<plntyk>
replacing "which" include/prereq.mk with "env which" might work
JohnA has quit [Remote host closed the connection]
Tusker has joined #openwrt-devel
<mangix>
i think using command -v needs to be reinvestigated
rmilecki has joined #openwrt-devel
jlsalvador has quit [Quit: jlsalvador]
jlsalvador has joined #openwrt-devel
dedeckeh has joined #openwrt-devel
nitroshift has joined #openwrt-devel
valku has quit [Quit: valku]
<jow>
nick[m]4: no, not really.
<nick[m]4>
jow: k. Thanks.
<jow>
rmilecki: never seen an "igmp_v3" option
tapper has joined #openwrt-devel
<jow>
rmilecki: and a grep -r in the netifd sources reveals no hits either, I would omit this option fro mthe migration, it appears to be something proprietary
<rsalvaterra>
Hm… they have a complete AP implementation?
<rsalvaterra>
I thought hostapd was the only game in town.
<mangix>
They have one. I don't know if open source yet
feriman has quit [Quit: WeeChat 3.1]
<rsalvaterra>
Ok, now I see the motivation for the mac80211.sh change… https://git.openwrt.org/?p=project/iwinfo.git;a=commitdiff;h=dd6d6d2dec3515e26847e6ff8e8950d71745d560
<barhom>
Hey, I got a routing dilemma that I can't seem to be able to solve. I got two routers that are MESH'd and bridged to LAN. Only One router has DHCP server running and that router has two uplinks. On the main router (the one doing the routing) I use physdev to see if packets come from SSID1 or SSID2 to route packets differently.
<barhom>
The issue is that on the mesh'ed router I do not find a way to route packets differently. I can't differentiate packets coming into the main router (i.e. to know which SSID it used on the mesh'ed sub-router)
_lore_ has quit [Read error: Connection reset by peer]
_lore_ has joined #openwrt-devel
feriman has joined #openwrt-devel
rmilecki has quit [Ping timeout: 480 seconds]
<russell-->
is there a convenient way of telling an mr24 not to send pause frames? "[18668.943990] eth0: link is up, 1000 FDX, pause enabled", apparently coming from drivers/net/ethernet/ibm/emac/core.c
<rsalvaterra>
russell--: Not through UCI yet. It's a huge pet-peeve of mine, not being able to enable/disable pause frames in UCI, especially now that we have a nice L2 configuration system.
goliath has joined #openwrt-devel
<rsalvaterra>
It needs to be implemented in netifd first.
<russell-->
is there a sysfs knob?
<rsalvaterra>
russell--: Not that I know of. You have to use ethtool.
ecloud has quit [Ping timeout: 480 seconds]
ecloud has joined #openwrt-devel
feriman has quit [Quit: WeeChat 3.1]
rejoicetreat has quit [Ping timeout: 480 seconds]
tapper has quit [Ping timeout: 480 seconds]
<barhom>
jow, you're always helpful with openwrt-style-firewall rules. Is this supported in openwrt-style-configs, "iptables -t mangle -A FORWARD -i br-lan -m physdev --physdev-in wlan0-1 --physdev-out mesh -j DSCP --set-dscp 63" ?
<Larhzu>
Hello! There seems to be a tiny bug in 21.02.0-rc2 in /lib/netifd/hostapd.sh: hostapd_common_add_bss_config() lacks "config_add_int bss_load_update_period chan_util_avg_period" and thus setting them via uci doesn't work.
dorf_ has quit [Remote host closed the connection]
dorf_ has joined #openwrt-devel
<dorf_>
when no updates are found in LuCI, might be better to display "No updates available" instead of "No information available".
<PaulFertser>
djStolen: are you sure something went wrong when you were overwriting the bootloader?
<PaulFertser>
djStolen: btw, another option is to program SPI NOR directly.
<PaulFertser>
djStolen: for JTAG I've seen a patch sent to the OpenOCD review system for many mtk SoCs, authored by the fellow who wrote that breed bootloader.
<djStolen>
PaulFertser: it doesnt respond to anything. neither when i try to boot it with reset button pressed, nor normal. so i assume it got corrupted.
<PaulFertser>
djStolen: and never prints anything on its own?
<djStolen>
nope
<PaulFertser>
djStolen: "Serial port works in one-way only (no input going through pin4)." what does this mean? They mark pin 4 as Tx and say no input goes there and provide boot logs?
<djStolen>
what makes me doubtful is that in the comment in the page i sent above it says that tha
<djStolen>
exactly
<PaulFertser>
djStolen: are you sure pin 4 is not Rx?
<PaulFertser>
Because that would explain it.
<djStolen>
actually, i asked that myself, and swapped the pins .-)
<djStolen>
:-)
<djStolen>
but still nothing
<PaulFertser>
djStolen: I think direct SPI flash programming (might be possible without desoldering even) is the way to go then.
<djStolen>
jup. seems like it.
<djStolen>
with OpenOCD, you would recommend?
<karlp>
some boards have some 0 ohm resistors not placed in the rx/tx paths as "protection from hacking"
<PaulFertser>
djStolen: no, OpenOCD doesn't program flash directly, you'd use flashrom.
<PaulFertser>
karlp: in xiaomi case it feels like they just lock down bootloader and host OS on software level, it just doesn't take any input unless you change some u-boot variable...
<karlp>
that too,
<karlp>
pcb doesn't look like it has any obvious pads like this, but the pictures not real great.
<djStolen>
PaulFertser: ok. got it.
<PaulFertser>
After seeing Xiaomi doing that nastiness (plus selling a completely different device as "mir3v2") I'm not recommending anything from them to anybody.
<djStolen>
karlp: actually, there are 2 empty spots near J1 4 x holes
<djStolen>
karlp: R87 and C1071
<djStolen>
and its connected to the TX line, if its to trust the above page
<djStolen>
PaulFertser: what do u mean by nastiness?
<karlp>
does one pad of r87 go to eithe rpin 2 or 4 of J1? you could _try_ shorting it, but... that's on you ;)
<PaulFertser>
djStolen: locking bootloader and abusing device names.
<djStolen>
karlp: R87 pin 1(lets name it like that) goes to TX, the other pin 2 goes to pin of C1071 and from there to J1-PIN1 (a.k.a. VCC)
<djStolen>
PaulFertser: so u have to change that u-boot var via hardware access to unlock it, to be able to do anything. did i get u right?
<mangix>
aparcar[m]: no I mean where is this from? who set it up?
dedeckeh has quit [Remote host closed the connection]
rmilecki has quit [Ping timeout: 480 seconds]
rejoicetreat has quit [Ping timeout: 480 seconds]
<aparcar[m]>
mangix: it's an automatic feature
anxy has quit [Quit: Page closed]
<greearb__>
what part of the logic names network devices 'lan1, lan2, ...' ? I'd prefer they be called 'eth1, eth2, ...' for my application.
Borromini has joined #openwrt-devel
<Pepes>
aparcar[m]: Well, I think mangix knows that it is automatic feature, but his question was rather, who enable it, etc. We know that we should not merge it.
<aparcar[m]>
Pepes: it's a service of GitHub which is enabled per default for every repository hosted on GitHub
<Pepes>
I would not say that it is enabled by default, though. It is configurable in settings.
<Pepes>
So, the still the question remains that who set it up, I see that was good intention, but maybe letting know others about that would be nice, tbh. And it is question if it is somehow usable for packages repository, but we will see.
Borromini has quit [Quit: Lost terminal]
tapper has quit [Quit: Tapper]
<greearb__>
isc-dhcp-server-ipv4 is no longer in menu-config as far as I can tell. Any idea how to get it back?
<greearb__>
ahh, have to enable bind-client and bind-server to get it to show up
Zaba2 has quit [Quit: Leaving.]
djStolen has quit [Remote host closed the connection]