In recent history someone was asking to do something via "netlink" of Linux; support for that had been added: cgit.freebsd.org/src/commit/?id=7e5…68495cc0a8c9793a338a8a02009a7f6dbb6

Title: src - FreeBSD source tree

I can take this to #poudriere but figured I'd ask here first... does anyone know where poudriere pulls it's broken ports data from?

It is currently marking shells/bash as broken but I can't find where it is marked as broken anywhere. freshports doesn't show it as such

That used be in the "ports/shells/bash/Makefile"; do not know if things has changed

The commit log for the Makfile show bash to be broken for static build on Sep 30; on Oct 1 that was fixed

But wait! There is more! The port is also broken if "readline" is used

yep, Makefile; "PORTS_READLINE_BROKEN= ld: error: undefined symbol: rl_trim_arg_from_keyseq"

parv, rtprio thanks for the heads up. So it looks like it is only broken if ports readline is in use then? I'll disable that and try again

Yeah, looky like that

aha, it's building now.

that makes sense, but I didn't know the broken definitions were within the makefile, so that's my new thing for today ;-)

👍

FreshPorts does not show bash port to be broken as "PORTS_READLINE" is not the current default option

thanks parv, that was my next line of tought. Thing is, I don't ever recall setting ports readline, so I wonder if that option changed sometime in the past

You could check the commit log

I've been building packages for this system since 10.2 or sometime, so I guess there's a wide window for that haha

I need to take a look at how htop does it's memory calculation on freebsd, because it is extremely flawed

Oh there is|are bugs over at GitHub: htop-dev/htop #1039 (on showing 16ZB of memory used out of 3.8 GB)

Title: Mem: 16.0Z on FreeBSD · Issue #1039 · htop-dev/htop · GitHub

1039 – two different spellings of "timezone" bugs.freebsd.org/bugzilla/show_bug.cgi?id=1039

I kind of get the gist of what it's trying to do, but it just doesn't work. when I dump a whole bunch of inact memory [like after unmounting something], it drops to below 0 bytes used, so rolls over to show 16 ZB used instead

haha read my mind parv!

I do not remember if it was you or someone else who had brought this up here not too long ago

I recall someone else mentioning it here, but I just started noticing it myself on my big mem systems too

Ok

.oO( 16 Z : 2 ** (70 +4) )

ahoi, if I do "sudo pkg fetch pkg" where is this file fetched to?

You mean where would the file be stored?

parv: yes

According to "pkg.conf(5)", thus "/usr/local/etc/pkg.conf", that would be in "/var/cache/pkg" directory

parv: thank you

hi, now I installed freebsd on my main machine that has a modern nvidia card. I installed nvidia-driver and added kld_list="nvidia" to my rc.conf. I want to use dwm so I created an .xinitrc with exec dwm but the xorg-server fails to start

how can I post the output of /var/log/Xorg.0.log here?

looks like it tries to load "nv" module instead of "nvidia" (I have no xorg.conf)

with my laptop and an integradet intel card I had no issues installing the driver

andmars: I don't use NVIDIA myself, but do you need to load the modeset driver at boot for the driver to work?

Eg: echo 'nvidia-modeset_load="YES"' >> /boot/loader.conf

edenist: fixed...I had to create a /usr/local/etc/X11/xorg.conf.d/driver-nvidia.conf stating that it's actually "nvidia" and change nvidia to nvidia-modeset in rc.conf

don't know why the handbook is stating this for all other cards but not nvidia?

i

me beeing lefthanded what's the easiest way to reverse left and right mouse button? Creating a /usr/local/etc/X11/xorg.conf.d/mouse0-buttons.conf

?

Hello there. I am looking for a new hardware on which I want to run freebsd. I think about AMD Ryzen 9 7950X. Did you try freebsd on this processor ? Does it work ?

Hi, I've installed freebsd 13.0 i386 and I am geting this message when I try to install a package: pastebin.mozilla.org/f7XR08ks

Title: Mozilla Community Pastebin/f7XR08ks (Bash)

any idea how to solve this ?

your OS is not on the same version as your package system

but I just installed the OS, it was blank, and the message appeared when I run pkg first time

I mean it was just the os without any package

Yes I have jsut realized that

i figured it out

I don't know If you should ignore OS version.

if I ignore OS version I am getting this when I try to run an app (installed from a package): ld-elf.so.1: /lib/libc.so.7: version FBSD_1.7 required by /usr/local/bin/wget not found

May be a missconfiguration in /usr/local/etc/pkg.conf ? this is where repos are configured freebsd.org/cgi/man.cgi?query=pkg.conf&sektion=5&format=html

Title: pkg.conf(5)

it is the default pkg.conf, I did not modify it

only wget return such error, mc seems to work

well, I'll try FreeBSD 12.3 instead of 13.0

you mees to updatw, or build ports rathwr than use packages

i wouldnt use an older version

*means

I know but I wanted to check if nanobsd works on 13.0

Midjak, it should work re: ryzen but it may not work correctly in all cases

I need this for programming. I will need to run VM on (for running docker image). I looked if I found something about incompatibilities with bhyve and this proc. It seems ok.

"limiting *memoryuse* may kill the machine due to thrashing." So should I set a limit on *vmemoryuse* instead

ì think I'm gonna give it a try . If it doesn't work I will switch on linux.

But I think it's ok.

meandrain: why do you want to use nanobsd

Some processes are being run in my UID, but not in my login class. Could this be related to the use of third-party tools to change from the root user to my user account?

This isn't a major issue, but it changes my algorithm for limiting resource use using rctls slightly - login classes no longer being an acceptable means of doing so.

which processes, and how are they started?

rtprio: because I have two small machine and I want to use those as wifi bridge

*machines

how small? does freebsd not run on that system?

rtprio: I am using CF cards as storage, so I kind of like nanobsd read only + mount in ram solution

I used freebsd but CF fails after 1-2 years even with noatime mount and logs disabled

oi

that hapened long time ago, then I switched to nanobsd and CF is still working after like 10 years

i messed with it briefly a long time ago but at the time i just openbsd on my cf-card based device (soekris 4501)

i am trying to upgrade from 13.1 to 14.0 (to see if I can get some hardware recognised/working) (it's a vm) but can't upgrade to 14.0 as it can't download the metadata. something going on with the servers?

rtprio: I have two old pcengine boards with AMD ALIX CPU

murmel: what are you doing to upgrade?

rtprio: bsd.to/9tE0/raw

Title: 9tE0

current is not a RELEASE, not supported with freebsd-update

so how would I do this then?

Title: Chapter 24. Updating and Upgrading FreeBSD | FreeBSD Documentation Portal

thanks :)

rtprio: so the isos are only snapshots basically? as I would always have to compile it from source?

correct; only snapshots; might be worth installing fresh with a snapshot to determine if things work

and judge the amount of effort to track current

rtprio: yeah will do that, I just thought there is an easy way to upgrade to current. oh well. hope I can get my wifi working :(

hi, how can I tell ranger to open images with sxiv instead of always opening gimp?

murmel: i thought you said it was a vm; isn't the wifi managed by the host system?

rtprio: it is, but this is a way to figure out if freebsd supports the wifi (tested it out on 13.1 as a host system) but I do need a running system, so it's an easy way as I can just redirect the wifi card into the vm without losing the whole host install

pretty roundabout way of doing that, eh? what virtualization passes the wifi onto the vm? why not just try the liveboot cd or memstick and see if the card is recognized?

rtprio: libvirt?

rtprio: well freebsd 13.1 complained about the wifi, ghostbsd (live) recognises the card, can scan for networks, but can't connect

ok, try the live current image unless you have several hours to idle while you build it all from source

rtprio: let me try that :). but I assume it will be the same as last time (i still have logs, but not reported)

just wanted to see if it's still the case so I can report it upstream. as the logs are already a month old or 2

a lot can happen in two months on -current

that's why I do hope that it works now :)

it works when you figured it out

or exploit the environment

until that, it - is just electronic circuits

oops wrong chan sry all

saltd: it was a mistake to teach sand to think

it is at a lower energy level

with their connection to universe

silicon cant follow logic w/o electronics

i have interesting theory, but it is soldable for 100,000$

OFFLINE ETA : 5y

possibly prolonging to another 5

so hurry up buy the tickets (when i die i've setuppd univers returns back to home)

*manuscriptS

I need to poke at my ZFS-on-root /boot/loader.conf from a rescue disk (long story), and am therefore looking at zpool import -f, then zfs mount. I intend to zpool export it when I'm done, but is that enough to get it mounted when (being optimistic here) I can boot from it again?

V_PauAmma_V: with zpool-import(8) with option -R and a specific altroot, yes?

<openzfs.github.io/openzfs-docs/man/8/zpool-import.8.html#R>

Title: zpool-import.8 — OpenZFS documentation

I'm on ZFS classic still, but let me check that.

Just to be crystal clear, you're talking about importing from the rescue disk, right?

From booting using the rescue disk, I mean.

Actually, forget that question. Alternative is to get the disk transferred to a USB enclosure, at which point it wouldn't matter, so I may as well risk it.

:-) for the benefit of other readers, the altroot hint is intended for use whilst in a rescue environment. For an import to not 'clobber' any existing mount point.

BTW, it worked, and I can now boot again, which is a relief.

Thanks.

Hi. Is anyone using wayland+swaywm with freebsd here? Does it work fine?

docs.freebsd.org/en/books/handbook/x11/#_the_sway_compositor says it works, but "fine" is in the mind of the beholder.

Title: Chapter 5. The X Window System | FreeBSD Documentation Portal

I guess I will have to tinker myself and see the boundaries. Like in Linux (Gentoo) I am a full time sway user without even using Xwayland.

Hope FreeBSD supports my 2020-ish hardware

Have a Ryzen Zen 2 with Renoir graphics

Ryzen is AMD, right?

Hmm. I'm not seeing any way to attach media (like an ISO image) to a running bhyve vm. Am I seeing that right?

cgull, this would really suck

It's not actually that important for me, I've managed to go years using it fairly heavily without noticing/caring. I can see ways to do it without explicit support, too, like ISCSI and maybe just even overwriting images.

cgull, but thats a dirty workaround

Any Rust developers been getting this error ... "error: No suitable version of LLVM was found system-wide or pointed

to by LLVM_SYS_120_PREFIX."

cgull, if you're seeing wrong, so am I.

So I managed to solve my spewing-keyboard problem with hint.atkbd.0.disabled="YES", but doing that also made my mouse (psm0) not move its pointer. :-(

s/not move/stop moving/

Might be due to sysmouse?

Hmm. Might be my event mask.

I think so, yes.

When did you get this error Dimonax ?

Midjak: Trying to build the Locutus code at github.com/freenet/locutus

Title: GitHub - freenet/locutus: Declare your digital independence

Did you install Rust from port or rustup ?

Hmm. It was 12. Changing it to 14 or 15 didn't help.

It's been too long for me to remember much else but that, I fear.

From port.

documentation say we need to instal LLVM, giving instruction for ubuntu. I think you need LLVM. freshports.org/devel/llvm

Title: FreshPorts -- devel/llvm: Meta-port for the default version of the LLVM Toolchain

I see that a lot of people are running into this problem, on different OSes, with different projects.

It's a little irritating that doing "make install clean" in the llvm12 directory craps out because of python dependencies.

On freebsd I think I remember using rust from rustup.

humm so you did it ?

I may just give up and go buy a cheap USB mouse when the stores open tomorrow morning.

(install llvm)

Dimonax, di you follow this guide docs.freenet.org/dev-guide.html ?

Title: Development Guide - The Locutus Book

I mean...I'm starting to wonder if language fanboys deliberately put deps for thier favorite lanuage just to make sure everyone MUST install it, even when there's no logical reason for it to be there.

VimDiesel: Yeah, I'm following it.

I don't think so :-) . Rust is well suited for the project.

Midjak: I didn't say Rust. I'm talking about llvm putting dependencies on Python3.

Did you manage to install llvm ?

in fact Rust use LLVM

Midjak: I did through ports. What's the command for rustup?

but I don't know why there are dependencies to python3

Midjak: /usr/ports/devel/llvm12

$ curl sh.rustup.rs -sSf | sh

$ curl sh.rustup.rs -sSf | sh

$ curl sh.rustup.rs -sSf | sh

curl sh.rustup.rs -sSf | sh

oups sorry...

You forgot to run that as root.

and use http://

Midjak: Something in the build script there is pulling in the py38-pygments-2.9.0 package.

it works on userspace. as far as i remember

Midjak: You saying THAT command outta download and install llvm12?

brb

Dimonax, so is llvm installed ? I ma not sure to understand.

No I think I had llvm installed. I am not really sure about rustup too...

So I cheated. Heh.

pkg install

skered you can download the script and look into, as well

Midjak: That's the standard joke when doing curl ... | sh

ah ah good joke (I think it's useful to recall how this kind of command is dangerous)

My favorite is naming a file "-rf".

(If you do that, don't try to rm it.)

humm kinky !

Seems like there should be a switch option for llvm, similiar to what OCaml does for it's compilers.

for what Dimonax ?

I have llvm12 installed, now I gotta figure the command to switch rust from llvm9 to llvm12.

ah ok...

Why didn't you install llvm9 instead ?

Midjak: It's already installed.

Midjak: Locutus depends on llvm12

ok

so cargo install is the next step no ?

did you still get the error ?

Midjak: Running it again, with backtracing.

interesting project BTW...

how well are recent gen AMD graphics card supported? Specifically say a 6750xt?

crb: id grab a nomadbsd installer on a usb, and test drm-kmod live (no need to commit to install on disk)

koobs: I don't have such a card, I have ancient nvidia cards and I'm looking to upgrade hence why I'm asking

crb: understood

a support matrix in the upstream repo woukld be useful :|

crb: freebsd/drm-kmod #163#issuecomment-1146660744

Title: Update to Linux 5.11 drivers by dumbbell · Pull Request #163 · freebsd/drm-kmod · GitHub