Tapper has quit [Read error: Connection reset by peer]
dangole has quit [Remote host closed the connection]
dangole has joined #openwrt-devel
<mangix>
rsalvaterra_: those initvals are done differently by qca8k
<rsalvaterra_>
mangix: Got it, thanks!
<rsalvaterra_>
Since you beat me to it, do you need testing on the 3600?
<rsalvaterra_>
You said the LAN ports were probably reversed.
<mangix>
you'll see it if they are
<mangix>
lan1 being lan4 in reality
<mangix>
that's just cosmetic
<mangix>
0x10 0x80000080 /* POWER_ON_STRAP */ <-- this line is potentially worrying. It's not present with the archer c7 setup. But maybe it works.
<rsalvaterra_>
Sure, but it would be nice to match the port labels to the internal ones. :)
<mangix>
rsalvaterra_: force pushed new commit
<mangix>
rsalvaterra_: you can fix those yourself if it's the case
<mangix>
wait a minute...
goliath has quit [Quit: SIGSEGV]
<mangix>
oh I see
<mangix>
they're not reversed
<mangix>
I guess the swconfig looks like that because there's only 1 cpu port.
<Tusker>
anyone know how to get a watchguard xtm330 to drop to uboot shell ?
<stintel>
Tusker: on the M300 it goes there automatically when the sdcard does not contain the stuff u-boot env expects, but not experience with the xtm330
<stintel>
what arch is that ?
<Tusker>
ppc 2020rdb
<Tusker>
maybe I'll open it up and remove the sdcard
<stintel>
cool, moar ppc
<stintel>
ah, also e500, so no altivec so musl no worky
<rsalvaterra_>
stintel: I thought that restriction was only for ppc64, not all of ppc. Oh, well…
<Tusker>
stintel[m]: hmm? is that a big deal, I have the aerohive ap370 working fine on openwrt ?
<stintel>
actually don't remember if that was ppc64 only
<rsalvaterra_>
stintel: And we have the mpc target (TL-WDR4900), which is ppc32 and works just fine… unless I'm missing something.
<stintel>
right, never mind me, p2020 is supported by mpc85xx
<Tusker>
no sdcard inside the xtm330, it's nand storage only it seems
<Tusker>
there are some jumpers, which may change behaviour though
<stintel>
oh :(
<rsalvaterra_>
Yikes, almost 1:30 here, time to hit the bed. Cheers, guys!
<Tusker>
yup, that did it
chadneufeld has quit [Ping timeout: 480 seconds]
<Tusker>
jumper1 switches the mode to xtm505 mode
<Tusker>
which contains a uboot prompt
chadneufeld has joined #openwrt-devel
<Slimey>
stintel still playing with my mpc85xx targets :P
<stintel>
;)
<grid>
stintel: i ordered a m300
<grid>
if they will ever ship it
<stintel>
grid: cool
<grid>
will of course try your branch
<stintel>
yeah, a Tested-by is always nice
<grid>
ya idk if i'll realistically use it, the p2041rdb is pretty chunky as it is, and the fans will be a bit much i think
<grid>
i took the p2041rdb out and put a quieter heatsink/fan on it. dunno if that's practical with this one, though it does kind of look like a matx board
<Tusker>
i've disconnected the fan of the xtm330 while it is sitting on the desk next to me... but the routers and switches are in the subfloor, so it's fine to have fans
<grid>
fancy
<stintel>
<1k EUR new, couldn't resist, as I had a period with loooooots of power outages here
<grid>
is xtm-330 t1042 or something?
<Tusker>
CPU0: P2020E, Version: 2.1, (0x80ea0021)
<grid>
oh
<stintel>
t1042 = m200
<stintel>
iirc
<grid>
the p2041 draws like 13w idle so i couldn't passively cool it even with a massive heatsink
<grid>
all though serdes lanes maybe
<grid>
all those*
<grid>
Tusker: did you see any random panics on the p2020?
<Tusker>
no... i've only had it powered on for an hour so far
<chadneufeld>
hi all. I am trying to figure out the WDS link connection issue on Ath9k and Ath10k. Is anyone able to help me out? Thanks
victhor has quit [Ping timeout: 480 seconds]
<Tusker>
chadneufeld: what is the issue ?
<chadneufeld>
After 21.02_rc4 the WDS Link I had setup stopped working.
<chadneufeld>
Both the client and the master show a link and a connection, but no data flows between the two.
<chadneufeld>
I've narrowed it down to one specific commit in NETIFD but I don't know where to go from there.
<chadneufeld>
This is the NETIFD commit that makes WDS not work. https://git.openwrt.org/?p=project/netifd.git;a=commit;h=f037b082923abc2dad0d14c8401ebe0afd816b5c
<Tusker>
hmm... the change looks sensible
<chadneufeld>
I got some help setting up the NETIFD tree so I can compile the NETIFD package along with my OPENWRT image.
<chadneufeld>
Something with that commit breaks WDS on Ath9/10k.
<chadneufeld>
Could it be something to do with DSA and that it's not rolled out to Ath9/10 yet?
<Tusker>
yeah, not really sure about the reasoning behind the failure. might have to wait for someone who is more familiar with that code base
<chadneufeld>
no problem. Thanks for the help.
Slimey has quit [Remote host closed the connection]
<chadneufeld>
Hey Tusker. I went through and manually undid the changes from the 3 files in that NETIFD commit. I undid everything related to "wireless_device_hotplug_event"
<chadneufeld>
My WDS links seem to work well now with the latest NETIFD and the device hotplug removed.
<Tusker>
ok, interesting :)
lmore377 has joined #openwrt-devel
chadn has joined #openwrt-devel
chadneufeld has quit [Read error: Connection reset by peer]
<chadn>
Yeah. I really have no idea why it would work or wouldn't work.
<chadn>
Hopefully someone else can explain why and propose a proper fix.
rmilecki has joined #openwrt-devel
DonkeyHotei has joined #openwrt-devel
robin_ has joined #openwrt-devel
<Tusker>
*sigh* why would 21.02.0-rc4 work on p2020rdb but in snapshot it hangs straight away ?
<dansan>
PaulFertser: Well a late thank you for the link!
<dansan>
Sadly I don't see documentation for how to use DUMP.
<dansan>
Maybe just DUMP=1 works, but it blew up the last time I tried -- then again, I had a lot of other changes in the tree.
<Tusker>
trying to build a 21.02 branch to see if it works locally ...
guidosarducci has quit [Remote host closed the connection]
<rsalvaterra_>
aparcar[m]: LGTM. I still have the APU in the office at 5.4, but ldir has been testing 5.10 for quite a while, so I think we're all good.
KGB-2 has quit [Quit: KGB-2]
<aparcar[m]>
rsalvaterra_: oh yea please scroll down and read the comment.
<aparcar[m]>
it's working for me too but people having trouble
<rsalvaterra_>
aparcar[m]: Wait, on QEMU…?
<rsalvaterra_>
WTF…!
<Tusker>
aparcar[m]: OK, maybe I will see if I can create a hard link or something... or bind mount it
KGB-2 has joined #openwrt-devel
<mangix>
implicit declaration of function 'dev_err'
<mangix>
rofl
<rsalvaterra_>
aparcar[m]: That's really strange… either there's a problem with our initramfs image generation, or QEMU is acting up.
<aparcar[m]>
I'm currently running another build but will test that in a bit
<rsalvaterra_>
aparcar[m]: Does it also happen on bare metal?
<aparcar[m]>
no the reporter tries it via qemu
<rsalvaterra_>
aparcar[m]: Sure, but if we knew it worked on bare metal, we could rule out any issues with our image generation (I seriously doubt it's a QEMU issue, though).
<rsalvaterra>
rmilecki: I'm bumping 5.10 to 5.10.65 and I had to massage your 800-bcma-get-SoC-device-struct-copy-its-DMA-params-to-th.patch a bit. I'll likely ask you to take a look, after I finish, since I don't have the hardware to test. :)
<rmilecki>
rsalvaterra: sure
<aparcar[m]>
Tusker: done?
nitroshift has quit [Remote host closed the connection]
<rsalvaterra_>
Hm. Looks like I got the Quassel core and Android client working. 😎
chadn has quit [Ping timeout: 480 seconds]
<stintel>
I threw that out, horrible experience
<rsalvaterra>
Still need to set up the nickserv identification…
<stintel>
for me it was constantly disconnecting, then causing a truckload of notifications on each reconnect
rsalvaterra has quit [Quit: Leaving...]
rsalvaterra_ has quit [Quit: https://quassel-irc.org - Converse confortavelmente. Em qualquer lugar.]
rsalvaterra has joined #openwrt-devel
dangole has quit [Ping timeout: 480 seconds]
victhor has joined #openwrt-devel
rsalvaterra has quit []
minimal has joined #openwrt-devel
rsalvaterra has joined #openwrt-devel
rsalvaterra has quit [Quit: https://quassel-irc.org - Converse confortavelmente. Em qualquer lugar.]
rsalvaterra has joined #openwrt-devel
victhor has quit [Quit: Leaving]
victhor has joined #openwrt-devel
victhor has quit []
victhor has joined #openwrt-devel
<rsalvaterra>
stintel: What kept disconnecting? The core, or the clients?
<stintel>
rsalvaterra: the android client
<rsalvaterra>
Well, I don't plan to keep the client always connected. However, I'm hoping things will be relatively stable over WireGuard, as it takes care of the roaming. 😉
<rsalvaterra>
But the font rendering of the macOS Qt client… God, it's atrocious.
<rsalvaterra>
(The macOS font rendering is atrocious, in general…)
<stintel>
s/font rendering //
<mrkiko>
I wanted to build an iamge without usb support for my RT3200 (mt7622) without usb support at all. But still I have CONFIG_USB_SUPPORT=y. What am I missing here?
<rsalvaterra>
stintel: Couldn't agree more!
<rsalvaterra>
mrkiko: USB support is… interesting. You need to disable it in the kernel configuration itself, otherwise it will pop up again in the OpenWrt configuration.
<stintel>
this piece of crap macbook frequently loses its connection, sometimes unplug/replug the usb-c ethernet/dp adapter fixes it, sometimes needs reboot
<stintel>
had similar issues with macbook air before I ditched apple completely 6 years ago or so
<mrkiko>
rsalvaterra: thanks! But it's not like that on all targets, or am I wrong?
<mrkiko>
rsalvaterra: because in the end I find the hci driver in the image still. No issues, just wondered why this happens
<rsalvaterra>
mrkiko: It's like that on ramips and ath79, for sure.
<rsalvaterra>
I don't know about other targets, though.
<rsalvaterra>
mrkiko: I guess it's all targets. https://git.openwrt.org/?p=openwrt/openwrt.git;a=blob;f=include/target.mk;h=7526224972e18148fec8a12318ca7f90a382475f;hb=HEAD#l276
<rsalvaterra>
Instead of the usb feature selecting CONFIG_USB in the kernel, it's the other way around.
<rsalvaterra>
The reasons? I have no idea. It was already like this when I arrived. :P
<mrkiko>
rsalvaterra: thanks
Tusker has quit [Quit: Time wasted on IRC: 11 hours 54 minutes 36 seconds]
rua has joined #openwrt-devel
* nitroshift
hands stintel a 10-pounds hammer
<nitroshift>
0/
<stintel>
;)
<blocktrron>
mangix: What do you mean with that?
<blocktrron>
Ah, i should read the following patches.
<blocktrron>
I did not came around to send them upstream yet.
<blocktrron>
AFAICT DSA currently has the issue of not being able to deal with the limitation of the Switch not being able to support hybrid port operation
<blocktrron>
But for regular bridge offloading, the patchset should be possible to rebase against latest kernel.
slh_ has joined #openwrt-devel
slh64 has quit [Read error: Connection reset by peer]
slh has quit [Remote host closed the connection]
<blocktrron>
I'd like to have the patches upstreamed prior utilizing them downstream.
slh has joined #openwrt-devel
wvdakker has quit []
wvdakker has joined #openwrt-devel
valku has joined #openwrt-devel
<rsalvaterra>
stintel: Any idea on how to make git am behave like rebase, when conflicts are found?
<rsalvaterra>
When doing git am, it just aborts on conflict.
<PaulFertser>
rsalvaterra: git am -3
<rsalvaterra>
Hm, no 3-way compares, heh? Let's see…
<rsalvaterra>
Oh, wait, you mean forcing a 3-way compare.
<rsalvaterra>
That actually worked, thanks!
<stintel>
that's good to know
<stintel>
TIL... thanks PaulFertser
<PaulFertser>
My pleasure :)
JohnA has joined #openwrt-devel
Slimey has joined #openwrt-devel
<rsalvaterra>
stintel: The more I learn about git, the more amazed/scared I get. :)
<stintel>
rsalvaterra: imagine having to learn mercurial after having used git exclusively for ~10y
<rsalvaterra>
That's unfortunate…
<rsalvaterra>
… but imagine if you had to use CVS at work, like I do.
<stintel>
wow
<rsalvaterra>
CVS. In 2021. I bullshit you not.
<stintel>
I recall svn somewhere 10y ago and replacing that with git
JohnA has quit [Quit: Leaving]
rua has quit [Quit: Leaving.]
JohnA has joined #openwrt-devel
nitroshift has quit [Quit: Gone that way --->]
<Slimey>
heh
chadn has joined #openwrt-devel
<rsalvaterra>
blogic_: I just noticed there's an upstream ath79 USB PHY driver, while we're still carrying patches to add ours. Is this known?
<rsalvaterra>
The device tree bindings are different, though (phy vs usb-phy).
chadn has quit [Ping timeout: 480 seconds]
chadn has joined #openwrt-devel
rua has joined #openwrt-devel
guifipedro_ has joined #openwrt-devel
dangole has joined #openwrt-devel
* Slimey
wraps DonkeyHotei with a 42m long worm
danitool has quit [Quit: Cubum autem in duos cubos, aut quadratoquadratum in duos quadratoquadratos]
jbowen has joined #openwrt-devel
slh has quit [Remote host closed the connection]
slh_ has quit [Quit: gone]
wvdakker has quit []
wvdakker has joined #openwrt-devel
wvdakker has quit []
wvdakker has joined #openwrt-devel
slh has joined #openwrt-devel
slh64 has joined #openwrt-devel
rua has quit [Quit: Leaving.]
linuxinside has joined #openwrt-devel
<linuxinside>
OpenWRT-tober of code event early October near Hamburg... maybe share ideas discuss how to move some interesting things forward..
<linuxinside>
Just thinking...
<rsalvaterra>
And now to see if my TL-WDR3600 is happy with DSA…
<linuxinside>
How happy is he?
<linuxinside>
Of course online event mostly, refering to OpenWRT-tober of Code
<rsalvaterra>
Hm. Came up, gave me an IP address, so the switch is working fine, but now I can't ssh into it.
<linuxinside>
route -n ?
<linuxinside>
ip link ?
<linuxinside>
using dhcp from the box?
<rsalvaterra>
Not good. Huge amounts of packet loss while pinging.
<rsalvaterra>
Serial console time.
crz1 has quit [Quit: WeeChat 3.2]
<rsalvaterra>
I had forgotten the pain that is to open this case up…
<Slimey>
heh what device?
<Slimey>
oh
Tapper has joined #openwrt-devel
JohnA has left #openwrt-devel [Leaving]
rmilecki has quit [Quit: Konversation terminated!]
rmilecki has joined #openwrt-devel
<rsalvaterra>
Yeah, this is weird. Nothing obvious in the dmesg. The interfaces came up, but I just don't receive the ping replies. Tcpdump on the router shows the requests and replies on the correct switch port.
<rsalvaterra>
The computer, however, sees nothing.
<rsalvaterra>
mangix: You there?
<rmilecki>
rsalvaterra: this drivers/bcma/main.c change is so trivial / obvious I can't see how it may be wrong :)
<rsalvaterra>
rmilecki: That's what I thought, but hey, just making sure. ;)
<rsalvaterra>
Thanks! :)
<rmilecki>
np
<russell_>
rsalvaterra: i have been installing a headphone jack on my WDR3600's for serial console, only need to open once! ;-)
<rsalvaterra>
russell_: I need to do that. My thumb is bleeding. :P
<russell_>
(swapping the flash for 16MB at the same time)
* russell_
stabs the tp-link case designers
<rsalvaterra>
They probably weren't expecting people to open them up… :)
russell_ is now known as russell--
* russell--
usually starts above the ethernet ports with a stiff guitar pick, if he remembers correctly
<rmilecki>
i have holes cut in all my routers for 3 serial cables ;)
danitool has joined #openwrt-devel
<stintel[m]>
I'm simply not buying devices with console port anymore ;)
<slh>
but given how they intentionally cut the UART traces, it's obvious that they want to keep users out
<russell-->
electrons want to be free
<olmari>
nah.. electrons just want to get enough positivuty to be neutral ;)
<FLD>
how to bring my .config files from the openwrt-19.07 branch to the openwrt-21.02 branch with git?
<Slimey>
heh i have the same problem FLD
<russell-->
FLD: i'd start with a scripts/diffconfig.sh and see what's different from stock
<linuxinside>
Nice touch, wire wrap, old school !
<Slimey>
hehe
<mangix>
rsalvaterra: ola
<rsalvaterra>
mangix: Olá outra vez! :)
<rsalvaterra>
mangix: So, the WDR3600 didn't like DSA at all. :P
<mangix>
Which tells me nothing :). No traffic?
jbowen has quit [Ping timeout: 480 seconds]
jbowen has joined #openwrt-devel
<aparcar[m]>
rsalvaterra: so based on the comment the issue was that lzma decrompression wasn't enabled in kernel
<mangix>
rsalvaterra: anyway. Only thing I can think of is changing phy0 to use sgmii instead of rgmii. (dmesg will tell you what it is)
<mangix>
There's also the issue of POWER_ON_STRAP. You'll have to ask Ansuel about that.
<aparcar[m]>
neoraider: ping
PaulFertser has quit [Ping timeout: 480 seconds]
minimal has quit []
PaulFertser has joined #openwrt-devel
jbowen has quit [Quit: leaving]
Tapper has quit [Ping timeout: 480 seconds]
<will[m]>
how should i interpret `ethtool --pause eth1 rx off tx off` having a result of remaining on... but if i add autoneg off then it all turns off? for eth0 i can turn them on/off independently
<rsalvaterra>
It's RGMII. Everything looks fine from the dmesg, unfortunately… :/
Grommish has joined #openwrt-devel
<rsalvaterra>
mangix: The router gives my laptop a DHCP lease, without issues. If I ping it, the laptop receives about 1 in every 50 replies (with normal sub millisecond latency). Running tcpdump on the router, listening on the port connected to the laptop, I see all ping requests and the respective replies. But not on the laptop.
<linuxinside>
filtering or ARP cache not working?
<rsalvaterra>
aparcar[m]: Any idea why we have OpenWrt-specific lzma magic in the kernel? And why lzma instead of xz? I think I know the answer to this one, lzma predates xz. ;)
linuxinside has left #openwrt-devel [#openwrt-devel]
<rsalvaterra>
linuxinside: ARP seems to be working fine too… caught a couple of them while testing. https://paste.debian.net/1211922/
<aparcar[m]>
rsalvaterra: no clue, aren't you fan of zstd?
Acinonyx has quit [Ping timeout: 480 seconds]
macdav has joined #openwrt-devel
macdav has quit [Remote host closed the connection]
<mangix>
rsalvaterra: need Ansuel's help looks like
dansan has joined #openwrt-devel
Tapper has quit [Ping timeout: 480 seconds]
<mangix>
rsalvaterra: so my reading of the code is that qca,mac6_exchange needs to specified in DTS. But how is there a mac6 on this device?
<mangix>
I'll update my patch soon, but this makes no sense