**** BEGIN LOGGING AT Sun Dec 09 03:00:01 2018
Dec 09 11:08:46 <RP> aehs29: not sure what happened to patch 1 of 2 :/
Dec 09 13:33:14 <didile> hi
Dec 09 13:33:47 <didile> the recipe make-mod-scripts of sumo doesn't work with my Linux Kernel 4.1.15
Dec 09 13:34:27 <didile> at first it says: kernel-sources is not clean, please run 'make mrproper'
Dec 09 13:34:52 <didile> and then:  dtc-lexer.l:40:10: fatal error: dtc-parser.tab.h: No such file or directory
Dec 09 13:35:50 <didile> https://git.yoctoproject.org/cgit.cgi/poky/plain/meta/recipes-kernel/make-mod-scripts/make-mod-scripts_1.0.bb
Dec 09 13:36:14 <didile> it works on morty
Dec 09 13:36:34 <didile> this seems to be a new recipe
Dec 09 17:20:00 <woky> Hello. Which DHCP server do you use for when testing a board? I'm currently using ISC dhcpd, but it remembers its leases. My board currently generates random MAC on boot. I've defined single IP range in the dhcpd server. So when I reset the board, the dhcp server has no free leases. Is there any lightweight dhcp server that just forgets its leases?
Dec 09 18:01:33 <moto-timo> woky: I use dnsmasq, although I have never needed to address your issue.
Dec 09 18:01:48 <moto-timo> woky: but this seems to be the tool for the job: http://manpages.ubuntu.com/manpages/bionic/man1/dhcp_release.1.html
Dec 09 18:02:47 <shauno> disregarding leases would be very niche, since it'd break if you had more than one dhcp client on the network (every device would be given the single address in your range)
Dec 09 18:03:13 <moto-timo> you might also be able to configure the leases to have a very short duration (just a few seconds)
Dec 09 18:03:16 <shauno> I use dnsmasq too though, with very short (3 minute) leases on the dynamic pool, and much longer leases on fixed assignments
Dec 09 18:03:25 <moto-timo> right
Dec 09 18:04:38 <moto-timo> I do the same, short lease for dynamic dhcp (which I use for device discovery) and then static ip for everything that is known (devices that do not change their MAC address)
Dec 09 18:05:07 <shauno> you can also do fixed assignments based on their hostname rather than their mac address, which might be an option for your wandering device
Dec 09 18:05:18 <moto-timo> I don't have any devices that change their MAC address... that seems like a path to madness (for an automated testing lab)
Dec 09 18:06:53 <shauno> I have a strange setup where we have a fixed set of devices, and a fixed set of mac addresses.  but no fixed relationship between the two.  madness doesn't quite cover the depth of depravity that lan can get to
Dec 09 18:07:17 <moto-timo> :P
Dec 09 18:07:56 <woky> thank you for hints
Dec 09 18:09:09 <moto-timo> woky: I know LAVA uses ISC by default and they have probably had to deal with your issue... you might find some help there http://lavasoftware.org
Dec 09 18:09:45 <moto-timo> woky: but I find dnsmasq much easier to configure and it does more with one tool
Dec 09 19:06:50 <khem> woky: do you use systemd ?
Dec 09 19:07:24 <woky> khem, I've already fixed my problem. The boot image indeed uses systemd, so I've just defined static network via systemd *.network file.
Dec 09 19:07:33 <woky> s/network/address/
Dec 09 19:07:54 <woky> I sure have to fix the MAC generating random address in u-boot.
Dec 09 19:08:08 <khem> usually boards publish dummy MACs and that results in differnet IPs
Dec 09 19:10:44 <khem> you can use a solution like the one I have here https://github.com/YoeDistro/meta-yoe/blob/master/recipes-core/systemd/systemd/wired.network#L10
Dec 09 19:10:55 <khem> you dont need static n/w
Dec 09 23:52:49 <RP> armpit: The sumo-next and thud-next patchsets seem to test ok, so we want to post for review?
Dec 10 00:02:31 <armpit> sounds fine
Dec 10 02:19:00 <aehs29> RP: why did you only merge my patch for py3 and not py2?
**** ENDING LOGGING AT Mon Dec 10 03:00:00 2018