[PATCH 00/10] GPST patches, trying again

[David Woodhouse]

On Tue, 2018-02-27 at 13:27 +0200, Daniel Lenski wrote:
> On Tue, Feb 27, 2018 at 10:29 AM, David Woodhouse  wrote:
> > and then are we going to refactor them into a more readable patch
> > series for review?
>
> The problem is that at this point the GP patches are interleaved with
> non-GP patches. Do you want me to refactor *all* the changes to add GP
> support into a more readable patch series on top of your master
> branch?

Are they? I thought I'd already picked out most of the generic
"preparation" patches, and what's left is fairly much all the new GP
code:

$ git diff master...gpst --stat
 Makefile.am             |    5 +-
 auth-globalprotect.c    |  464 ++++++++++++++++++++++++++++++++++++++
 esp.c                   |  118 +++++++++-
 gpst.c                  | 1146 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 hipreport.sh            |  185 ++++++++++++++++
 http.c                  |   22 +-
 library.c               |   18 ++
 main.c                  |    4 +-
 mainloop.c              |    2 +-
 openconnect-internal.h  |   24 +-
 openconnect.8.in        |   55 +++--
 www/Makefile.am         |    4 +-
 www/features.xml        |    2 +-
 www/globalprotect.xml   |   82 +++++++
 www/hip.xml             |   89 ++++++++
 www/mail.xml            |    4 +-
 www/menu2-protocols.xml |    1 +
 17 files changed, 2186 insertions(+), 39 deletions(-)

The bits in esp.c can live elsewhere now too; see commit 69226f400 for
inspiration.

I think at first glance, most of the commits from master..gpst can
probably be folded back into the first three commits?

> > Do you think it's going to pass review now? :)
>
> Well… I've gone through and tried to fix memory handling bugs, dodgy
> string handling, add comments, standardize the code formatting, ensure
> that it will work with the NM plugin… and generally take a critical
> eye towards the code style. I'm not sure what else, if anything, you
> think needs another review but I'll try to address anything that comes
> up promptly.

Sounds good. So at this point I just need to be pointed at a series of
commits that I can read like a story, and review as I go.

Maybe the answer is that there *isn't* a way to do that, and I just
have to bite the bullet and read it one C file at a time. But I think
we can at least manage to collapse it into those three (or four, if you
count the user/pass strncmp one) initial commits. 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 5213 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/openconnect-devel/attachments/20180227/aff71855/attachment-0001.bin>