or an unstable link
<jaafarrc> A whatwho?
<midfavila> Traceroute. it's a command that tracks the path of a packet throughout the internet
<midfavila> so find the host that the package is on, then traceroute $host
<midfavila> where $host is... well, the host of the package
<jaafarrc> What do you mean $host?
<jaafarrc> You mean LAN IP address?
<jaafarrc> The hostname?
<midfavila> the hostname, yeah
<jaafarrc> I think it's fine now, anyway.
<midfavila> cool
<midfavila> anyway, in the future, if you want to play with traceroute -
<jaafarrc> OH, I thought you meant *my* hostname.
<midfavila> you would traceroute ftp.gnu.org
<midfavila> and it'll display a reading of your path to it
<acheam> another cool one is mtr
<midfavila> "mtr"?
<acheam> its a hybrid of traceroute and ping
<midfavila> Weird.
<midfavila> That's not on my system.
<acheam> its in community
<acheam> you can get the same info from traceroute and ping, but MTR presents it in a way that makes debugging much quicker
<dilyn> BUT
<dilyn> is it written in rUsT?
<midfavila> doesn't appear to be.
<dilyn> smh
<dilyn> what's even the point
<jaafarrc> I actually like Rust.
<midfavila> "how can I even ping in a memory-safe fashion"
jaafarrc has quit [Quit: Quit]
midfavila has quit [Remote host closed the connection]
<rio6> who needs memory safty when you can just crash and let the OS clean up your grabage
midfavila has joined #kisslinux
illiliti has quit [Quit: leaving]
gtms has quit [Remote host closed the connection]
m3g has joined #kisslinux
Guest28 has quit [Ping timeout: 246 seconds]
claudia has joined #kisslinux
<claudia> dilyn: Falkon works. plays videos.
<claudia> This error message is shown and repeated http://ix.io/3rpL
midfavila has quit [Remote host closed the connection]
midfavila has joined #kisslinux
<midfavila> >mfw bash, dash and other packages fail with unable to create executables when built with kiss, but manual building still works
<midfavila> it's another one of those days, I guess
<claudia> check you env
<claudia> inside a pkg
<dilyn> claudia: ty! I believe those errors have always been present in falkon
Guest28 has joined #kisslinux
<claudia> no problemo :D
<Guest28> In order to play HDMI/DP audio from RX 580 do I need to enable EDID in kernel?
riteo has quit [Ping timeout: 265 seconds]
<acheam> dilyn: how would you feel about measuring bootup time to the millisecond?
smartin has quit [Ping timeout: 265 seconds]
<acheam> it just takes one teeny tiny change in the bootup script
<acheam> IFS=' ' instead of IFS=.
riteo has joined #kisslinux
buffet0 has joined #kisslinux
paradigm has quit [Ping timeout: 244 seconds]
buffet has quit [Read error: Connection reset by peer]
buffet0 is now known as buffet
paradigm has joined #kisslinux
<midfavila> man, I might just fork the kiss rootfs at this point
<midfavila> manually ripping out all the bits I don't like every time I build a new root is a real drag
mahmutov has quit [Ping timeout: 272 seconds]