00:07:18 <polyex> ?
00:16:28 <nicholaus04> Howdy, folks!
00:17:16 <polyex> o/
00:18:22 <nicholaus04> Managed to change my VT font to be the tom-thumb 4x6 font.
00:18:36 <nicholaus04> At least i can display more characters now.
00:19:06 <polyex> lol
00:19:47 <nicholaus04> Though, i've only set it via vidfont, but not in a way where it's set on (re)boot.
00:19:56 <nicholaus04> It looks cool though.
00:31:17 <polyex> antranigv you ok?
00:31:36 <antranigv> polyex yes, sorry, been trying to fix this AMDGPU issue...
00:31:42 <saper> Tried OCR-ing a screenshot of a failed kernel boot log with 8x8 font and it was very bad
00:31:44 <polyex> well so did it work?
00:31:52 <antranigv> polyex I was not able to get it to work on Jails. but I think there's a workaround
00:31:57 * antranigv is thinking
00:31:58 <polyex> what's the workaround?
00:32:00 * antranigv is reading the code
00:32:27 <polyex> the only workaround i found was not using mac, then setting reserved low port down to 0
00:32:34 <polyex> but that's a blunt weapon
00:34:06 <antranigv> Has anyone ever got Venus XT [Radeon HD 8870M / R9 M270X/M370X] to work with Xorg?
00:36:47 <polyex> antranigv so that's 1 limitation in jails i found. i found 1 other wanna know what it is?
00:37:48 <antranigv> polyex possibly could be a simple fix, but still going over the code. MAC is a bit new for me.
00:37:55 <antranigv> polyex what's the other one?
00:38:48 <polyex> ok well https://cgit.freebsd.org/ports/tree/net/wireshark/pkg-message#n13 doesn't work. making the rules perm i mean. so i have to put them in /etc/rc.local it's so dirty
00:38:50 <VimDiesel> Title: pkg-message « wireshark « net - ports - FreeBSD ports tree
01:01:33 <occ> how to restart sound sub-system in FreeBSD?
01:03:49 <polyex> antranigv ok you won't believe this. so i have the devfs unmounts going on my jails and it works, no error window after bsdinstall finishes. but then i comment out the code running mount and grep, and also the sleep 60 command i was using to make the mount output readable, and the busy error came back!
01:04:04 <polyex> so now i'm trying it with the mount line commented out, but sleep 60 uncommented
01:07:40 <nicholaus04> Managed to set up some external speakers, and install lynx and openmpt123.
01:08:20 <nicholaus04> Works pretty darn well, other than the external speakers having stereo support being a bit goofed up.
01:09:31 <nicholaus04> But lucky, i don't mind that much.
01:15:40 <polyex> antranigv ok with sleep 60 in, error went away again wtf
01:17:19 <nicholaus04> Yep, xmp works as well. (And so does every other program that uses libxmp by that extent)
01:40:32 <polyex> antranigv reduced sleep to 5, error came back. trying 30s now
02:22:39 <polyex> antranigv 15s causes error, trying 20 now
02:29:54 <polyex> 20 errors, trying 25 now
02:32:40 <GoSox> i have a question that is really pushing the limits of "on topic", however....
02:32:56 <GoSox> as a life long mac user, when i need a computer, I buy a Mac at a place that sells macs.
02:33:09 <GoSox> But if I'm setting up a freeBSD server. what do i pick for hardware?
02:33:43 <GoSox> It needs to be a 1U server, ideally new/newer, but at the same time, I really don't need more power than what you might find in a DDR3 era computer system
02:33:57 <GoSox> so something relatively low end, but reliable and affordable
02:43:39 <polyex> 25 errors trying 29 now
02:48:57 <rtprio> GoSox: i run freebsd on both dell 2u and supermicro servers
02:49:25 <rtprio> but my first several freebsd machines were hand-me-down hardware franken-servers
03:03:35 <polyex> antranigv 29 seconds worked. running 2 more times to verify
03:06:50 <SponiX> GoSox: Anything 1U is going to be stupid loud
03:07:25 <SponiX> If only looking for a mid-range system. Might as well get a used workstations like the HP Z440, or a NUC
03:07:32 <polyex> antranigv ok 2nd run of 29s failed. gonna try 30 again
03:30:46 <polyex> in unattended bsdinstall installerconfig, i service start a jail, config it with jexec, service stop it, jls confirms it's stopped. but zfs umount ... errors: cannot unmount ... pool or dataset is busy. why pls? i unmount the jail's devfs and it kinda works, but only when i umount then 'sleep' for >30s. (29s sometimes worked)
03:31:34 <polyex> >=30
03:32:21 <rtprio> well, you could always run sync and forcably unmount it
03:33:00 <polyex> run sync after unmounting devfs?
03:33:10 <rtprio> yeah.
03:33:32 <polyex> so you're thinking i'm hitting some indeterminism and the sleep is just a bandaid?
03:33:34 <rtprio> you can't run .. fstat or anything to see if anything else has this open?
03:33:53 <lemoi> offtopic: anything left out of #not-##freebsd @freenode around here?
03:33:58 <polyex> it's a jail's mountpoint
03:34:21 <rtprio> i understand that
03:34:42 <polyex> rtprio btw i do verify that devfs is gone because i run mount | grep dev
03:34:42 <rtprio> but if zfs-unmount thinks something has it open, then you're best to figure out what that is
03:35:17 <polyex> you think devfs can be unmounted but have the jail's mountpoint still be busy?
03:37:21 <rtprio> yes, like if you have a process that's cd'd into that directory
03:38:54 <polyex> so fstat from the jail host obviously, after the service jail stop myjail and unmount of its devfs?
03:39:50 <rtprio> right
03:40:07 <rtprio> how many places are you deploying this installer with the jail
03:40:40 <polyex> pass FS path or mountpoint to fstat? like fstat /jails/jail/myjail ??
03:40:58 <polyex> eventually a bunch but for now just local bhyve vm
03:41:19 <rtprio> perhaps read the man page
03:43:00 <polyex> file path
03:43:52 <polyex> how does that work? i have a text file open in ~/foo, but fstat ~/foo has no records
03:48:29 <rtprio> i would not rely on ~ expansion
03:49:24 <rtprio> also the parameter is 'fstat -f /usr/home'
03:56:20 <polyex> btw ok i ran bsdinstall 6 times with 30s sleep. 1 error 5 success
03:56:27 <polyex> gonna run fstat now
03:56:57 <polyex> maybe the devfs unmount solution wasn't even a solution, we'll find out
04:28:18 <rtprio> ok, what happened
04:48:54 <remiliascarlet> jbo: I used to use DWM for a long time, but at one point you'll patch it so hard in order to get rid of the Linuxisms, you'll end up with a buggy mess. So I ended up switching to CWM instead, which already works equally well on OpenBSD, FreeBSD, NetBSD, and Linux.
04:51:04 <remiliascarlet> saper | CrtxReavr: CDE is abandonware, but there is NsCDE which tries to replicate the look and feel of CDE using fvwm.
04:51:25 <SponiX> I was considering running that for the lolz
04:55:21 <polyex> rtprio so i ls -la the jail path to make sure it's really there. then fstat -f /path/to/jail but none listed
04:55:33 <polyex> the column headers like user and cmd, but no entries
04:56:32 <polyex> how do i try the sync idea you had?
04:59:12 <polyex> oh just run 'sync'?
04:59:46 <rtprio> sync; sync; sync; zfs umount -f
05:01:33 <polyex> why 3 times?
05:05:45 <kevans> old habits die hard, perhaps
05:06:38 <polyex> kevans you ever ran into this issue?
05:07:21 <kevans> nope
05:07:37 <polyex> dang
05:07:44 <kevans> but nothing is mounted inside?
05:08:00 <polyex> in the jail?
05:08:02 <kevans> yeah
05:08:15 <kevans> no nullfs mounts, no devfs mounts, no datasets
05:08:54 <polyex> devfs i guess. but does it matter when the service jail stop myjail has been confirmed with jls?
05:09:54 <kevans> yes, because i'm pretty sure jail(8) isn't doing any mounting or unmounting on its own
05:10:16 <kevans> ah, there are mount.* flags now
05:10:48 <polyex> i have mount.devfs for jails
05:11:09 <kevans> you should still confirm manually that the mount isn't there anymore
05:11:21 <kevans> especially since you're having a problem unmounting
05:11:25 <polyex> with mount | grep dev on the jail host?
05:12:00 <kevans> yes, or just grep the mountpoint to see if maybe there's something else
05:15:43 <polyex> ok so i'll test service jail stop myjail, then i umount /path/to/jail/dev, then zfs snapshot, then mount | grep, then zfs umount /path/to/jail
05:20:01 <polyex> says pool or dataset is busy
05:20:20 <polyex> so what am i doing wrong? i unmount the devfs for the jail
05:24:03 <polyex> oh ill try putting sync before the zfs umount
05:28:54 <polyex> ok tried sync, still get error
05:30:01 <mzar> maybe some process is running
05:30:10 <polyex> in the jail?
05:30:27 <mzar> in this directory, try to forcibly umount
05:31:03 <polyex> ok took out sync, trying -f on zfs umount ... now
05:36:43 <polyex> ok well with -f the error went away
05:36:54 <polyex> there any other solution than forcing tho?
05:37:27 <mzar> find this process
05:37:58 <polyex> the process would be in the jail?
05:38:32 <mzar> running shell with cwd to mounted dir will also prevent umount
05:38:43 <polyex> but i don't do that
05:39:21 <polyex> jls confirms the jail isnt running, but it could still have a process running?
05:40:07 <mzar> probably not
05:40:18 <mzar> ps -J 1,2,3,4
05:40:31 <mzar> maybe jid was higher than 4 ?
05:40:34 <polyex> so what's keeping pool/dataset busy?
05:40:57 <mzar> one of the processes on the host or maybe mounted subid
05:41:38 <mzar> but if you have umounted nested shares then the process is preventing umount
05:44:21 <mzar> procstat -a pwdx | grep mountpoint
05:45:12 <mzar> polyex: ^^^
05:46:14 <mzar> polyex: do you run only FreeBSD or other BSDs too ?
05:48:30 <polyex> only freebsd
05:48:38 <polyex> i'll try procstat line
05:55:53 <polyex> ya procstat outputs nothing matching the mount point
05:58:04 <mzar> that's good
05:58:38 <polyex> any other ideas?
05:59:55 <mzar> there are still plenty of ideas, for example directory was shared via NFS
06:00:03 <polyex> no
06:00:59 <mzar> I have to get back to work, thanks for the conversation polyex, have a nice troubleshooting !
06:01:10 <polyex> ty
06:43:13 <polyex> why do i use umount on the jail's devfs? should i be using zfs unmount on it?
06:43:29 <polyex> ah no
06:45:45 <polyex> ok well i can't figure it out i'm just gonna use zfs unmount -f i guess
07:07:59 <saper> remiliascarlet: Linuxisms? I use x11-wm/dwm port with my config.h (124 lines)
07:14:13 <saper> remiliascarlet: Depending on the definition of abandonware - The OpenGroup officially released the sources (if you mean the code died as proprietary), last new release is from November 2023 (if you mean nobody cares)
07:55:19 <remiliascarlet> saper: There are patch makers that employ Linuxisms unfortunately.
08:04:47 <saper> remiliascarlet: you mean those extra patches from people using dwm, not the dwm itself?
08:06:38 <darwin> i use XFCE across both FreeBSD UNIX and Slackware GNU/Linux (since 1997 when I started  using them, FreeBSD was still called UNIX in official documentation and I also read & post to nntp:comp.unix.bsd.freebsd.* )
08:06:53 <darwin> i noticed recently in FreeBSD I'm missing some settings in XFCE after this
08:08:26 <saper> darwin: slackware has more features than the FreeBSD port?
08:11:49 <darwin> no
08:11:57 <darwin> some standard settings are now missing
08:13:07 <darwin> for example desktop
08:36:15 <remiliascarlet> saper: Yes.
08:37:10 <saper> darwin: I am sorry, I don't know what you mean (not using xfce)
08:37:32 <saper> remiliascarlet: now I get it. I don't use any, that's probably why I am happy with dwm as it is ...
08:37:50 <remiliascarlet> darwin: I always found Xfce to be lacking as a DE, although seems to be the only DE left that isn't trying so hard to push Wayland onto everybody.
08:39:02 <remiliascarlet> Even MATE is going all on in Wayland, which then sparked an X11-only fork called "OpenMATE".
08:40:07 <darwin> settings->desktop
08:40:12 <darwin> changing background, etc.
08:48:39 <saper> darwin: what does "pkg info -x xfce" say? see if I can reproduce
08:50:39 <darwin> libxfce4menu-4.18.6 libxfce4util-4.18.2 xfce-icons-elementary-0.19 xfce4-appfinder-4.18.1_2 xfce4-conf-4.18.3 xfce4-notifyd-0.9.4_1
08:50:39 <darwin> xfce4-panel-4.18.6 xfce4-power-manager-4.18.4 xfce4-session-4.18.4 xfce4-settings-4.18.5 xfce4-terminal-1.1.3 xfce4-tumbler-4.18.2_2 xfce4-wm-4.18.0_3 xfce4-wm-themes-4.10.0_2
08:52:24 <saper> darwin: thx will try
08:54:29 <darwin> i used ZFS to run it on both.  I wrote the instructions for that on the docs.slackware.com site
08:54:36 <darwin> i'll be back later
09:13:53 <saper> darwin: on a fresh install, it works, but of course this is expected
09:16:13 <saper> darwin: looks like you might be missing /usr/local/bin/xfdesktop-settings which come via xfce4-desktop package. You might want to "pkg install xfce" again.
09:43:12 <saper> one of the big problems with so-called desktop environments is that stderr of the programs started gets usually lost
11:00:12 <polyex> saper i had to resort to using -f in zfs unmount -f /path/to/jail to get busy error to go away
12:34:39 <rtprio> yep
12:35:48 <darwin> thanks
17:12:50 <saper> kevans: I should have cc'd you on https://lists.freebsd.org/archives/freebsd-xen/2024-June/000204.html - the other address was bugs in the copy
17:12:51 <VimDiesel> Title: 13.2-RELEASE-p11 dom0 fails to boot unlike -p10
17:56:07 <antranigv> ended up using SCFB
17:56:16 <antranigv> Somehow AMDGPU is not working
17:56:27 <antranigv> I keep getting "Returned 2" or "returned 22"
20:26:21 <ascreen> i have a weird issue with boot environments.. bectl list nicely shows them, but when i activate a different one, it only seems to load that kernel, but not the / (using 14.1-RELEASE and -STABLE)?
20:26:48 <ascreen> and all datasets are canmount=noauto
20:33:18 <polyex> saper i had to resort to using -f in zfs unmount -f /path/to/jail to get busy error to go away
20:33:26 <polyex> antranigv ^
20:34:31 <polyex> hey about jails, would anyone want an "after" directive that orders a jail after 1 or more others? because depend means DO NOT start unless dependent jail(s) are started. after would mean DO order the startup of a jail after 1 or more others
20:34:39 <polyex> because i REALLY want that feature
20:38:17 <antranigv> polyex what do you mean?
20:38:26 <antranigv> polyex if you start app0, make sure db0 is up
20:38:35 <antranigv> but if you start db0, make sure app0 is up?
20:39:03 <antranigv> while I love the idea, it's a mess in real-life, what if you're doing changes in app0? you see what I mean?
20:40:13 <polyex> depend: app0 depends on db0. after: start logprocessor0 after app0, db0. but there's no reason it can't start if they aren't started
20:40:32 <polyex> depend and after are saying 2 different things that only overlap on the question of startup order
20:41:19 <polyex> dependency is local to a jail. but after is local to an rc context and not a jail
20:44:39 <ascreen> ah lol, the new BE had a explicit mountfrom= in /boot/loader.conf -.-
20:52:00 <darwin> should I enable HDD/SSD & RAM power down modes if I'm running FreeBSD (or similar)?
21:03:43 <saper> ascreen: Today spent half a day fighting boot environments - I suspect one of my disks (a mirror) was missing gptzfsboot or something like that - temporary boot (activate -t) didn't work at all
21:11:05 <saper> polyex: this is actually one of the problems of systemd - it starts "jobs" starting services, but it does not know or care if things are really up
21:14:54 <polyex> saper "this is"... what's this?
21:18:58 <saper> polyex: sorry I misread your "after" directive description
21:20:06 <polyex> so do you like the idea?
21:20:43 <saper> no, depends was enough so far here
21:21:02 <polyex> so an idea is only good to you if you need and use it?
21:26:35 <rtprio> i would just make those services more robust to tolerate not having a database connection
21:53:33 <polyex> what services? did you even read? there was 1 service that depended on the db
21:54:35 <polyex> and anyway you made my point. 'after' directive would be a useful replacement for 'depend' often times
22:02:44 <levitating__> is it normal for there to be a /entropy file in /
22:03:29 <levitating__> does that serve a purpose similar to /dev/random or whatever?
22:22:48 <polyex> levitating_ i have it
22:30:54 <antranigv> dvl what's your latest smb4.conf? can you share it? the old one is not working with samba419
23:29:18 <levitating_> how do I configure upowerd? my laptop is not boosting
23:30:28 <polyex> powerd?
23:31:12 <polyex> read man powerd yet?
23:33:32 <levitating_> polyex yes but I can't find anything about boosting/turboing specifically
23:33:50 <levitating_> this cpu should have a 1400Mhz to 17000Mhz range
23:34:07 <levitating_> but the ability to temporarily boost much higher
23:34:09 <polyex> set max profile first
23:34:22 <polyex> mode
23:35:16 <polyex> -r too, not sure what it does tho
23:38:35 <levitating_> I am not that sure if I am measuring it correctly either
23:38:52 <levitating_> But when I run btop on Linux I can clearly see my cpu being above 1.7ghz basically all the time
23:38:58 <levitating_> And on freebsd it sticks to 1700
23:40:30 <levitating_> I read on the forum setting the frequency to 1 beyond the actual max frequency enables boosting, but sysctl refuses to accept the setting
23:45:24 <rtprio> uh, what frequency?
23:45:28 <rtprio> levitating_: and how are you setting it?
23:45:29 <levitating_> 1701
23:46:01 <levitating_>  sysctl dev.cpu.0.freq=1701
23:46:46 <rtprio> i don't think that will work
23:47:15 <levitating_> I thought it was weird as well but I saw this forum post
23:47:16 <levitating_> https://forums.freebsd.org/threads/how-to-run-an-intel-cpu-at-full-speed.82553/
23:47:18 <VimDiesel> Title: Solved - How to run an Intel CPU at full speed? | The FreeBSD Forums
23:47:44 <rtprio> does that sysctl exist on your system?
23:48:12 <rtprio> elliot@phil:~/$ sysctl dev.cpu.0.freq
23:48:13 <rtprio> sysctl: unknown oid 'dev.cpu.0.freq'
23:48:24 <levitating_> yes I can read it
23:48:44 <levitating_> root@r-mint ~# sysctl dev.cpu.0.freq=1701
23:48:44 <levitating_> dev.cpu.0.freq: 1700 -> 1700