Re: Fwd: [tip:x86/urgent] x86/platform/UV: Fix critical UV MMR address error

[Greg KH <gregkh@linuxfoundation.org>]

On Mon, Apr 09, 2018 at 03:48:58PM -0700, Mike Travis wrote:
> 
> 
> On 4/9/2018 1:03 PM, Greg KH wrote:
> > On Mon, Apr 09, 2018 at 10:57:56AM -0700, Mike Travis wrote:
> > > This was sent to stable@vger.kernel.org almost 2 weeks ago and I haven't
> > > heard anything back yet.  Is there any update on the status?  (sorry for the
> > > bad formatting, this is the only way I can forward the email.  I can only
> > > resend the original email before the commit email from Thomas.)
> > 
> > How did you send it?
> > 
> > Ugh, you bounced it?  No wonder I didn't do anything with it at all, it
> > looked exactly like a "this is being committed to the TIP tree" type
> > message.
> > 
> > How was I supposed to know what to do with it?
> > 
> > Please be more specific next time, read:
> >      https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
> > for how to do this properly.
> > 
> > A simple "please commit git id XXXXX to the X.X tree" is all that is
> > needed.
> 
> I'll definitely do this in the future.
> 
> For my info which "X.X" tree is significant?  In this case I'm guessing that
> it would be the version where the breakage occurred?  Or is it some other
> specific major or minor version change?

That's up to you to decide looking at where the breakage happened.  If
you do not know this, then asking for a patch to be backported is a bit
odd :)

> The end goal is for the distros to pull them in and they usually monitor
> certain stable trees for new commits for their own releases.  But they
> generally don't share with us those details.

The "end goal" for the stable list is to get patches into the stable
kernel trees.  If you want to deal with the distros, work with the
distros directly.

Remember, it's guessed that only about 1/3 of the systems out there are
actually running the "enterprise distros", the other 2/3 are either
running kernel.org kernels or Debian.  So if you want the majority of
the systems to have your patches, please just worry about the stable
kernels.

> > Wait, in looking closer at this:
> > 
> > > -------- Forwarded Message --------
> > > Subject: [tip:x86/urgent] x86/platform/UV: Fix critical UV MMR address error
> > > Date: Wed, 28 Mar 2018 11:25:04 -0700
> > > From: tip-bot for mike.travis@hpe.com <tipbot@zytor.com>
> > > Reply-To: andrew.banman@hpe.com, hpa@zytor.com, tglx@linutronix.de,
> > > mingo@kernel.org, russ.anderson@hpe.com, travis@sgi.com,
> > > linux-kernel@vger.kernel.org, mike.travis@hpe.com, dimitri.sivanich@hpe.com
> > > To: linux-tip-commits@vger.kernel.org
> > > CC: dimitri.sivanich@hpe.com, mike.travis@hpe.com,
> > > linux-kernel@vger.kernel.org, mingo@kernel.org, hpa@zytor.com,
> > > tglx@linutronix.de, andrew.banman@hpe.com, russ.anderson@hpe.com,
> > > travis@sgi.com
> > > 
> > > Commit-ID:  bd47a85acd727e27b7283daff557865ad04c59f6
> > 
> > Ok, that commit showed up in 4.16:
> > 	$ git describe --contains bd47a85acd727e27b7283daff557865ad04c59f6
> > 	v4.16~2^2
> > 
> > But the commit this fixes:
> > 
> > > Gitweb: https://git.kernel.org/tip/bd47a85acd727e27b7283daff557865ad04c59f6
> > > Author:     mike.travis@hpe.com <mike.travis@hpe.com>
> > > AuthorDate: Wed, 28 Mar 2018 12:40:11 -0500
> > > Committer:  Thomas Gleixner <tglx@linutronix.de>
> > > CommitDate: Wed, 28 Mar 2018 20:19:45 +0200
> > > 
> > > x86/platform/UV: Fix critical UV MMR address error
> > > 
> > > A critical error was found testing the fixed UV4 HUB in that an MMR address
> > > was found to be incorrect.  This causes the virtual address space for
> > > accessing the MMIOH1 region to be allocated with the incorrect size.
> > > 
> > > Fixes: 673aa20c55a1 ("x86/platform/UV: Update uv_mmrs.h to prepare for UV4A
> > > fixes")
> > 
> > Is in 4.16-rc1:
> > 	$ git describe --contains 673aa20c55a1
> > 	v4.16-rc1~174^2~9
> > 
> > And is not backported to any other stable tree that I see:
> > 	$ fr 673aa20c55a1
> > 	4.16
> > 	$ alias fr
> > 	alias fr='~/linux/scripts/fix_in_what_release'
> > 
> > So what exactly am I supposed to do with this commit?  What do you want
> > to see done with it?
> > 
> > totally confused,
> > 
> > greg k-h
> > 
> 
> I am too.  The earlier "breakage" was committed on Jan 8th and has been
> pulled into earlier releases by the distros.

Again, I don't care about the "distros", that's not what you are asking
me to do here.

> It was the initial patch for enabling UV4A HUB systems, so it was not
> a fix to software, but an update in response to fixed hardware (to
> allow NVDIMMS and Icelake CPUs).
> 
> Should this have been pushed to "stable" as well?

I don't know, does it match the rules for stable kernels that you have
already read now?

Again, does this need to go into any released stable kernel?  If so,
which ones.  If not, there's nothing I can do here.

thanks,

greg k-h