Re: [PATCH 5.16 138/200] mptcp: keep track of local endpoint still available for each msk

[Greg Kroah-Hartman]

On Tue, Feb 01, 2022 at 02:23:55PM -0800, Mat Martineau wrote:
> On Tue, 1 Feb 2022, Greg Kroah-Hartman wrote:
> 
> > On Tue, Feb 01, 2022 at 03:30:02PM +0100, Greg Kroah-Hartman wrote:
> > > On Mon, Jan 31, 2022 at 11:36:32AM -0800, Mat Martineau wrote:
> > > > On Mon, 31 Jan 2022, Greg Kroah-Hartman wrote:
> > > > 
> > > > > From: Paolo Abeni <pabeni@xxxxxxxxxx>
> > > > > 
> > > > > [ Upstream commit 86e39e04482b0aadf3ee3ed5fcf2d63816559d36 ]
> > > > 
> > > > Hi Greg -
> > > > 
> > > > Please drop this from the stable queue for both 5.16 and 5.15. It wasn't
> > > > intended for backporting and we haven't checked for dependencies with other
> > > > changes in this part of MPTCP subsystem.
> > > > 
> > > > In the mptcp tree we make sure to add Fixes: tags on every patch we think is
> > > > eligible for the -stable tree. I know you're sifting through a lot of
> > > > patches from subsystems that end up with important fixes arriving from the
> > > > "-next" branches, and it seems like the scripts scooped up several MPTCP
> > > > patches this time around that don't meet the -stable rules.
> > > 
> > > I think these were needed due to 8e9eacad7ec7 ("mptcp: fix msk traversal
> > > in mptcp_nl_cmd_set_flags()") which you did tag with a "Fixes:" tag,
> > > right?  Without these other commits, that one does not apply and it
> > > looks like it should go into 5.15.y and 5.16.y.
> > > 
> 
> Ok, I see what happened there (there's a new variable present in 5.17 that's
> not relevant for the earlier kernels). 8e9eacad7ec7 needs manual
> backporting, I'll work on that and send to the stable list.
> 
> > > Note, just putting a "Fixes:" tag does not guarantee if a commit
> > > will go into a stable tree.  Please use the correct
> > > "Cc:stable@xxxxxxxxxxxxxxx" tag as the documentation asks for.
> > > 
> 
> I will make sure to use the Cc:stable@xxxxxxxxxxxxxxx tag in the future.
> 
> 
> > > I'll go drop all of these mptcp patches, including 8e9eacad7ec7 ("mptcp:
> > > fix msk traversal in mptcp_nl_cmd_set_flags()") for now.  If you want
> > > them added back in, please let us know.
> > 
> > To be specific, I have dropped the following commits from the queues
> > now, which was more than just these three that you asked:
> > 
> > 602837e8479d ("mptcp: allow changing the "backup" bit by endpoint id")
> > 59060a47ca50 ("mptcp: clean up harmless false expressions")
> > 86e39e04482b ("mptcp: keep track of local endpoint still available for each msk")
> > 8e9eacad7ec7 ("mptcp: fix msk traversal in mptcp_nl_cmd_set_flags()")
> > a4c0214fbee9 ("mptcp: fix removing ids bitmap setting")
> > 9846921dba49 ("selftests: mptcp: fix ipv6 routing setup")
> > 
> > If you want them back, please let us know.
> > 
> 
> Of the above, there's only one suitable for stable as-is:
> 
> 9846921dba49 ("selftests: mptcp: fix ipv6 routing setup")
> 
> Please add that one to the 5.16 and 5.15 queues.

Now queued up, thanks.

greg k-h