Re: [Qemu-devel] Re: [PATCH V5 07/29] pci/bridge: clean up of pci_bridge_initfn()

[Michael S. Tsirkin]

On Tue, Oct 13, 2009 at 06:12:17PM +0300, Blue Swirl wrote:
> On Tue, Oct 13, 2009 at 4:20 PM, Isaku Yamahata <address@hidden> wrote:
> > On Fri, Oct 09, 2009 at 08:53:10AM +0200, Michael S. Tsirkin wrote:
> >> On Fri, Oct 09, 2009 at 03:28:40PM +0900, Isaku Yamahata wrote:
> >> > - use symbolic constant
> >> > - use helper function pci_set_xxx()
> >> >
> >> > Signed-off-by: Isaku Yamahata <address@hidden>
> >> > ---
> >> >  hw/pci.c |   23 ++++++++++++-----------
> >> >  1 files changed, 12 insertions(+), 11 deletions(-)
> >> >
> >> > diff --git a/hw/pci.c b/hw/pci.c
> >> > index a66e3de..eaf471a 100644
> >> > --- a/hw/pci.c
> >> > +++ b/hw/pci.c
> >> > @@ -923,17 +923,18 @@ static int pci_bridge_initfn(PCIDevice *dev)
> >> >      pci_config_set_vendor_id(s->dev.config, s->vid);
> >> >      pci_config_set_device_id(s->dev.config, s->did);
> >> >
> >> > -    s->dev.config[0x04] = 0x06; // command = bus master, pci mem
> >> > -    s->dev.config[0x05] = 0x00;
> >> > -    s->dev.config[0x06] = 0xa0; // status = fast back-to-back, 66MHz, 
> >> > no error
> >> > -    s->dev.config[0x07] = 0x00; // status = fast devsel
> >> > -    s->dev.config[0x08] = 0x00; // revision
> >> > -    s->dev.config[0x09] = 0x00; // programming i/f
> >> > -    pci_config_set_class(s->dev.config, PCI_CLASS_BRIDGE_PCI);
> >> > -    s->dev.config[0x0D] = 0x10; // latency_timer
> >> > -    s->dev.config[PCI_HEADER_TYPE] =
> >> > -        PCI_HEADER_TYPE_MULTI_FUNCTION | PCI_HEADER_TYPE_BRIDGE; // 
> >> > header_type
> >> > -    s->dev.config[0x1E] = 0xa0; // secondary status
> >> > +    pci_set_word(dev->config + PCI_COMMAND,
> >> > +                 PCI_COMMAND_MEMORY | PCI_COMMAND_MASTER);
> >>
> >> BTW, I think this is a wrong reset value: should be disabled
> >> by default. Fixing this needs some testing though, so
> >> I am not suggesting we do it in this patch. Have some time
> >> to fix this?
> >
> > Hmm, the user of it is only apb_pci.c
> > I guess other magic values came from the real machine.
> > So one possible fix is to create apb_pci specific initialization function
> > and to move those initialization code into apb_pci.c leaving to
> > sparc guys. So we can avoid breakage.
> > What do you think of it?
> 
> Current code is buggy. According to manual (805-1251)

Link? Maybe we should put spec links in comments
in relevant code ...

> the reset value
> should be zero, unless the boot pin is tied high (which is true) and
> thus it should be PCI_COMMAND_MEMORY.
> 
> I think the default value should be zero and this should be overridden
> later in apb_pci.c.

Makes sense.

-- 
MST