[Lenny400] Patches for linux-stable
H. Nikolaus Schaller
hns at goldelico.com
Fri Sep 1 22:55:54 CEST 2017
> Am 01.09.2017 um 22:25 schrieb Paul Boddie <paul at boddie.org.uk>:
> On Friday 1. September 2017 19.45.47 H. Nikolaus Schaller wrote:
>> Hi Paul,
>> great and that quick...
> Well, it was all sitting around, so a quick diff against the mainline
> changeset was all it took.
I also took the time to add an untested draft for:
* keyboard matrix
* touch buttons
* i2c peripherals (rtc, power controller, audio codec)
* letux_defconfig to add drivers for matrix keyboard, rtc chip, wifi chip
What may be very difficult is to get Ethernet working. The RTL8201CL
is said to have no driver...
What I think is missing in DT is:
* LCD and backlight
* detecting AC and Battery charging
* i2c and i2s drivers on SoC side
* some pinmux
* power control for WiFi on/off
But I would say we have an untested and undebugged pile of ideas which
already cover 70% of the device...
>>> Am 01.09.2017 um 19:31 schrieb Paul Boddie <paul at boddie.org.uk>:
>>> I have attached a single, combined diff for all of the changes I have
>>> done so far against linux-stable (changeset
>>> There are probably mistakes made in this code, even though it compiles,
>>> and so it isn't so interesting to provide the change history because I
>>> definitely made mistakes and forgot things as I stumbled through the
>>> process of writing it. And the objectives of writing each piece of code
>>> are pretty clear, anyway, so there's no benefit in seeing it in "slow
>>> motion" with commentary. :-)
>> No problem. I am currently working on a tool to split commit sets by file
>> name... So making multiple patches out of this single one isn't very
>> difficult (as soon as the tool works - which *is* difficult).
> It depends how you want to do it. Taking a combined diff just involves
> splitting on some recognisable separator. Indeed, I created the list of
> filenames by doing this:
> grep '^+++ b/' filename.diff | sed 's/+++ b\///'
finding what has changed is even simpler:
git diff --names-only <from> <to>
> That is a bit of a hack, but it is the consequence of using git diff, as
> opposed to something else, to get the initial diff.
Well, in this case it might be simple. My goal is to try to keep history as
good as possible... This involves mixing git diff and blame and forming a series
of new commits by git commit -C. It roughly works, but seems to loose some deletions
and file renames. So if I merge things back it still differs.
>> Here I wonder if we shouldn't have a jz4730 subdirectory?
>> But if differences are small it may be the better way.
> In the 2.6 kernel, there are jz4730, jz4740 and jz4750 directories, and you
> can imagine what other Ingenic code releases do. :-) But I think the
> differences are small enough, and I believe that the maintainers have tried to
> accommodate jz4770 and jz4780 within this jz4740-only arrangement.
Ok, if that is the "standard", we should keep it that way.
>> Everything looks marvelous!
> At the moment... :-)
Well, I did not compile and try to run and watch the smoke signs :)
>>> I will keep looking at the other areas that will need updating, but I
>>> thought that it might be useful to get some feedback now.
>> I will try to apply/merge it to the Letux tree.
>> At the moment I am trying to get my cross-compiler working. It stumbles
>> about not finding some sgidefs.h and byteswap.h. The typical include path
>> problems :)
> One thing I forgot to mention about my experiments with David's work is that I
> needed to apply changes to get my modern GCC working with the old 2.6 kernel
> code. Here, however, GCC 6.x should work just fine, and I am fortunate enough
> to be using Debian unstable which provides mipsel cross-compilers, as should
> the latest Debian stable release.
Well, I have a different setup with a multi-architecture-cross-gcc-4.9.
The ARM side works fine for years... Only the mips side comes up with such hickups.
But more time will make it working :)
More information about the Lenny400