Am 25.06.2012 17:50, schrieb ron minnich:
Hence, people's thinking move to MP auto generation from ACPI.
And nothing like that ever happened (I think that's good).
I'm wondering if we want more effort going into MP tables when they're basically obsolete?
MP tables are a relatively easy problem. The next step could be reusing the data for PIR (and extending the devicetree model where necessary).
Once our devicetree is suitable for these two, it should also be good enough for ACPI. That's one thing less to worry about when extending acpigen.
Or does the hardware somehow require it?
Hardware doesn't require it, but some software is happier when having meaningful legacy tables.
Patrick
On Mon, Jun 25, 2012 at 9:53 AM, Patrick Georgi patrick@georgi-clan.de wrote:
Am 25.06.2012 17:50, schrieb ron minnich:
Hence, people's thinking move to MP auto generation from ACPI.
And nothing like that ever happened (I think that's good).
yeah, I guess I do too.
ron
Hi all,
Sorry for the delay. I was busy with other stuff recently. I like it, just please note that ACPI requires to aggregate the _PRT objects under every PCI bridge device.
The root bridge has to have _PRT table (usually it contains routing for BUS 0). The other bridges (_PCI devices) contains aggregate _PRT routing entries for bus directly behind the bridge.
Plz check src/mainboard/asus/m2v-mx_se/dsdt.asl for details.
Thanks Rudolf