From mboxrd@z Thu Jan 1 00:00:00 1970 From: Trent Piepho Date: Thu, 24 May 2018 16:29:12 +0000 Subject: [Buildroot] [PATCH] util-linux: Bump to version 2.32 In-Reply-To: <871se1dogl.fsf@dell.be.48ers.dk> References: <20180521191623.7049-1-tpiepho@impinj.com> <1526938480.31570.89.camel@impinj.com> <877enwis89.fsf@dell.be.48ers.dk> <857898562.2094800.1527079548616.JavaMail.zimbra@datacom.com.br> <871se1dogl.fsf@dell.be.48ers.dk> Message-ID: <1527179352.31570.103.camel@impinj.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net On Thu, 2018-05-24 at 18:10 +0200, Peter Korsgaard wrote: > > > > > > "Carlos" == Carlos Santos writes: > >> > Nevermind, I see I missed this change already being on buildroot/next. > >> > If anyone wonders why mke2fs blocks for 2 seconds getting a filesystem > >> > UUID on recent kernels on low boot entropy systems, it's because of the > >> > problem described here. > >> > >> I actually also hit this on 4.14.39 recently. Perhaps we should include > >> this bump (or better just the getrandom changes) on master and on > >> 2018.02.x? > >> > >> Care to send a patch for master? > > > Isn't it simpler to cherry-pick commit 4b54bd3ae8, instead? > > First of all, 4b54bd3ae8 is already on master so the question is just > what we do for 2018.02.x? > > Just cherrypicking the bump to 2.32 is certainly the easiest, but > perhaps also riskier than just cherry-picking the getrandom() changes: > > git shortlog v2.31.. -- lib/randutils.c > Carlo Caione (1): > lib/randutils: Do not block on getrandom() > > Karel Zak (1): > lib/randutils: don't break on EAGAIN, use usleep() > > On the other hand, util-linux bumps are normally quite safe and we have > not had any issues reported, so perhaps we should just go for 2.32 for > 2018.02.x as well? If you bump util-linux to 2.32, then systemd < 238 doesn't build without systemd commit 227b8a762fea1458547be2cdf0e6e4aac0079730 core: don't include libmount.h in a header file (#8580)