Re: [Qemu-stable] [Qemu-ppc] [PATCH] spapr_pci: Fix interrupt leak in rtas_ibm_change_msi() error path

[Greg Kurz]

On Fri, 8 Feb 2019 10:06:47 +1100
David Gibson <address@hidden> wrote:

> On Thu, Feb 07, 2019 at 06:28:37PM +0100, Greg Kurz wrote:
> > Now that IRQ allocation has been split in two (first allocate IRQ numbers,
> > then claim them), if the claiming fails, we must release the IRQs.
> > 
> > Fixes: 4fe75a8ccd80 "spapr: split the IRQ allocation sequence"
> > Signed-off-by: Greg Kurz <address@hidden>  
> 
> Applied to ppc-for-4.0, thanks.
> 

Oops I've just realized there's an off-by-one error...

> > ---
> >  hw/ppc/spapr_pci.c |    6 ++++++
> >  1 file changed, 6 insertions(+)
> > 
> > diff --git a/hw/ppc/spapr_pci.c b/hw/ppc/spapr_pci.c
> > index 41d81f4a8500..6fe3c10c8d4c 100644
> > --- a/hw/ppc/spapr_pci.c
> > +++ b/hw/ppc/spapr_pci.c
> > @@ -393,6 +393,12 @@ static void rtas_ibm_change_msi(PowerPCCPU *cpu, 
> > sPAPRMachineState *spapr,
> >      for (i = 0; i < req_num; i++) {
> >          spapr_irq_claim(spapr, irq + i, false, &err);
> >          if (err) {
> > +            if (i) {
> > +                spapr_irq_free(spapr, irq, i + 1);

... here. It should actually be:

+                spapr_irq_free(spapr, irq, i);

Can you fix this in your tree or should I post a v2 ?

> > +            }
> > +            if (!smc->legacy_irq_allocation) {
> > +                spapr_irq_msi_free(spapr, irq, req_num);
> > +            }
> >              error_reportf_err(err, "Can't allocate MSIs for device %x: ",
> >                                config_addr);
> >              rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
> >   
>

pgpO0VOfnxlY_.pgp
Description: OpenPGP digital signature