mangix has quit [Remote host closed the connection]
mangix has joined #openwrt-devel
danitool has quit [Quit: Cubum autem in duos cubos, aut quadratoquadratum in duos quadratoquadratos]
fda has quit [Read error: Connection reset by peer]
fda has joined #openwrt-devel
fda has quit [Read error: Connection reset by peer]
fda has joined #openwrt-devel
Tapper has joined #openwrt-devel
Tapper has quit [Ping timeout: 480 seconds]
Tapper has joined #openwrt-devel
goliath has joined #openwrt-devel
Tapper has quit [Ping timeout: 480 seconds]
Tapper has joined #openwrt-devel
Tapper has quit [Ping timeout: 480 seconds]
Tapper has joined #openwrt-devel
Tapper has quit [Ping timeout: 480 seconds]
goliath has quit [Quit: SIGSEGV]
mrkiko has joined #openwrt-devel
Tapper has joined #openwrt-devel
Tapper has quit [Ping timeout: 480 seconds]
Tapper has joined #openwrt-devel
goliath has joined #openwrt-devel
Tapper has quit [Ping timeout: 480 seconds]
Tapper has joined #openwrt-devel
decke has joined #openwrt-devel
FPSUserename has joined #openwrt-devel
FPSUserename has quit [Remote host closed the connection]
FPSUserename has joined #openwrt-devel
rsalvaterra_ has joined #openwrt-devel
rsalvaterra has quit [Ping timeout: 480 seconds]
danitool has joined #openwrt-devel
figgyc has quit [Read error: Connection reset by peer]
figgyc has joined #openwrt-devel
rsalvaterra_ has quit []
rsalvaterra has joined #openwrt-devel
Namidairo has quit [Ping timeout: 480 seconds]
aleasto has joined #openwrt-devel
<rsalvaterra>
hauke: Ping. Quick question, how do you edit/refresh gcc patches? The wiki seems outdated/wrong in that regard.
<FPSUserename>
PaulFeetser do you know when you'll be looking into that dump? Just so I know when to come back online on IRC
<Habbie>
FPSUserename, you should really get a bouncer or a VM with an irc client :)
<Habbie>
you missed a whole conversation last night
<FPSUserename>
oh lol
<FPSUserename>
I know that KDE plasma has a build in IRC client, but I'm not using my laptop at the moment. I'll check if I can find a client for bindows
<FPSUserename>
Any highlights?
<Habbie>
hurricos also looked at the dump
<Habbie>
and figured the password might be double MD2 instead of MD5
<Habbie>
but nobody has cracked it yet
<Habbie>
23:24Z <hurricos> it looks like the whole function is all one, too, which makes it hard to get into u-boot by shorting SPI pins
<Habbie>
23:24Z <hurricos> because it will simply not be able to load the prompt
<Habbie>
23:24Z <hurricos> pretty securely written
<FPSUserename>
Would it be possible to do a complete overwrite with a uboot from another device, or would that cause issues because a uboot is device specific
<FPSUserename>
That would probably be done from jtag or directly written onto the nand
<Habbie>
tmn505, oh, neat
<Habbie>
FPSUserename, if you can overwrite like that, you can do it even as an experiment
<Tusker>
yeah, I think I would attack it by just writing the kernel and filesystem directly
<Tusker>
and leave uboot as is
<FPSUserename>
oh that's neat regarding the logging
<FPSUserename>
I'll check how matrix connects to this
<stintel>
oh, that's supposed to be advertised in the topic
<stintel>
looks like I don't have chanserv access, can't add it
<FPSUserename>
Doesn't look like matrix connects to this channel though. It's separated https://openwrt.org/contact
<Habbie>
matrix also bridges to irc
<stintel>
#_oftc_#openwrt-devel:matrix.org
stintel[m] has joined #openwrt-devel
<stintel>
see :)
<stintel>
not liking that ident though
<FPSUserename>
hmm
<FPSUserename>
No results for "#_oftc_#openwrt-devel:matrix.org"
<FPSUserename>
and it's the OFTC room
<Habbie>
stintel, 'what color?' 'stijn'
<stintel>
eh ?
<Habbie>
stijn-tint
<stintel>
ahhhhh
<stintel>
heh
<Tusker>
FPSUserename: if you set your tftp server to 192.168.11.2 (as per that mtdblock2) do you see any tftp traffic if you boot with the reset button pressed ?
<FPSUserename>
will check, but how do I connect to this irc chat through matrix (element). The channel stintel forwarded yields no results
<FPSUserename>
also I haven't setup a tftp server yet
<Tusker>
or even just run wireshark and see whether you see any traffic coming from 192.168.11.1 during boot
<FPSUserename>
but why would 11.1 show any traffic? The whole network connection is on 2.x, right?
<Habbie>
after boot, yes
<Habbie>
the uboot settings refer to 11
<Tusker>
in mtdblock2 it shows the recovery boot file linux.trx-recovery and mentions those IP addresses, which is probably the recovery mode default server IP
<FPSUserename>
ah okay. Will check wireshark
<Tusker>
or easier to just filter based on mac address
<Tusker>
in case the IP is different on your uboot
<Habbie>
like on the MCU in the new IKEA air quality sensor
<fpsusername[m]>
Doubt it would work, since the chip is a lot wider
<Habbie>
fpsusername[m], right! a set like that arrived here last week
<Habbie>
well, it depends on whether you need those pins
<fpsusername[m]>
<Habbie "well, it depends on whether you "> Exactly this. I didn't look into the pinout, but if you need a connection on pin 1 and 8 for example, this clip won't work
<Habbie>
it's 8 and 15
<Habbie>
and a few more
<Habbie>
so, no
<fpsusername[m]>
Rip, time to buy a clip and see y'all in 4 weeks haha
<Habbie>
by then we'll have bricked five devices from a root shell in our impatience ;)
<Habbie>
i wonder if the boot loader is signed
<Habbie>
i wonder if the mt7621 even has such a concept
<fpsusername[m]>
I'll keep an eye out on this chat
<fpsusername[m]>
I mean, if there's no way around it besides using a clip to get past the boot security
<fpsusername[m]>
And if the root shell access exploit won't be a viable choice, because of vulnerabilities
Rentong has joined #openwrt-devel
<fda>
dangole: i tested with frequencies @1.0V. 125 and 300 are working. if i enable in dts only 125 / only 300 / 125 + 300 there are about 30transitions.
<fda>
dangole: with 125+300 enabled, but set minfreq to 300 there are < 1 transition and the smallest temperature
Rentong has quit [Remote host closed the connection]
<rmilecki>
Press the [1], [2], [3] or [4] key and hit [enter] to select the debug level
<rmilecki>
rm: can't remove '/tmp/tmp.hHKjHE': No such file or directory
<rmilecki>
ah, it's a result of mktemp that may be used for locking
Rentong has joined #openwrt-devel
<fpsusername[m]>
Bought the clip and some jumper wires, since I didn't have enough female to female wires
<dangole>
fda: that's very good news. do you feel like posting patches making the changes in dts to linux-dts? and then also submit that mbox as patch in target/linux/mediatek/patches-5.10 to OpenWrt via openwrt-devel?
Rentong has quit [Remote host closed the connection]
Rentong has joined #openwrt-devel
Rentong has quit [Remote host closed the connection]
<fda>
dangole: if you want to commit it quickly, you could use the refreshed patches on my server i send you. see the one with 8450 in name. im currently still on fixing some random things
<fda>
dangole: you wrote about SDK: this is not exactly what im searching for :( i just want to have hosttools and toolchain in an archive to not have build them after i ran distclean. when i know openwrt better, i try to pack the correct paths
<fda>
i have a problem with ipv6 (i have only ULA here): odhcpd gives ipv6 adresses for every prefix ("hint") to every (v)lan
<fda>
so there are multiple ipv6 on every client. config: ula_prefix='fda1::/48' - ip6assign='64' - ip6ifaceid='::1' - ip6hint='<depends on vlan>'
<fda>
dnsmasq gives ipv4s correctly depending by (v)lan
eigma has joined #openwrt-devel
fda has quit [Remote host closed the connection]
fda has joined #openwrt-devel
<fda>
on the router the ipv6 are assigned correctly by interface
fda has quit [Read error: Connection reset by peer]
fda has joined #openwrt-devel
<fda>
or: when i disable all ipv6-dhcp settings, except for 1 interface, all interfaces get wrong ipv6s from this interface
<rsalvaterra>
Guys, can anyone tell me how to refresh gcc patches? I've tried make toolchain/gcc/{clean,refresh}, as specified in the wiki, but it fails. (I also tried gcc/{minimal,initial,final}/{clean,refresh}, which I don't even know if it makes sense, with the same result.)
danitool has joined #openwrt-devel
<fda>
maybe its better to use dnsmasq for dhco of ipv6 - it works by interface
<fda>
rsalvaterra: if the patches dont apply, you have not changed them properly. there should be an error message which tells you more
<rsalvaterra>
fda: They do apply, of course. I just don't know if they're fuzzy.
<fda>
rsalvaterra: a bump without the patches does not make sense - as this is the most work if they dont apply
<rsalvaterra>
fda: What? Have you looked at the commit? The patches are there, they apply, but they haven't been refreshed.
<fda>
yep
<fda>
rsalvaterra: why "Run-tested (built images):" if you cant apply the patches and so are not abe to build gcc
<rsalvaterra>
fda: Did I wrote that I couldn't apply the patches?
<fda>
rsalvaterra: " tell me how to refresh gcc patches"
<rsalvaterra>
refresh != apply
<fda>
dont know, check the output make
<rsalvaterra>
If I couldn't apply the patches, I couldn't be running this…
<fda>
maybe prepare is the correct target
<rsalvaterra>
[ 0.000000] Linux version 5.10.54 (rui@crystalwell) (arm-openwrt-linux-muslgnueabi-gcc (OpenWrt GCC 11.2.0 r17220+52-ae1c5d0d6a) 11.2.0, GNU ld (GNU Binutils) 2.37) #0 SMP Wed Jul 28 14:56:21 2021
<rsalvaterra>
… could I?
Tapper has quit [Ping timeout: 480 seconds]
<rsalvaterra>
fda: That's the thing, I tried all possibilities I know of, to know avail… I wouldn't be asking for help othewise. :)
Tapper has joined #openwrt-devel
ServerStatsDiscoverertraveler4 has joined #openwrt-devel
MatrixTravelerbot[m] has joined #openwrt-devel
<fda>
what is "refresh"?
<rsalvaterra>
That's Quilt terminology. Patches are applied in chunks, each chunk as an associated context.
<rsalvaterra>
When the context changes position (for example, then code was added before the patch context), Quilt is still "smart" enough to find it and apply it correctly (most of the time).
<rsalvaterra>
In this situation, we say the patch contains "fuzz". It's still appliable, but it's relying on smarts. :)
<rsalvaterra>
Refreshing is the procedure by which the patches are updated to match the current context (eliminating the said fuzz).
<rsalvaterra>
This is the way I understand it (haven't digged too deep in the Quilt documentation), so please correct me if I'm wrong, veteran devs. :)
<stintel>
grid: is there in your pile of docs something that can help me translate the RCW value in the link above?
<grid>
you can import it into the QCVS or whatever it's called eclipse extension
<stintel>
grid: also, it seems that the OEM u-boot is actually uploading the FMan microcode, so this should probably not be the reason why I couldn't get network to work
<rsalvaterra>
stintel: Are you writing the device tree by trial and error? :P
<stintel>
kind of
<stintel>
ow
<stintel>
I have a ubuntu 14.04 still around
<stintel>
party
<grid>
oxygen.3 march 2018 is the newest eclipse that'll work with the plugin, it needs to be even older to install it (eclipse-SDK-4.2.2-linux-gtk-x86_64.tar.gz i think)
<rsalvaterra>
## Flattened Device Tree blob at 00fc0000
<stintel>
rsalvaterra: well I started by decompiling the OEM DTB to DTS
<rsalvaterra>
Right, that's where I'd start. :)
<stintel>
and combining that with what's in the upstream kernel, as some T2081 is supported in vanilla kernel
<rsalvaterra>
stintel: I believe adding it as a mpc85xx subtarget isn't the way to go, though… that's ppc64, we probably should create a new target, no?
<stintel>
I'm not building anything u-boot related though
<grid>
k
<stintel>
hmm maybe I need to read about that pbl thing
rsalvaterra_ has joined #openwrt-devel
rsalvaterra has quit [Ping timeout: 480 seconds]
<stintel>
ERROR: please fix package/feeds/packages/apparmor/Makefile - see logs/package/feeds/packages/apparmor/dump.txt for details
<stintel>
hmmm
<stintel>
seems to trip over include ../../../packages/lang/python/python3-package.mk
<Slimey>
hmm i have a dlink dgs-1100-08P with 25q16jvsiq flashrom and rtl8370n that looks interesting and it has a serial header ;)
<Slimey>
grid what is that pic of
<grid>
Slimey: the imgur? NXP QCVS eclipse plugin
<stintel>
bah can't install it
<stintel>
complains about touchpoint missing
<Slimey>
i have some P1010 and T1023E that have a custom OEM uboot im attempting to get working with OpenWrt
<stintel>
ah wait
<Slimey>
based wifi aps that is
<grid>
i don't remember if you also need com.freescale.eclipse.3.7.updater.custom.updatesite.zip
<stintel>
had to install things one by one
minimal has quit []
<fda>
@rsalvaterra_ i know this as "autofix"
<rsalvaterra_>
fda: Hm… never heard of it, to be honest. :/
<fda>
but i dont know if this the "official" name
<hauke>
I would like to see this realtek change in openwrt 21.02 branch: https://git.openwrt.org/ad712c71cea3b26d8a7c53200c742181d1d040e7, but it conflicts. It would be nice if someone of the realtek switch experts could prepare and test a branch for the openwrt 21.02 relase
<hauke>
I am also fine with backporting bigegr changes there
<fda>
has the openwrt toolchain no tool to do this automatically?
<stintel>
grid: after importing. changes are hilighted how?
aleasto has quit [Remote host closed the connection]
<grid>
it may be loading a hard-coded RCW. depends on the strapping pins.
<PaulFertser>
fda: why freetz people didn't use OpenWrt?
<fda>
i do ^^
<fda>
PaulFertser: drivers for many things, like dsl, dect, pots (internal + external) dont work with openwrt
<fda>
as they are closed source
<fda>
and so the kernel version could even not be changed
<PaulFertser>
fda: ah, of course. So it's like a vendor firmware "mod" rather than real OS, I see.
<PaulFertser>
fda: still better than vendor version I guess, makes sense.
<fda>
yes, freetz = avm + modification
<stintel>
grid: ok I see now, default has BOOT_LOC PCIe1
<fda>
with freetz you could remove some avm things, some to change and build additional packages
<stintel>
aha now I see yellow after closing project, not saving, opening and importing again
<stintel>
ah once I expand other things the yellow disappears
aleksander has joined #openwrt-devel
<fda>
i removed CONFIG_PACKAGE_odhcpd-ipv6only + CONFIG_PACKAGE_dnsmasq and replaced with CONFIG_PACKAGE_dnsmasq-dhcpv6 and now ipv6 are assigned by interface correctly
<fda>
the interesting part is that there are 0 visible changes in luci
<stintel>
grid: seeing quite some differences actually
<stintel>
interesting stuff, thanks a lot for pointing me to that QCVS
Borromini has left #openwrt-devel [#openwrt-devel]