Some new hardware with 9.1 does not reboot easily

[Willem Jan Withagen]

On 2012-11-29 16:08, Andriy Gapon wrote:
> on 29/11/2012 10:53 Willem Jan Withagen said the following:
>> On 2012-11-29 9:41, Patrick Lamaiziere wrote:
>>> Le Mon, 26 Nov 2012 12:19:07 +0200,
>>> Andriy Gapon <avg at FreeBSD.org> a écrit :
>>>
>>> Hello,
>>>
>>>> on 26/11/2012 12:10 Patrick Lamaiziere said the following:
>>>>> As far I can see it fails because there is no getnewvnode_reserve()
>>>>> / get_newvnode_drop_reserve() in 9.1.
>>>>
>>>> The patch is for stable/9.
>>>
>>> Ok, thanks.
>>>
>>> I've made a new diff because Willem's patch does not apply
>>> on 9.STABLE (fails on opensolaris_lookup.c) :
>>>
>>> This one is for 9-stable rev 243569:
>>> http://user.lamaiziere.net/patrick/9-STABLE-r243569-patch-zfs-reboot
>>>
>>> It solves the reboot problem on my server. I'm testing it on my
>>> workstation with few "poudriere bulk". It looks to work.
> 
> Thank you very much for testing!
> 
>> Andriy sent me his diff aginst HEAD, which I reworked against the code I
>> have here locally. I did an svn diff against the checkout of that
>> morning. And I'm pretty sure i'm tracking stable/9
>>
>> So I'm not sure where I went wrong, but thanx for fixing it.
> 
> I MFC-ed some ZFS changed from HEAD last weekend.

Would that mean that the regular checkout of stable/9 contains enough
code to allow "painless" rebooting...

Reason I ask:
	I'm about to ship the server to the customer

But there is still 45 minutes before UPS comes to pick it up.
And I'd rather run a regular checkout, instead of the patched one.

Which then gets me to the svn question:
	How do I easily get ride of the changeds?
--WjW