[01:05] so is kabi broken only if an exported fn's args change, or even when the length of the fn changes? [01:06] noob q === JanC is now known as Guest29977 === JanC_ is now known as JanC [07:10] hallyn, when a function signature changes yes, an arg or the type of that arg or the consituents of that arg (in the case of a struct) [07:10] or its return value of course === amitk_ is now known as amitk [14:04] hello, I Was trying to install ubnutu server 16.04 as KVM guest of Proxmox 4.4-13. In the VM settings I choosed "automatically allocate memory" with a maximum memory and minimum memory acceptable values in mbytes (512 and 4096) [14:05] but I get kernel panics, ther kernel starts tu kill proccesses and simetimes kills systemd. [14:06] AFAIR proxmox pve uses kvm ballooning and the guest, unbutu 16.04 should have drivers for kvm ballooning. [14:06] Is memory ballooming supported un unbuntu server? [14:09] https://pve.proxmox.com/pve-docs/pve-admin-guide.html#qm_memory [14:09] ^ reference doc [15:13] apw: but just changing the length of the fn shouldn't do it right? [15:14] i'm confused by https://bugs.centos.org/view.php?id=13265 the patch therein [15:15] does not seem capable of changing the kabi .. i guess i may have to look at their kabichecker [15:15] groan [15:35] hallyn, can you point me to the patch ? [15:40] apw: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=4e52365f279564cef0ddd41db5237f0471381093 [15:42] hallyn, it is hard to see how the heck that changes the kabi, you could confirm by reverting it on our tree and seeing it if breaks our abi [15:45] just by building without doing the skipabi=true in debian/rules should suffice? [15:51] hallyn, yeah something like that [15:51] hallyn, our abi checks function signatures, which includes any structures used as parameters === ejat is now known as fenris === fenris is now known as ejat === stgraber_ is now known as stgraber