Re: [PATCH] rules: fix empty COMMITCOUNT/AUTORELEASE

2021-01-29 Thread Ilya Lipnitskiy
On Fri, Jan 29, 2021 at 10:23 AM Paul Spooren wrote: > > On Fri, Jan 29, 2021 at 10:43, Petr Štetiar wrote: > > I guess, that the actual error from that dump.txt file would be more > > useful in > > this case :) > > Sorry, it was late: > Makefile:48: *** Package/px5g-wolfssl is missing the

Re: dnsmasq-dhcp[1503]: DHCP packet received on eth0.102 which has no address

2021-01-29 Thread Enrico Mioso
Sorry guys, should have been more careful to explain myself at least. The message is pretty clear, still I expected packets from that interface to be ignored, since the device is in the wan interface. # uci show dhcp.wan dhcp.wan=dhcp dhcp.wan.interface='wan' dhcp.wan.ignore='1' ... thanks!

[PATCH] build,package-defaults: AUTORELEASE as fallback

2021-01-29 Thread Paul Spooren
This commits changes the package defaults so that $(AUTORELEASE) is used whenever no explicit PKG_RELEASE is set and $(COMMITCOUNT) whenever neither is given. Previously an explicit definition of PKG_RELEASE=$(AUTORELEASE) would be required. Signed-off-by: Paul Spooren ---

Re: [PATCH] rules: fix empty COMMITCOUNT/AUTORELEASE

2021-01-29 Thread Paul Spooren
On Fri, Jan 29, 2021 at 10:43, Petr Štetiar wrote: I guess, that the actual error from that dump.txt file would be more useful in this case :) Sorry, it was late: Makefile:48: *** Package/px5g-wolfssl is missing the VERSION field. Stop. VERSION is defined as PKG_VERSION-PKG_RELEASE, or

dnsmasq-dhcp[1503]: DHCP packet received on eth0.102 which has no address

2021-01-29 Thread Enrico Mioso
Hello all!! Just observing something strange happening on my router. I am running openwrt 19.07.5 with updated dnsmasq package, opkg says dnsmasq - 2.80-16.3 I had to isolate one lan port to use it like a WAN one with PPPoE. My network config is: config interface 'loopback' option

Re: [PATCH] umdns: bump to 2021-01-26

2021-01-29 Thread Petr Štetiar
Paul Spooren [2021-01-28 09:11:50]: > On Wed, Jan 27, 2021 at 22:37, Petr Štetiar wrote: > > Can't this be a default which could be overriden when needed? > > Sure, see patch below. As it's a new feature I wanted it optionally rather > than treewide for the beginning. I didn't wrote treewide.

Re: [PATCH] rules: fix empty COMMITCOUNT/AUTORELEASE

2021-01-29 Thread Petr Štetiar
Paul Spooren [2021-01-28 23:25:57]: Hi, > ERROR: please fix package/utils/px5g-wolfssl/Makefile - \ > see logs/package/utils/px5g-wolfssl/dump.txt for details I guess, that the actual error from that dump.txt file would be more useful in this case :) Cheers, Petr

[PATCH] rules: fix empty COMMITCOUNT/AUTORELEASE

2021-01-29 Thread Paul Spooren
Packages that are in-tree only often lack a PKG_VERSION and only use the PKG_RELEASE to mark changes. Using COMMITCOUNT/AUTORELEASE variables causes an issue as both variables are empty during the metadata DUMP phase. Instead of leaving these variables empty and causing an error message like

Re: ntpd problem if procd-ujail is enabled

2021-01-29 Thread e9hack
Am 27.01.2021 um 17:54 schrieb Daniel Golle: On Wed, Jan 27, 2021 at 05:17:48PM +0100, e9hack wrote: It should be executed (i've tried), but as user 'ntp', hence the added ACLs. I've recently reworked support for non-root hotplug calls via ubus, please see my staging tree and let me know if the