This bug was introduced in the 2.6.24 i386/x86_64 tree merge, where

MSI-X vector allocation will eventually fail.  The cause is the new

bit array tracking used vectors is not getting cleared properly on

IRQ destruction on the 32-bit APIC code.
This can be seen easily using the ixgbe 10 GbE driver on multi-core

systems by simply loading and unloading the driver a few times.

Depending on the number of available vectors on the host system, the

MSI-X allocation will eventually fail, and the driver will only be

able to use legacy interrupts.
Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>

---
 arch/x86/kernel/io_apic_32.c |    1 +

 1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/arch/x86/kernel/io_apic_32.c b/arch/x86/kernel/io_apic_32.c

index 4ca5486..f239b30 100644

--- a/arch/x86/kernel/io_apic_32.c

+++ b/arch/x86/kernel/io_apic_32.c

@@ -2477,6 +2477,7 @@ void destroy_irq(unsigned int irq)

 	dynamic_irq_cleanup(irq);
 	spin_lock_irqsave(&vector_lock, flags);

+	clear_bit(irq_vector[irq], used_vectors);

 	irq_vector[irq] = 0;

 	spin_unlock_irqrestore(&vector_lock, flags);

 }
--

When sending patches to -stable, please add "CC: stable

<stable@kernel.org> to the patch itself in the signed-off-by area, that

way we get notified automatically when it goes into Linus's tree so that

we can then know to apply it to the tree.  Otherwise it's a tough job to

try to track all of these knowing when to watch out for them to be

committed.
Can you please resend them to stable@kernel.org when they do go into

Linus's tree?
thanks,
greg k-h

--

Sure thing.
Thanks Greg,
-PJ Waskiewicz

--