@proxy-maintainers, please : 1/ update the tree with the attached ebuilds for the 3.9.2 and 3.8.13 releases. 2/ drop 3.8.3 ebuild. Reproducible: Always
Created attachment 348156 [details] ck-sources-3.9.2.ebuild Ebuild for the 3.9.2 release. Based on : - Sources for the linux 3.9 branch - genpatches rev 5 - 3.9-ck1 patchset including bfs-430 Note that as announced under misc. medias, we stop validating/qualifying early releases (that is to say releases < first gentoo-sources stable release) under x86 architectures. As a consequence of this, the x86 keyword was dropped from the 3.9.2 ebuild. Real users who would be annoyed are welcome to voice their concern in this bug, the gentoo forums or by mail.
Created attachment 348158 [details] ck-sources-3.8.13.ebuild Ebuild for the 3.8.11 release. Based on : - Sources for the linux 3.8 branch - genpatches rev 16 - 3.8-ck1 patchset including bfs-428
(In reply to comment #1) > Created attachment 348156 [details] > ck-sources-3.9.2.ebuild > > Ebuild for the 3.9.2 release. > > Based on : > > - Sources for the linux 3.9 branch > - genpatches rev 5 > - 3.9-ck1 patchset including bfs-430 > > Note that as announced under misc. medias, we stop validating/qualifying > early releases (that is to say releases < first gentoo-sources stable > release) under x86 architectures. > As a consequence of this, the x86 keyword was dropped from the 3.9.2 ebuild. > > Real users who would be annoyed are welcome to voice their concern in this > bug, the gentoo forums or by mail. Could you explain that? I see no reason to drop the ~x86 unless the ck patches explicitly break x86 support.
(In reply to comment #3) > (In reply to comment #1) > > Note that as announced under misc. medias, we stop validating/qualifying > > early releases (that is to say releases < first gentoo-sources stable > > release) under x86 architectures. > > As a consequence of this, the x86 keyword was dropped from the 3.9.2 ebuild. > > > > Real users who would be annoyed are welcome to voice their concern in this > > bug, the gentoo forums or by mail. > > Could you explain that? I see no reason to drop the ~x86 unless the ck > patches explicitly break x86 support. a/ Before submitting a bump request, the releases are scrupulously qualified. The process we follow for this is quite huge and takes considerable time. b/ I have observed that the users of the ck-sources package who are eager to get early releases are running amd64 arches while x86 users stay on "stable" (I mean releases of the 3.4 branch or releases based on revisions of the genpatches used in stable gentoo-sources). => Not qualifying the first releases of new kernel versions under x86 arches enables me to submit earlier a new version of the kernel to (I believe) nobody's expenses. c/ There are actually no known reason for the 3.9-ck1 patchset to break x86 support however, I was told on sunrise that no actual testing on a particular arch => no keyword for that arch, (which sounds sensible) => I dropped the ~x86 keyword for the 3.9.2 In addition to this, regarding this particular release of the ck-sources, users are likely to face issues with power off/reboot/suspend/hibernate. I will honestly not be in a position to efficiently support any user facing this sort of troubles under an x86 arch.
That's fine but given the base code (linux kernel) is well supported in x86 i see no reason to drop the keyword even if we can't test it ourselves. Moreover, the ck-sources is yet another kernel patchset, and like every other kernel patchset, users are expected to know what they are doing when they use it. So please, restore the ~x86 keyword.
Created attachment 348176 [details] ck-sources-3.9.2.ebuild (KEYWORDS="~amd64 ~x86) ~x86 arch restored.
+*ck-sources-3.9.2 (15 May 2013) +*ck-sources-3.8.13 (15 May 2013) + + 15 May 2013; Markos Chandras <hwoarang@gentoo.org> +ck-sources-3.8.13.ebuild, + +ck-sources-3.9.2.ebuild, -ck-sources-3.8.3.ebuild: + Version bump thanks to Eric F. GARIOUD <eric-f.garioud@wanadoo.fr>. Remove + old. Bug #469686 +
Thank you Markos.