mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
<russell-->
i'm getting an lzma decode error trying to tftp boot on a gl-inet gl-ar750
<russell-->
needed to add CONFIG_TARGET_INITRAMFS_COMPRESSION_LZMA=y
Tator has quit [Remote host closed the connection]
zatwai has quit [Quit: ZNC 1.8.2+deb2+b1 - https://znc.in]
zatwai has joined #openwrt-devel
valku has quit [Quit: valku]
kistlin has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
srslypascal is now known as Guest3484
srslypascal has joined #openwrt-devel
goliath has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
Guest3484 has quit [Ping timeout: 480 seconds]
mzvd has joined #openwrt-devel
srslypascal is now known as Guest3487
srslypascal has joined #openwrt-devel
Guest3487 has quit [Ping timeout: 480 seconds]
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
goliath has quit [Quit: SIGSEGV]
<SwedeMike>
rsalvaterra: Dave Täht responded re CAKE and you were indeed right, CAKE "does a better job" but it takes more CPU (he mentioned 2.5x slower)
robimarko has joined #openwrt-devel
<russell-->
how is "2.5x slower" consistent with "does a better job"?
<stintel>
just run some flent rrul tests and see for yourself
fblaese has joined #openwrt-devel
frwol has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
csrf has quit [Ping timeout: 480 seconds]
mzvd has quit [Read error: Connection reset by peer]
<rsalvaterra>
SwedeMike: Right, like I interpreted.
<rsalvaterra>
russell--: Speed isn't everyting. Just like efficiency isn't effectiveness.
<stintel>
for DoT you can probably just use tcp:853
<stintel>
DoH or DoQ is going to be another story
zexon has quit [Quit: Page closed]
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
borek has joined #openwrt-devel
<karlp>
what's the Q stand for? that one's new to me
<Habbie>
QUIC
<Habbie>
udp:853, so not a big problem
<stintel>
ah that's not using 443 - great
<Habbie>
yep
<karlp>
is that not just http2 now?
<stintel>
quic = http3
<Habbie>
there was some discussion about numbers but 443 was not on the menu
<Habbie>
karlp, no
<karlp>
stintel: thanks
<Habbie>
stintel, also no
floof58 has quit [Ping timeout: 480 seconds]
<Habbie>
http/3 runs -over- quic
<Habbie>
DoQ runs over quic; no http involved
<rsalvaterra>
stintel: I know, but matching on tcp:853 is still too broad for my taste… :/
<Habbie>
(http2 runs over tcp)
<rsalvaterra>
Maybe QUIC (HTTP/3) is the final push for fascist sysadmins to enable UDP on their networks… :)
<stintel>
rsalvaterra: at some point you need to tell your OCD to fuck off :P
<stintel>
tcp:853 should really be enough :)
* stintel
is using it too ;)
<rsalvaterra>
*sigh*
<rsalvaterra>
You're not wrong. :P
<stintel>
or use a different IP for every service and do it IP based - there's plenty of IP's available ;)
<rsalvaterra>
Did you edit /etc/qosify/00-defaults.conf?
<stintel>
no I added 99-local.conf
<rsalvaterra>
Fair enough, it does seem a better way, yes.
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
<stintel>
the thing with qosify is both the tcp and udp map have an entry for each possible port, this design won't scale with IP:port
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
<Tapper>
rsalvaterra
<Tapper>
stintel If you have made changes to /etc/qosify/00-defaults.conf can you share? I set mine up from the forums. I don't really know if it's that good. I will past bin it. Here's what I have. https://pastebin.pl/
<robimarko>
Yeah, I hit the same issue multiple times
<robimarko>
Its rather annoying when trying to debug
mzvd has joined #openwrt-devel
<stintel>
robimarko: thank you btw for pointing that out
<stintel>
that's a 10 year old patch, we should probably investigate if we can ditch it in favor of CONFIG_NO_STDOUT_DEBUG=y
<stintel>
for now I'm going to build with WPA_MSG_MIN_PRIORITY=0
mzvd has quit [Read error: Connection reset by peer]
<robimarko>
CONFIG_NO_STDOUT_DEBUG should do the same as the patch
<robimarko>
But in a more sane way, we should probably get rid of the patch
<stintel>
although CONFIG_NO_STDOUT_DEBUG exists since 2009, so there must have been a reason for adding that patch
<stintel>
might no longer be valid though
<robimarko>
AFAIK, CONFIG_NO_STDOUT_DEBUG should remove all debug prints
<stintel>
yes, the custom patch makes that level configurable
<stintel>
we should really try to reduce the amount of patches we carry for hostapd though
<stintel>
I'm trying to get approval to work on upstreaming all the ubus related stuff
<stintel>
because at least qsdk and mediatek sdk are based on openwrt so they use ubus, so I'm 99.9% sure there are waaaay more users of hostapd ubus than hostapd dbus, yet ubus not upstream and dbus is
<stintel>
and having the ubus code upstream will expose it to broader review, etc
<stintel>
and reduce maintenance burden
<robimarko>
Isnt dbus wpa_supplicant only
<stintel>
that could be, either way, there are so many hostapd ubus users in the world that it makes sense to upstream it, imo
<robimarko>
Yeah, it makes a lot of sense to upstream it
<kistlin>
there i see that the data status is "connected"
<kistlin>
what are the next steps that i see an interface to ping something on the internet through the lte module?
<kistlin>
a config for the module is in /etc/network/config
<kistlin>
well i see a wwan0 but cant set it to up nor do i get an IP
<stintel>
nick[m]1234: I guess you're interested in adding MBO reason codes to DAWN :)
<nick[m]1234>
stintel: yes ^^ but I am still a bit unsure what exactly is the benefit. I try to find some doucments explaining the new stuff, but I can not find anything really useful :/
<stintel>
nick[m]1234: yeah the WFA document is members-only I'm afraid
<nick[m]1234>
stintel: sad. I am not a member. :(
<stintel>
either way, use of these codes will make client steering cleaner. the effective usefulness will still depend on how clients handle it, as usual :)
<ynezz>
eduble: back in the days, I was using following base-files-nfs package http://sprunge.us/klV5zh
danitool_ has quit [Quit: Cubum autem in duos cubos, aut quadratoquadratum in duos quadratoquadratos]
<eduble>
ynezz: thanks, I'm looking at it
valku has joined #openwrt-devel
mzvd has joined #openwrt-devel
<jow>
stintel: that patch adds a configurable cutoff limit for debug messages
<jow>
like "compile with only debug messages having priority >= x"
<stintel>
28|14:47:41 < stintel> yes, the custom patch makes that level configurable
<stintel>
did we ever try to upstream that ?
<jow>
no, because it's one of these optimizations that only make sense on weird embedded systems
<stintel>
weird embedded systems are the biggest users of hostapd
<jow>
;)
<stintel>
I might try upstreaming some hostapd patches at some point, using that as argument
<jow>
I like that argument
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
mva has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
<mrkiko>
kistlin: so, dhcp doesn't work over that interface?
mzvd has quit [Read error: Connection reset by peer]
<mrkiko>
kistlin: some LTE modules provide an emulation of dhcp or something like that, some others need you to configure the IP statically after getting parameters from the QMI interface; don't know how openwrt / uqmi is supposed to handle this case
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
<kistlin>
mrkiko: hmm, i managed to ping google with AT commands so that internet connection should work.regarding dhcp do you mean i should run something?
mzvd has quit [Read error: Connection reset by peer]
<mrkiko>
kistlin: in general all should happen automatically, but still - what happens if you run udhcpc against wwan0 ?
frwol has quit [Quit: leaving]
<kistlin>
mrkiko: it assigned an ip ...
<mrkiko>
kistlin: ok, so now you should be able to ping the Internet, right?
<kistlin>
mrkiko: yeah it works. thanks.
<mrkiko>
kistlin: but it should happen automatically, don't know why; try adding such an entry to etc/config/network
<kistlin>
maybe it doesn't yet work automatically because if it starts up it goes into 3G only but not LTE
<kistlin>
mrkiko: do you mean "config interface wwan0" and there an option proto 'dhcp'?
<mrkiko>
kistlin: yeah
<mrkiko>
kistlin: but I have very very little, near to zero, experience with this thing in OpenWrt itself, I usually use ModemManager
<kistlin>
mrkiko: hmm there i have the options proto 'qmi'
<kistlin>
what i don't get is wwan0 shows up with "ip addr" but in the config it has a different name
<mrkiko>
kistlin: which is effectively right ... ohm, sorry, I was saying something wrong
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
Gaspare has joined #openwrt-devel
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]