Date: Thu, 17 Jul 2014 08:28:50 +0300 From: Kimmo Paasiala <kpaasial@icloud.com> To: "freebsd-stable@freebsd.org" <freebsd-stable@freebsd.org>, kasahara@nc.kyushu-u.ac.jp Cc: delphij@FreeBSD.org Subject: Re: problem updating 10-STABLE Message-ID: <87FFF1B1-669A-487A-B4AD-ED4F800AF2B3@icloud.com> In-Reply-To: <mailman.51.1405512001.14763.freebsd-stable@freebsd.org> References: <mailman.51.1405512001.14763.freebsd-stable@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
--Apple-Mail=_5A9C004A-FE44-4B6B-B20F-585B8C5E90DA Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=windows-1252 >=20 > Message: 5 > Date: Wed, 16 Jul 2014 19:06:32 +0900 (JST) > From: Yoshiaki Kasahara <kasahara@nc.kyushu-u.ac.jp> > To: stable@freebsd.org > Subject: problem updating 10-STABLE > Message-ID: > <20140716.190632.1014073092745429063.kasahara@nc.kyushu-u.ac.jp> > Content-Type: Text/Plain; charset=3Dus-ascii >=20 > Hi, >=20 > Today I tried to update my 10-STABLE machine (full ZFS including root) > from r268480 (built on Jul 10) to r268734 (built today), but after > make installkernel and reboot into single user mode (to do make > installworld and mergemaster), I couldn't proceed further because "zfs > mount -a" crashed immediately with segfault. Because only read-only > root fs was available, I couldn't do any debugging (/usr resides in a > separate fs). Actually any zfs or zpool commands seemed to crash > immediately. >=20 > There were several ZFS related commits between two builds, but I > believe old userland should be usable with a newer kernel (or I cannot > follow the standard install procedure). >=20 > % uname -a > FreeBSD elvenbow.cc.kyushu-u.ac.jp 10.0-STABLE FreeBSD 10.0-STABLE #0 = r268480: Thu Jul 10 15:25:15 JST 2014 = root@elvenbow.cc.kyushu-u.ac.jp:/usr/obj/usr/src/sys/ELVENBOW amd64 > % what /boot/kernel.new/kernel > /boot/kernel.new/kernel: > FreeBSD 10.0-STABLE #0 r268734: Wed Jul 16 12:36:48 JST 2014 >=20 > Any suggestions? I don't want to do make installworld during multi > user mode because I cannot ensure new zfs userland will work with new > kernel (if not, my system will become unusable). Maybe try installing > new /rescue only and try it with the new kernel? >=20 > I'm considering to set up BE to avoid this kind of problem, but I need > some exercise... >=20 > Regards, > --=20 > Yoshiaki Kasahara > Research Institute for Information Technology, Kyushu University > kasahara@nc.kyushu-u.ac.jp >=20 >=20 I=92m also seeing the same behavior on my amd64 system, with the updated = kernel none of the zpool(8)/zfs(8) tools work but segfault. I reverted = back to the kernel from my previous build that is r268600 and that one = still works. The one that didn=92t work was r268733. I=92m just guessing = here but the suspect commit could be: = http://svnweb.freebsd.org/base?view=3Drevision&sortby=3Drev&sortdir=3Ddown= &revision=3D268658 Adding the committer delphij to this discussion, can you please comment = on this matter? -Kimmo --Apple-Mail=_5A9C004A-FE44-4B6B-B20F-585B8C5E90DA Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename=signature.asc Content-Type: application/pgp-signature; name=signature.asc Content-Description: Message signed with OpenPGP using GPGMail -----BEGIN PGP SIGNATURE----- Comment: GPGTools - https://gpgtools.org iQEcBAEBCgAGBQJTx18WAAoJEFvLZC0FWRVp5kMH/12PzI5Jafe8ODCn272lZjEW ZuWpjdDIrj8SiM1EbJ8j1wEZx6yPqT76TwaPqwY9NbbShXsK4tNEtkW/MVazd/LH D9L6hUhVy3mVhuKFIEsXkj4tu6y82NpK9JOgSnWbrzyG1ObIHfJg3wvwqMTuCeet jOiX7d6OQsv22yedpQ2Ow50Ardgp9uPTOmueFDJWcVFBeRwJMPmz3j/5QoEiwrvh uC9AR5wwjvZieaQTscONqfebGJdD0yHNqZxNrA2E4vKkZAUxPyIzXou2W7QP0pc7 JGoeYodkSFX9KvBd++TO+IoBSEgC6PjiUwuKKsfc9uv0saYfn3Tc3t+9AQlu9rg= =p08G -----END PGP SIGNATURE----- --Apple-Mail=_5A9C004A-FE44-4B6B-B20F-585B8C5E90DA--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?87FFF1B1-669A-487A-B4AD-ED4F800AF2B3>