Re: [PATCH 5.4.y] virtio-net: Add validation for used length

[Greg KH <gregkh@linuxfoundation.org>]

On Wed, Mar 26, 2025 at 07:02:38AM +0530, Harshit Mogalapalli wrote:
> Hi Greg,
> 
> On 26/03/25 06:47, Greg KH wrote:
> > On Wed, Mar 26, 2025 at 06:32:19AM +0530, Harshit Mogalapalli wrote:
> > > Hi Larry,
> > > 
> > > 
> > > On 26/03/25 05:14, Larry Bassel wrote:
> > > > From: Xie Yongji <xieyongji@bytedance.com>
> > > > 
> > > > commit ad993a95c508 ("virtio-net: Add validation for used length")
> > > > 
> > > 
> > > I understand checkpatch.pl warned you, but for stable patches this should
> > > still be [ Upstream commit ad993a95c508417acdeb15244109e009e50d8758 ]
> > > 
> > > Stable maintainers, do you think it is good idea to tweak checkpatch.pl to
> > > detect these are backports(with help of Upstream commit, commit .. upstream,
> > > or cherrypicked from lines ?) and it shouldn't warn about long SHA ?
> > 
> > Nope!  Why would you ever run checkpatch on a patch that is already
> > upstream?
> > 
> 
> Ah right, not in this case but it might help when the backport is a bit
> different from the upstream patch(i.e after conflict resolution if the line
> in code exceeds 80 chars) -- checkpatch.pl might help us do it in the right
> way ? (only in a case where there are changes between current upstream code
> and the stable branch where we are backporting to)

Then run checkpatch like normal to catch that if you feel you need it,
you know to ignore foolish warnings from checkpatch, that's just normal.

People doing backports better be experienced kernel developers as this
is NOT a task for newbies for obvious reasons.  Which is maybe why no
one has ever brought this up in the past 15+ years we have had stable
kernels?  :)

thanks,

greg k-h