04:25:31 Does anyone know why LLVM on FreeBSD isn't built with Polly support or why the port doesn't have an option to enable it? Does it not compile on FreeBSD? 04:32:55 PaddyMac: Don't know, but it might be worth looking at https://releases.llvm.org/download.html to see if they ship it themselves. 04:34:23 Polly is shipped as part of the LLVM code base. It's an additional project that can be built. 04:34:50 https://polly.llvm.org/get_started.html 04:34:51 Title: Polly - Getting Started 04:40:59 if it can be built independently of llvm proper then it probably should be a separate port 04:41:43 hello I am new intallation of freebsd on a Dell Vostro 200 an old computer and I cannot get the xorg to work 04:41:57 what kind of GPU? 04:42:27 it's integrated with the chip I would say there's no special gpu Intel 04:42:47 Intel duo core i386 04:42:53 pciconf -lv should tell you the vendor and chip ids even for integrated GPUs 04:43:15 ok just a second let me bring this laptop down the stair 04:43:22 what's the actual CPU type as shown in /var/run/dmesg.boot ? 04:44:12 (there's about 8-12 lines of CPU info starting with the CPU: line) 04:44:54 https://termbin.com//o7s37 04:45:38 device = '82G33/G31 Express Integrated Graphics Controller' vendor=0x8086 device=0x29c2 <-- so that's the GPU, let me look it up 04:46:02 This is still a development branch and, at any given time, the sources for FreeBSD-STABLE may or may not be suitable for general use 04:46:11 this this true? what should i use then? 04:46:48 thank you RhodiumToad how do you look it up out of curiosity it's a sauna in this part of the room it's literally 29 degrees celsius :D 04:46:50 uuidNuniq: what do you want to do? 04:47:09 I should have just ssh'ed into the computer :D 04:47:28 paul2: personally I look these things up in the source code of the drm-kmod port 04:47:50 RhodiumToad: i see, i think i need -RELEASE 04:47:51 interesting 04:48:49 i am from linux world, -STABLE sounds like a LTS to me :) 04:49:24 but actually, -RELEASE is suitable for daily use. this is somehow confused me 04:50:14 uuidNuniq: RELEASE versions are for general use, they get security and errata patches in binary form via freebsd-update (in addition to getting them in git) 04:50:48 yeah, i am reading https://www.freebsd.org/where/ 04:50:49 Title: Get FreeBSD | The FreeBSD Project 04:51:02 now i see, but why do they call a dev branch -STABLE? 04:51:04 uuidNuniq: STABLE branches are development for the next RELEASE of a given major version, they have a stable ABI and API, but new features are added as well as bugs fixed 04:51:32 uuidNuniq: CURRENT branch is the development of the next major release, it's where all the real work happens 04:52:03 (new features that appear in STABLE are always taken from CURRENT development) 04:52:56 (CURRENT's ABI is not stable) 04:53:09 FreeBSD-13.2-RELEASE-arm64-aarch64-bootonly.iso.xz it is. thanks RhodiumToad 04:53:59 both STABLE and CURRENT are only for people who are OK with compiling the system themselves and getting security and bug fixes only in source form 04:55:16 paul2: so it looks like the intel gpu driver in drm-kmod should recognize that GPU 04:55:29 paul2: what have you tried so far to get xorg working? 04:56:52 so much stuff literally I tried making .conf files, adding the kld load i915kms or whatever it is, removing the driver adding the driver, adding vesa driver, xorg -configure, etc.. 04:56:56 a lot of stuff and no luck 04:57:27 installing other desktop environments lxde mate enlightenment etc.. 04:57:50 what freebsd version? 04:58:09 13.2 release i386 04:59:03 ok. first make sure the drm-kmod port is installed 04:59:15 how? 04:59:24 Did you install xf86-video-intel? 04:59:26 what does pkg info drm-kmod say? 05:00:27 https://termbin.com/u6gj 05:01:07 xf86-video-intel no package 05:02:19 Did you check /var/log/Xorg.0.log 05:02:48 If the X server is trying to start but fails, that should tell you why or at least give you a clue. 05:03:23 Lines with (EE) especially. 05:03:57 Anywa, I wish I stay and help, but I'm already up much later than I ought to be. 05:03:59 https://termbin.com/9kkp xorg.0.log 05:04:17 no screens found 05:04:37 sometimes fonts missing and the start lxde is a different error 05:05:03 cannot open display 05:05:49 don't jump around, we'll do things systematically 05:06:07 thank you 05:06:07 Good luck and good night 05:06:14 remove any configuration files for xorg that you've added (e.g. in xorg.conf[.d]) 05:06:52 make sure i915kms is loaded in the kld_list entry in /etc/rc.conf and not mentioned in /boot/loader.conf 05:07:22 remove or disable anything you've set up to run X on startup, e.g. a display manager 05:08:13 ok all done 05:08:27 then reboot, and check that the screen mode changes during startup (after the kernel is loaded, during normal multiuser startup, you should see messages from the DRM driver and it should change the display mode) 05:08:52 (don't try and start X yet) 05:08:56 k 05:09:09 give me a few minutes 05:09:30 just got to go downstairs I think and push f1, etc.. 05:16:37 ok back all done 05:16:55 did you see the display mode change? 05:17:17 I didn't check but I can I am in via ssh 05:17:28 it's 29 degrees downstairs and humid 05:17:45 just a second 05:17:50 if you look at the output from dmesg, what does it say about drm? 05:18:34 it's going to be a bit hard to work through this if you're not actually looking at the display :-) 05:23:38 I looked at the display there's no appreciable difference 05:23:51 ok, what's in the dmesg output? 05:24:29 also what's the output of find /dev -group video -ls 05:24:38 https://termbin.com/4nsh dmesg | grep drm 05:25:16 https://termbin.com/py4kw dmesg 05:30:07 https://termbin.com/zwi9 find /dev -group video -ls 05:30:18 qemu on macOS (SoC M1) uses 400% CPU, while the installed freebsd idle is 99% 05:31:22 paul2: that all looks good so far. were you trying startx as root, or as your own user? 05:31:42 as root 05:31:49 there is no user 05:32:25 ok, just remember that if you want to run any significant X stuff as a non-root user you'll need to add them to the video group 05:33:01 next step, if you do startx as root, what ends up in the Xorg.0.log 05:38:00 going to try it now 05:38:21 just did the groupmod video -m paul2 05:46:20 https://termbin.com/f1odr 05:49:12 xorg.0.log 05:57:13 that worked, then? 05:57:45 no it didn't 05:57:52 but I tried Xorg -configure 05:57:58 will let you know if that works 05:57:58 don't do that 05:58:08 what actually happened with startx then? 05:58:18 ok 05:58:29 it just gives me 3 screens 05:58:44 yes, that's what it's supposed to do 05:58:46 and then it didn't work and startlxde 05:58:57 ok that's it? 05:59:11 the 3 tiny screens? 05:59:18 if you want to start a desktop environment, that's the next stage 05:59:22 ok 05:59:49 startx is just to start a (by default) very basic set of apps, but the fact that it worked means that the X server itself is working 06:00:06 ok but I had that before but interesting 06:00:33 the next step is either to set up a display manager, or to use whatever desktop environment's startup command instead of startx 06:00:47 ok I did startlxde 06:01:20 (lxsession:1372): Gtk-WARNING **: 02:00:59.941: cannot open display: 06:03:12 I don't want the display manager I actually remember the dos and windows from back in the day and although windows all the time was cool I like the idea to go back to that era 06:03:30 you did startlxde from outside X? 06:03:36 where a terminal is a valid option as well but desktop is an option 06:03:41 yes 06:03:52 it works this way on my laptop somehow 06:04:12 it looks like it expects to run from within X, e.g. from .xinitrc 06:04:43 yes I might have done something back in the laptop I don't remember 06:05:01 I'm not familiar with lxde myself 06:05:43 .xinitrc contains commands that are run by startx once the X server is up. if it doesn't exist, you get those three terminal windows 06:05:49 I have mate also but lxde is the best one Lightweight X Desktop Environment 06:07:00 interesting 06:07:01 man startx has some examples of what you can do with it, it looks like you'd run exec startlxde as the last thing in the file 06:07:17 what about for mate? 06:07:23 no idea 06:07:46 look at the docs for mate? this is not freebsd-specific 06:08:34 (personally I prefer a display manager, and I don't like "desktop environments", so I just use xdm and twm) 06:11:42 ok I found it in the docs with the help of the added understanding I gained from you 06:11:46 just now 06:12:55 but the thing I don't understand is if I do startlxde how does that work? 06:13:03 it also needs a line in xinitrc? 06:13:37 https://wiki.mate-desktop.org/introduction/installation/#bsd 06:13:38 Title: Installation | MATEwiki 06:16:27 from the error message you showed from startlxde, it expects the X server to be already up, so you should be doing it from within .xinitrc 06:17:34 I added the line to xinitrc and it didn't help 06:17:41 for mate 06:18:13 generally what you put in .xinitrc is commands to initialize the server to your preferences, e.g. using xrdb to load a .Xresources, and then end it with an exec of whatever window manager or desktop environment you're using 06:18:29 what exactly did you add? 06:19:54 https://termbin.com/kvnk 06:21:10 root@kagan:~ # cat ~/.xinitrc 06:21:12 exec ck-launch-session mate-session 06:21:29 that's ~/.xinitrc and you tried running startx then? 06:21:32 what happened? 06:21:39 same as before 3 screens 06:22:00 running startx as which user? 06:22:09 paul2 06:22:29 and ls -l ~paul2/.xinitrc shows what? 06:23:32 no such file or directory 06:23:48 so... where did you put it? 06:24:30 in ~root/.xinitrc 06:24:36 probably 06:24:49 ~/.xinitrc as root 06:24:52 well, obviously ~/.xinitrc is per-user 06:25:16 so create the correct file and try again :-) 06:27:22 ok I loaded it in root and it's working but I added the line to paul2 as well 06:27:42 it seems like it's working I mean the comp is a little slow so it was loading 06:28:17 thank you I will just confirm it in a second 06:33:08 confirmed it works 06:33:59 thank you but a question is what do I need to do to get lxde to work cause on this laptop for example I can either startx or startxfce4 I think 06:34:28 so do I need to edit xinitrc as well? 06:34:50 startxfce4 isn't like startlxde, you use it in place of startx 06:35:01 ok 06:35:34 so I would need to change it in the xinitrc file in order to run it basically? 06:35:44 no, what did I just say? 06:36:05 for startlxde I am asking 06:36:12 right now I am doing mate 06:36:26 but I was wondering if I wanted to right lxde instead of mate 06:36:42 oh, right; if you want startlxde then put that in place of mate, yes 06:36:56 I see thank you 06:38:44 I find it weird that my old laptop from 1997 seemed to run windows 95 easily but now the desktop environment on a 20xx something computer it hardly works for :D which mystifies me but I will leave that question hanging I suppose 06:39:30 going to wish you a good night 06:39:43 farewell 06:39:49 I ran X with twm and various apps on a machine of similar vintage without issues (until the machine died) 06:40:02 interesting 06:40:06 I cannot understand that 06:40:08 but twm is kind of old-school 06:40:16 I stupidly gave away my old machine 06:40:23 (literally, i used it at university in 1988) 06:41:06 I was stupid I gave the old machine away I have regrets doing so but it's done 06:41:27 ls 06:51:01 thank you good night 06:51:24 mine was from high school when I did my best work 09:21:18 I want to change mouse pointer (to a larger size and possible to a different color). Everytime I set it, and open a new app in awesomewm, it reverts it back to the normal/original one. How can i do it? Currently I am doing it like this `xsetroot -xcf /usr/local/share/icons/redglass/cursors/top_left_arrow 64` I use freebsd 13.2 but I am hopeful this 09:21:19 is same in Linux as well as I also use ubuntu. 09:23:24 sorry for offtopic and advertisment, but it means something to me because i was there in the 90s - https://www.gofundme.com/f/rettet-die-rave-the-planet-parade 09:23:25 Title: Fundraiser by Ellen Dosch-Roeingh : Rettet die Rave The Planet Parade! 09:23:27 save the loveparade! 09:24:08 Beladona: does that still occur when you set that line in ~/.xinitrc? 09:24:31 TommyC yes. at start it ok, but when I launch an app, it resets 09:24:58 awesome is reseting things it seems 09:25:34 Beladona: seems that way, I don't use awesomewm so I can't help you there 09:27:46 mouse pointer in general changes according to what window it's over, anyway 09:28:16 RhodiumToad yes. so how do i change per window everytime, all the time 09:28:27 TommyC ^ 09:55:02 same issue here like I have in awesomewm https://bbs.archlinux.org/viewtopic.php?id=264003 09:55:03 Title: [SOLVED] AwesomeWM cursor not changing (LXAppearance) / Applications & Desktop Environments / Arch Linux Forums 10:02:29 Re bbs.archlinux thread, Damn thing has "[SOLVED]" label but nothing (else) is there that gives such an indication 10:03:15 it has this link: https://wiki.archlinux.org/title/Cursor_themes#Configuration 10:03:16 Title: Cursor themes - ArchWiki 10:03:32 which seems to be the thing 10:04:58 Saw that; OP did not say what had actually worked. 🤷🏽‍♂️ Oh well 12:10:21 RhodiumToad should I follow https://wiki.archlinux.org/title/Cursor_themes#Configuration 12:10:22 Title: Cursor themes - ArchWiki 12:11:56 most of it, obviously not the bits that are specific to archlinux 12:41:09 which desktop env does freebsd uses? kdm or ? 12:41:36 whichever one you install 12:41:54 none by default, I use KDE 12:49:38 "desktop environments" are just packages that you can install and use like any other software 12:53:35 when I used FreeBSD, I used Xfce 12:54:27 * RhodiumToad doesn't use a desktop, twm suffices for me 12:56:22 CmdLnKid paulf RhodiumToad  Reinhilde if I use awesomwm, then what I have? 12:57:01 you.. have awesomewm, if you use awesomewm 13:01:10 ... handbook 13:07:39 Hilighting multiple people to get an answer is generally frowned upon. 13:30:39 :( 13:35:06 just use a DE that you like and fits your needs 14:16:46 meena: Regarding the Oracle payment error, I got an email from them now saying they have "identified the issue and made some adjustments." Will give it another go when I become less comatose. 14:18:19 did you email them complaining, or did they proactively mail you? 14:20:15 I complained 14:21:01 ok cool 14:21:17 I'll try again, and if it doesn't work, also complain 14:27:46 whatcha gettin from oracle ? 14:27:51 ;-) 14:29:16 Headaches, probably. But I'm hoping to get a test builder for aarch64 14:30:28 lol thats about normal 14:49:15 i cannot mount a file system with label "ext2fs/root". it simply complains about 'invalid argument' and returns /dev/xxx address 14:49:43 ...when i type mount -t ext2fs /dev/[address] [mountpoint address] 14:50:29 the builtin ext2fs is not very good (doesn't support ext4 for writing, for example) 14:50:52 i just want to read it 14:50:54 the usual recommendation is to install the package for the fuse-based ext[234] filesystem 14:51:09 then add -r (or -o readonly) 14:51:22 debdrup the are responding to me and me mentioning their name is ok 14:51:45 s/the/they 14:52:37 anyone else having issues with the new pkg version (1.20.1)? I'm seeing ludicrously excessive memory usage 14:53:43 it seems there is no single command or script to change icon and size 14:54:08 is it more of an issue with awesome or xorg? 14:54:29 RhodiumToad: nothing out of the ordinary for me 14:55:38 99796 root 20 0 8162M 607M swread 0 2:13 0.79% pkg-static 14:56:05 that's in a poudriere build, doing pkg-static add -A /packages/All/qt5-gui-5.15.8p157_1.pkg 14:56:27 RhodiumToad I think , as you said the cursor is for each window, I should add the cursor set script on each launch of app by awesom 14:56:29 8g sheesh 14:56:50 that's not even a very large package 14:56:59 beladona, /join #awesome 14:57:05 already there 14:57:09 since morning 14:57:16 Beladona: no, you shouldn't do that 14:57:23 I see 14:57:52 Beladona: it's the app that controls the cursor appearance, so you have to convince each app to do the right thing 14:58:06 which is what all that cursor theme stuff is for. 14:58:21 I see 15:00:02 RhodiumToad same reason why this gets reset ? `xsetroot -xcf /usr/local/share/icons/redglass/cursors/top_left_arrow 64`   anything to be changed in it? 15:05:11 that only applies when the cursor is over the root window 15:06:25 just noticed /tmp is mounted to my zroot pool by default. is there any practical reason it's not using tmpfs? i.e., any reason i shouldn't change it? 15:07:14 RhodiumToad: i have most of my personal data on a hard drive partition which is ext2fs and it works perfectly without fusefs. but this time even fusefs fails... 15:07:39 most likely that it isn't safe in most circumstances to judge how much ram a computer has 15:08:31 or what it will be used for 15:08:41 hubertm: have you checked you have the device right? and that it's a valid filesystem? 15:09:20 yes 15:10:55 RhodiumToad ok I will try to fix via this again https://wiki.archlinux.org/title/Cursor_themes#Configuration 15:10:56 Title: Cursor themes - ArchWiki 15:11:28 you realise thats for arch linux right ? 15:12:21 CmdLnKid: arch wiki is good 15:12:27 ya seem to be missing some basics here. trying to understand the disconnect ... your french right ? 15:12:35 the stuff for Xcursor and GNOME cursor themes should work in this case 15:12:43 meena, oh no don't mean it in that way 15:13:17 "taking a guess" 15:29:24 Beladona: sure, but if they know the answer, and have been helping you, they'll respond whether you hilight them or not. 15:34:58 hello I am trying to get lxde to work on a seperate computer 15:36:47 having problems in the startx I followed instructions 15:37:11 basically screen will start and then crash nothing 15:38:33 what's in your .xinitrc 15:38:34 sounds like your environmennt has not been setup to export DISPLAY... "sounds like" 15:42:30 vkarlsen: I've now completely relaxed my cookie policy in my browser and Oracle's website is still complete bunk 15:42:57 deep reset 15:43:42 I had a website flatly refuse to recognize my (tablet, not freebsd) browser recently, first time I've seen that in recent years 15:44:16 "your browser is too unusual for us, go away" or something like that 15:45:03 i found "clear_tmp_enable" for rc.conf ... i think i'll make use of that rather than switch to tmpfs. purging /tmp on reboot was my main concern. 15:45:56 markmcb: i woulda thought that defaults YES 15:46:32 cat ~sixpiece/.xinitrc 15:46:34 ck-launch-session dbus-launch --exit-with-session startlxde 15:46:38 meena: me too 15:47:55 clearing /tmp on boot is too dangerous to be on by default 15:47:58 https://termbin.com/8oe8 15:48:16 that one is Xorg.0.log 15:48:29 was there any output from startx? 15:56:53 was there any output from startx? 15:56:56 wow sorry it wasn't working for a while my computerjust stopped 15:57:02 there was no output to speak of 15:57:58 I mean at first I had a few issue I added xf86-video intel legacy drivers, I gave the permission to paul2 to own the .xinitrc and a few other things and then I am at the blank screen now 15:58:03 RhodiumToad: huh? why is clearing `/tmp` too dangerous? the name kind of implies the state? 15:58:11 I added a line to procfs I think 15:59:04 fstab 15:59:36 proc /proc procfs rw 0 0 15:59:44 I added that line 15:59:45 why would you need to give permissions? the user should own their own .xinitrc 16:00:12 there was some issue with sessions I ended up installing lxsessions 16:00:31 the xf86-video-intel driver is optional, you shouldn't need it but it shouldn't do any harm either 16:00:38 yes because I wrote it with root so I had to change the permissions 16:01:18 basically chown paul2: ~paul2/.xinitrc 16:01:19 you should change the ownership instead, and if that was wrong, you probably also need to check everything else in the user's homedir 16:01:30 that's what I meant 16:02:15 when you say "blank screen", how blank? mouse? 16:02:50 no mouse black but possibly some small amount of lighting so maybe brownish black 16:03:17 ie. it doesn't look off 16:03:40 any response to keys? 16:03:48 but it's black enough that I could say I made a mistake 16:04:14 no response from keys but at the very initial split second transient there is a mouse cursor that dissapears 16:04:25 does ctrl-alt-F1 (or F2) switch to another vt? 16:04:37 that would be a good idea 16:04:41 vt is one? 16:05:04 I would have done ctrl alt delete but that sounds like a better idea 16:05:17 vt is terminal? 16:05:21 ctrl-alt-Fn switches to virtual screen n 16:05:53 not sure I can try it 16:06:06 by default 10 of them are set up for login prompts, but X may have taken over one 16:06:25 (or it may be using the 11th) 16:07:26 ok I get message 16:08:07 session is lxde de is lxde log directory : hom/paul2/.cache/lession/LXDE 16:08:19 log path same but run.log 16:08:57 allso xauth file /home/paul2/.serverauth.1127 does not exist 16:09:17 Xorg X Server , etc.. .the log file etc.. 16:10:20 ctrl alt f2 switches to another terminal 16:12:35 then i killed it 16:17:49 any idea? 16:18:16 so what if anything is in the log? 16:24:14 https://termbin.com/30wk 16:24:26 it had this https://termbin.com/30wk 16:24:55 but then I installed openbox and now it works but it's not the lxde that I am used to 16:25:39 can't help you with that 16:25:45 never used lxde 16:27:55 it's working though thank you I just have to figure out how to switch it from openbox-lxde to the better one 16:29:27 debdrup thats my point. They will help despite me highlighting.   Highlighting someone makes sure the message was for him/her. 16:31:12 it didn't close properly as it said 16:31:41 Anyone lately seen koobs? 16:32:28 haven't seen him in here for months 16:32:32 (at least) 16:32:34 He is Last seen : (about 34 weeks ago). Does he has another nick? 16:32:49 why? 16:32:57 not sure 16:33:21 you must have a reason :-) 17:04:33 thank you RhodiumToad 17:04:48 I got it nicely working on LxQt 17:06:32 koobs is busy with real-life 17:48:53 did chrome drop out? 17:48:55 ??? 17:49:03 upgraded all these packages and now no chrome 17:49:05 ??? 17:49:06 wtf 17:52:15 $ pkg search chromium 17:52:15 chromium-bsu-0.9.16.1_2 Arcade-style, top-scrolling space shooter 17:52:15 ungoogled-chromium-114.0.5735.198 Google web browser based on WebKit sans integration with Google 17:52:28 mason beat me to it 17:52:43 I'm quick, like rabbit. 17:55:50 * debdrup offers mason a carrot 17:56:03 hrm. so i have a 3 disk zpool, and one disk just went into FAULTED. it was not the disk i was worried was dying, so i have a second suspect drive 17:56:04 * mason improves his eyesight. 17:56:28 Demosthenex: Do you have back-ups? 17:56:40 mason: yes, i have many layers of backups ;] 17:56:44 good good 17:56:46 but i'm in the process of copying data off this server 17:56:54 i canceled a scrub to lower the io 17:57:02 i just don't want a cascading failure. 17:57:15 is there a way to bring that disk back online and rejoin the pool 17:57:25 most of the data is intact, seems it just hit a bad spot 17:57:28 Depends on why it failed. 17:57:40 smartd says 90 incorrectable sectors 17:57:57 Demosthenex: Might be worth verifying that the cabling is okay just in case it's not the drive itself. 17:58:11 yeah, i really don't want to take it down 17:58:52 what happned to make chromium need an un google? 17:58:55 so zfs says "number of io errors ... zfs device exceeded acceptable levels, marked as faulted" 17:58:57 or un googling 17:59:01 Demosthenex: If you have back-ups, could you perhaps minimize churn with a new snapshot and a then-minimal send to get anything unique off the box? 17:59:01 as it were 17:59:09 best idea: dont use zfs 17:59:13 lol 17:59:26 mogileFS solves the scaling problem better 17:59:42 That doesn't help Demosthenex just now. 17:59:54 And anyway, I'd keep on using ZFS. 18:00:24 zfs gave me so many problems with 3 usb drives 18:00:30 it kinda drove em crazy 18:00:38 USB isn't ideal. 18:00:48 liek it would not find them in order or something 18:00:53 and require long checks 18:00:58 since drives lik 4-6 t 18:02:21 so if i zpool clear, and that disk is at all accessible, it'll come back online? but then do i have to go thru a full resilver before it's stable again? 18:02:28 that io load may not be the best idea 18:03:08 actually, phew. nvm, it was the disk i was monitoring that went offline 18:03:17 Demosthenex: If the pool's still running and you have back-ups, I'd think a minimal send would be the best bet. 18:03:38 I could be wrong, but I don't think you can bring a faulted drive back into the pool without a scrub. 18:03:41 i'm rsyncing data off right now. 18:03:45 yeah, i dont' want to scrub/resilver 18:03:54 but i think 2 disks is ok, none of the errors have errors 18:04:03 Ah, unfortunate. If you are able to switch to zfs send/receive it's way lighter weight than even rsync. 18:04:17 yeah, i'm aware. long term goals 18:04:19 Won't help now, but might be good in the future. 18:05:33 new server is much better setup ;] 18:05:38 Good good. 18:14:29 i'm proud of this new one. dual SSDs with a properly mirrored freebsd installation. 4x HDD zraid for bulk storage, and jails for every application 18:14:46 Nice. 18:15:32 and puppet managing it all =] 18:36:52 puppet yikes 18:42:08 so, i shouldn't do usb drive raid with puppet? 18:44:20 I'm curious what's happening here: https://bpa.st/NSFUQ 18:44:21 Title: View paste NSFUQ 18:44:45 Host is 13.2, new jail, also 13.2. 18:45:03 Minimal packages installed predate 13.2. 18:51:02 Never mind. Forced everything to update and it seems content now. 18:51:49 mason, Running -STABLE? __FreeBSD_version has been bumped for whatever reason, ignoring that (IGNORE_OSVERSION=yes) may or may not work. Encounter that often on -CURRENT. After some point drm*kmod had stopped working with IGNORE_OSVERSION=yes & installing the package; needed to do world building 18:51:58 parv: Always -RELEASE here. 18:53:57 mason, "Wow" is all can come up with then as packages are build with last supported release. Either that policy should change or hold back the package updates (Makefile, etc) which require version bump 18:55:53 mason: `running kernel: 1300139` suggests that you aren't running what you think you are 18:56:15 are you running your own pkg / poudriere setup here? 18:57:24 because package `1301000` suggests that packages are being pulled from a 13.0-RELEASE, maybe an old stable? I suspect 18:58:39 mason: what does `sysctl kern | fgrep .os` yield? 13.2-RELEASE-p1 should give: 18:58:52 https://www.irccloud.com/pastebin/PAcyKdIY/osrel.txt 18:58:53 Title: Snippet | IRCCloud 19:01:30 dch: That's interesting, because: https://bpa.st/K7Y2O 19:01:31 Title: View paste K7Y2O 19:02:17 Not running a local Poudriere yet, although I'm going to set up a new one before too long. 19:02:52 I might be able to reproduce this. Would be interesting to explore. I'll do that. Half a sec. 19:04:37 "Anyone lately seen koobs?" <- He's relatively active in Discord. Thirty-something messages from him on 29th June (not an exact count). 19:04:49 dch: https://bpa.st/W6ZRW 19:04:50 Title: View paste W6ZRW 19:05:38 mason: can you run that with `pkg -DDD ...` 19:06:01 pkg: illegal option -- D 19:06:20 oh, sorry 19:06:21 half a sec 19:06:36 Hm, no, it doesn't like -DDD 19:08:42 aahm what did I use here 19:09:36 dch: With -d: https://bpa.st/QYFXU 19:09:37 Title: View paste QYFXU 19:09:40 not much more there though 19:09:53 try ddd 19:10:00 could have sworn its capital 19:10:31 That's got a bit more: https://bpa.st/GLBEY 19:10:32 Title: View paste GLBEY 19:11:00 What jumps out at me is that it's somehow getting the kernel wrong. 19:13:01 reading pkg src `ucl_parser_register_variable(p, "OSVERSION", myosversion);` 19:14:22 anything unusual in /usr/local/etc/pkg.conf ? 19:14:30 or actual env vars 19:15:08 none of those worked for me, but thats a better guess than nothing 19:16:34 more to the point mason that should show up in -ddd output 19:17:16 here's the ctx.version which is later used for the actual check https://github.com/freebsd/pkg/blob/master/libpkg/pkg_config.c#L77 19:17:17 Title: pkg/libpkg/pkg_config.c at master · freebsd/pkg · GitHub 19:20:03 kenrap: hi, what about the Reddit post? (AMDGPU fails to boot) 19:21:28 https://github.com/freebsd/pkg/blob/master/libpkg/pkg_config.c#L1027-L1030 and thats where it gets picked up and assigned 19:21:29 Title: pkg/libpkg/pkg_config.c at master · freebsd/pkg · GitHub 19:24:08 any suggestions for a light http/https forwarding proxy for logging traffic 19:25:00 mason: you can run `pkg --jail ..` and it will do the right thing inside the (active) jail. In your case, `pkg --chroot` should pick up the details from /usr/local/etc/pkg.conf in the jail. I think. 19:27:26 dch: Hm, I chrooted in, so it shouldn't have had access to anything outside the jail. 19:27:33 And the jail's not yet running in this case. 19:28:10 dch: My assumption is that what I tried was valid and uncovered something that shouldn't have happened, but I'm not sure I'd assign super high confidence to that. 19:32:37 mason: if you can (because zfs snapshot) see what was in /var/jail/test/.zfs/snapshot/2023070600/usr/local/etc/pkg.conf prior? 19:33:15 dkeav: haproxy is awesome and has amazing logging of initial connections. but if you want all the data in between, no idea. tcpdump ? 19:33:42 does it do forward, or only reverse 19:33:57 i guess squid is still around, i can probably make that work 19:33:59 dkeav: like squid? no idea 19:34:00 dch: Yeah, note that I just reproduced this from a ZFS snapshot, so I'll snag that from the original. 19:34:13 mason, it should be fine, see the path 19:35:36 dch: https://bpa.st/5UUWU 19:35:37 Title: View paste 5UUWU 19:47:11 #ABI = "freebsd:10:x86:64"; # Autogenerated 19:47:42 thats the same on my 13.2-RELEASE as well 19:47:52 so TLDR I have no idea where this comes from 19:48:20 so unless its somehow compiled into pkg(8) IDK 19:48:45 Yeah, that'd be strange but would explain it. 19:49:56 I recall bapt saying something about how pkg automagically identifies the ABI of the location its installing to 19:50:05 but I haven't found that in the code 19:50:19 RT would be sure to know 19:50:46 dch: Well, I shouldn't be nuking this snapshot any time soon, so I should be able to reproduce the issue for some time yet if it becomes interesting enough to show up in a bug. 19:51:02 Forcing it seemed to return it to its senses. 19:52:01 maybe sha256 of the pkg(8) in the jail would be of interest 19:52:13 the rest is probably not relevant 19:52:53 SHA256 (/usr/local/sbin/pkg) = d69b314dfee8a7e0677a5fa37e69f1a3b5298d51ca6764abc6753271f248f3d4 19:53:13 Identical to the host system, FWIW. 20:18:47 ABI is determined by reading the uname file (not its output) IIRC 20:19:02 or maybe that's just in the build system hmmm 21:31:16 grahamperrin: the gpu-firmware-kmod, doesn't provide all the capable amdgpu firmware that is supported in drm-515-kmod package. After I had the same problem as OP with my rx 6600 xt, I looked into /var/log/messages after rebooting and it said I was missing a firmware bin "dimgrey_cavefish_sos.bin". What I did to solve that error was I git-cloned from the official linux-firmware repository and grabbed the firmware bin from there to supplement the missing 21:31:18 firmware. After that, I loaded the amdgpu module and it just worked beautifully. 21:34:09 The OP of that reddit post likely needs a different firmware bin for their rx 6900 xt to function on 14-CURRENT. They have to check the messages logs first during a boot without loading the amdgpu module 22:06:24 kenrap: habe you seen https://freshbsd.org/freebsd/src/branch/main?q=fwget ? 22:06:25 Title: FreeBSD / src - FreshBSD 22:58:29 meena: I wasn't aware of it, I appreciate it. I've looked at its manpage and I'm not sure if the tool is about pulling in linux firmware that hasn't been officially packaged yet. I tried running `fwget -vn pci` and only prints usage info. So I'm a bit confused. 23:03:15 kenrap: i'd say it might still be a bit rough around the edges 23:03:40 Gotcha 23:05:51 but, if your reading of the man page and the output you're getting doesn't match up, that's either a bug in the software or in the docs, and worth reporting 23:06:13 developers need this kind of feedback 23:08:19 Also, the fact that it's not here yet: https://man.freebsd.org/cgi/man.cgi?query=fwget(8)&manpath=FreeBSD+14.0-CURRENT is mildly disappointing 23:08:20 Title: fwget(8) 23:08:53 What's the best way to give the developers bug reports/feedback? 23:11:37 If it's a bug report, the bugzilla. I'm not really sure about other feedback, except that the appropriate mailing list is probably a pretty good way to go. 23:11:39 open a bug for base / bin, cc'ing the committers. or write a mail to current@ if you think it's more PEBKAC than bug 23:21:25 delete the jvm 23:22:00 propote picolisp.cm happstack.com gnu clog aidaweb.si hek werc.cat-v.org and varnish cache and ha-proxy 23:22:08 picolisp.com that is 23:22:29 whats fwget? parallel wget? 23:23:00 any warriors useing naviserver? 23:23:22 promote-- 23:23:33 my mind faster than my fingers 23:33:18 concrete_houses: Firmware Get 23:34:25 Alright, I signed up for the bugzilla and subscribed to the current@ mailing list. I suppose I send this as feedback to the latter.