[U-Boot] [PATCH] patman: Don't request full names from get_maintainer

Simon Glass sjg at chromium.org
Thu May 22 03:01:42 CEST 2014


Hi Doug,

On 20 May 2014 12:51, Doug Anderson <dianders at chromium.org> wrote:
> Simon,
>
> On Tue, May 13, 2014 at 11:17 AM, Simon Glass <sjg at chromium.org> wrote:
>> Hi Doug,
>>
>> On 18 April 2014 15:32, Doug Anderson <dianders at chromium.org> wrote:
>>> Simon,
>>>
>>> On Fri, Apr 18, 2014 at 1:43 PM, Simon Glass <sjg at chromium.org> wrote:
>>>> Hi Doug,
>>>>
>>>> On 17 April 2014 12:47, Doug Anderson <dianders at chromium.org> wrote:
>>>>>
>>>>> The Linux get_maintainer.pl can often produce a whole lot of results.
>>>>> As a result you'll sometimes blow your CC field over 1024 characters
>>>>> and that can cause listservs to reject your message.
>>>>>
>>>>> As a stopgap, call get_maintainer.pl with "--non" so it doesn't
>>>>> include real names.  This will dramatically reduce the number of
>>>>> characters.
>>>>>
>>>>> Signed-off-by: Doug Anderson <dianders at chromium.org>
>>>>> ---
>>>>>
>>>>>  tools/patman/get_maintainer.py | 2 +-
>>>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>>>>
>>>>> diff --git a/tools/patman/get_maintainer.py b/tools/patman/get_maintainer.py
>>>>> index 00b4939..a5160bc 100644
>>>>> --- a/tools/patman/get_maintainer.py
>>>>> +++ b/tools/patman/get_maintainer.py
>>>>> @@ -43,5 +43,5 @@ def GetMaintainer(fname, verbose=False):
>>>>>              print "WARNING: Couldn't find get_maintainer.pl"
>>>>>          return []
>>>>>
>>>>> -    stdout = command.Output(get_maintainer, '--norolestats', fname)
>>>>> +    stdout = command.Output(get_maintainer, '--norolestats', '--non', fname)
>>>>>      return stdout.splitlines()
>>>>> --
>>>>> 1.9.1.423.g4596e3a
>>>>>
>>>>
>>>> Good to avoid this problem, but I wonder if we should check the size
>>>> and re-run the command if too long? That way we can keep names on the
>>>> thread when it is possible.
>>>>
>>>> Also, why is there a limit on CC - is that a limitation described in
>>>> the RFC, or just certain mailers? Can we get around it by specifying
>>>> multiple Cc tags?
>>>
>>> I believe that we're actually fighting a heuristic that the mailing
>>> list servers have to avoid spammers.  In the past I've seen my mail
>>> get through to list servers that weren't run by kernel.org but _not_
>>> the the ones at kernel.org.  Kind of like how if you want to write
>>> code that's for generic exynos 5 hardware you need to use exynos_5yyy
>>> and _not_ replace the y with x (because that would be adult content!)
>>>
>>> We could possibly only run the heuristic if the CC list was > 800
>>> characters?  I'd hesitate to getting too close to 1024 since I think
>>> there are instances where git will itself add a CC and that might blow
>>> us out.
>>
>> Yes that seems reasonable - that way we get the names most of the time.
>>
>>>
>>> ...another option is to just forget this patch and force people to add
>>> this to ".get_maintainer.conf"
>>
>> I think the previous option is better.
>
> I started coding this up and then realized one corner case that
> wouldn't be handled: the cover letter case.  This is the most likely
> message to go over the limit, even if individual patches are not over
> the limit.  That's because we add all addresses to the cover letter in
> (3118725 patman: Add all CC addresses to the cover letter).
>
> If we want to try to handle this dynamically, we probably need to
> handle it at a higher level and manually strip off the full names
> right before passing it off to git.
>
> What do you think?

It looks like it should be doable in series.MakeCcFile unless I am
misunderstanding the problem. It creates separate CC lists for each
commit, including the cover letter.

Regards,
Simon


More information about the U-Boot mailing list