On Wed, Nov 27, 2019 at 08:08:23AM +0100, Gerd Hoffmann wrote:
> On Mon, Nov 25, 2019 at 07:25:26PM -0800, Your Real Name wrote:
> > From 3bf91481863ec504d113aa6b94827bf92840e291 Mon Sep 17 00:00:00 2001
> > From: gaobin <gao...@amazon.com>
> > Date: Thu, 19 Sep 2019 11:23:04 -0700
> > Subject: [PATCH 2/4] pci: Allow scanning pci bus number up to 255 in CSM
> > mode
> >
> > On real hardware especially server platforms, there are many pci
> > devices, bridges, either SoC integrated or addon. They can exhaust
> > all the pci bus numbers. So when scanning the pci bus, we need to
> > allow the bus number up to 255.
> >
> > Signed-off-by: gaobin <gao...@amazon.com>
> > ---
> > src/hw/pcidevice.c | 6 ++++++
> > 1 file changed, 6 insertions(+)
> >
> > diff --git a/src/hw/pcidevice.c b/src/hw/pcidevice.c
> > index 8853cf7..acf15b4 100644
> > --- a/src/hw/pcidevice.c
> > +++ b/src/hw/pcidevice.c
> > @@ -26,6 +26,12 @@ pci_probe_devices(void)
> > struct hlist_node **pprev = &PCIDevices.first;
> > int extraroots = romfile_loadint("etc/extra-pci-roots", 0);
> > int bus = -1, lastbus = 0, rootbuses = 0, count=0;
> > +
> > + // On real hardware especially server platforms, the bus number
> > + // could run up to the top value, i.e. 0xff
> > + if (CONFIG_CSM)
> > + extraroots = 0xff;
>
> What exactly you are trying to fix here?
>
> seabios should find all pci devices behind bridges without problems.
The problem is from the following statement:
while (bus < 0xff && (bus < MaxPCIBus || rootbuses < extraroots)) {
extraroots always returns 0 in csm mode, and "bus < MaxPCIBus" returns false
very early.
So the while loop ends early, leaving many pci devices un-scanned.
Without the patch, I got:
"Found 33 PCI devices (max PCI bus is 02)"
With the patch, I got:
"Found 159 PCI devices (max PCI bus is b2)"
>
> cheers,
> Gerd
>
_______________________________________________
SeaBIOS mailing list -- seabios@seabios.org
To unsubscribe send an email to seabios-le...@seabios.org
