RE: [ACPI] ACPI owner_id limit too low

[Moore, Robert]

If you felt ambitious, you could take a look at making the change to not
allocate owner IDs for the static tables (tables that cannot be
unloaded). This would really take the pressure off the Owner ID.
Requires a change to the shutdown mechanism to make sure that the
namespace is completely deleted.


> -----Original Message-----
> From: Alex Williamson [mailto:alex.williamson@xxxxxx]
> Sent: Thursday, December 08, 2005 2:21 PM
> To: Moore, Robert
> Cc: Brown, Len; linux-kernel@xxxxxxxxxxxxxxx; acpi-
> devel@xxxxxxxxxxxxxxxxxxxxx
> Subject: RE: [ACPI] ACPI owner_id limit too low
> 
> On Thu, 2005-12-08 at 14:03 -0800, Moore, Robert wrote:
> > We have increased the number of owner IDs to 255 in the most recent
> > version of ACPICA, 20051202. This should hit Linux soon.
> >
> > Additionally, we plan to conserve OwnerIDs by not using them for
tables
> > that can never be unloaded, to be implemented in a future release.
> > However, 255 Ids should be plenty for now.
> >
> > Here is the text from the release memo:
> >
> > Increased the number of available Owner Ids for namespace object
> > tracking from 32 to 255. This should eliminate the OWNER_ID_LIMIT
> > exceptions seen on some machines with a large number of ACPI tables
> > (either static or dynamic).
> 
> Hi Bob,
> 
>    Sorry if I wasn't clear, I'm worried about what happens in the
> interim.  The problem will be fixed in ACPICA 20051202, but we have at
> least one, likely two stable kernels that will be tagged before that
> ACPICA version hits the upstream kernel.  We can hit the owner_id
limit
> fairly easily on a few development systems.  How many stable kernels
do
> we want out in the wild with such a low owner_id limit?  Bumping it up
> to 64, while not ideal, is sufficient for our current usage, and I
think
> the patch is trivial enough that it could be included quickly.
Thanks,
> 
> 	Alex
> 
> --
> Alex Williamson                             HP Linux & Open Source Lab

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/