Created attachment 422020 [details] mcelog-129.ebuild mcelog has seen quite a few releases since our last one in portage, mostly just to support new cpus & hardware. Though this recent commit requires an adjustment to one of our patches: https://github.com/andikleen/mcelog/commit/b056a172dba6ba13c8badda1f897e98d70db581b Since it now correctly uses $(CC) instead of gcc, our mcelog-1.0_pre3_p20120918-build.patch does more than it needs to. I'm attaching a new patch, and a new 129 ebuild. I have bumped the ebuild to EAPI6, hope that's ok. I can revise the ebuild to stay on EAPI5 if there is any reason for that. Thanks!
Created attachment 422022 [details, diff] mcelog-129-debugflags.patch
Created attachment 432472 [details] mcelog-136.ebuild Here is my most recent ebuild for 136. I bump these ebuilds in my overlay pretty promptly at https://github.com/benkohler/iamben-overlay/tree/master/app-admin/mcelog but yeah it's time we got a newer one in ::gentoo. I will also attached a diff of this ebuild versus the latest in tree, let me know if you see any other changes that are needed.
Created attachment 432474 [details, diff] ebuild diff for 122 -> 136
*** Bug 581476 has been marked as a duplicate of this bug. ***
Imported this into my overlay for testing, stumbled over mcelog-1.0_pre3_p20120918-bashism.patch not applying correctly since it lacks a leading entry in the patch path. Added a/ & b/ in the patch: applies, builds, works. \o/ Afterwards I saw that your git repo contains the fixed version, that should probably be attached here as well. This update makes me happy because stopping & restarting mceclient now works properly; the version in portage does not cleanly restart without manual removal of a stale socket. This version still leaves the /run/mcelog-client socket behind, but then starts correctly anyway. Should we delete the /run/mcelog-client socket on stop() in the start-stop-script while we're at it?
any stoppers in making this version added?
I guess we were just waiting on a proxy maintainer to commit it? Not sure. Anyway, I bumped it to 148 after applying Ben's EAPI=6 changes, and updating the two patches. commit 1c87fc4663268b5dddb1c933cdbd249d0ef4460a Author: Matt Turner <mattst88@gentoo.org> Date: Tue Mar 7 17:58:13 2017 -0800 app-admin/mcelog: Version bump to 148.
