<Znevna>
doesn't light at boot, doesn't blink at upgrade(?)
<Znevna>
it doesn't do anything :
<Znevna>
if I add the label, it works
<karlp>
label, colour, function are all just mangled to end up with sysfs filenames.
<karlp>
somethign _else_ has an assumption on the led "name" but it's not unreasonable that you would need to change these together.
<Znevna>
well I don't know how to dig deeper in this, I've tried last night to make the wifi leds work and I'm calling it quits
zer0def has joined #openwrt-devel
robimarko has joined #openwrt-devel
MaxSoniX has quit [Remote host closed the connection]
MaxSoniX has joined #openwrt-devel
embargo has quit [Ping timeout: 480 seconds]
<aiyion>
arinc does not read here, does he?
<robimarko>
nbd: You plan to push 6.1 backports instead of 5.15 ones or that is just for your use?
<nbd>
i'd like to push it eventually, but i guess it needs more testing
<nbd>
i'm only testing mt76 myself at the moment
<robimarko>
master seems like a great place for that
<robimarko>
I can tell you that ath11k works fine, better than with 300 backports
<nbd>
great
<nbd>
i will update it in the next few days and push it to master
<robimarko>
Awesome
MaxS0niX has joined #openwrt-devel
<\x>
is this something new?
<\x>
Wed Nov 30 18:10:14 2022 daemon.debug hostapd: phy5g-lan: STA e0:d4:e8:XX:XX:XX IEEE 802.11: start SAE authentication (RX commit, status=126 (SAE_HASH_TO_ELEMENT))
<\x>
a windows laptop seems to be doing this
<stintel>
that's just indicator of WPA3?
<\x>
hash to element?
<\x>
oh
<\x>
nvm, googled it
<Znevna>
any pointers on how can I make the wifi leds blink on this thing? x.x
<robimarko>
Znevna: What HW?
<Znevna>
rt-ax53u
<\x>
stintel: started seeing it just today that I made SAE/FT-SAE on main lan and just made a dedicated AP for legacy stuff https://rimgo.bus-hit.me/a/7Azzm4l or yeah maybe I only noticed it now. seems a lot are FT-SAE capable
<robimarko>
Znevna: What is the WLAN HW?
<Znevna>
MT7915e
<robimarko>
LED-s are connected to GPIO or?
<Znevna>
I've tried all gpios, and no
<robimarko>
Oh, so it could be connected to the radio
<robimarko>
I have no experience with MTK radios and that unfortunately
<Znevna>
we have this but I have no ideea on how to use it
<Znevna>
I have to set CONFIG_MT76_LEDS=y somewhere? and the leds are somehow muxed to gpio 18 and 26
<stintel>
interesting ... testing with CoDeL, iperf and iperf3 (UDP test with 2Gbps rate), notice that with iperf3 all packets bypass the qdisc (due to TCQ_F_CAN_BYPASS and empty queue), but with iperf (v2), all packets enter the qdisc
<stintel>
nbd: do you have an explanation for that?
<nbd>
multiple parallel streams?
<nbd>
or are you doing a single-stream test?
<stintel>
nbd: no
<stintel>
not using --parallel in either iperf
<nbd>
what kind of system?
<stintel>
OpenWrt on RPi4
<nbd>
and what effective transmit rate does it reach on them?
embargo has joined #openwrt-devel
<stintel>
nbd: iperf3 ~660Mbps, iperf2 ~500Mbps (from server report)
<stintel>
ok but I think I see it
<stintel>
I should look at the client (sender) rate
<stintel>
with iperf3 this starts at ~800, then drops to ~600
<stintel>
with iperf2 it claims 956Mbps
<stintel>
which is enough to saturate the gigE link most likely
<stintel>
hence queuing needs to happen
<stintel>
so the "throttling" is just iperf3 saturating a full CPU core
<stintel>
geez
<stintel>
nbd: thanks, I feel dumb now
robimarko has quit [Remote host closed the connection]
robimarko has joined #openwrt-devel
embargo has quit [Ping timeout: 480 seconds]
dangole has quit [Ping timeout: 480 seconds]
<dhewg>
nbd: your 6.1-rc4 backport patch works just fine for me on mt7621 and ipq4019, at least I haven't noticed any regressions so far
<Borromini>
dhewg: what MediaTek radios do you have?
dangole has joined #openwrt-devel
<dhewg>
In use is the mt7515e on a wac124 and a mt7921u attached to the ipq4019
<Borromini>
ok, ty
<dhewg>
there's a 2g mt7603e too on the wac124, but I don't use that
<Borromini>
:) I'm interested in MT7915.
<Znevna>
:<
<dhewg>
the 2g and 5g radio's are in use on the ipq, but those are non mt ofc
<Borromini>
it should be pretty trivial to backport to 22.03 since it shares the same wireless stack with master atm, but i'd be running a complete FrankenWrt I guess
<Borromini>
and I have ath9k/ath10k users as well, would be too messy. can't backport mac80211 per target/arch afaik :P
MaxSoniX has joined #openwrt-devel
<dhewg>
erm, mt7615e on wac, not mt7515e ofc
MaxS0niX has quit [Ping timeout: 480 seconds]
<dhewg>
and interessted how? as in buying a device with 7915?
<Borromini>
no i have a few running 22.03 (can't afford to switch those to master)
<Borromini>
yes, i gathered that was a typo :)
Tapper has quit [Ping timeout: 480 seconds]
G10h4ck has joined #openwrt-devel
embargo has joined #openwrt-devel
cbeznea has joined #openwrt-devel
MaxSoniX has quit [Remote host closed the connection]
MaxSoniX has joined #openwrt-devel
<\x>
dhewg: how is your ipq4019 handling that mt radio? when I tried with an intel radio its like when I run iperf3 server on the ipq4019 itself, I can get 850Mbps+, but when I run iperf3 server on another host bridged to it, I get like 550Mbps. does the same thing happen with the MTK radio?
embargo has quit [Read error: Connection reset by peer]
Tapper has joined #openwrt-devel
embargo has joined #openwrt-devel
torv has quit [Remote host closed the connection]
torv has joined #openwrt-devel
G10h4ck has quit [Read error: Connection reset by peer]
minimal has joined #openwrt-devel
<dhewg>
you mean the bitrate as reported by iperf3?
dangole has quit [Ping timeout: 480 seconds]
Mangix has quit [Read error: Connection reset by peer]
<dhewg>
because with a mt7921e connected to the mt7921u I do get "1200.9 Mbit/s, 80 MHz, HE-MCS 11, HE-NSS 2" rates, but iperf3 only gets ~200 MBits/sec
Mangix has joined #openwrt-devel
<dhewg>
doesn't matter if to that router itself or to a bridged host
<dhewg>
haven't looked into it yet why that is though, call me lazy
<dhewg>
maybe because of me being lazy and using irqbalance :P
<\x>
dhewg: yup, bench results on iperf3
<dhewg>
nice, I wonder why it's such a big difference for me
<dhewg>
will look into it some day, too many other things to do...
<\x>
try it, run the iperf3 server right inside that ipq4019 AP
<\x>
then iperf3 -R from a host connected to it
<\x>
then try iperf3 server on something bridged to it via ethernet
<dhewg>
what I did notice though, that this 7921<->7921 connection also get 40mhz rx and 80mhz tx
<dhewg>
need to try that -R later though
<nick[m]12>
hauke: is there some reference or simple steps explained to generate my own backports package for openwrt?
valku has joined #openwrt-devel
<nick[m]12>
I tried to compile drivers with from the kernel directly (with 6.1 kerbek). However, I face so many issues.
<nick[m]12>
So I thought about using the same commit hash to generate also a backport
<nick[m]12>
And just use mac80211 package as normal
<karlp>
nick[m]12: nbd has already gota 6.1 backports here: https://git.openwrt.org/?p=openwrt/staging/nbd.git;a=commit;h=6f3cac77439b6b57e5f9e7cc4882dcac0e443822
<karlp>
"nbd | i will update it in the next few days and push it to master"
<stintel>
hmmmz, there's no CONFIG_ option to set the default LAN IP address?
<aiyion>
I think it's fine for led_status to be permanently on, as its invoked by various aliases including led_running at the top of the dts.
<aiyion>
Why the rssi LEDs are lit, I do not understand at all.
<aiyion>
Setting triggers does not affect them either, apparently.
<aiyion>
On is on, one is off, one is heartbeat.
<Znevna>
those gpios seem oddly high
<Znevna>
how did you end up with the values?
<aiyion>
I followed the instructions on the wiki, detemined the base value (416) and toggled gpios in a binary search manner, until I could control them.
<Znevna>
I'm curios how the lan/wan leds work without being defined x.x
<Znevna>
the switch controls them directly?
<Habbie>
i was just reading the RTL8367S datasheet and it has pins for the port LEDs
<Habbie>
so, without having read any context for your question, it could be that simple
<Znevna>
so adding them in the dts would just provide the ability to turn them off
<Znevna>
or do something else fancy with them
<aiyion>
If possible at all; might be uncontrollable as the Power LED or the 4G LED at the bottom.
<aiyion>
I have looked quite a lot for a gpio I'm still missing (the rest line for the LTE modem) I hope I would've noticed if I had lit them up at one point.
<aiyion>
The stock devicetreeblob did not contain more than these four LEDs either.