On Wed, Apr 30, 2014 at 4:53 PM, Pavel Machek [off-list ref] wrote:

Hi!

quoted

quoted

quoted

quoted

quoted

Apparently, socfpga CAN needs 32-bit accesses and different raminit
sequence.

Signed-off-by: Pavel Machek <redacted>

...

quoted

quoted

quoted

Or ... let me upload my tree so that you can generate the diff.

You should be able to get it using:

git pull git://git.denx.de/linux-denx-pavel.git my_net_master

quoted

I also like your updates to my patch as shown below - much cleaner.

quoted

quoted

Since you based this patch on my original code, feel free to add my

signed-off.

Will do.

quoted

I pulled in your branch
(git://git.denx.de/linux-denx-pavel.gitmy_net_master) and tested it on
the Altera Development Kit. It works well -
I don't see any issues at a number of different bitrates (80K, 200K, 800K,
1Mb).

Thanks for testing. I'll also add your Tested-by: header, if you don't
mind.

Hi Pavel. That would be fine.

Thor

quoted

I didn't investigate what the issues were with applying the patch though. I
just used your branch.

The way I generated that branch... you picked up some extra code in
other areas of the kernel, but it should not affect the testing.

Thanks!
                                                                Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html