Re: [patch] x86_64, irq: check remote IRR bit before migrating level triggered irq

[Eric W. Biederman]

"Yinghai Lu" <yhlu.kernel@xxxxxxxxx> writes:

> On 5/18/07, Eric W. Biederman <ebiederm@xxxxxxxxxxxx> wrote:
>> We can solve the problem without doing that, and keeping the same
>> vector number during migration keeps x86 from scaling.
>
> I mean ioapic level irq couls be limited. new device could use MSI or
> HT irq directly and less irq routing problem.

Possibly.  It really doesn't buy us anything until most irqs are MSI
which they are not yet.

>> Personally I would prefer to disallow irq migration.
> ? typo?
> For amd platform with different hypertransport chain on different
> nodes, irq migration is needed.

irqs not on cpu0 are needed.  irq migration is less necessary, and I periodically
think we are insane for supporting it.

Eric
-
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/