qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Qemu-devel] Re: [PATCH 3/3] pci/pcie: make pci_find_device() ARI aware.


From: Michael S. Tsirkin
Subject: [Qemu-devel] Re: [PATCH 3/3] pci/pcie: make pci_find_device() ARI aware.
Date: Wed, 26 Jan 2011 15:40:02 +0200
User-agent: Mutt/1.5.21 (2010-09-15)

On Wed, Jan 26, 2011 at 06:45:20PM +0900, Isaku Yamahata wrote:
> make pci_find_device() ARI aware.
> 
> Signed-off-by: Isaku Yamahata <address@hidden>
> ---
>  hw/pci.c  |    6 ++++++
>  hw/pcie.c |   33 +++++++++++++++++++++++++++++----
>  hw/pcie.h |    2 +-
>  3 files changed, 36 insertions(+), 5 deletions(-)
> 
> diff --git a/hw/pci.c b/hw/pci.c
> index ac16029..daba310 100644
> --- a/hw/pci.c
> +++ b/hw/pci.c
> @@ -1601,12 +1601,18 @@ PCIBus *pci_find_bus(PCIBus *bus, int bus_num)
>  
>  PCIDevice *pci_find_device(PCIBus *bus, int bus_num, int devfn)
>  {
> +    PCIDevice *d;
>      assert(devfn < PCI_DEVFN_MAX);
>      bus = pci_find_bus(bus, bus_num);
>  
>      if (!bus)
>          return NULL;
>  
> +    d = bus->parent_dev;
> +    /* ARI: See the comment above the pcie_check_slot() for details */
> +    if (d && !pcie_check_slot(d, devfn)) {
> +        return NULL;
> +    }
>      return bus->devices[devfn];
>  }
>  
> diff --git a/hw/pcie.c b/hw/pcie.c
> index 6a113a9..ae4b531 100644
> --- a/hw/pcie.c
> +++ b/hw/pcie.c
> @@ -424,13 +424,38 @@ void pcie_cap_ari_reset(PCIDevice *dev)
>      pci_long_test_and_clear_mask(devctl2, PCI_EXP_DEVCTL2_ARI);
>  }
>  
> -bool pcie_cap_is_ari_enabled(const PCIDevice *dev)
> +/*
> + * 6.13 Alternative routing-ID Interpretation(ARI)
> + * 7.8.16 Device control 2 register
> + *   ARI forwarding Enable
> + *
> + * With PCI Express Endpoints, there's a single device behind
> + * each downstream port bus, and bits 3:7 of the function number get
> + * encoded in the slot number (the Express spec calls it the Device
> + * Number). This allows > 8 functions, but
> + * these extended functions are only accessible when the
> + * Alternative routing-ID Interpretation (ARI)
> + * capability is enabled in the root/downstream port. With that capability
> + * disabled the port enforces the Device Number field being 0.
> + */
> +bool pcie_check_slot(const PCIDevice *dev, int devfn)
>  {
> +    uint8_t type;
> +
>      if (!pci_is_express(dev)) {
> -        return false;
> +        return true;
>      }
> -    if (!dev->exp.exp_cap) {
> -        return false;
> +
> +    type = pcie_cap_get_type(dev);
> +    if (!(type == PCI_EXP_TYPE_ROOT_PORT ||
> +          type == PCI_EXP_TYPE_DOWNSTREAM)) {
> +        return true;
> +    }
> +
> +    if (!PCI_SLOT(devfn)) {
> +        /* With ARI, this means function < 8.
> +           functions < 8 are always accesible. */
> +        return true;
>      }
>  
>      return pci_get_long(dev->config + dev->exp.exp_cap + PCI_EXP_DEVCTL2) &
> diff --git a/hw/pcie.h b/hw/pcie.h
> index bc909e2..0c84662 100644
> --- a/hw/pcie.h
> +++ b/hw/pcie.h
> @@ -119,7 +119,7 @@ void pcie_cap_flr_write_config(PCIDevice *dev,
>  
>  void pcie_cap_ari_init(PCIDevice *dev);
>  void pcie_cap_ari_reset(PCIDevice *dev);
> -bool pcie_cap_is_ari_enabled(const PCIDevice *dev);
> +bool pcie_check_slot(const PCIDevice *dev, int devfn);

I think it's better to make it static inline.
This way pci.c depends on pcie.h but not pcie.c

>  
>  /* PCI express extended capability helper functions */
>  uint16_t pcie_find_capability(PCIDevice *dev, uint16_t cap_id);
> -- 
> 1.7.1.1



reply via email to

[Prev in Thread] Current Thread [Next in Thread]