[PATCH v2] ACPI / PCI: fix acpi_pci_irq_enable() memory leak

[PATCH v2] ACPI / PCI: fix acpi_pci_irq_enable() memory leak

[Wenwen Wang]

In acpi_pci_irq_enable(), 'entry' is allocated by kzalloc() in
acpi_pci_irq_check_entry() (invoked from acpi_pci_irq_lookup()). However,
it is not deallocated if acpi_pci_irq_valid() returns false, leading to a
memory leak. To fix this issue, free 'entry' before returning 0.

Fixes: e237a5518425 ("x86/ACPI/PCI: Recognize that Interrupt Line 255 means
"not connected"")

Signed-off-by: Wenwen Wang <wenwen@cs.uga.edu>
---
 drivers/acpi/pci_irq.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/acpi/pci_irq.c b/drivers/acpi/pci_irq.c
index d2549ae..dea8a60 100644
--- a/drivers/acpi/pci_irq.c
+++ b/drivers/acpi/pci_irq.c
@@ -449,8 +449,10 @@ int acpi_pci_irq_enable(struct pci_dev *dev)
 		 * No IRQ known to the ACPI subsystem - maybe the BIOS /
 		 * driver reported one, then use it. Exit in any case.
 		 */
-		if (!acpi_pci_irq_valid(dev, pin))
+		if (!acpi_pci_irq_valid(dev, pin)) {
+			kfree(entry);
 			return 0;
+		}
 
 		if (acpi_isa_register_gsi(dev))
 			dev_warn(&dev->dev, "PCI INT %c: no GSI\n",
-- 
2.7.4

Re: [PATCH v2] ACPI / PCI: fix acpi_pci_irq_enable() memory leak

[Rafael J. Wysocki]

On Wed, Aug 21, 2019 at 5:44 AM Wenwen Wang <wenwen@cs.uga.edu> wrote:
>
> In acpi_pci_irq_enable(), 'entry' is allocated by kzalloc() in
> acpi_pci_irq_check_entry() (invoked from acpi_pci_irq_lookup()). However,
> it is not deallocated if acpi_pci_irq_valid() returns false, leading to a
> memory leak. To fix this issue, free 'entry' before returning 0.
>
> Fixes: e237a5518425 ("x86/ACPI/PCI: Recognize that Interrupt Line 255 means
> "not connected"")
>
> Signed-off-by: Wenwen Wang <wenwen@cs.uga.edu>

Bjorn, any more comments?

> ---
>  drivers/acpi/pci_irq.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/acpi/pci_irq.c b/drivers/acpi/pci_irq.c
> index d2549ae..dea8a60 100644
> --- a/drivers/acpi/pci_irq.c
> +++ b/drivers/acpi/pci_irq.c
> @@ -449,8 +449,10 @@ int acpi_pci_irq_enable(struct pci_dev *dev)
>                  * No IRQ known to the ACPI subsystem - maybe the BIOS /
>                  * driver reported one, then use it. Exit in any case.
>                  */
> -               if (!acpi_pci_irq_valid(dev, pin))
> +               if (!acpi_pci_irq_valid(dev, pin)) {
> +                       kfree(entry);
>                         return 0;
> +               }
>
>                 if (acpi_isa_register_gsi(dev))
>                         dev_warn(&dev->dev, "PCI INT %c: no GSI\n",
> --
> 2.7.4
>

Re: [PATCH v2] ACPI / PCI: fix acpi_pci_irq_enable() memory leak

[Bjorn Helgaas]

On Mon, Sep 02, 2019 at 11:19:58PM +0200, Rafael J. Wysocki wrote:
> On Wed, Aug 21, 2019 at 5:44 AM Wenwen Wang <wenwen@cs.uga.edu> wrote:
> >
> > In acpi_pci_irq_enable(), 'entry' is allocated by kzalloc() in
> > acpi_pci_irq_check_entry() (invoked from acpi_pci_irq_lookup()). However,
> > it is not deallocated if acpi_pci_irq_valid() returns false, leading to a
> > memory leak. To fix this issue, free 'entry' before returning 0.
> >
> > Fixes: e237a5518425 ("x86/ACPI/PCI: Recognize that Interrupt Line 255 means
> > "not connected"")
> >
> > Signed-off-by: Wenwen Wang <wenwen@cs.uga.edu>
> 
> Bjorn, any more comments?

Nope, looks fine to me.

> > ---
> >  drivers/acpi/pci_irq.c | 4 +++-
> >  1 file changed, 3 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/acpi/pci_irq.c b/drivers/acpi/pci_irq.c
> > index d2549ae..dea8a60 100644
> > --- a/drivers/acpi/pci_irq.c
> > +++ b/drivers/acpi/pci_irq.c
> > @@ -449,8 +449,10 @@ int acpi_pci_irq_enable(struct pci_dev *dev)
> >                  * No IRQ known to the ACPI subsystem - maybe the BIOS /
> >                  * driver reported one, then use it. Exit in any case.
> >                  */
> > -               if (!acpi_pci_irq_valid(dev, pin))
> > +               if (!acpi_pci_irq_valid(dev, pin)) {
> > +                       kfree(entry);
> >                         return 0;
> > +               }
> >
> >                 if (acpi_isa_register_gsi(dev))
> >                         dev_warn(&dev->dev, "PCI INT %c: no GSI\n",
> > --
> > 2.7.4
> >

Re: [PATCH v2] ACPI / PCI: fix acpi_pci_irq_enable() memory leak

[Rafael J. Wysocki]

On Tue, Sep 3, 2019 at 12:34 AM Bjorn Helgaas <helgaas@kernel.org> wrote:
>
> On Mon, Sep 02, 2019 at 11:19:58PM +0200, Rafael J. Wysocki wrote:
> > On Wed, Aug 21, 2019 at 5:44 AM Wenwen Wang <wenwen@cs.uga.edu> wrote:
> > >
> > > In acpi_pci_irq_enable(), 'entry' is allocated by kzalloc() in
> > > acpi_pci_irq_check_entry() (invoked from acpi_pci_irq_lookup()). However,
> > > it is not deallocated if acpi_pci_irq_valid() returns false, leading to a
> > > memory leak. To fix this issue, free 'entry' before returning 0.
> > >
> > > Fixes: e237a5518425 ("x86/ACPI/PCI: Recognize that Interrupt Line 255 means
> > > "not connected"")
> > >
> > > Signed-off-by: Wenwen Wang <wenwen@cs.uga.edu>
> >
> > Bjorn, any more comments?
>
> Nope, looks fine to me.

Thanks, queuing it up for v5.4 then.