ivan | could whoever pushes please push to ubuntu/ubuntu-xenial.git, it's missing Ubuntu-4.4.0-71 | 01:50 |
---|---|---|
hallyn | stupid question. If I have livepatch enabled, how do I know whether my kernel (a) is in fact being livepatched and (b) is uptodate with the most recent panicy cve-patching kernel release? | 03:23 |
hallyn | does "fully-patched: true" mean that latter? | 03:23 |
hallyn | (as a response to 'sudo canonical-livepatch status') | 03:24 |
stgraber | hallyn: canonical-livepatch status --verbose | 04:56 |
stgraber | hallyn: that should show you the CVE that got patched too | 04:56 |
smb | ivan, xenial tree is updated now | 08:06 |
smb | though only the master branch right now | 08:06 |
ivan | smb: thanks | 10:43 |
ivan | is there anything in 4.4.0-64 -> 4.4.0-69 (proposed, yes) that could have regressed network device renaming? I used to see eth0,eth1 -> em1,em0 renaming but now absolutely no renaming | 10:43 |
ivan | on just one machine. of course. | 10:44 |
ivan | running a tg3 0000:03:00.0 eth0: Tigon3 [partno(N/A) rev 5720000] (PCI Express) | 10:44 |
ivan | oh wow a SIOCSIFNAME ioctl does this so I need to look outside the kernel too | 10:52 |
smb | ivan, I believe the whole renaming game is in some way tied into systemd nowadays (formerly done by udev rules) | 10:55 |
ivan | thanks | 10:55 |
ivan | I need a Really Stable Device Names that just brings up all the network devices and sees which gets which IP over DHCP | 10:56 |
smb | ivan, Maybe "man systemd.link" is helpful for you. Should be working nowadays by creating a *.link file in /etc/systemd/network with a match section for the mac address and a link section with the desired name. | 10:58 |
smb | Then update-initramfs -u (not 100% its needed but should not hurt either) | 10:58 |
ivan | ah thank you | 11:04 |
ivan | it looks like I had an empty (with comments) /etc/udev/rules.d/80-net-setup-link.rules that was inhibiting network device renames | 11:54 |
ivan | but now they're eno1,eno2 instead of em1,em0. whatever, I'll live with it | 11:57 |
ivan | _oh_, em0/1 aren't stable names, that was a dumb thing to assume | 11:59 |
=== cachio_afk is now known as cachio | ||
ivan | oh, em = embedded NICs and they're supposed to be stable, I guess the kernel stopped thinking my NICs were embedded NICs (?), whatever | 12:08 |
hallyn | stgraber: hm, it doesn't. i wonder whether the latest cve was not patchable | 13:09 |
=== jjohansen1 is now known as jj-cloaked | ||
=== jj-cloaked is now known as jjohansen | ||
ivan | oh, I see, things changed because I purged biosdevname | 13:51 |
=== ben___ is now known as ben_r | ||
manjo | I need to add a module for ethernet and PHY for arm64 vendor.. it is applicable only to ARM64 and a given vendor .. should I add that to d-i/modules/arm64/nic-modules ? d-i/modules/arm64/modules simply includes generic nic-modules so asking | 15:12 |
manjo | apw, ^ any thoughts ? | 15:16 |
manjo | apw, this for di install of that system | 15:16 |
rtg | manjo, looks right | 15:25 |
manjo | rtg, add it to arm64 specific nic-modules correct ? | 15:26 |
rtg | manjo, yep | 15:27 |
manjo | awesome thanks | 15:27 |
=== cmagina_ is now known as cmagina | ||
=== JanC is now known as Guest23500 | ||
=== JanC_ is now known as JanC | ||
lostgoat | hey guys, trying to make a ppa with some local kernel changes + 4.11-rc4 | 16:24 |
lostgoat | I've got my kernel + debian patches, and a source package I created using "debuild -S -I -i" | 16:25 |
lostgoat | that uploaded nicely to launchpad, but resulted in error building after 5 hours | 16:26 |
lostgoat | /bin/bash: line 3: /<<BUILDDIR>>/linux-4.11.0-rc4+losgoat/debian/build/udeb-meta-packages: No such file or directory | 16:26 |
lostgoat | I can look for the solution to this, but I was mostly curious if anyone knows any other pitfalls I should be trying to avoid | 16:27 |
lostgoat | don't want to have to wait another 5 hours for a silly mistake | 16:27 |
lostgoat | Or if anyone knows how to perform the same build locally that would be great | 16:27 |
* lostgoat is not super experienced using debuild | 16:28 | |
apw | lostgoat, did you clean the package before you built it ? | 17:22 |
apw | lostgoat, the specific error you have implies you have debian installer bits turned on but no d-i config | 17:24 |
apw | lostgoat, if you arn't making old style CDs from your kernel you might want to set disable_d_i=true in your buidl | 17:24 |
lostgoat | apw: yeah I had a local build at some point, but I ran fakeroot debian/rules clean before starting the debuild command. (and also a git clean -fd before that). I might have messed this up somehow though | 18:28 |
lostgoat | I'll look into disable_d_i and turn it off as you suggested. I don't need CDs | 18:28 |
lostgoat | apk: thanks for the tips | 18:28 |
lostgoat | *apw | 18:28 |
lostgoat | apw: I've also read about disabling abi checks. Is that something else I should be doing? | 18:52 |
apw | yes most likely | 18:54 |
apw | skipabi and skipmodule something like that | 18:55 |
lostgoat | thanks, will do | 19:05 |
=== Guest86599 is now known as RAOF |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!