fakuivan has quit [Remote host closed the connection]
fakuivan has joined #openwrt-devel
fakuivan has quit [Remote host closed the connection]
fakuivan has joined #openwrt-devel
fakuivan has quit [Remote host closed the connection]
fakuivan has joined #openwrt-devel
noltari has quit [Ping timeout: 480 seconds]
rua has quit [Quit: Leaving.]
fakuivan has quit [Remote host closed the connection]
fakuivan has joined #openwrt-devel
<Rondom>
dwfreed: thanks for investigating. I was suspecting that commit to be the culprit all the time, without ever spending the time to investigate :-/
<Rondom>
I think it does work for reading if the u-boot env is specified via NVMEM in device-tree though. There is no write-support for u-boot env via NVMEM, unfortunately. Otherwise we could specify everything in DTS instead of u-boot-envtools scripts
<enyc>
It seems more like something is going OOM bezerk
rua has quit [Quit: Leaving.]
rua has joined #openwrt-devel
<KanjiMonster>
dwfreed: Rondom: b89b19b ("base-files: execute uboot-env script before calling config_generate") attempts to populate fw_env.config at an earlier time (a few commits before the one using it)
<KanjiMonster>
dwfreed: Rondom: 7a120ee72e06217d87dfdff819052a84db32fd2b that is (I'll never get used to the two different buffers in linux ...)
<enyc>
f00b4r0: colo in #openwrt talking and noticing modprobe's running well after boot ...........
n3ph has joined #openwrt-devel
robimarko has joined #openwrt-devel
<enyc>
f00b4r0: now removed many kmod- packages not needed here and seeing ow that changes things, etc etc
rsalvaterra_ has joined #openwrt-devel
rsalvaterra_ is now known as rsalvaterra
rsalvaterra_ has joined #openwrt-devel
rsalvaterra_ is now known as rsalvaterra
rsalvaterra_ has joined #openwrt-devel
rsalvaterra_ is now known as rsalvaterra
n3ph_ has joined #openwrt-devel
n3ph has quit [Ping timeout: 480 seconds]
Znevna has quit [Quit: ZNC 1.8.1+deb1~bpo10+1 - https://znc.in]
Znevna has joined #openwrt-devel
gromero has joined #openwrt-devel
gromero has quit [Remote host closed the connection]
<dwfreed>
KanjiMonster: it clearly doesn't work though; when I boot a watchguard firebox m300, I get an error about fw_env.config missing in preinit, which is probably before overlay is even mounted
minimal has joined #openwrt-devel
<KanjiMonster>
dwfreed: I suspect as much as well, but the issue is that creating fw_env.config before config_generate doesn't seem to work, not that it isn't available in general (in theory at least)
<dwfreed>
ack
madwoota has quit [Server closed connection]
madwoota has joined #openwrt-devel
mcbridematt has quit [Server closed connection]
mcbridematt has joined #openwrt-devel
rsalvaterra has quit []
rsalvaterra has joined #openwrt-devel
frytaped has quit [Server closed connection]
frytaped has joined #openwrt-devel
xes has quit [Read error: Connection reset by peer]
<stintel>
robimarko: thanks for the RUST kernel config symbol fix
<robimarko>
:)
tohojo is now known as Guest15709
tohojo has joined #openwrt-devel
Guest15709 has quit [Ping timeout: 480 seconds]
<opty>
i don't use luci so it seemed that i don't need libustream-mbedtls but i didn't find much about the library, i also didn't find any other 'core' package depended although it belongs to default packages which makes me doubt, any idea?
<dwfreed>
opty: opkg needs it to be able to access package lists via https
<dwfreed>
it's also used by "wget" for https
<dwfreed>
because the "wget" in openwrt is using libustream as well
valku has joined #openwrt-devel
<opty>
dwfreed: thanks here too :)
goliath has quit [Quit: SIGSEGV]
sorinello has joined #openwrt-devel
n3ph_ has quit [Ping timeout: 480 seconds]
FiniteField has joined #openwrt-devel
FiniteField has quit []
<f00b4r0>
enyc: looks like the slab leak alright
n3ph has joined #openwrt-devel
bluew has joined #openwrt-devel
<enyc>
f00b4r0: So, what is needed tohelp with this, how will it get fixed?
<f00b4r0>
i don't know. If I knew how to fix it I would have already
<enyc>
f00b4r0: what, from my log, confirms slab-leak?
<enyc>
f00b4r0: and how does the funny business of modprobe processes stuck running relate to anything?
<opty>
enyc: btw, what module is it trying to load?
<f00b4r0>
that's nearly 140MB of slab unreclaimable memory
<f00b4r0>
(these are 4K pages figures)
<enyc>
f00b4r0: humm uerr aerr uerr in what way can this be kernel issue, and in what way user issues ?
<f00b4r0>
by definition it's a kernel problem since slab memory is handled by the kernel. But it can be that the kernel is being asked to do something stupid by e.g. hostapd.
<f00b4r0>
in my case, killing hostapd when this occurs makes slab go back to "normal" in a few seconds.
<KanjiMonster>
so ... valgrind on hostapd? ;p
<opty>
another btw, nice once: '[31737.313175] Out of memory: OOM victim 10822 (modprobe) is already exiting. Skip killing the task', probably never seen o_O
<f00b4r0>
KanjiMonster: it's probably something in a patch we're carrying. I somehow doubt nobody else would have noticed otherwise
* enyc
ROARs
<opty>
enyc: if you can 'cat /proc/`pidof modprobe`/cmdline | tr \\0 \ && echo'...
<f00b4r0>
I don't know what you're referring to regarding modprobe being stuck, but when the kernel is in that OOM state, anything can happen. I wouldn't waste time on that.
<enyc>
opty: booted with less kmod's and no rcpd/uhttpd started, just seeing what happens!
<opty>
enyc: you can also watch slab and restart hostapd in case of a rise to prevent oom
drussell_ has joined #openwrt-devel
<f00b4r0>
"wifi up" is apparently enough. At least in my experience
<opty>
even better
<opty>
because it does more than just hostapd restart
<enyc>
;O
minimal has quit [Quit: Leaving]
n3ph has quit [Ping timeout: 480 seconds]
matoro has quit [Remote host closed the connection]
matoro has joined #openwrt-devel
goliath has joined #openwrt-devel
drussell_ is now known as Guest15721
drussell_ has joined #openwrt-devel
Guest15721 has quit [Ping timeout: 480 seconds]
GATfly has joined #openwrt-devel
<enyc>
GATfly here asking after dev/support status for Gli-Net AX3000. "It's currently not supported but it does look like it is potentially being worked on?"
<enyc>
Encouraged them not to close out IRC and hopefully somebody may point them in right direction/information =).
<GATfly>
enyc: Thanks for the shout here too!
_whitelogger has joined #openwrt-devel
dryimpact4 has joined #openwrt-devel
dryimpact has quit [Read error: Connection reset by peer]