Re: cover letter not considered Fwd: [GIT PULL 0/7] KVM: s390: Fixes for 7.2

On Mon, Aug 03, 2026 at 09:34:37AM +0200, Christian Borntraeger wrote:
> Am 31.07.26 um 17:02 schrieb Derek Barbosa:
> > Hi Christian, sorry for the delay
> > 
> > On Thu, Jul 30, 2026 at 06:42:10PM +0200, Christian Borntraeger wrote:
> > > Folks,
> > > 
> > > I am a bit puzzled. Looking at https://sashiko.dev/#/?list=org.kernel.vger.linux-s390
> > > I do see  1,2,3,4,5,6,7 but not 0 for the PULL request above
> > >   Clicking on on of those mails then also has the 0 mail so its there
> > > https://sashiko.dev/#/patchset/20260727095720.9150-2-borntraeger%40linux.ibm.com
> > > 
> > > but during review this seems to have been lost. This is unfortunate as the cover letter might
> > > help to find the right tree and with 3 failed applies several already fixed
> > > things (in those 3 patches) are reported instead.
> > 
> > IIUC in the current code, cover letters are not considered patches in
> > the DB and therfore do not get forwarded to the reviewer processes.
> > 
> > I have been working on improving the concept of a "patch series summary" that
> > propagates a small snippet of "context" to all reviewer threads. Currently, the
> > individual worker threads only get a range from base..HEAD and use git log
> > --format=%s to get insight into other patches in a series.
> > 
> > So the cover letter never really makes it's way into the "working context" of a
> > single worker in isolation.
> > 
> > Is your concern not being able to quickly view the '0'-indexed mail? or that it
> > may help with review accuracy if it were to make it's way into the review
> > process?
> My concern is more that we get reports about issues that are fixed in a subsequent patch.

Got it. I've had something sitting on a local branch for some time. I will see
if I can push it up when Roman returns. [0]

> But maybe this is just due to random email delay that sashiko receives some mails later?
> Maybe we could delay review for 5 minutes or so if this seems to be part of a series.
> But I think we already detect that?
> 
I don't think that is the issue. I could be wrong.

[0] https://github.com/sashiko-dev/sashiko/issues/392

-- 
Derek <[email protected]>