Dongchun,

On Thu, Jul 02, 2020 at 07:06:05PM +0800, Dongchun Zhu wrote:

Hi Sakari,

Sorry I just sent email using outlook where default format is HTML, now
I use evolution, one Linux mail client that I used to send upstream
patch previously.

On Thu, 2020-07-02 at 08:34 +0300, Sakari Ailus wrote:

quoted

Dongchun,

Please don't send HTML e-mail to Linux kernel related mailing lists.

On Thu, Jul 02, 2020 at 03:48:56AM +0000, Dongchun Zhu (朱东春) wrote:

quoted

 Hi Sakari,

 Sorry to bother you again, but I am so confused about the questions you raised.
 I just synced mainline: 5.8-rc3 tarball from https://www.kernel.org/, on which I ran the git am <patch> command.
 The patch-applying process shows no error.
 -----------------8<-------------------
 [mtk15013@mtkslt307 linux]$git apply --check media-i2c-Add-support-for-DW9768-VCM.patch
 [mtk15013@mtkslt307 linux]$git am media-i2c-Add-support-for-DW9768-VCM.patch
 Applying: media: dt-bindings: media: i2c: Document DW9768 bindings
 Applying: media: i2c: dw9768: Add DW9768 VCM driver
 -----------------8<-------------------

 On the other hand, I also compared dongwoon,dw9768.yaml file with other media device dt-bindings(like imx219.yaml and ov8856.yaml).
 It seems there are no apparent differences between them.
 Especially, the sentence '# SPDX-License-Identifier: (GPL-2.0 OR BSD-2-Clause)' shall be common.
 I dunno why here dongwoon,dw9768.yaml reports trailing whitespace warnings while ov8856.yaml is silent.

 For the patch failed on MAINTAINERS, I am still curious what's wrong.
 In fact, I locally have run parse-maintainers.pl script to check MAINTAINERS file before submitting patch.
 The result also reports no errors.
 -----------------8<-------------------
 [mtk15013@mtkslt307 linux]$perl scripts/parse-maintainers.pl
 [mtk15013@mtkslt307 linux]$ls
 -----------------8<-------------------

 As to Base64 encoding, I checked each patch file again. They are all encoded in UTF-8.
 As https://www.base64encode.org/ says, for an example, '77' in ASCII format would be changed to 'T' in Based64-encoded format.
 This means there shall be messy code if we adpoting Based64-encoded format.
 But I cannot see garbled messages in the current patches.

 The DW9768 serials-patch is attached.
 @Tomasz @Andy @Rob could anyone help try to see whether the patch can be cherry-picked on Linux master branch or not?
 Patchwork link:
 https://patchwork.kernel.org/cover/11633291/

Both of the patches appear to contain only ASCII characters.

I did some research on this. It seems that the base64 encoded message body
does have carriage returns, in both cases. Git am does not attempt to
remove them in that case, but Patchwork does. Hence the files are fine if
you download them from Patchwork --- where the mbox files have neither
carriage returns nor base64 encoding.

What does the file command say about the patch files produced by git
format-patch? My guess is that something in between your local computer and
LMML (and other mail servers) base64 encodes the message body. But where
are the carriage returns added? Nevertheless they seem to be added before
the base64 conversion.

Hm... I used the file command to check the diff patch generated from git
format-patch and that downloaded from Patchwork, they are both ASCII
text.

That's because Patchwork appears to be removing the carriage returns. git
does not if the content is base64 encoded.

Your e-mail setup simply appears to be broken. I'd suggest trying to send
the patches encoded in base64 as a workaround. git send-email uses
sendemail.transferEncoding configuration option for this.

-- 
Sakari Ailus