rmilecki has quit [Read error: Connection reset by peer]
rmilecki has joined #openwrt-devel
rmilecki has quit []
rmilecki has joined #openwrt-devel
slh has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
schwicht has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
tSYS has quit [Quit: *squeak*]
tSYS has joined #openwrt-devel
Daanct12 has joined #openwrt-devel
schwicht has joined #openwrt-devel
Daanct12 has quit [Quit: WeeChat 4.0.4]
Daanct12 has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
schwicht has joined #openwrt-devel
tidalf has quit [Remote host closed the connection]
tidalf has joined #openwrt-devel
rua has quit [Quit: Leaving.]
rua has joined #openwrt-devel
pekster has joined #openwrt-devel
GNUmoon has quit [Remote host closed the connection]
GNUmoon has joined #openwrt-devel
dangole has quit [Ping timeout: 480 seconds]
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
fakuivan has joined #openwrt-devel
SpectreDev_01 has joined #openwrt-devel
<SpectreDev_01>
Hi, I'm trying to port Openwrt to Linksys MX4200 (qcom WiFi 6), I want to know how I can get the DTS file, so far I only have one DTS file and I also want to know how I can get ath10k firmware blobs
<SpectreDev_01>
And if someone could help me port Openwrt that would be great since you guys know more than I do
dgcampea has quit [Remote host closed the connection]
dgcampea has joined #openwrt-devel
torv has quit [Remote host closed the connection]
torv has joined #openwrt-devel
slh64 has joined #openwrt-devel
torv has quit [Remote host closed the connection]
torv has joined #openwrt-devel
KGB-2 has quit [Remote host closed the connection]
KGB-2 has joined #openwrt-devel
hitech95 has joined #openwrt-devel
G10h4ck has joined #openwrt-devel
danitool has joined #openwrt-devel
robimarko has joined #openwrt-devel
goliath has joined #openwrt-devel
schwicht has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
schwicht has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
schwicht has joined #openwrt-devel
GNUmoon has quit [Quit: Leaving]
rua has quit [Quit: Leaving.]
rua has joined #openwrt-devel
ptudor has joined #openwrt-devel
GNUmoon has joined #openwrt-devel
<stintel>
robimarko: do you have experience with T-mobile HR ZTE MF283+ 4G router?
danitool has quit [Remote host closed the connection]
<robimarko>
stintel: Not really, I only used one of the Alcatel USB 4G modems
Daanct12 has quit [Quit: WeeChat 4.0.4]
ptudor has quit [Quit: Strict-Transport-Security: max-age=48211200; preload]
ptudor has joined #openwrt-devel
<SpectreDev_01>
robimarko: could you help port Openwrt to mx4200 it's very similar to AX3600
<stintel>
robimarko: ok. getting lag spikes over its wifi of > 10000ms. plugged in and activated my starlink instead :)
<stintel>
couldn't get into the web interface, and there are only limited password attempts (5 iirc)
<robimarko>
SpectreDev_01: What do you need help with, also its not really similar to AX3600
<robimarko>
Even the SoC is different, its one of the stupid Oak models
<stintel>
first time I try the starlink with the ethernet adapter and an OpenWrt router behind it
<robimarko>
stintel: That sounds like the usual shit vendor FW that then ISP locks down
<stintel>
DHCPv6 PD seems to work fine, have a public IPv6 on my laptop
<stintel>
robimarko: yeah :(
<stintel>
I'm in Istria at the moment, but leaving tomorrow
<stintel>
have time for a beer tomorrow or saturday?
<robimarko>
stintel: How was the weather?
<robimarko>
On news it looked like crap
<stintel>
horrible :P
<stintel>
today is better, but was totally bad for 3 days
<stintel>
oh well, I hadn't seen bad weather since June or so, can't complain
<robimarko>
You missed out on the lovely 38+ degrees weather we had here for around a week
<stintel>
and my gf was complaining it was too hot anyway :P
<stintel>
oh we had a couple of days like that in Sofia and also in Montenegro
<robimarko>
I am free on Saturday if you are passing by
<stintel>
now it looks like OpenWrt also PDs the ULA from the starlink :/
<stintel>
where was it again?
<SpectreDev_01>
robimarko: ah the soc is different I didn't think 0ak made a difference, didn't know till now, all I need help is to get it to build and boot, never done Openwrt ever before, a little guide to get me on the right track should be enough (I think anyways), I mostly do android stuff
<robimarko>
I live in Osijek, so its like 80-ish km detour
<robimarko>
SpectreDev_01: Best docs are previous boards that were added.
<robimarko>
That should pretty much explain how to add a board
<stintel>
robimarko: ok if I PM you?
<robimarko>
Obviously there will be differences, I can probably help once you get stuck
<robimarko>
stintel: Sure
tidalf has joined #openwrt-devel
<SpectreDev_01>
robimarko: ok, thanks I'll have a look
tidalf has quit [Ping timeout: 480 seconds]
dangole has joined #openwrt-devel
<SpectreDev_01>
Can't seem to find a commit to add board support with a guide
<robimarko>
There is no guide, but look at commits adding boards for this target
torv has quit [Remote host closed the connection]
tidalf has joined #openwrt-devel
<robimarko>
There is not a lot of variety and changes that are needed
torv has joined #openwrt-devel
rz has joined #openwrt-devel
rz has quit []
rz has joined #openwrt-devel
rz has quit [Quit: rz]
rz has joined #openwrt-devel
rz has quit []
rz has joined #openwrt-devel
dangole has quit [Remote host closed the connection]
dangole has joined #openwrt-devel
rz has quit [Quit: rz]
<SpectreDev_01>
Ah ok
rz has joined #openwrt-devel
rz has quit []
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
tidalf has quit [Ping timeout: 480 seconds]
minimal has joined #openwrt-devel
<minimal>
with 23.05.0-rc3 (x86 64) I noticed that the boot partition is mounted *twice*. Both "df" and "mount" show this.
<minimal>
I haven't yet checks any other OpenWRT versions so this may not be specific to 23.05.0-rc3
schwicht has joined #openwrt-devel
<colo>
minimal: mounted where to?
<minimal>
/dev/sda1 mounted to /boot twice
<minimal>
I just checked 22.03.0 and see the same behaviour
<colo>
ok, so to the same mountpoint? I don't think that's cause for alarm
<minimal>
well at the very least it is strange to see 2 read-write mounts of the same filesystem, even if they are to the same mountpoint
hitech95 has quit [Ping timeout: 480 seconds]
hitech95 has joined #openwrt-devel
<colo>
minimal: was a pretty common sight on many distros for / (rootfs) when they slightly botched the initrd->"real userland" handover, iirc
<colo>
I don't think it causes problems
<colo>
(since its all mediated through the same VFS instance of one single kernel instance, and not multiple machines :))
<minimal>
I had a look and /etc/fstab is empty, haven't yet searched around to see when filesystems are mounted during boot
<Mangix>
I don't know what the long term solution is to this honestly.
<Habbie>
for re2? or for openwrt?
<Mangix>
the latter
<Habbie>
ack
<Habbie>
now i wonder if oe/yocto fixed this in some way
<Mangix>
I think cross-compilation is supposed to be done in a chroot or something
<Mangix>
so paths don't leak
<Habbie>
that makes sense to me, yes
<Mangix>
oe/yocto already do that AFAIK
<Habbie>
right
<Habbie>
aha! dnsdist is the only user of re2
<Mangix>
makes sense
<Habbie>
i've adopted a few other libs that only dnsdist uses
<Habbie>
guess i should adopt this one too then
<Mangix>
other project that used it was libfolly, which I nuked after the facebook employee that introduced it became unresponsive
<Habbie>
ah
<Mangix>
Habbie: actually since dnsdist is the only user, you opposed to static linking?
<Habbie>
that is, really, a great question
<Habbie>
unrelated to the problem we have now, btw
<Habbie>
but a good question
<Habbie>
it avoids a package and, theoretically, allows the linker to throw away the parts of the lib we don't use
<Mangix>
yes, LTO is already enabled too
<Habbie>
globally on openwrt?
<Mangix>
no on dnsdist
<Habbie>
ah right
<Mangix>
interesting that dnsdist still uses autotools
<Habbie>
a coworker is converting us to meson
<Habbie>
it is a big task
<Habbie>
he's been working on it for a few months now
<Habbie>
(doing other things in between too, but still)
<Mangix>
last I remember dnsdist was using submodules and whatnot
<Mangix>
I would think it's ideal to do it gradually
<Habbie>
submodules?
<Habbie>
as in git?
<Mangix>
yeah
<Habbie>
ah
<Habbie>
no, never did
<Habbie>
only for pdns-builder, for building deb/rpm packages
<Habbie>
which is not relevant for openwrt
<Habbie>
it's always been autotools
<Mangix>
ah ok
<Habbie>
also, moving to meson gradually is not really a thing, sadly
<Mangix>
libdir=/usr/lib
<Mangix>
includedir=/usr/include
<Mangix>
confirmed
<Habbie>
ack
<Habbie>
'good' :)
<Habbie>
re2 has zero issues open on github
<Habbie>
so it looks like nobody reported it yet
<Habbie>
current version of the file is the same as in the diff you found
<Mangix>
Habbie: it's interesting. that commit fixes compilation in eg: nix but breaks it on openwrt. I'll ask mr mson what he thinks later. I'm of the opinion that prefix should be used.
<Habbie>
who is mr mson?
<Mangix>
eli-schwartz on github
<Mangix>
core developer
<Habbie>
of re2?
<Mangix>
meson
<Habbie>
sorry, i'm confused now
<Mangix>
ah /s/mson/meson
<Habbie>
sure
<Habbie>
but the commit was about re2, right
<Mangix>
yeah so there are two ways to find re2, pkgconfig and cmake
<Habbie>
right
<Habbie>
and dnsdist uses pkgconfig
<Mangix>
the latter probably works fine with no hacks. but it's needlessly complicated. one of the reasions why meson uses pkgconfig first and falls back to cmake
<Mangix>
one other nasty thing about cmake is its pkgconfig support is barebones. re2's pkgconfig file is incomplete and potentially buggy.
<Habbie>
ack
hitech95 has quit [Remote host closed the connection]
hitech95 has joined #openwrt-devel
<Habbie>
good news
<Habbie>
my local build worked
<Habbie>
bad news
<Habbie>
it's because packages.git was at 21 Aug :D
<yolo>
building trunk with luci enabled and failed after 'make package/index' as it could not find usign from host/bin? I did enable usign but it's only for the target
<yolo>
the Makefile does call HostBuild but there is no usign under build/host/bin, my previous build had it
<Mangix>
yolo: hostpkg/bin
<yolo>
it's not there either
<yolo>
it used to be under host/bin/usign, something might have changed recently?
<Mangix>
hmm nvm
<Mangix>
I have staging_dir/host/bin/usign
<yolo>
my trunk Feb15 built has it, today's newest pull made it disappear
<Habbie>
Mangix, i'll see if i can fix/sed re2 next week; if you happen to do an attempt before that, please let me know :)
robimarko has quit [Quit: Leaving]
<Mangix>
I’ll ask ChatGPT later
<Habbie>
ugh :)
<SpectreDev_01>
robimarko: I was able to build an initramfs img successfully but it won't boot, it gets to starting kernel and then cuts out
mrkiko has quit [Remote host closed the connection]
mrkiko has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
bluew has quit [Remote host closed the connection]
bluew has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<slh>
dangole: "system-linux: switch to new ETHTOOL_xLINKSETTINGS API" seems to break (FTBS) netifd, http://paste.debian.net/1290625/
<dangole>
slh: oh, that's odd. can you dump /tmp/pkg/openwrt/build_dir/target-x86_64_musl/netifd-2023-08-31-f429bd94/ethtool-modes.h somewhere, so I can see the generated header which seems to break the build?
<slh>
sure, just a moment (just finishing the build reverted to the commit before)
schwicht has joined #openwrt-devel
<dangole>
slh: ... because most likely some of that sed/grep'ery is blowing up there...
<dangole>
looks like the "cut -d' ' -f2" at the end of the pipe in make_ethtool_modes_h.sh is not having the desired effect of removing the prefixed numbers which are there only to maintain sorting order by speed anyway...
<slh>
/usr/bin/cut would be provided by coreutils 9.1-1
<dangole>
slh: we can just replace it with another "sed -r ...". please try https://termbin.com/z0bc
<dangole>
slh: but strangly also the 'sort -g' just before doesn't seem to have any effect in the output you share, so I somehow doubt this will fully fix it