<dwfreed>
in makefiles, variables are generally always all uppercase
<slh>
uppercase is kind of a convention for shell scripts, no technical necessity, but it's commonly done that way
dangole has joined #openwrt-devel
cmonroe has quit [Ping timeout: 480 seconds]
<dwfreed>
there's no rules, people do whatever they want
<slh>
just follow what's already the coding style within that file and you won't be yelled at, maybe ;)
dangole_ has quit [Ping timeout: 480 seconds]
<Gramdalf>
All right, sounds good ;)
<schmars[m]>
karlp: regarding ucode-mod-uloop earlier, i was having it poll itself to death on a bad file descriptor because i didn't uloop.init() - all good :) AND i was running master ucode packages on a 22.03 image...
<Gramdalf>
Couple more questions:
<Gramdalf>
1. Is /etc/fstab controlled by uci? My guess is no, from what I can tell on the wiki.
<dwfreed>
block-mount uses a uci /etc/config/fstab
<Gramdalf>
2. Is it bad to invoke `uci show` or similar in a `/etc/init.d` script? Is uci even functional at that point?
<dwfreed>
it is functional, but there's proper means to load uci data in init.d
<dwfreed>
namely config_load
<dwfreed>
(note there's no uci daemon; everything is just reading the config files themselves)
ptudor has quit [Quit: Strict-Transport-Security: max-age=48211200; preload]
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Gramdalf has quit [Remote host closed the connection]
hanetzer3 has joined #openwrt-devel
hanetzer4 has joined #openwrt-devel
hanetzer3 has quit [Ping timeout: 480 seconds]
Borromini has joined #openwrt-devel
cbeznea has joined #openwrt-devel
<cc0_>
Hi, i just installed openwrt 22.03.3 on my Archer C6 v2 (EU) router, and everything was working fine. Then I noticed the ...notice at the top of the page:
<cc0_>
> Some users report much improved long-term stability with 5 GHz by using the closed-source ath10k firmware instead of the open-source ath10k-ct firmware. To do this remove the packages “kmod-ath10k-ct” and “ath10k-firmware-qca9888-ct” and instead install “kmod-ath10k” and “ath10k-firmware-qca9888”.
<cc0_>
So I did just that, and power cycled my device. But now phy0 (the physical radio for 5GHz) does not appaear anymore in iw list, it's gone. Only phy1 (for 2.4GHz) is listed now.
<cc0_>
In my kernel logs i can see:
<cc0_>
Tue Mar 28 23:24:38 2023 kern.err kernel: [ 17.530896] ath10k_pci 0000:00:00.0: failed to fetch board data for bus=pci,bmi-chip-id=0,bmi-board-id=20 from ath10k/QCA9888/hw2.0/board-2.bin
<cc0_>
Tue Mar 28 23:24:38 2023 kern.err kernel: [ 18.146271] ath10k_pci 0000:00:00.0: failed to fetch board-2.bin or board.bin from ath10k/QCA9888/hw2.0
<cc0_>
Tue Mar 28 23:24:38 2023 kern.err kernel: [ 18.156024] ath10k_pci 0000:00:00.0: failed to fetch board file: -12
<cc0_>
Tue Mar 28 23:24:38 2023 kern.err kernel: [ 18.179262] ath10k_pci 0000:00:00.0: could not probe fw (-12)
<cc0_>
Do you have any hints on how I could fix this? I just followed the instructions and removed the two packets listed above and installed the replacement ones
dgcampea has quit [Remote host closed the connection]
dgcampea has joined #openwrt-devel
<cc0_>
seems like there are multiple topics on the forums from people with the same issue. the notice should probably be removed because it does not work (and a full reinstallation is needed to make wifi work again)
robimarko has joined #openwrt-devel
dangole has joined #openwrt-devel
danitool has joined #openwrt-devel
cbeznea has quit [Read error: Connection reset by peer]
<mrkiko>
cc0_: no one of these firmware is open source actually
cbeznea has joined #openwrt-devel
<olmari>
Dunno how exactly archer c7 v2 differs, awesome perf and stability with ct-ht(t?)-full ath10k firmware variant
<stintel>
have you tried enabling 802.11w connecting an apple device :P
<stintel>
+and
<olmari>
Tho I also compilef full iw and full wpad-openssl "stack"
<enyc>
olmari: hrrm is that the same a BT-HHv5a ath10k?
<olmari>
stintel: don't owe such rotten tomatoes.. We do have 2 SSID's tho... WPA2 and WPA3 variants.. Latter requiring 11w and earlier optional
<olmari>
enyc: I have no clue what is BT-HHv5a
<olmari>
...(because combined compat mode wpa2/3 just does not work for every old device anyways)
<stintel>
olmari: it's not about not connecting, it's about the firmware shitting itself every 2 seconds effectively rendering your 5GHz network unusable
<enyc>
olmari: BT HomeHub V5 Type A which has some sort of ath9k+ath10k in it. https://openwrt.org/toh/bt/homehub_v5a ... in praticular I'm not sure if all ath10k are equivalent in practice including -ct firmware etc.
bluew_ has quit [Ping timeout: 480 seconds]
cbeznea has quit [Quit: Leaving.]
cbeznea has joined #openwrt-devel
cbeznea has quit [Ping timeout: 480 seconds]
cbeznea has joined #openwrt-devel
aiyion has joined #openwrt-devel
fakuivan_ has joined #openwrt-devel
aiyion_ has quit [Ping timeout: 480 seconds]
Ryncewynd has joined #openwrt-devel
fakuivan has quit [Remote host closed the connection]
<aiyion>
norris: Have you seen the two updated patches for the Google device?
<stintel>
enyc: now there's a reason for a new point release ^
<hauke>
nbd: do you have some time and could take care of this?
<hauke>
hopefully some people from hostapd or linux wireless are already working on this
cbeznea has quit [Ping timeout: 480 seconds]
<nbd>
hauke: i have to discuss this with johannes
<hauke>
nbd: thanks
schwicht has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
schwicht has joined #openwrt-devel
schwicht_ has joined #openwrt-devel
schwicht has quit [Read error: Connection reset by peer]
<enyc>
stintel: heh I was thinking that ;O
<enyc>
stintel: however, it looks like will take awhile to be fathemed out properly I suspect.
<stintel>
usually such issues get addressed rather quickly
<robimarko>
Its weird, everybody is quiet about it
<stintel>
it's in my Google feed today and someone of the local hacker lab shared it with me earlier
<stintel>
probably just out?
<stintel>
(didn't look yet, day job)
<Borromini>
you have a day job? o_O
<stintel>
3 more days
<stintel>
and then I'm taking a break for an indefinite time
<Borromini>
:)
<Habbie>
quit your job?
<stintel>
layed off
<Habbie>
ugh
<Borromini>
stintel: that sucks, sorry to hear
<stintel>
laid* off
<Borromini>
from OpenWifi?
<stintel>
RIF everywhere
<stintel>
Meta, I already wasn't working on OpenWiFi for a while anymore
<Borromini>
ok
<stintel>
planned some travel, track day, etc
<robimarko>
So they are laying off people now to then try and rehire them in a year
<Borromini>
robimarko: same everywhere I guess...
<Borromini>
i work in travel, when COVID hit tons of people got sacked (because, no travel)
<Borromini>
now they can't rehire people quickly enough and lo and behold lots of them just left the line of work altogether
<robimarko>
Well, its the classic corporate model
<stintel>
I'll probably take the opportunity to do a startup
<Borromini>
stintel: visiting friends and relatives or also some leisure travel planned?
<stintel>
already have a working prototype / mvp for some idea I had for some years
<robimarko>
Huge hiring boom, then layofs at first sight of profits droping
<Borromini>
robimarko: yeah. Squeeze us like an orange
<Borromini>
stintel: sounds cool. Hope it works out :)
<stintel>
Borromini: both - I consider track day as leisure for example ;p
<Borromini>
:D
<stintel>
also meeting with some ex colleagues, pitch my idea, see if there's actual interest
<Borromini>
:)
madwoota- has joined #openwrt-devel
madwoota has quit [Ping timeout: 480 seconds]
madwoota- is now known as madwoota
minimal has joined #openwrt-devel
<Borromini>
question about 23.xx branching: will there be a cutoff date for targets not having 5.15? Or will we wait around until all active targets have been switched?
<Borromini>
seems like some targets have very little interest when it comes to testing/migrating to 5.15
<Slimey>
test and test often :)
schwicht_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
rsalvaterra has quit []
rsalvaterra has joined #openwrt-devel
<f00b4r0>
and here I thought I could keep my 19.07 devices around a while longer :P
<stintel>
f00b4r0: just revert to OEM firmware, you'll have 14.something for free!
<f00b4r0>
;P
<Borromini>
:P
<f00b4r0>
I should just give up on wifi and only trust the noodly ethernet appendage
<stintel>
for some reason I get ~9 between 2 R515 with AMD Opteron(tm) Processor 4228 HE, but between one of those AMD and the newer Xeon(R) CPU E5-2673 v4 @ 2.30GHz in my workstation, I get only ~5Gbps
<f00b4r0>
stintel: jumbo frames?
<stintel>
f00b4r0: hell no, you can't use jumbo frames almost ever :P
<f00b4r0>
I do, almost always :)
<f00b4r0>
they help even gige
<stintel>
and you don't have MTU/MSS issues?
<Slimey>
havent had jumbo frames ever even in out DCs
<Slimey>
our
<f00b4r0>
no, I use openwrt's default clamping rule and everything Just Works™
<stintel>
ime it causes more trouble than it's worth
<Slimey>
that
<stintel>
maybe it's the cable - the one between switch and workstation is 30M
<Slimey>
i still have nightmares of that time we where migrating away from cat6500's to nexus and someone wanted to test with using jumbo frames...
<stintel>
CAT6a S/FTP
<stintel>
should be able to do 10GbE at 100m
<stintel>
guess it's low quality
<Slimey>
everything here is MM/SM or twinax, no copper twisted pair to servers
<f00b4r0>
do you have a switch in between?
<stintel>
no, straight from switch to workstation
<f00b4r0>
should be easy to check for Xmit errors
<f00b4r0>
yes that's what I meant
<stintel>
ah yes of course :P
<f00b4r0>
so your switch should be able to tell you if something is wrong
valku has joined #openwrt-devel
<f00b4r0>
if no errors, the bottleneck isn't your network ;)
<djfe>
I'm reusing established code (just a different target) for wifi path migration on sysupgrade.
<djfe>
ok I just read the above about the wifi attack vector, uff.
danitool has quit [Ping timeout: 480 seconds]
<norris>
aiyion: yes, patches are in my inbox. I was planning to give them a quick run myself, but got caught up in other things (plus hitting build flakiness). I'll review today I hope
<aiyion>
wasn't meant to make you go faster, just wanted to make sure you saw them. Please take your time :)
<aiyion>
norris: ^
<aiyion>
If you've got an idea, how I'd get the LED to be white instead of blue, that would be interesting in order to match the stock lighting pattern
<norris>
aiyion: no worries. Today's a good day to check if my flaky builds succeeded, at a minimum ;)
<aiyion>
^^
<tom->
aiyion: higher voltage? :D
kenny has joined #openwrt-devel
Daaanct12 is now known as Danct12
Borromini has quit [Quit: Lost terminal]
<norris>
aiyion: oh, I missed your question. I think you can enable all the colors to make white?
<norris>
It's been a while since I played with them
<aiyion>
norris: that's true; but how would I do that with led_running?
<norris>
aiyion: no clue off hand. I'm really not that much of an expert with openwrt specifics. But I don't think it knows how to handle multi color LEDs much
<aiyion>
Then I'll wait for your review. Thanks so far!
<norris>
Also, who cares about stock configurations? Have a little fun with color in your life! I did a rainbow pattern for OnHub, which has like 27 (3x9) LEDs :D
<nbd>
hauke: got some patches for the wifi issues from johannes
<nbd>
working on integrating them now
<nbd>
they only fix iwlwifi, but not other drivers
<nbd>
so i have to do some more work
<nbd>
actually, they fix other drivers as well, but in a heavy handed way that i'd like to avoid
dgcampea has joined #openwrt-devel
dgcampea has quit []
dgcampea has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
schwicht has joined #openwrt-devel
<hauke>
nbd: thanks
<Ansuel>
btw did we made some changes to how ccache is used with external toolchain? looking at github CI it looks like ccache works only for release but doesn't for master
<Ansuel>
(master ccache size is 1-5 mb but release ccache are 170-250mb (a correct number for ccache run for a kernel compile))
schwicht has quit [Ping timeout: 480 seconds]
danitool has joined #openwrt-devel
<philipp64>
tmn505: I was trying to add him to the list of maintainers on a PR and for some reason Github didn't seem to recognize it...
danitool_ has quit [Ping timeout: 480 seconds]
<tmn505>
indeed, maybe some new privacy option in profile settings inhibits that.
<philipp64>
noahm: any thoughts on issue #20468? should we be using `procd_add_reload_interface_trigger` or `procd_add_raw_trigger` to bind to watch for interface state changes?
<philipp64>
In this case, `isc-dhcp-dyndns` should require +bind-server, +bind-client, and (isc-dhcp-server-ipv4 OR isc-dhcp-server-ipv6) but I'm not sure what the proper expression of this should be.
<hauke>
Ansuel: I see this already for many months
<hauke>
the full build generates ~150MB chache data
<hauke>
the kernel build only 1MB
<hauke>
as far as I see it
schwicht has joined #openwrt-devel
<Ansuel>
hauke maybe adding some ccache stats may help in bisecting this... i wonder if the problem is that with external toolchain setup ccache is not used at all???
<Ansuel>
but for sdk based toolchain yes?
<Ansuel>
(release use toolchain generated from a sdk)
<hauke>
ok
<hauke>
I haven't looked into the details and I am also not very familiar with ccache
schwicht has quit [Ping timeout: 480 seconds]
schwicht has joined #openwrt-devel
schwicht has quit [Read error: Connection reset by peer]
<Ibarakasen>
Hello, I would like to build python3 shapely package for OpenWRT. Currently it depends on numpy and requires its headers during compilation. How can I tackle the problem?
<Ibarakasen>
Currently, the package requires numpy's headers without which it will not compile
hanetzer has joined #openwrt-devel
hanetzer4 has quit [Ping timeout: 480 seconds]
cbeznea has quit [Quit: Leaving.]
noahm has quit [Read error: Connection reset by peer]
noahm has joined #openwrt-devel
<Ibarakasen>
Is there a way to pass parallel build to cmake when building OpenWRT?
Borromini has joined #openwrt-devel
valku1 has joined #openwrt-devel
valku has quit [Remote host closed the connection]
valku1 is now known as valku
<Ibarakasen>
One last thing: On a computational machine with Arch Linux, the target/linux/install goal fails where make does not find fwtool.
Habbie has quit [Ping timeout: 480 seconds]
bluew has joined #openwrt-devel
zer0def has quit [Ping timeout: 480 seconds]
GNUmoon has quit [Ping timeout: 480 seconds]
zer0def has joined #openwrt-devel
GNUmoon has joined #openwrt-devel
Habbie has joined #openwrt-devel
Borromini has quit [Quit: leaving]
robimarko has quit [Quit: Leaving]
Ibarakasen has quit [Quit: Leaving]
<Tapper>
How come onley 6 people at the OpenWrt Meeting?