Re: Created files in patch comment intended?

[Mauro Carvalho Chehab <mchehab@infradead.org>]

Guennadi Liakhovetski wrote:
> On Fri, 18 Dec 2009, Mauro Carvalho Chehab wrote:
> 
>> Hi Guennadi,
>>
>> Guennadi Liakhovetski wrote:
>>> Hi Mauro
>>>
>>> Looking at how my mediabus patches got committed into the mainline, I 
>>> noticed, that the add-mediabus patch contains a list of added files 
>>> between the patch description and the Sob's:
>>>
>>>      create mode 100644 drivers/media/video/soc_mediabus.c
>>>      create mode 100644 include/media/soc_mediabus.h
>>>      create mode 100644 include/media/v4l2-mediabus.h
>>>
>>> Is this intended, and if yes - why? If not, maybe you'd like to fix this 
>>> in your hg-git export scripts.
>>>
>> No, this is not intentional. The scripts have a logic to identify the description
>> body of a mercurial commit and of a patch received by email. The logic should
>> just import whatever description is provided on -hg.
>>
>> By looking on your commit for this patch on mercurial, we have:
>>
>> $ hg log -r 13658 -v
>> changeset:   13658:2c60bd900a7a
>> user:        Guennadi Liakhovetski <g.liakhovetski@gmx.de>
>> date:        Fri Dec 11 15:41:28 2009 +0100
>> files:       linux/drivers/media/video/Makefile linux/drivers/media/video/soc_mediabus.c linux/include/media/soc_mediabus.h linux/include/media/v4l2-mediabus.h linux/include/media/v4l2-subdev.h
>> description:
>> v4l: add a media-bus API for configuring v4l2 subdev pixel and frame formats
>> From: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
>>
>> Video subdevices, like cameras, decoders, connect to video bridges over
>> specialised busses. Data is being transferred over these busses in various
>> formats, which only loosely correspond to fourcc codes, describing how video
>> data is stored in RAM. This is not a one-to-one correspondence, therefore we
>> cannot use fourcc codes to configure subdevice output data formats. This patch
>> adds codes for several such on-the-bus formats and an API, similar to the
>> familiar .s_fmt(), .g_fmt(), .try_fmt(), .enum_fmt() API for configuring those
>> codes. After all users of the old API in struct v4l2_subdev_video_ops are
>> converted, it will be removed. Also add helper routines to support generic
>> pass-through mode for the soc-camera framework.
>>
>> Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
>> Acked-by: Hans Verkuil <hverkuil@xs4all.nl>
>> ---
>>  drivers/media/video/Makefile       |    2 +-
>>  drivers/media/video/soc_mediabus.c |  157 ++++++++++++++++++++++++++++++++++++
>>  include/media/soc_mediabus.h       |   65 +++++++++++++++
>>  include/media/v4l2-mediabus.h      |   61 ++++++++++++++
>>  include/media/v4l2-subdev.h        |   19 ++++-
>>  5 files changed, 302 insertions(+), 2 deletions(-)
>>  create mode 100644 drivers/media/video/soc_mediabus.c
>>  create mode 100644 include/media/soc_mediabus.h
>>  create mode 100644 include/media/v4l2-mediabus.h
>>
>>
>> As you see, you added those comments at the end of the patch, together with a diffstat.
>> While the script has a logic to remove diffstats, it doesn't contain anything to remove
>> the "create mode" lines that you've added at the end of the patch description.
> 
> No, _I_ didn't add those, git did. This is the standard output from "git 
> format patch." And the patch format for submission to the kernel is 
> roughly
> 
> <subject>
> 
> <description>
> 
> <Sob, ack,...>
> ---
> <ignored lines>
> <diff -u>
> 
> So, anything, beginning with "---\n" and the patch must be ignored. That's 
> also where you provide any comments, that should not be included in the 
> commit text.

You added it on your -hg tree. See:

	http://linuxtv.org/hg/~gliakhovetski/v4l-dvb/rev/2c60bd900a7a

Probably, the scripts you're using to generate your -hg trees are wrong. They
should be discarding anything after "---\n", but they aren't doing it.

Mercurial never adds such comments on their descriptions. For example, this patch:

http://linuxtv.org/hg/~gliakhovetski/v4l-dvb/rev/dae3ac9449c8

Also creates a file, but you won't see there any diffstat lines on it.

Cheers,
Mauro.