On 6/6/07, David Waybright <david_at_iximd.com> wrote:
> Most of the time the theory was the hard reset wouldn't work and you had to let
> the PDA sit and drain it's battery for like a week before powering it up
> again.
>
> I think your right about the 16K, what happens is the first 16K (or so) of ROM
> is wrote to RAM then kicks over to that RAM copy to boot up far enough to
> continue off of ROM. I thought the 16K corruption was already fixed for the
> suspend to work. If not I think we need to try and fix it so the power
> mangement (off & suspend) work better, possibly by telling either Haret, LAB,
> Linux or the like NOT to touch that portion of RAM.
Suspend/resume/power management has worked fine for a long time, at
least on older kernels, by patching the bootloader in SRAM to work
with Linux. The problem happens when something (anything in the
kernel) goes awry and scribbles on the SRAM.
Likely the suspend/resume problems with HEAD kernels is due to
hamcop_base not being converted to a platform driver yet.
Matt
Received on Wed Jun 06 2007 - 14:12:04 EDT
This archive was generated by hypermail 2.2.0 : Wed Jun 06 2007 - 14:12:34 EDT