lucascastro has quit [Remote host closed the connection]
n3ph has joined #openwrt-devel
gromero has quit [Ping timeout: 480 seconds]
<zetagon>
hello I'm back again. is there any conventional wisdom around making procd services for a shell script? I'm struggling with properly terminating the script, presumably because shell scripts often spawn subprocesses that procd can't terminate
<zetagon>
is it just a bad idea to make a procd service for a shell script?
<KanjiMonster>
zetagon: you could maybe add a signal handler to the shell script where you terminate those subprocesses yourself
<robimarko>
Anybody knows the idea behind the "Default" profiles is?
<robimarko>
Cause, we have an issue for Lantix XRX200 where IB when used with the Default profile will simply use the image recipe for the first device in subtarget.mk image recipes
<robimarko>
And fail since that device is broken and marked with DEFAULT := n so the kernel it needs is not built
<KanjiMonster>
"Default" is usually "all images". It's from pre per-device rootfs times
<robimarko>
Ahh
<robimarko>
Where is the "recipe" for its generation stored?
<robimarko>
So, we can at least make it do what IB does and not show profiles that cannot be built
<KanjiMonster>
the association of devices to profiles is done via the PROFILES variable, e.g. bmips has "PROFILES = Default $$(DEVICE_NAME)" in its Device/Default
<KanjiMonster>
the "DEFAULT := n" is probably only applied when using multi-device profiles
<robimarko>
Well, lantiq only has: PROFILES = Default
<robimarko>
I see there is a mix
<robimarko>
Some have Default $$(DEVICE_NAME) while some Default and there is even one Generic $$(DEVICE_NAME)
danitool has joined #openwrt-devel
sorinello has quit [Ping timeout: 480 seconds]
merbanan has quit [Remote host closed the connection]
merbanan has joined #openwrt-devel
SirLouen has quit [Read error: Connection reset by peer]