Re: [Qemu-devel] [PATCH] hw/pci: reserve IO and mem for pci express downstream ports with no devices attached

[Marcel Apfelbaum]

On Thu, 2014-06-19 at 17:21 +0300, Michael S. Tsirkin wrote:
> On Thu, Jun 19, 2014 at 04:52:17PM +0300, Marcel Apfelbaum wrote:
> > commit c6e298e1f12e0f4ca02b6da5e42919ae055f6830
> >     hw/pci: reserve IO and mem for pci-2-pci bridges with no devices 
> > attached
> > 
> > introduced support for hot-plugging devices behind pci-2-pci bridges.
> > Extend hotplug support also for pci express downstream ports.
> > 
> > Signed-off-by: Marcel Apfelbaum <address@hidden>
> > ---
> >  src/fw/pciinit.c  | 23 +++++++++++++++++++++--
> >  src/hw/pci_regs.h |  2 ++
> >  2 files changed, 23 insertions(+), 2 deletions(-)
> > 
> > diff --git a/src/fw/pciinit.c b/src/fw/pciinit.c
> > index 0ad548f..edb3fe9 100644
> > --- a/src/fw/pciinit.c
> > +++ b/src/fw/pciinit.c
> > @@ -636,6 +636,25 @@ pci_region_create_entry(struct pci_bus *bus, struct 
> > pci_device *dev,
> >      return entry;
> >  }
> >  
> > +static int pci_bus_hotplug_support(struct pci_bus *bus)
> > +{
> > +    u8 shpc_cap = pci_find_capability(bus->bus_dev, PCI_CAP_ID_SHPC);
> > +    u8 pcie_cap = pci_find_capability(bus->bus_dev, PCI_CAP_ID_EXP);
> > +    int downstream_port = 0;
> > +    int slot_implemented = 0;
> > +
> > +    if (pcie_cap) {
> > +        u16 pcie_flags = pci_config_readw(bus->bus_dev->bdf,
> > +                                          pcie_cap + PCI_EXP_FLAGS);
> > +        u16 port_type = ((pcie_flags & PCI_EXP_FLAGS_TYPE) >>
> > +                        PCI_EXP_FLAGS_TYPE_SHIFT) & 0xf;
> 
> 0xf is not needed here: PCI_EXP_FLAGS_TYPE is enough.
> Also should be u8.
OK, thanks
> 
> > +        downstream_port = (port_type == PCI_EXP_TYPE_DOWNSTREAM) ||
> > +                          (port_type == PCI_EXP_TYPE_ROOT_PORT);
> 
> This is correct, the spec says:
>     Configuration Space, and other capabilities.
>     8 Slot Implemented – When Set, this bit indicates that the Link
>     HwInit associated with this Port is connected to a slot (as compared to
>     being connected to a system-integrated device or being
>     disabled).
>     This bit is valid for Downstream Ports. This bit is undefined for
>     Upstream Ports.
> 
> Maybe add a comment in case people will wonder.
Sure

> 
> > +        slot_implemented = !!(pcie_flags & PCI_EXP_FLAGS_SLOT);
> 
> I would do
>         return downstream_port && slot_implemented;
> here, avoid testing shpc for express devices.
> 
> Also allows moving downstream_port && slot_implemented here.
Makes sense

> 
> > +    }
> > +    return shpc_cap || (downstream_port && slot_implemented);
> > +}
> > +
> >  static int pci_bios_check_devices(struct pci_bus *busses)
> >  {
> >      dprintf(1, "PCI: check devices\n");
> > @@ -678,7 +697,7 @@ static int pci_bios_check_devices(struct pci_bus 
> > *busses)
> >              continue;
> >          struct pci_bus *parent = &busses[pci_bdf_to_bus(s->bus_dev->bdf)];
> >          int type;
> > -        u8 shpc_cap = pci_find_capability(s->bus_dev, PCI_CAP_ID_SHPC);
> > +        int hotplug_support = pci_bus_hotplug_support(s);
> >          for (type = 0; type < PCI_REGION_TYPE_COUNT; type++) {
> >              u64 align = (type == PCI_REGION_TYPE_IO) ?
> >                  PCI_BRIDGE_IO_MIN : PCI_BRIDGE_MEM_MIN;
> > @@ -687,7 +706,7 @@ static int pci_bios_check_devices(struct pci_bus 
> > *busses)
> >              if (pci_region_align(&s->r[type]) > align)
> >                   align = pci_region_align(&s->r[type]);
> >              u64 sum = pci_region_sum(&s->r[type]);
> > -            if (!sum && shpc_cap)
> > +            if (!sum && hotplug_support)
> >                  sum = align; /* reserve min size for hot-plug */
> >              u64 size = ALIGN(sum, align);
> >              int is64 = pci_bios_bridge_region_is64(&s->r[type],
> > diff --git a/src/hw/pci_regs.h b/src/hw/pci_regs.h
> > index e5effd4..6a71569 100644
> > --- a/src/hw/pci_regs.h
> > +++ b/src/hw/pci_regs.h
> > @@ -426,6 +426,8 @@
> >  #define PCI_EXP_DEVCTL2            40      /* Device Control 2 */
> >  #define  PCI_EXP_DEVCTL2_ARI       0x20    /* Alternative Routing-ID */
> >  
> > +#define PCI_EXP_FLAGS_TYPE_SHIFT 4
> > +
> 
> It's generally not a good idea to randomly add macros
> in random places.
> In this case this is not needed. Just call __ffs on
> PCI_EXP_FLAGS_TYPE.
I didn't see any usage of __ffs method on seabios code, so
I wondered if adding an include to string.h would be ok.
Would it?

Thanks,
Marcel

> 
> >  /* Extended Capabilities (PCI-X 2.0 and Express) */
> >  #define PCI_EXT_CAP_ID(header)             (header & 0x0000ffff)
> >  #define PCI_EXT_CAP_VER(header)            ((header >> 16) & 0xf)
> > -- 
> > 1.8.3.1