Re: A patch against FreeBSD-SA-05:15.tcp

[Noritoshi Demizu <demizu@xxxxxxxxxxxxxx>]

> From: Matthew Dillon <dillon@xxxxxxxxxxxxxxxxxxxx>
> To: bugs@xxxxxxxxxxxxxxxxxxxxxxx
> Subject: Re: A patch against FreeBSD-SA-05:15.tcp
> Date: Mon, 29 Aug 2005 23:00:51 -0700 (PDT)
>
> :Could you let me know the procedure of MFC?
> :
> :I guess:
> :  1. Prepare a patch against FreeBSD-SA-05:15.tcp to
> :     for DragonFly_RELEASE_1_2_Slip.
> :  2. Wait for a sign from Matt.
> :  3. Commit the prepared patch to DragonFly_RELEASE_1_2_Slip
> :     in the same way as the commit to HEAD.
> :
> :Is this correct?
> :
> :Regards,
> :Noritoshi Demizu
>
>     Minimum 1 week wait before something goes into the release branch.
>
>     The release branch is DragonFly_RELEASE_1_2.  That is what you would
>     commit into.  You don't have to worry about it after that.
>
>     Generally I'm the only one who slips the slip tag.  It doesn't become
>     truely official until I bump the release sub-version and synchronize
>     the slip tag on the release branch.  The idea here is that we might get
>     a dribbling of MFCs into the release branch, and we want committers to
>     be able to MFC those, but we don't want to bump the subversion each time
>     or create confusion by rolling lots of subversions/official-updates.  So
>     the slip tag will not typically be synchronized any faster then once a
>     month, depending, and I'm usually the only one who does it.
>
> 					-Matt
> 					Matthew Dillon
> 					<dillon@xxxxxxxxxxxxx>

Thank you.  I misunderstood the relation between DragonFly_RELEASE_X_Y
and DragonFly_RELEASE_X_Y_Slip.  Now, I understand that the relation
is just the same as the relation between HEAD and PREVIEW.

So, what I should do would be to commit the patch for FreeBSD-SA-05:15.tcp
to DragonFly_RELEASE_1_2 on September 5th or later.  But, unfortunatelly,
after the ABI change, I do not have a DragonFly_RELEASE_1_2-compatible box
with me.  What I can do is to check whether the kernel with the patch can
be built without errors or significant warnings.  I belive it will work,
but I cannot test whether it really prevents attacks.
What should I do? ...

Regards,
Noritoshi Demizu