Le 22/06/2012 22:41, Yinghai Lu a écrit :
> On Fri, Jun 22, 2012 at 1:38 PM, Brice Goglin <brice.goglin@xxxxxxxxx> wrote:
>> Le 22/06/2012 19:28, Yinghai Lu a écrit :
>>> On Fri, Jun 22, 2012 at 12:14 AM, Brice Goglin <brice.goglin@xxxxxxxxx> wrote:
>>>
>>>>> BIOS _PXM return 0 for both pci root buses?
>>>>>
>>>> Here's dmesg.
>>>> I can't say for sure whether _PXM returns 0 since I don't know how to
>>>> read all this :) But Linux puts the first socket cpumap in
>>>> /sys/bus/pci/devices/*/local_cpus (and that's wrong according to the
>>>> motherboard manual and according to the performance we see).
>>>>
>>> hi, looks like you system bios does not provide _PXM for the root bus.
>>>
>> So why does Linux say that all buses are close to socket 0 instead of
>> close to everything as usual?
> if (bus && node != -1) {
> #ifdef CONFIG_ACPI_NUMA
> if (pxm >= 0)
> dev_printk(KERN_DEBUG, &bus->dev,
> "on NUMA node %d (pxm %d)\n", node, pxm);
> #else
> dev_printk(KERN_DEBUG, &bus->dev, "on NUMA node %d\n", node);
> #endif
> }
>
> so can you boot with "debug ignore_loglevel" ?
>
I can't find anything relevant in the debug log. The code has changed a
lot since my old SLES11 but I don't see any old message either. I guess
"bus && node != -1 && pxm>=0" evaluates to false.
Brice
--
To unsubscribe from this list: send the line "unsubscribe linux-pci" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
