Sorry for a delay. I've run my test with 16 messages per channel (can0, can1 concurrently) with a cycle time 2.8ms, 1Mbps. And it works now properly.

Thank you all participated on this issue

Richard

-----Urspr?ngliche Nachricht-----
Von: Marc Kleine-Budde [mailto:mkl at pengutronix.de] 
Gesendet: Freitag, 17. Juni 2016 11:04
An: tthayer at opensource.altera.com; socketcan at hartkopp.net; wg at grandegger.com; Richard Andrysek [off-list ref]
Cc: anilkumar at ti.com; linux-can at vger.kernel.org; linux-arm-kernel at lists.infradead.org; tthayer.linux at gmail.com
Betreff: Re: [RFC] can: c_can: Update D_CAN TX and RX functions to 32 bit.

On 06/16/2016 06:10 PM, tthayer at opensource.altera.com wrote:

From: Thor Thayer <redacted>

When testing CAN write floods on Altera's CycloneV, the first 2 bytes 
are sometimes 0x00, 0x00 or corrupted instead of the values sent. Also 
observed bytes 4 & 5 were corrupted in some cases.

The D_CAN Data registers are 32 bits and changing from
16 bit writes to 32 bit writes fixes the problem.

Testing performed on Altera CycloneV (D_CAN).
Requesting tests on other C_CAN & D_CAN platforms.

Reported-by: Richard Andrysek <redacted>
Signed-off-by: Thor Thayer <redacted>

Applied to can with stable on Cc.

Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |