[PATCH] ARCH 2.6.24.y: Fix 32-bit x86 MSI-X allocation leakage

[PATCH] ARCH 2.6.24.y: Fix 32-bit x86 MSI-X allocation leakage

[PJ Waskiewicz]

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 232fdeb..9994c52 100644
--- a/arch/x86/kernel/io_apic_32.c
+++ b/arch/x86/kernel/io_apic_32.c
@@ -2478,6 +2478,7 @@ 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);
 }

Re: [PATCH] ARCH 2.6.24.y: Fix 32-bit x86 MSI-X allocation leakage

[Ingo Molnar]

* PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com> wrote:

> 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.

Greg, please fix the commit message - this is a separate bug that has 
nothing to do with the unification.

	Ingo

Re: [stable] [PATCH] ARCH 2.6.24.y: Fix 32-bit x86 MSI-X allocation leakage

[Greg KH]

On Tue, Apr 29, 2008 at 12:22:35PM +0200, Ingo Molnar wrote:
> 
> * PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com> wrote:
> 
> > 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.
> 
> Greg, please fix the commit message - this is a separate bug that has 
> nothing to do with the unification.

What should I fix the commit message to be?  I copied the same commit
message that went into Linus's tree.  We should be consistent :)

thanks,

greg k-h

Re: [stable] [PATCH] ARCH 2.6.24.y: Fix 32-bit x86 MSI-X allocation leakage

[Ingo Molnar]

* Greg KH <greg@kroah.com> wrote:

> On Tue, Apr 29, 2008 at 12:22:35PM +0200, Ingo Molnar wrote:
> > 
> > * PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com> wrote:
> > 
> > > 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.
> > 
> > Greg, please fix the commit message - this is a separate bug that 
> > has nothing to do with the unification.
> 
> What should I fix the commit message to be?  I copied the same commit 
> message that went into Linus's tree.  We should be consistent :)

instead of:

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

use:

   This bug was introduced in the 2.6.24 lguest merge,

or just:

   This bug was introduced in 2.6.24,

:-)

	Ingo

Re: [stable] [PATCH] ARCH 2.6.24.y: Fix 32-bit x86 MSI-X allocation leakage

[Greg KH]

On Wed, Apr 30, 2008 at 12:01:15AM +0200, Ingo Molnar wrote:
> 
> * Greg KH <greg@kroah.com> wrote:
> 
> > On Tue, Apr 29, 2008 at 12:22:35PM +0200, Ingo Molnar wrote:
> > > 
> > > * PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com> wrote:
> > > 
> > > > 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.
> > > 
> > > Greg, please fix the commit message - this is a separate bug that 
> > > has nothing to do with the unification.
> > 
> > What should I fix the commit message to be?  I copied the same commit 
> > message that went into Linus's tree.  We should be consistent :)
> 
> instead of:
> 
>    This bug was introduced in the 2.6.24 i386/x86_64 tree merge,
> 
> use:
> 
>    This bug was introduced in the 2.6.24 lguest merge,

Ok, I've used this one for both .24 and .25 stable releases.

thanks,

greg k-h