On 7/29/19 4:17 PM, Mark Balantzyan wrote:
---------- Forwarded message ---------- Hi all, sorry for the duplicate message Guenter, wanted to be sure my message is transferred: Thank you for your reply, Guenter! Sorry there were issues applying the patch, I used git format-patch to produce the patch and pasted the main contents into a plaintext email client so I thought it would work.. May I please request clarification on which functions are no longer needed?
All functions dealing with file accesses directly. open, read, write, ioctl. Don't you see that when you compile the code ?
Sorry about forgetting about that last misc_deregister(). Will do more tests, if that’s alright with you. In effect, may it be best to start the watchdog from the “init” function?
The watchdog should only be started from its start function, when it is opened. There are alternatives, such as informing the watchdog core that the watchdog has been started even though it is not closed. I am repeating myself here, since I don't recall an answer: Do you have the hardware ? If not, it does not make sense to continue this work; it is too risky to make all those changes without testing them on real hardware (or, at the very least, with qemu, but qemu doesn't support this chip). Thanks, Guenter
