Re: [PATCH] Drivers: hv: vmbus: finally fix hv_need_to_signal_on_read()

From: gregkh@xxxxxxxxxxxxxxxxxxx
Date: Thu Jan 26 2017 - 05:10:00 EST


On Thu, Jan 26, 2017 at 09:31:32AM +0000, Dexuan Cui wrote:
> > From: gregkh@xxxxxxxxxxxxxxxxxxx [mailto:gregkh@xxxxxxxxxxxxxxxxxxx]
> > > > From: gregkh@xxxxxxxxxxxxxxxxxxx [mailto:gregkh@xxxxxxxxxxxxxxxxxxx]
> > > > > > > It's interesting v4.4.44 is not impacted, but actually it needs both the 2
> > > > patches:
> > > > > > > i.e. this patch, and the previous one:
> > > > > > > Commit a389fcfd2cb5 ("Drivers: hv: vmbus: Fix signaling logic in
> > > > > > hv_need_to_signal_on_read()")
> > > > > >
> > > > > > That patch does not apply to the 4.4-stable tree, which is why it was
> > > > > > not included there. If you feel it should be included, please provide a
> > > > > > backport and send it to the stable@xxxxxxxxxxxxxxx mailing list.
> > > > > >
> > > > > > greg k-h
> > > > >
> > > > > Thanks! I'll do the backport after this patch goes in the mainline.
> > > >
> > > > Why wait?
> > >
> > > I thought a patch must first appear in Linus's tree, before it can be
> > > back-ported to the stable tree?
> >
> > Commit a389fcfd2cb5 _IS_ in Linus's tree, it was released in the 4.7-rc1
> > kernel, which happened in May of 2016.
> >
> > greg k-h
>
> Sorry, it looks I didn't make it clear. :-)
>
> v4.4.44 needs 2 patches, i.e. a389fcfd2cb5, and this patch (which is not in
> Linus's tree yet). Only backporting the first patch immediately is not enough
> and is also improper IMO, because the first patch introduces a new issue,
> which is being resolved by this patch. So my understanding is that I should
> backport the 2 patches together.

Ok, that makes a bit more sense, thanks. I'll wait for your "real"
patch to hit Linus's tree then.

greg k-h