<xback>
hauke: Hi, looks like the gdb 15.x bump broke arm32 debugging (ipq40xx). I'll try to figure out some details
<xback>
hauke: gdb 14 broke mips debugging in the past, hence I had a local patch that let gdb stay at 13.x which seems to work on all used arches here
rua has quit [Quit: Leaving.]
robimarko has joined #openwrt-devel
torv has quit [Remote host closed the connection]
torv has joined #openwrt-devel
robimarko_ has joined #openwrt-devel
robimarko__ has joined #openwrt-devel
swegener has quit [Read error: Connection reset by peer]
swegener has joined #openwrt-devel
robimarko__ has quit [Read error: Connection reset by peer]
<xback>
hauke: I just locally reverted the gdb 15.x bump and I can debug again with gdb
Tapper has quit [Ping timeout: 480 seconds]
n3ph has joined #openwrt-devel
n3ph has quit [Quit: WeeChat 4.4.2]
n3ph has joined #openwrt-devel
Tapper has joined #openwrt-devel
<f00b4r0>
so I have this random problem with dnsmasq that stops replying to dns queries after PPP redial. Rings a bell to anyone?
<f00b4r0>
it seems it never recovers from accumulated queries during the redial period: "Maximum number of concurrent DNS queries reached (max: 150)"
n3ph has quit [Ping timeout: 480 seconds]
n3ph has joined #openwrt-devel
<nbd>
i'm so looking forward to replacing dnsmasq with something sane at some point in the future
n3ph has quit [Ping timeout: 480 seconds]
n3ph has joined #openwrt-devel
gromero has joined #openwrt-devel
<f00b4r0>
so apparently it's not a bug, it's a "feature", if I read this comment correctly: https://thekelleys.org.uk/gitweb/?p=dnsmasq.git;a=blob;f=src/forward.c;h=32f37e4054baae04d1de035e1497c134386aceb2;hb=b6769234bca9b0eabfe4768832b88d2cdb187092#l246
<dpawlik>
@nbd "looking forward to replacing dnsmasq with something sane at some point in the future" => Knot Resolver - seems to be interesting - never checked.
Daanct12 has joined #openwrt-devel
n3ph_ has joined #openwrt-devel
n3ph has quit [Ping timeout: 480 seconds]
<nbd>
dpawlik: interesting, thanks for the pointer
Stat_headcrabbed has joined #openwrt-devel
Daanct12 has quit [Quit: WeeChat 4.4.2]
<schmars[m]>
keeping the split dns feature would be important :-) i.e. the ability to forward requests for certain domains to a separate resolver
gromero_ has joined #openwrt-devel
gromero__ has joined #openwrt-devel
<f00b4r0>
never heard of Knot, but if it's good enough for Cloudflare it's a good start ;)
gromero has quit [Ping timeout: 480 seconds]
rua has joined #openwrt-devel
gromero_ has quit [Ping timeout: 480 seconds]
<\x>
iirc knot/kresd is the default for turris devices
minimal has joined #openwrt-devel
<Habbie>
f00b4r0, the fine people that write Knot Resolver find it very important that you know that what Cloudflare runs no longer resembles it
fossdd has quit [Remote host closed the connection]
fossdd has joined #openwrt-devel
Lynx- has joined #openwrt-devel
Lynx- is now known as Guest6455
Lynx- has joined #openwrt-devel
Guest6455 has quit [Ping timeout: 480 seconds]
<\x>
is it just me or ftm responder doesnt really work on ath10k_pci ? works well on ath1*k_ahb (ipq4019 / ipq6018)
Nuckles has joined #openwrt-devel
Nuckles has quit []
asmodehn has joined #openwrt-devel
Lynx- has quit [Read error: Connection reset by peer]
Lynx- has joined #openwrt-devel
plappermaul has joined #openwrt-devel
Svanto24 has joined #openwrt-devel
<Svanto24>
Hi all
<Svanto24>
I'd been thinking... I've read somewhere on the forum that you can get devices running even if they have secure boot enabled by changing the bootloader
<Svanto24>
does that work even if they have the fuses in qfprom fused or no?
kenny has quit [Quit: WeeChat 4.4.1]
<robimarko>
No, if you flash an unsigned bootloader you will brick it
<robimarko>
If you can acess the bootloader CLI now and they havent removed any usefull commands you can boot unsigned images anyway
<plappermaul>
I'm trying desperately to add mtd/nand/ubifs tools to my initramfs build. Seems as if NAND_SUPPORT config can do the trick. But how to add it to my target?
<dpawlik>
Habbie could you send url for the PR to follow? Do you have any guide how to deploy? I don't see any guide so far on wiki (https://openwrt.org/docs/guide-user/services/dns/start) but for sure pdns might be interesting to test :)
<Svanto24>
robimarko: Oh wow, that's really ass... Is that a guarantee or a risk? I got 4 of these, I can probably sacrifice one to find out.
<Svanto24>
They have removed some commands. There is a system update command. I'm about to try flash an openwrt image on it to see if I can get something to come up, even with a flawed device tree
<Svanto24>
there is an update uboot command however
<Habbie>
dpawlik, i'll try to remember. i realised today we have zero documentation on that
<schmars[m]>
system update sounds much like signature check
<schmars[m]>
did you get a GPL code dump? sometimes helpful vendor developers include the signing keys
<Svanto24>
I could not find any. I tried binwalking it but I'm not nearly good enough to do so successfully so all I got were corrupted files.
<Svanto24>
are signing keys baked into the SoC?
<schmars[m]>
in the firmware, so they can be used to check signatures
<robimarko>
Svanto24: Look, the public key hash is fuses into the SoC
<robimarko>
So, if you flash an unsigned or signed with incorrect keys bootloader, it will refuse to run it
<robimarko>
And then you can only recover by flashing the original backup to the boot media
<Svanto24>
robimarko: understood, thank you. Yeah in that case, it would be really nice if I could get my hands on a GPL dump.
<Svanto24>
schmars[m]: do you know how I might get a code dump? I feel like on enterprise gear they're probably only gonna give them to customers which I mean, I do have a customer account, but I'm not a "huawei partner" or whatever
<Svanto24>
*GPL code dump
<schmars[m]>
ah, huawei, yeah then that's difficult
<dpawlik>
Habbie: indeed and I suggest to add - even scratch would be helpful. I'm using dnsmasq + dnscrypt-proxy-2, but if pdns can provide same with smaller size -> I will try. Got few device with small storage. Do I understand correct that pdns != pdns-recursor?
<Habbie>
dpawlik, pdns, pdns-recursor, dnsdist are 3 separate packages
<Habbie>
dpawlik, in this case, dnsdist is the one, but the size likely won't be smaller
<schmars[m]>
otherwise write email to their legal dpt with device model and serial number, and that you'd like code under the GPL
<Svanto24>
schmars{m]: yess... not only huawei but also enterprise hardware so it's not on there
<Svanto24>
What should I ask for? "all source code pertaining to the Huawei AP7050DE under GPL and other open source licenses"?
<dpawlik>
Habbie dnscrypt-proxy2 has 4 MB, if pdns combo would be smaller -> I will try :) Or even no matter of the size, I will take a look when I have few min. Thanks (some wiki guide would be welcome!)
<schmars[m]>
yeah something like that
<Habbie>
dpawlik, understood :)
<Svanto24>
schmars[m]: if they do send me keys, do those only apply to a single serial number?
<Habbie>
GPL sources tend to not include keys, i believe
<schmars[m]>
yeah very rarely do they accidentally get included
<Habbie>
schmars[m], right, lol
<Svanto24>
right-o, no custom uboot for me then hah
<Svanto24>
I'm gonna give foss@huawei.com a shot
<Svanto24>
There is a possibility for envvars though, so if you can think of anything that might work, please do let me know.
* f00b4r0
wishes there was a way to tell udhcpc to refuse private IPs and retry until it gets a public one
<Habbie>
f00b4r0, what fresh hell have you gotten yourself into? :)
<f00b4r0>
apparently some LTE sticks will "provide" a private IP to a DHCP request when the LTE link hasn't been established
<Habbie>
ah
<Habbie>
fun
<f00b4r0>
not.
<f00b4r0>
;P
<Habbie>
yeah
<f00b4r0>
and it seems the lease time is also $long. So udhcpc "never" retries and gets stuck on this useless IP with no connectivity
<Habbie>
hmm. how about reducing lease time? can udhcpc do that?
<f00b4r0>
doesn't look like it
Lynx- has quit [Quit: Going offline, see ya! (www.adiirc.com)]
bookworm has quit []
bookworm has joined #openwrt-devel
ssterlingg has joined #openwrt-devel
ssterling has quit [Ping timeout: 480 seconds]
Lynx- has joined #openwrt-devel
Stat_headcrabbed has quit [Quit: Stat_headcrabbed]
<Svanto24>
I think I bricked it lmao. I put a boot argument and now I'm in a boot loop.
<Svanto24>
Saving POST results: Done POST test: End Image: Current Bootup is B Press f or F to stop Auto-Boot in 3 seconds: 0 get_emt_file_from_flash fail! eth0 Waiting for PHY auto negotiation to complete... TIMEOUT ! Auto-neg error, defaulting to 10BT/HD Using eth0 device ARP Retry count exceeded; starting again eth0 Waiting for PHY auto negotiation to complete... TIMEOUT ! Auto-neg error, defaulting to 10BT/HD Using eth0 device ARP Retry count exc
<dwfreed>
sounds like it's attempting to auto-recover by pulling a file from tftp
<tersono>
I have a package with two build variants, and I want to have a different PKG_BUILD_DEPENDS dependency for each. The documentation says "if you want to build bar if package FOO is selected, use PACKAGE_FOO:bar", so I tried to use "PACKAGE_foo:dep1 PACKAGE_foo-alt:dep2", but this ends up forcing both dep1 and dep2 to be built even when only one variant of my package is enabled. Is there a way to get the dependencies to only be built if neede
n3ph has quit [Ping timeout: 480 seconds]
cmonroe has joined #openwrt-devel
<tersono>
Hm maybe my problem is that the same dependencies are already being picked up as unconditional build deps from DEPENDS
<Svanto24>
I'm going to bed. Maybe I can fix it tomorrow. Or not, but then I'd at least have an excuse to get a nand flasher
Tusker has joined #openwrt-devel
Svanto24 has quit [Quit: Page closed]
<Tusker>
heya guys, trying to boot from uboot on a t30-w (powerpc p1011e) - and the default kernel fails to uncompress. How can I calculate what load address should I use ? "GUNZIP: uncompress, out-of-mem or overwrite error - must RESET board to recover"
<Tusker>
so, probably the 6.6 kernel is too big for the reserved space
felix has quit [Read error: Connection reset by peer]
felix has joined #openwrt-devel
skynet2 has quit []
skynet2 has joined #openwrt-devel
<tersono>
Looks like this could be fixed by having package-metadata.pl emit the Make dependencies guarded by an $(if $(CONFIG_PACKAGE_%s),...) based on the package itself actually being selected (at least for cases where a Makefile contains more than one package)
slh has quit [Remote host closed the connection]
slh64 has quit [Quit: gone]
slh has joined #openwrt-devel
Tapper has quit [Ping timeout: 480 seconds]
tersono has quit [Ping timeout: 480 seconds]
<Tusker>
is there a loader that I could use that can work with a kernel on ext2 ?
tersono has joined #openwrt-devel
<Tusker>
I think the main issue is uboot is limited to about 8m of malloc() - "Reserving 8200k for malloc() at: 3f72e000"