Re: [PATCH] pci-assign: Use PCI-2.3-based shared legacy interrupts

["Michael S. Tsirkin" <mst@redhat.com>]

On Tue, Nov 09, 2010 at 03:55:00PM +0100, Jan Kiszka wrote:
> Am 09.11.2010 15:42, Michael S. Tsirkin wrote:
> > On Mon, Nov 08, 2010 at 12:25:47PM +0100, Jan Kiszka wrote:
> >> @@ -417,6 +423,17 @@ static void assigned_dev_pci_write_config(PCIDevice *d, uint32_t address,
> >>  
> >>      if (address == 0x4) {
> >>          pci_default_write_config(d, address, val, len);
> >> +        pci_dev->intx_masked = val & PCI_COMMAND_INTX_DISABLE;
> >> +#ifdef KVM_CAP_PCI_2_3
> >> +        memset(&assigned_dev_data, 0, sizeof(assigned_dev_data));
> >> +        assigned_dev_data.assigned_dev_id  =
> >> +            calc_assigned_dev_id(pci_dev->h_segnr, pci_dev->h_busnr,
> >> +                                 pci_dev->h_devfn);
> >> +        if (pci_dev->intx_masked) {
> >> +            assigned_dev_data.flags = KVM_DEV_ASSIGN_MASK_INTX;
> >> +        }
> >> +        kvm_assign_set_intx_mask(kvm_context, &assigned_dev_data);
> >> +#endif
> >>          /* Continue to program the card */
> >>      }
> >>  
> >> @@ -494,6 +511,10 @@ do_log:
> >>          else if (address == 6)
> >>              val &= ~0x10;
> >>      }
> >> +    if (address == PCI_COMMAND) {
> > 
> > Could also be a single-byte write into the high byte
> > of this word.
> 
> Yes, will fix.
> 
> Jan

I think assigned devices call default write config, don't they?
If yes you can just check
	pci_get_word(d->config + PCI_COMMAND) & PCI_COMMAND_INTX_DISABLE

> -- 
> Siemens AG, Corporate Technology, CT T DE IT 1
> Corporate Competence Center Embedded Linux