On 01/10/15 03:45 PM, Fabio Estevam wrote:
Hi Sinan,
On Thu, Oct 1, 2015 at 4:39 PM, Sinan Akman sinan@writeme.com wrote:
Hi Fabio, I just wanted to point out that with this revert we don't only break imx again
We are not breaking imx by doing the revert. The reset still works. 623d96e89aca64c2 appeared only in 2015.10-rc4.
(whatever the initial bug was for this commit) but also we are having ls1021atwr working accidentally, just because writew(WCR_WDE, &wdog->wcr) clears SRS bit which is the only requirement for reset if watchdog is not running.
I don't have any strong opinion on this but i just wanted to make it clear that we are leaving both imx6 and ls1021atwr not properly implemented.
I think it is the best/safest we can do at -rc4 to avoid the reset regression on ls1021.
Then a proper implementation can be done for 2016.01.
Do you agree?
Hi Fabio, yes this seems to be the best thing to do for now. Let's implement then this thing properly soon after.
Thanks Sinan Akman