Date:   Sat, 11 Jul 2020 05:55:30 +0900
From:   Stafford Horne <shorne@...il.com>
To:     "Alexander A. Klimov" <grandmaster@...klimov.de>
Cc:     Greg KH <gregkh@...uxfoundation.org>, stern@...land.harvard.edu,
        linux-usb@...r.kernel.org, usb-storage@...ts.one-eyed-alien.net,
        linux-kernel@...r.kernel.org, Jonathan Corbet <corbet@....net>,
        David Miller <davem@...emloft.net>,
        Linus Torvalds <torvalds@...ux-foundation.org>
Subject: Re: [PATCH] Replace HTTP links with HTTPS ones: USB MASS STORAGE
 DRIVER

On Fri, Jul 10, 2020 at 09:36:03PM +0200, Alexander A. Klimov wrote:
> 
> 
> Am 10.07.20 um 12:36 schrieb Stafford Horne:
> > On Thu, Jul 09, 2020 at 08:14:09AM +0200, Greg KH wrote:
> > > On Wed, Jul 08, 2020 at 08:41:54PM +0200, Alexander A. Klimov wrote:
> > > > 
> > > > 
> > > > Am 08.07.20 um 12:39 schrieb Greg KH:
> > > > > On Wed, Jul 08, 2020 at 11:55:00AM +0200, Alexander A. Klimov wrote:
> > > > > > Rationale:
> > > > > > Reduces attack surface on kernel devs opening the links for MITM
> > > > > > as HTTPS traffic is much harder to manipulate.
> > > > > > 
> > > > > > Deterministic algorithm:
> > > > > > For each file:
> > > > > >     If not .svg:
> > > > > >       For each line:
> > > > > >         If doesn't contain `\bxmlns\b`:
> > > > > >           For each link, `\bhttp://[^# \t\r\n]*(?:\w|/)`:
> > > > > > 	  If neither `\bgnu\.org/license`, nor `\bmozilla\.org/MPL\b`:
> > > > > >               If both the HTTP and HTTPS versions
> > > > > >               return 200 OK and serve the same content:
> > > > > >                 Replace HTTP with HTTPS.
> > > > > > 
> > > > > > Signed-off-by: Alexander A. Klimov <grandmaster@...klimov.de>
> > > > > 
> > > > > Your subject lines are very odd compared to all patches for this
> > > > > subsystem, as well as all other kernel subsystems.  Any reason you are
> > > > > doing it this way and not the normal and standard method of:
> > > > > 	USB: storage: replace http links with https
> > > > > 
> > > > > That would look more uniform as well as not shout at anyone.
> > 
> > I would agree.  The OpenRISC patch for this series says:
> >    "OPENRISC ARCHITECTURE:..."
> > 
> > Here it would just be "openrisc:..." I think fixing the whole series is needed.
> > Greg is not the only on complaining.
> > 
> > Ideally, I think, it would be good to have this sent out as a series i.e [PATCH 3/55]
> > rather than individual patches so this could be discussed as a whole.
> 1) To who? As right now? As right now plus Torvalds, KH, Miller, etc.?
>    As right now, but all-to-all?

Make sure you have a cover letter explaining what you expect.

You can ask maintainers to pick up individual patches by mentioning that in the
cover letter.

You can use `git send-email --cc-cmd` so each patch goes only to the
maintainers, for example:

  send-email --to linux-kernel@...r.kernel.org --cc-cmd scripts/get_maintainers.pl`

> 2) Apropos "series" and "as whole"... I stumbled over

I stumble over "apropos". :)

>    `git log --oneline |grep -Fwe treewide`
>    and am wondering:
>    *Shouldn't all of these patches even begin with "treewide: "?*
>    E.g.: "treewide: Replace HTTP links with HTTPS ones: GCC PLUGINS"

As Greg said that is not what patch subjects loo like.

  - GCC PLUGINS: is not correct, remove it.
  - treewide: may work, but as you want individual maintainers to pick up the patches put
    a subsystem in the subject as maintainers like.
  - The rest of the text should be lowercase "replace http links with https"

Have a look at other patch subject lines based on the file you are editing.  For example:

  $ git log --oneline -- Documentation/kbuild/gcc-plugins.rst
  2020-03-10 2b4cbd5c9505 Jonathan Corbet  docs: move gcc-plugins to the kbuild manual

  $ git log --oneline -- scripts/Makefile.gcc-plugins
  2019-03-04 81a56f6dcd20 Kees Cook        gcc-plugins: structleak: Generalize to all variable types
  2018-12-29 668c35f69cc7 Linus Torvalds   Merge tag 'kbuild-v4.21' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild
  2018-12-12 189af4657186 Ard Biesheuvel   ARM: smp: add support for per-task stack canaries
  2018-12-01 ce2fd53a10c7 Masahiro Yamada  kbuild: descend into scripts/gcc-plugins/ via scripts/Makefile
  2018-09-04 10e9ae9fabaf Alexander Popov  gcc-plugins: Add STACKLEAK plugin for tracking the kernel stack
  2018-07-24 7ccb95e8fe91 Kees Cook        gcc-plugins: Regularize Makefile.gcc-plugins
  2018-07-02 c17d6179ad5a Masahiro Yamada  gcc-plugins: remove unused GCC_PLUGIN_SUBDIR
  2018-06-11 59f53855babf Masahiro Yamada  gcc-plugins: test plugin support in Kconfig and clean up Makefile
  2018-06-11 8034c2fb1225 Masahiro Yamada  gcc-plugins: move GCC version check for PowerPC to Kconfig
  2018-06-11 5aadfdeb8de0 Masahiro Yamada  kcov: test compiler capability in Kconfig and correct dependency


So you will have:

  docs: replace http links with https
  gcc-plugins: replace http links with https

-Stafford

> > 
> > -Stafford
> > 
> > > > > thanks,
> > > > > 
> > > > > greg k-h
> > > > > 
> > > > Hi,
> > > > 
> > > > I'm very sorry.
> > > > 
> > > > As Torvalds has merged 93431e0607e5 and many of you devs (including big
> > > > maintainers like David Miller) just applied this stuff, I assumed that's OK.
> > > > 
> > > > And now I've rolled out tens of patches via shell loop... *sigh*
> > > > 
> > > > As this is the third (I think) change request like this, I assume this rule
> > > > applies to all subsystems – right?
> > > 
> > > Yes, you should try to emulate what the subsystem does, look at other
> > > patches for the same files, but the format I suggested is almost always
> > > the correct one.  If not, I'm sure maintainers will be glad to tell you
> > > otherwise :)
> > 
> > 

Powered by blists - more mailing lists