schwicht_ has quit [Quit: Textual IRC Client: www.textualapp.com]
csrf has joined #openwrt-devel
Ansuel has quit [Quit: Probably my PC decided to sleep or I decided to sleep.]
<\x> patches/008-Enabling_vht_capability_for_2G.patch only needed this on top of pr-2522 it compiled fine, ill test later
<\x> oops Ansuel and robi isnt here
<\x> 009-ath11k-Enable-VHT-for-2G.patch *
Tapper has quit [Ping timeout: 480 seconds]
csrf has quit [Quit: Leaving]
danitool has quit [Quit: Cubum autem in duos cubos, aut quadratoquadratum in duos quadratoquadratos]
srslypascal has joined #openwrt-devel
nixuser has quit [Remote host closed the connection]
nixuser has joined #openwrt-devel
csrf has joined #openwrt-devel
guidosarducci has joined #openwrt-devel
srslypascal is now known as Guest1644
srslypascal has joined #openwrt-devel
Guest1644 has quit [Ping timeout: 480 seconds]
cbeznea has joined #openwrt-devel
srslypascal is now known as Guest1645
srslypascal has joined #openwrt-devel
Guest1645 has quit [Ping timeout: 480 seconds]
srslypascal is now known as Guest1646
srslypascal has joined #openwrt-devel
Guest1646 has quit [Ping timeout: 480 seconds]
bookworm_ has quit []
bookworm has joined #openwrt-devel
crzidea has joined #openwrt-devel
<crzidea> Hi, I had a PR for package ss-rules here: https://github.com/openwrt/packages/pull/19872
<crzidea> It has been 6 days, need help to review this PR.
cbeznea has quit [Quit: Leaving.]
<dhewg> crzidea: maybe because the prefix is off? s/ss-rules/shadowsocks-libev/
madwoota has quit [Ping timeout: 480 seconds]
csrf has quit [Ping timeout: 480 seconds]
madwoota has joined #openwrt-devel
<crzidea> Do you mean I should change the prefix to shadowsocks-libev?
<dhewg> yeah, I didn't know wtf ss-rules is, but shadowsocks rings a bell
<dhewg> also see its commit history
<crzidea> OK, I will try, thanks.
<crzidea> Updated.
crzidea has quit [Quit: Page closed]
<dhewg> also, 6 days are nothing: https://github.com/openwrt/openwrt/pull/4861
goliath has joined #openwrt-devel
mrnuke_ has joined #openwrt-devel
mrnuke has quit [Read error: Connection reset by peer]
Tapper has joined #openwrt-devel
Piraty has joined #openwrt-devel
danitool has joined #openwrt-devel
torv is now known as Guest1655
torv has joined #openwrt-devel
Piraty has quit [Remote host closed the connection]
Guest1655 has quit [Remote host closed the connection]
Piraty has joined #openwrt-devel
srslypascal is now known as Guest1661
srslypascal has joined #openwrt-devel
Guest1661 has quit [Ping timeout: 480 seconds]
<karlp> dhewg: for better or worse, while PRs on github for core open are accepted, it's third class citizen at best, if you want them addressed, they need to go to the mailing list.
<dhewg> it was like that in the past, but nowadays I'd say they're first class?
robimarko has joined #openwrt-devel
srslypascal has quit [Quit: Leaving]
<\x> hi robimarko
<\x> only need this on top of pr-2522 009-ath11k-Enable-VHT-for-2G.patch
<\x> already has the changes for netifd and hostapd and mac80211, thanks for sharing the resource
<robimarko> Nice that it works
MaxSoniX has joined #openwrt-devel
rua has quit [Quit: Leaving.]
<\x> very much thanks yeah ;) hope this gains some traction and 2522 can be considered again
<\x> I actually dont have any handheld wifi6 devices so this is quite the use for me!
srslypascal has joined #openwrt-devel
<aiyion> good morning. I've got the internal usb working and can detect my LTE modem (Quectel EC200T-EU) in lsusb as "2c7c:6026 Android Android"
<aiyion> where AndrewZ suggests configurig the modem like this: https://openwrt.org/docs/guide-user/network/wan/wwan/ethernetoverusb_cdc
rua has joined #openwrt-devel
<aiyion> cdc_ether does not feel responsible for the device though.
<f00b4r0> robimarko: made a fun discovery today: it seems the ipq40xx ethernet driver (non-DSA) ignores MTU setting for incoming packets *and* allows replying to DF ping packets with fragmentation: https://paste.debian.net/1260883/
<robimarko> Honestly, I wouldnt me shocked
<aiyion> This is the current output of `lsusb` and `cat /sys/kernel/debug/usb/devices` any idea what I might be missing?
<f00b4r0> robimarko: they have no respect for anything, do they? ;)
<robimarko> f00b4r0: Well, its the driver:
<robimarko> AR40XX_MAX_FRAME_SIZE_MTU, 9018 + 8 + 2);
<robimarko> ar40xx_rmw(priv, AR40XX_REG_MAX_FRAME_SIZE,
<robimarko> During setup
<f00b4r0> splendid
<karlp> aiyion: do you need usb mode switch?
<robimarko> Does swconfig even have an OP for MTU?
<f00b4r0> no idea. never needed to fiddle with swconfig to set MTU tbh
<aiyion> karlp: I'm not sure how I'd determine that.
<aiyion> but currently it is installed. Would it help to try without?
<aiyion> kmod-usb3 kmod-usb-net-cdc-ether usb-modeswitch usbutils
<robimarko> f00b4r0: Well, in theory setting MTU should work for the ethernet adaptor itself
<aiyion> karlp: these are the related packages I'm currently trying out.
<f00b4r0> robimarko: tbh I'm surprised that the upper layer accept a 9000 packet when the incoming interface has 1500 MTU (and happily fragment the reply to match said MTU)
<robimarko> f00b4r0: Me as well
<f00b4r0> (all that with DF set, too ;P)
<robimarko> Cause, switch wont drop it as its set to allow jumbo packets and EDMA has jumbo packets as max by default
<robimarko> But manually setting it should work still
<f00b4r0> heh. sounds like I really should move to DSA, this thing is utterly broken.
<karlp> aiyion: I'm no expert on modems, but do you have a particular reason to try and use it in cdcether like this? did you not get it working any other way? is this an external lte modem ?
<\x> robimarko: https://pastebin.com/raw/Anw1GntR my question is that, once those 256 and 512MB profile be fixed and usable, can I do mx match of this? like for example, I use 16MB for NSS and 40MB for WLANQ6
<karlp> the cdc ether bit is basicalyl saying, "if the modem supports running as it's own computer, and providing a gadget usb ethernet, let me use it"
<karlp> which is a very different mode of operation than "use this as a usb modem"
<\x> mix match*
<robimarko> \x: 256MB will never be usable, just forget that
<\x> ah
<aiyion> karlp: neither am I, help is greatly appreciated. The reason I'm doing it like this is strictly following https://openwrt.org/docs/guide-user/network/wan/wwan/ethernetoverusb_cdc
<robimarko> \x: As for mix and matching, well you can only try and see what happens
<\x> so I guess im only looking forward for 30MB, from 1G profile to 512M
<aiyion> root@OpenWrt:~# opkg install kmod-usb-net-cdc-ether usb-modeswitch
<aiyion> that's the line I'm referring to.
<aiyion> I'll remove the modeswitch and report back. brb.
<\x> but tbh, even with the 1GB profile currently, its running well with zram, but i havent done some real long uptime testing since theres always something new being fixed along the way :D
<karlp> aiyion: right, but why are you following that guide?
<aiyion> The Forum Post I mentioned at 11:21:52
<karlp> basically, if you install cdc-ether, and then replug the modem, and it doesn't give you a new ethernet interface, then it's not a relevant method, as your modem doesn't work that way...
<aiyion> I've got a EC200T-EU and the post is about EC200T-CN and seems to have worked out. Was hoping they'd be similar.
<\x> so the only mix match that is cool is like, 8M for NSS and 55M on Q6, that gains 38MB from 1G profile but meh, still not a lot
<aiyion> karlp: just to be clear, replugging the modem is not an option, though its using usb its all onboard.
<dhewg> \x: do you have #2522 rebased with your patch somewhere to fetch? or maybe open a new PR? :)
<\x> im not really knowledgeable with github
<karlp> aiyion: if it's a fixed onboard modem, I wouldn't expect cdc-ether to be the method it uses to operate.
<\x> or git. I just massage the changes locally
<karlp> but hey, people do all sorts of thigns.
<dhewg> hehe okay
<\x> but yes, I have tried it on many devices, ipq40xx, all work even as a client with ath10k-ct
<\x> intel cards, 7260 and newer, tested on 8260, 9260, ax200
<\x> ipq60xx too, just today
<karlp> aiyion: so it's still a card module though, after you isntalled cdc-ether, if you can't re-plug, you'd probably need to reboot then.
<\x> dhewg: this is two ipq4019 connected to each other https://i.imgur.com/VW5dy3D.png
<\x> they are doing WiFi 5 / 11ac over 2.4GHz, and as a bonus MU-MIMO gets enabled, but im not sure how much helpful that is seeing this only has a 2x2 2.4GHz radio
<\x> dhewg maybe this is what you mean? https://pastebin.com/raw/JQ72HauW
<\x> then you just copy package/kernel/mac80211/patches/subsys/600-mac80211-allow-vht-on-2g.patch
<Namidairo> i can't get this 7260 to do it. but then again i probably need a client patch.
<\x> dhewg: maybe Ansuel will make a pr, since dengqf6 (2522 op) seems inactive
<\x> Namidairo: i have not tried client
<\x> it works as AP
<Namidairo> you're using a 7260 as AP?
<\x> you can on 2.4 for sure
<\x> here is 8260 Namidairo https://i.imgur.com/j0AcJyP.png
<\x> no driver patch needed, just 2522 and enable vendor_vht on wifi config
<Namidairo> I... don't feel like patching mac80211 on here
<\x> then can only go up to 11n on 2.4GHz, 11ac can be done on 5Ghz but will need the hostapd patch from tildearrow
<\x> very hacky
<\x> Namidairo: you can get an extra usb and install your modified openwrt on that, boot it and then try ;) that way you dont have to patch linux that youre using.
<Namidairo> or I could just take your word for it
<\x> it werks on 8260 atleast!
<\x> in theory any 256-qam capable card from intel can do this so yeah 7260 included
<Namidairo> i think most 11ac cards should probably work
<\x> most phones made with an SoC 2019 onwards also does it too, even on hotspot/tether mode, mediatek and qualcomm does it
<Namidairo> i believe my broadcom-toting 11ax capable phone does it too
<\x> yup you can check, just enable hotspot mode, then scan it with iw scan or just use another phone and check it on aruba tools
<\x> aruba utilities im sorry, its like this https://i.imgur.com/Jhs38kq.png
<Namidairo> it's broadcom so I think it's only doing it via the broadcom vendor IE
<Namidairo> the one where they copy paste the VHT capa and operation bits in
<Namidairo> Vendor specific: OUI 00:90:4c, data: 04 08 bf 0c 32 78 81 0f fa ff 00 00 fa ff 00 00 c0 05 00 01 00 00 00 c3 02 00 68
bluew has quit [Ping timeout: 480 seconds]
cbeznea has joined #openwrt-devel
<\x> wifiman from ubiquiti also works for this
<\x> itll show the max speed
<\x> 2x2 40 11n doesnt have that speed
<dhewg> \x: nice, worked with a ipq4019 router: 192.7 Mbit/s, 20 MHz, VHT-MCS 9, VHT-NSS 2, Short GI
<\x> yup ;)
<\x> i hope this thing gets considered again or something, even as a hack
<\x> dhewg: it should also enabled the beamformer, check it, grep -i beamf /var/run/hostapd-phy*
<\x> but maybe only useful if you have multiple 1x1 clients hehe
<dhewg> nope, not in there
<\x> weird, it works on my ipq4019 device on current master
<\x> i just reflashed like yesterday
Albert has joined #openwrt-devel
<dhewg> did you add that? I didn't, and it did not automatically, if I understood you correctly
<\x> i didnt have to touch anything with beamforming options, basically i just added option vendor_vht '1'
<dhewg> yeah, that's what I did
<dhewg> I also have an older router with AR9381, that prolly doesn't support vendor_vht eh?
<Albert> Help: When I use mwan3, Fast Path doesn't seem to work, only 300M throughput
<\x> device here is R619AC
<\x> im using ath10k-ct
rua has quit [Remote host closed the connection]
floof58 is now known as Guest1669
floof58 has joined #openwrt-devel
rua has joined #openwrt-devel
<dhewg> oh wait, BEAMD* is in there, I grepped for lowercase
<dhewg> *BEAMF*
<\x> ah yeah
<\x> i did say grep -i there you mightve missed hehe
<dhewg> I don't have greenfield, what's that again?
<\x> but yup, its basically ac wav2 over 2.4GHz
<\x> greenfield ignores 11b traffic
Guest1669 has quit [Ping timeout: 480 seconds]
<dhewg> ah that one
<\x> it treats it as noise something like that
Albert has quit [Quit: Page closed]
<\x> dhewg: what clients do you have that work with it?
<\x> Sometime ago Ansuel wanted to compile a list of devices where this works, even client ones to convince upstream something like that
<\x> and nbd will discuss it with someone upstream
<dhewg> \x: only tried a phone so far, poco x3 pro
<\x> yup, basically i notice its like any soc that does ac with a linux 4.14 or newer has it on qualcomm side
<\x> I have an old phone that is ac capable on 5ghz but doesnt do it on 2.4, it has an sd617 on it on linux 3.10 something
<\x> mediatek also has this enabled, but i cant "profile" them since I only seen them from guests, I never let my fam to get a non snapdragon phone since im always the one getting hand-me-downs and I want custom rom support heh
<dhewg> does it work from standard debian with wpa_supplicant?
<dhewg> wpa_cli set vendor_vht=1
<dhewg> Selected interface 'wlp2s0'
<dhewg> FAIL
<\x> dhewg: you need to patch the kernel
<dhewg> damn :P
<\x> package/kernel/mac80211/patches/subsys/600-mac80211-allow-vht-on-2g.patch
<dhewg> yeah, I know, but I won't do that on this dev
<\x> yeaah
<\x> i dont do that too hehe
<aiyion> karlp: blocktrron suggested using kmod-usb-net-rndis instead; which seems to be working, as it acts as driver for the usb device opposed to kmod-usb-net-qmi-wwan or kmod-usb-net-cdc-ether, I tried before.
<\x> it does work on ipq4019 with ct dhewg, it does vht/2.4 as a client
<aiyion> Following the datasheet I should now be able to send AT commands to the interface.
<aiyion> I just do not see yet, why the cdc-ether kmod did not work as a driver.
<aiyion> EC200T's datasheet suggests it does support ECM.
<dhewg> \x: I'm using this on non -ct ipq4019 because of mesh
<\x> ah yeah theres a patch for non-ct hehe
<\x> but ofcourse as long as it works, das good. a lot of WiSoC and most 11ac radios being ran on 2.4GHz can benefit from this
<karlp> yeah, net-rndis or net-cdc-ether ar eboth "the device wants to be a mini computer of it's own, and it will provide you with an IP directly" just different types. qmi is "it's a modem, and we manage it"
<karlp> some dongles offer both modes, via usb mode switch, some one or the other...
Tapper has quit [Ping timeout: 480 seconds]
minimal has joined #openwrt-devel
MaxSoniX has quit [Quit: Konversation terminated!]
MaxSoniX has joined #openwrt-devel
MaxS0niX has joined #openwrt-devel
MaxSoniX has quit [Ping timeout: 480 seconds]
madwoota has quit [Ping timeout: 480 seconds]
<aiyion> the mode switch packages is meant to switch between rndis and ecm?
<aiyion> karlp: ^
<aiyion> so if rndis worked without the modeswitch package, I'd expect the cdc-ether one to work with modeswtich, right?
MaxSoniX has joined #openwrt-devel
madwoota has joined #openwrt-devel
<karlp> aiyion: normally mode switch would be between qmi and <the other one> be that rndis or cdc-ecm/ncm
MaxS0niX has quit [Ping timeout: 480 seconds]
<karlp> usb modeswitch can normally go in both ways, what it defaults to I have no idea :)
<aiyion> ah
<karlp> more classically, it was to mode switch between "appear as a usb mass storage device with the drivers" and "appear as the modem itself"
<karlp> but it now gets used for all sorts of back and forths.
<aiyion> thanks
<aiyion> Well. The datasheet suggest one can switch between rndis and ecm `AT+QCFG="usbnet",1` or three respectively.
<aiyion> the datasheet encourages to use rndis, "It is recommend to use RNDIS driver because it has better compatibility than ECM drive"
<aiyion> has ECM any benefits, that would be worth switching the mode actively? Or should I just be fine using rndis, as that's what the vendor uses as well?
<karlp> ecm is faster, rndis is built into windows and runs without drivers... depending on how fast the modem is the "speed" is likely irrelevant...
<aiyion> mhm. I mean for my usage in OpenWrt.
<karlp> whichever works is fine...
<aiyion> thanks :)
Piraty has quit [Remote host closed the connection]
Piraty has joined #openwrt-devel
Piraty has quit [Remote host closed the connection]
Piraty has joined #openwrt-devel
Tapper has joined #openwrt-devel
Piraty has quit []
Piraty has joined #openwrt-devel
Piraty has quit [Remote host closed the connection]
Piraty has joined #openwrt-devel
floof58 is now known as Guest1682
floof58 has joined #openwrt-devel
Guest1682 has quit [Ping timeout: 480 seconds]
MAbeeTT has joined #openwrt-devel
MAbeeTT5 has quit [Ping timeout: 480 seconds]
Misanthropos has quit [Ping timeout: 480 seconds]
Misanthropos has joined #openwrt-devel
MaxS0niX has joined #openwrt-devel
valku has joined #openwrt-devel
MaxSoniX has quit [Ping timeout: 480 seconds]
<\x> robimarko: can I use these? https://i.imgur.com/m4gEd87.png
<\x> like as firmware
<\x> this is from the stock rootfs
<\x> vendor*
<robimarko> \x: No idea really, I suppose they are rather old
Ansuel has joined #openwrt-devel
danitool has quit [Ping timeout: 480 seconds]
cbeznea has quit [Quit: Leaving.]
<russell--> hmm, what happened to gpio direction file in sysfs?
Ansuel has quit [Read error: Connection reset by peer]
MaxS0niX has quit [Remote host closed the connection]
MaxS0niX has joined #openwrt-devel
Borromini has joined #openwrt-devel
MaxS0niX has quit [Quit: Konversation terminated!]
MaxS0niX has joined #openwrt-devel
MaxS0niX has quit []
MaxS0niX has joined #openwrt-devel
Lynx- has joined #openwrt-devel
<Lynx-> easy way to randomly shuffle bash array elements?
danitool has joined #openwrt-devel
<KGB-1> https://tests.reproducible-builds.org/openwrt/openwrt_kirkwood.html has been updated. (100.0% images and 99.6% packages reproducible in our current test framework.)
zatwai has quit [Quit: ZNC 1.8.2+deb2+b1 - https://znc.in]
zatwai has joined #openwrt-devel
MaxSoniX has joined #openwrt-devel
Lynx- has quit [Ping timeout: 480 seconds]
MaxS0niX has quit [Ping timeout: 480 seconds]
nixuser has quit [Ping timeout: 480 seconds]
Tapper has quit [Ping timeout: 480 seconds]
MaxSoniX has quit [Quit: Konversation terminated!]
minimal has quit [Quit: Leaving]
Lynx- has joined #openwrt-devel
<russell--> cat /sys/class/gpio/foo/direction --> in ; echo out > /sys/class/gpio/foo/direction ; cat /sys/class/gpio/foo/direction --> in
<russell--> wtf?
<robimarko> You sure that pin is bidirectional?
<russell--> it was 6 months ago
<robimarko> Sorry, but I am not following
<russell--> this is for an asiarf mt76x8-based ramips board, as of an openwrt version 6 months ago, the gpio direction worked, i rebased my stuff recently (a couple weeks ago) and noticed this change
<stintel> I should really work on implementing nftable set support in vallumd
Tapper has joined #openwrt-devel
<russell--> looks like a hardware problem...
<aiyion> how do I set the loglevel in /etc/init.d/network ?
<aiyion> "Add the switch -l <level> to /etc/init.d/network"
<aiyion> where do I put that switch?
<schmars[m]> aiyion in init_switch() before or after setup_switch i would guess
<aiyion> or in setup_switch? :D
<aiyion> nevermind got enough verbosity in logread
Lynx- has quit [Quit: Going offline, see ya! (www.adiirc.com)]
robimarko has quit [Quit: Leaving]
<KGB-1> https://tests.reproducible-builds.org/openwrt/openwrt_x86.html has been updated. (100.0% images and 99.6% packages reproducible in our current test framework.)
minimal has joined #openwrt-devel
ynezz has quit [Ping timeout: 480 seconds]
Borromini has quit [Quit: leaving]
Tapper has quit [Ping timeout: 480 seconds]
ynezz has joined #openwrt-devel
ynezz is now known as Guest1713
csrf has joined #openwrt-devel
csrf has quit []
* stintel curses libnftnl
<owrt-snap-builds> Build [#390](https://buildbot.openwrt.org/master/images/#builders/72/builds/390) of `imx/cortexa9` completed successfully.
Misanthr- has joined #openwrt-devel
Misanthropos has quit [Ping timeout: 480 seconds]
Tapper has joined #openwrt-devel
Guest1713 has quit [Ping timeout: 480 seconds]