Re: PATCH: ecng for 6.x and 7.x

William wrote:
> On 9/11/07, *Nate Lawson* wrote:
> 
>     William Grzybowski wrote:
>     > I tested this patch on my acer notebook (intel chipset) and i did not
>     > notice any changes, unless some errors on dmesg, like:
>     > acpi_ec0: EcCommand: no response to 0x84
>     > acpi_ec0: GPE query failed: AE_NO_HARDWARE_RESPONSE
>     > acpi_ec0: EcCommand: no response to 0x82
>     > acpi_ec0: EcCommand: no response to 0x80
>     > ACPI Error (psparse-0626): Method parse/execution failed
>     > [\\_TZ_.THRM._TMP] (Node 0xc3bbdcc0), AE_NO_HARDWARE_RESPONSE
>     > ACPI Error (psparse-0626): Method parse/execution failed
>     > [\\_SB_.ACAD._PSR] (Node 0xc3bc02a0), AE_NO_HARDWARE_RESPONSE
> 
>     As I noted before, your system enters the poll loop with the status
>     appearing to be already complete.  Can you get back to me on my previous
>     questions, especially whether forcing polled mode works for you?  I
>     didn't see any errors in that dmesg case.
> 
>     I've updated the patches to do one final check if the interrupt-driven
>     mode gets a timeout.  If the status is complete, it will force the
>     system back into polled mode since interrupt mode doesn't work.  It also
>     has a case for polled mode during boot where the status appears to be
>     already complete.  It waits a short while before actually checking the
>     status, just in case the EC is really slow and hasn't gotten to work on
>     the new request yet.
> 
>     Give it a try also, with no tunables set. 
>
> Nate,
> 
> Yesterday I send to you privately the answers which you asked for, annoy
> me if you didn't receive...
> Today I recompiled the kernel from a today's cvs without modules with
> KTR and recompiled the acpi module without any patches,
> without any debug option on boot, the system can initialize the battery
> and the thermal, when i enable any debug (polled or burst or even both)
> the battery get ready after 2 tries.
> 
> Tonight I will remake some testes with your patches to make sure about
> things which I told you.

There was one other change needed.  I was too aggressive in trying to
hold the lock over the _Qxx method evaluation.  Some BIOSes apparently
try to read/write from EC space in that method.  So I reverted that part
of the patch and added a comment explaining it to future maintainers.

Attached are revised patches, version D.  Please try them without any
debug.acpi.ec tunables set in loader.conf.  If your battery status fails
with this patch, add:
    debug.acpi.ec.polled="1"

You should also keep jkim's OsdSynch patch in your tree.  It fixes
different problems but ones that are needed to test my EC patches.

Thanks,
Nate