On Fri, Apr 18, 2014 at 06:30:37PM +0200, Andreas Färber wrote:
> Am 18.04.2014 15:41, schrieb Kirill Batuzov:
> > acpi_pcihp_get_bsel implements functionality of object_property_get_int for
> > specific property named ACPI_PCIHP_PROP_BSEL, but fails to decrement
> > object's
> > reference counter properly. Replacing it with generic
> > object_property_get_int
> > serves two purposes: reducing code duplication and fixing memory leak.
> >
> > Signed-off-by: Kirill Batuzov <batuz...@ispras.ru>
> > ---
> > hw/acpi/pcihp.c | 23 ++++++-----------------
> > 1 file changed, 6 insertions(+), 17 deletions(-)
> >
> > diff --git a/hw/acpi/pcihp.c b/hw/acpi/pcihp.c
> > index f80c480..ff44aec 100644
> > --- a/hw/acpi/pcihp.c
> > +++ b/hw/acpi/pcihp.c
> > @@ -61,24 +61,11 @@ typedef struct AcpiPciHpFind {
> > PCIBus *bus;
> > } AcpiPciHpFind;
> >
> > -static int acpi_pcihp_get_bsel(PCIBus *bus)
> > -{
> > - QObject *o = object_property_get_qobject(OBJECT(bus),
> > - ACPI_PCIHP_PROP_BSEL, NULL);
> > - int64_t bsel = -1;
> > - if (o) {
> > - bsel = qint_get_int(qobject_to_qint(o));
> > - }
> > - if (bsel < 0) {
> > - return -1;
> > - }
> > - return bsel;
> > -}
> > -
> > static void acpi_pcihp_test_hotplug_bus(PCIBus *bus, void *opaque)
> > {
> > AcpiPciHpFind *find = opaque;
> > - if (find->bsel == acpi_pcihp_get_bsel(bus)) {
> > + if (find->bsel == object_property_get_int(OBJECT(bus),
> > + ACPI_PCIHP_PROP_BSEL, NULL))
> > {
> > find->bus = bus;
> > }
> > }
>
> I note that the wrapper function was changing negative values up to be
> -1, which is getting dropped here. Not sure if it matters.
I think this was to ensure that we don't get an overflow.
I'm not sure why didn't I validate against ACPI_PCIHP_MAX_HOTPLUG_BUS
too.
How about making acpi_pcihp_get_bsel call object_property_get_int
and validate that value is between 0 and ACPI_PCIHP_MAX_HOTPLUG_BUS?
> > @@ -185,7 +172,8 @@ void acpi_pcihp_device_plug_cb(ACPIREGS *ar, qemu_irq
> > irq, AcpiPciHpState *s,
> > {
> > PCIDevice *pdev = PCI_DEVICE(dev);
> > int slot = PCI_SLOT(pdev->devfn);
> > - int bsel = acpi_pcihp_get_bsel(pdev->bus);
> > + int bsel = object_property_get_int(OBJECT(pdev->bus),
> > + ACPI_PCIHP_PROP_BSEL, NULL);
> > if (bsel < 0) {
> > error_setg(errp, "Unsupported bus. Bus doesn't have property '"
> > ACPI_PCIHP_PROP_BSEL "' set");
> > @@ -210,7 +198,8 @@ void acpi_pcihp_device_unplug_cb(ACPIREGS *ar, qemu_irq
> > irq, AcpiPciHpState *s,
> > {
> > PCIDevice *pdev = PCI_DEVICE(dev);
> > int slot = PCI_SLOT(pdev->devfn);
> > - int bsel = acpi_pcihp_get_bsel(pdev->bus);
> > + int bsel = object_property_get_int(OBJECT(pdev->bus),
> > + ACPI_PCIHP_PROP_BSEL, NULL);
> > if (bsel < 0) {
> > error_setg(errp, "Unsupported bus. Bus doesn't have property '"
> > ACPI_PCIHP_PROP_BSEL "' set");
>
> These ones seem to just check for < 0, so look okay from reading the
> patch. CC'ing mst.
Hmm int is 32 bit and object_property_get_int can return a 64 bit one.
> The alternative would be to leave the wrapper around and just replace
> the ..._get_qobject() with the ..._get_int() inside.
Yes, I'd prefer that, and extra validation there too.
> Regards,
> Andreas
>
> --
> SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
> GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
