nlowe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
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]
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
schwicht has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
user95 has quit [Ping timeout: 480 seconds]
mzvd has quit [Read error: Connection reset by peer]
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]
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
<slh>
stintel: ipq807x doesn't /need/ to have high idle power figures, e.g. the xiaomi ax3600 (ipq8071a) is extremely modest at ~6 watts idle - but ipq8074a with quad-ax-radio and 6 GHz can get into the 25 watts/ idle range
danitool has quit [Quit: Cubum autem in duos cubos, aut quadratoquadratum in duos quadratoquadratos]
Ansuel has quit [Ping timeout: 480 seconds]
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]
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
Ansuel has joined #openwrt-devel
Ansuel has quit [Ping timeout: 480 seconds]
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
schwicht has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
ekathva 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]
mzvd has joined #openwrt-devel
user95 has joined #openwrt-devel
valku has quit [Quit: valku]
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
mzvd has quit [Remote host closed the connection]
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]
arinc9 has joined #openwrt-devel
arinc9_ has joined #openwrt-devel
<stintel>
slh: thanks for the clarification :)
mzvd has joined #openwrt-devel
f5- has quit [Quit: Leaving]
mzvd has quit [Read error: Connection reset by peer]
<nick[m]1234>
I hopefully fixed all packages that were not compiling with iptables 1.8.8 (rtpengine). Can we now merge iptables 1.8.8 https://github.com/openwrt/openwrt/pull/9886
mzvd has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
<Harm_>
Habbie: remember the black MT7621 device of which you could not extract the u-boot password? I looked into the u-boot dump, the hash is just a plain md5 hash (not md2!). I changed a few bytes to skip the password check, I can now get into the u-boot shell :D
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
frwol has joined #openwrt-devel
<frwol>
hi
mzvd has quit [Read error: Connection reset by peer]
<Harm_>
No, this patches fhe function that asks for your password to a 'return 0', so it skips that whole part
<Habbie>
i mean, no checksum over the entire uboot to deal with
<Habbie>
also, are you fpsusername[m]? or another person? :)
<Harm_>
Ah, no checksum apparently. This just booted
<Habbie>
neat
<Harm_>
I'm someone else
<Habbie>
cool
<Habbie>
how did you get write access?
<Harm_>
I lifted some pins of the flash chip and connected them to my raspberry pi to flash
<Harm_>
I tried brute-forcing the md5 hash but my poor GTX980 could not find it after 1.5 days :(
<Habbie>
cool :) i still haven't gotten around to that trick
<Habbie>
however, i have root access once booted
<Habbie>
it's just that i feared bricking uboot if i patched it
<Habbie>
so, nice work :)
<Harm_>
Yeah, it requires a bit of soldering experience :). With root access you should be able to flash. Next step for me is to gain root in Linux. Shouldn't be hard to patch the squashfs
<Habbie>
but clearly you'll manage without that :)
<Harm_>
Yeah I've seen that. My firmware is newer so I couldn't use that
<Habbie>
ack
<Harm_>
nice work on that btw :D
<Habbie>
and also not the other exploits from Tenable I presume?
<Habbie>
thanks!
<Habbie>
btw, now that uboot is unprotected, do you have TFTP? or at least read/write via the uboot shell?
<dwfreed>
Harm_: an 8 character password could take up to a week to brute force on a 980
<dwfreed>
based on this random benchmark I found online
<Habbie>
and it might be longer than 8
<Harm_>
Habbie: I've not checked for other exploits so far (apart from concluding I could not overflow some data by entering a very long password in u-boot)
<Habbie>
hehe
<Harm_>
dwfreed: iirc I stopped at the point where it was trying 10 character passwords and HashCat estimated 1-2 weeks for that length so it was a bit quicker than that
<aiyion>
Good morning. Regarding LAN/WAN LEDs at the front, is there already a syntax to support "ROUTER" LEDs?
<aiyion>
Either as in 'blink if there's traffic on lan or wan' or 'blink if there's traffic at all being routed'?
<dwfreed>
you could use a iptables led target to accomplish that, but you'd have to do much of that manually
<dwfreed>
s/a/an/
<dwfreed>
and it would break in software and hardware flow offloading scenarios
<dwfreed>
(more accurately, it would only identify the packet that creates the flow, not additional packets part of the flow)
<aiyion>
would ucidef_set_led_netdev be extensible to support two interfaces instead of one?
<dwfreed>
I don't think netdev in general supports more than one interface
<aiyion>
it was md2, not md5 the last time around iirc
<Habbie>
aiyion, somebody suspected md2 based on some of the code in the uboot block i rcall
<Habbie>
*recall
<Habbie>
it didn't sound conclusive
<stintel>
so just plain md5 ?
<Habbie>
that is what Harm_ says
danitool has joined #openwrt-devel
<Habbie>
i don't know
<aiyion>
just wanted to leave, what I remembered, to eventually spark something.
<stintel>
I'll let it run for a bit :)
<Habbie>
aiyion, it is appreciated :)
<Habbie>
ah, hurricos spotted the s-table for md2, in that log
<Harm_>
There is an md2 sbox in the code but I've emulated the password check function and it is not touched at all
<Habbie>
right :)
<aiyion>
nice :)
<Harm_>
I created some breakpoints in the code and dumped the output of the hash it generated, when I entered a well known password (i.e. password) and searched for the hash online i found that it was an md5 hash
Marcin_ has quit [Remote host closed the connection]
mzvd has quit [Read error: Connection reset by peer]
mzvd has joined #openwrt-devel
nlowe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
mzvd has quit [Remote host closed the connection]
nlowe has joined #openwrt-devel
mzvd has joined #openwrt-devel
nlowe has quit []
nlowe has joined #openwrt-devel
mzvd has quit [Ping timeout: 480 seconds]
ekathva has quit [Ping timeout: 480 seconds]
mzvd has joined #openwrt-devel
nlowe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
nlowe has joined #openwrt-devel
mzvd has quit [Read error: Connection reset by peer]
GraphicHealer has joined #openwrt-devel
<GraphicHealer>
Does anyone here know where I can get openwrt for the Orbi routers?
<GraphicHealer>
RBR50v2
<GraphicHealer>
And RBS50v2
nlowe has quit [Read error: Connection reset by peer]
<GraphicHealer>
Which are the same hardware, but the RBS50v2 is the remote point with the wan port set as lan in software
<GraphicHealer>
They are both netgear
Ansuel has joined #openwrt-devel
<GraphicHealer>
And they are on Atheros chips, which are some of the best for OpenWRT
<GraphicHealer>
I looked it up and there are at least 10 other routers with openWRT that have the exact same processors.
<GraphicHealer>
And yet these Orbi V2 routers dont
<PaulFertser>
GraphicHealer: if OpenWrt already supports boards with the same SoC it's possible that you only need to write a suitable DTS file to describe your target. You'll want to have serial console to debug it.
<Ansuel>
jow learnd that the bad way... but i wonder if a good solution would be ""unify"" a configuration from stage1 and stage2. This problem wasn't present with local compilation and presented only with CONFIG_ALL. Wonder if we can think of a way to select the package (and enable the config flag) but skip the compilation. But I guess all this would be a big workaround to problematic Makefile that should be written correctly in the
<Ansuel>
first place
<Ansuel>
(wonder if we have other case like that... i mean we notice this just because the package had an hash for abi... if it was simply libwolfssl we wouldn't even notice)
danitool has quit [Quit: Cubum autem in duos cubos, aut quadratoquadratum in duos quadratoquadratos]
mzvd has quit [Read error: Connection reset by peer]
ekathva has joined #openwrt-devel
ekathva has quit [Remote host closed the connection]
<Lechu>
Ansuel: I checked the register mapping between AR9331 and AR9344, and some registers were shifted around. Fields seem to have stayed in place, at least. Color me surprised.
<Ansuel>
to me ar9344 seems more similar to qca8k than ar9311
<Lechu>
Building now with a hack, let's see how it goes.
<Ansuel>
the tagging scheme looks identical
<Lechu>
Yeah, maybe it'll be easier to attach it to qca8k actually
frwol has quit [Quit: leaving]
<Lechu>
also ar9331 is in dire need of refactoring of all that regmap stuff with a ton of #defines
<Ansuel>
theory time ar9311 -> ar9344 ->qca8k (and ar9344 is in the middle as they were changing switch implementation
<Lechu>
I need to compare the datasheet against ar8327 then as well.
<Ansuel>
about regs i had this crazy theory of converting eerything to regfield implementation but that would slow down the driver considering it's used on low spec devices
<Lechu>
Also, I very much like the idea of modularizing existing drivers - obviously, for ability to convert individual devices
<Ansuel>
again problem is that you will quickly fall in making the driver slower
<Ansuel>
(and also unreadable making another qca ssdk that supports 7 different switch with regs all over the include)
<Lechu>
Yeah, probably :|
<Ansuel>
bu yhea feature wise the 3 driver are very similar... even in how the function but creating a unified driver produce major problems (perf regression, bigger size....)
<Lechu>
I also think that this should be done the other way around - separate drivers, sharing some common code.
<Ansuel>
that will be done for ipq40xx internal driver
<Ansuel>
problem is when the regs are different
<Lechu>
dereferencing a register map, even a const one is costly :|
<Ansuel>
that can be handled by regfield implementation but add tons of overhead to the driver (regfield implementation you declare the regmap and on the function side you reference it with a define)
<Lechu>
Ansuel: no traffic other than tagged egress from CPU, same as before. I'll go digging through regmap then, I guess.
<Lechu>
hanetzer: yeah
<Lechu>
hanetzer: that driver is using regmap, but with direct register addressing, with offsets in #defines, which is PITA when trying to support a second, similar device
<hanetzer>
gotcha, so its a bastard hybrid instead of pure define/readl/writel vs the regmap api
<Ansuel>
Lechu what i would first check are the 0x000 regs (the one that configure the gmac and the initial configuration)
<Ansuel>
i bet the regs are different
<Lechu>
Ansuel: for sure they are, I caught at least 3 different addresses
<Lechu>
haven't yet looked into the fields as much tho
<Ansuel>
your goal is to get the switch to send tagged packet.
valku has quit [Read error: Connection reset by peer]
<jow>
enyc: it would solve the problem, but would also shift a lot of responsibility to the server side (to "reassemble" the repos there)
<jow>
enyc: so people building openwrt themselves would either require additional infrastructure to create their own repos or we'd need to maintain different implementations (repo generation in buildroot and repo generation on the server from uploaded ipks)
<jow>
not sure if it makes sense
Ansuel has joined #openwrt-devel
Gaspare has joined #openwrt-devel
<mangix>
Ansuel: welcome back
arinc9 has quit [Remote host closed the connection]