Hello Aric,
Saturday, June 24, 2006, 3:59:10 PM, you wrote:
[]
> We had a discussion about this about a year ago when I put it in. The
> flash comes out of suspend unwritable, and if something tries to write
> to it before we have a chance to unlock it, chaos ensues. We tried this
> elsewhere, but it seemed that other code could always write before it
> was unlocked.
> But I'm not familiar with ll_pm (low level power management?)
> handlers. How do they work?
I guess they were introduced after that. pm.c:142,161. I.e. those
handlers are essentially called immediately before and after sleep.S
code. They are used in quite a few ports now to support of resume with
native bootloader (including hx4700, recently added by Matt). So, it
seems like good place to move flash unlock code, as it is guaranteed
to be called before restarting different queues and unfreezing
processes.
>>> Are there other places I have to look out for? A quick grep for _H4700
>>> only shows the old SDRAM_ROWS issue[1] in cpu-pxa.c
>>>
>>> thanks
>>> Philipp
>>>
-- Best regards, Paul mailto:pmiscml_at_gmail.comReceived on Sat Jun 24 2006 - 09:50:19 EDT
This archive was generated by hypermail 2.2.0 : Sat Jun 24 2006 - 12:29:27 EDT