Re: [PULL REQUEST] Please pull rdma.git

[Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>]

[-- Attachment #1: Type: text/plain, Size: 5219 bytes --]

Side note: it's very annoying to have Time Warner Cable decide to take
your internet down in the middle of writing a reply :-/

On 09/08/2015 11:33 PM, Linus Torvalds wrote:
> On Tue, Sep 8, 2015 at 8:08 PM, Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> wrote:
>>
>> With a comment that said "I can carry this merge forward, no further
>> action is necessary on your part".  That combined with my lack of deep
>> internal knowledge of what it is that Stephen is doing made me go "Ok,
>> he says don't do anything, so I won't change it."
> 
> So quite frankly, Stephen does a really good job at merging and most
> of his merges are very on point. He's been doing a lot of them as part
> of linux-next, and has seen more conflicts than just about anybody
> else.
> 
> But I think to him it's mostly just an issue of "get the right end
> result". I don't think he goes: "this merge conflict is a result of a
> breakdown of the development process".
> 
> Conversely, to me, one of the main reasons I want to do those merges
> is exactly because I think conflicts are more about the development
> process issues than about "just getting the right end result". Yes,
> obviously I want to get the rigth end result too, but I very much
> react to how/why the conflict happened in the first place. The end
> result is _almost_ secondary, although 99% of the time the primary
> issue doesn't really even raise its head.
> 
> So I'm upset not because the conflict is hard to resolve (it isn't),
> but because I feel this was really badly handled.
> 
> Yes, the fact that Mellanox people sent two different patches to two
> different maintainers that did the same thing in two different ways is
> odd. Matan and Jiri are cc'd, and I think that whole thing just smells
> really bad.
> 
> But at the same time, I expect more of maintainers, and I don't see
> any sign that David and the networking people were notified about the
> _other_ patch to _their_ subsystem.

Me telling Matan to Cc: netdev on patches related to their subsystem:
http://marc.info/?l=linux-rdma&m=143398479529819&w=4

Above was v5 of the patchset.  Both the v6 and v7 of the patchset had
netdev Cc:ed on the cover letter and at least the first three patches.

Here's the v7 cover and the first three patches in the netdev archives:
http://marc.info/?l=linux-netdev&m=143827046913908&w=4
http://marc.info/?l=linux-netdev&m=143827047413915&w=4
http://marc.info/?l=linux-netdev&m=143827052913936&w=4
http://marc.info/?l=linux-netdev&m=143827055013953&w=4

Here I am saying I've pulled the patchset in in the netdev archives:
http://marc.info/?l=linux-netdev&m=143834704701705&w=4

Here Jason Gunthorpe is asking me if I was going to take netdev stuff
without an ack from netdev in the netdev archives:
http://marc.info/?l=linux-netdev&m=143836033706501&w=4

And here is my response to Jason, again, in the netdev archives:
http://marc.info/?l=linux-netdev&m=143836450808174&w=4

Here is the initial notification from Stephen on the merge conflict:
http://marc.info/?l=linux-next&m=144072523831647&w=4

Here's is Jiri's response (on linux-netdev, but according to my records
also direct Cc: Dave Miller):
http://marc.info/?l=linux-netdev&m=144074372902869&w=4

> The fact that you weren't aware of the other patch in the networking
> subsystem is kind of to be expected. You're not the network
> maintainer, so why would you? But exactly because you're not the
> networking maintainer, I would have expected you to check with him
> when you apply patches to generic networking code.

See above.  This was done.

> This time it conflicted, and I noticed, and I went "this is not how
> kernel development is supposed to go".

I didn't switch up for the newer patch, that didn't help.  But
otherwise, I did what was needed.

> But say that the other networking patch hadn't even existed: in that
> case I *still* shouldn't have gotten a patch to net/core/dev.c from
> you without any sign that David had ack'ed it (or at the very least
> been notified, even if he hadn't reacted).

And such a thing wouldn't have happened.

> See?

I see perfectly well Linus.  I am not new to engineering.  You have
stated several times that I'm missing the point, or do I see your point,
please understand mine: I *did* the things you are assuming I didn't do.
 The record is in the public archives.  I didn't make a big deal about
it because it *isn't* a big deal unless you assume that I'm running
around nilly-willy throwing shit at the wall and hoping to make a
painting.  As I knew very well that's not what I'm doing, *I* didn't
have those assumptions, and therefore I didn't think this merge fixup
was a big deal.  As the efforts to bring this to the network
maintainer's attention is in the record above, you will likewise find if
you talk to Greg K-H that I worked with him both publicly and privately
on the changes I made in the drivers/staging area.  I do now how to keep
responsible parties informed when I'm working in their domain.

-- 
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
              GPG KeyID: 0E572FDD



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 884 bytes --]