Hi, On Sun, 13 Dec 2015 22:12:24 +0100 gregor herrmann <gre...@debian.org> wrote: > Control: tag -1 - pending > > On Mon, 14 Dec 2015 08:05:22 +1100, Ian Wienand wrote: > > > The patch has not been applied because it is not clear it is correct > > > > There is a thread that starts at [1] and ended inconclusively at [2] > > > > It's unfortunate that the number of people who understand power memory > > ordering > > issues at the level required can probably be counted on one hand. But I > > don't > > think we should apply patches upstream feel don't address the issue > > > > -i > > > > [1] > > http://www.hpl.hp.com/hosted/linux/mail-archives/gc/2014-January/005825.html > > [2] > > http://www.hpl.hp.com/hosted/linux/mail-archives/gc/2014-February/005852.html > > Thanks for this additional information. > I've cancelled the aforementioned NMU.
The last version had failed on armel, powerpc and ppc64el. I gave it back and it built on armel and powerpc, but it failed on ppc64el (gave it back twice there): https://buildd.debian.org/status/package.php?p=libatomic-ops Dunno if you want to close this bug or retitle it. Has there been any progress on a fix for ppc64el? Regards, Emilio