freebsd stable - Jan 2016 - [HEADS UP] "options ZFS" broken on stable/10; please use zfs(4) module for the time being

Hi,

Did you merge the changes made
in ./sys/cddl/compat/opensolaris/sys/nvpair.h and
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=201356 ? Those twos
should repair the problem.

Thanks for doing all the work around merging it.

Cheers,
Mariusz

On 3 January 2016 at 05:32, NGie Cooper <yaneurabeya at gmail.com> wrote:
> Hi all,
>         Marcel brought to my attention that ?options ZFS? is currently
> broken on stable/10 as of r292973. It builds on head however. I?m trying to
> determine what MFCs are missing in order to fix stable/10. I?ll send out an
> update email once I figure it out and fix the issue.
>         I apologize for the breakage.
> Thanks,
> -NGie

> On Jan 3, 2016, at 06:42, Mariusz Zaborski <oshogbo at FreeBSD.org>
wrote:
> 
> Hi,
> 
> Did you merge the changes made in
./sys/cddl/compat/opensolaris/sys/nvpair.h and
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=201356 ? Those twos should
repair the problem.
> 
> Thanks for doing all the work around merging it.
Hi Mariusz!
	I figured out the missing revisions to get things working again on stable/10
(the revisions I listed below were the sticking point to getting this working
100%). There are a lot of changes around pci(9) / nv(9) that I?ll be putting out
for review soon, so the backport of pci_iov(9) and nv(9) can be completed (I
need to do some more testing soon and some careful inspection to ensure KBI
guarantees and the like are maintained).
	All in all I?m hoping that this can be fixed ?soon? (sometime in the next week
or so).
Thanks!
-NGie

r293093 | ngie | 2016-01-03 01:11:55 -0800 (Sun, 03 Jan 2016) | 21 lines

MFC r279437,r284107:

r279437 (by rstone):

Allow Illumos code to co-exist with nv(9)

Suggested by:                   pjd

r284107 (by avg):

compat nvpair.h: make sure that the names are mangled only for kernel

Currently there is no good reason to mangle the userland API.
The change was introduced in eac1d566b46edef765754203bef22c75c1699966,
r279437.  Also see https://reviews.freebsd.org/D1881.

I am still convinced that nv should not have introduced intentionally
conflicting API.

Discussed with: rstone