[ath9k-devel] [PATCH v2 0/6] ath10k: pci cleanup

[Kalle Valo <kvalo@qca.qualcomm.com>]

Michal Kazior <michal.kazior@tieto.com> writes:

> On 25/04/13 09:56, Kalle Valo wrote:
>> Michal Kazior <michal.kazior@tieto.com> writes:
>>
>>>> Applying: ath10k: remove pci probe retrying
>>>> fatal: sha1 information is lacking or useless (drivers/net/wireless/ath/ath10k/pci.c).
>>>> Repository lacks necessary blobs to fall back on 3-way merge.
>>>> Cannot fall back to three-way merge.
>>>> Patch failed at 0003 ath10k: remove pci probe retrying
>>>>
>>>> I didn't want to start manually fixing the conflict, so can you please
>>>> rebase? And please double check how you submit patches, git still
>>>> complains about invalid ids. Of course it might be something on my end,
>>>> but your patches are the only ones with which I have noticed this
>>>> problem.
>>>
>>> I've made sure to send the v2 with no private patches around. When
>>> rebasing I had to resolve the conflict on this patch myself too. It
>>> was in conflict with the hif_type variable removal.
>>>
>>> I use this command:
>>>
>>> 	git send-email --annotate --cover-letter [--subject-prefix=..]
>>> [--in-reply-to=..] commit_id..commit_id
>>
>> Ok, so you don't use 'git format-patch' at all? I wonder if that makes
>> the difference here. I would assume not.
>>
>>> I also rebase against your github/master branch.
>>
>> What commands do you use to rebase? Have you verified that there are no
>> private commits, eg. merges?
>
> ; git rebase github/master
> (github is my remote name for your github ath10k repository)
>
> If I have some private commits around I do a temporary branch, rebase
> interactively and drop those commits.
> ; git rebase -i github/master

Looks good to me.

> I then do `git send-email` with `github/master..` so I'd probably see
> more patches than I should (after being dropped into editor for
> annotate/cover letter) if there were private commits still around.

Yeah.

This is so weird as with Bartosz' patches I don't have any problems. For
example the last patchset I applied had three patches conflicting and
3-way was always used.

-- 
Kalle Valo