Re: [PATCH 19/23] PCI: add /sys/bus/pci/rescan_root

From: Bjorn Helgaas
Date: Tue Mar 06 2012 - 23:31:33 EST


On Tue, Mar 6, 2012 at 12:13 AM, Yinghai Lu <yinghai@xxxxxxxxxx> wrote:
>
> It will be used to rediscover removed pci root buses.
>
> Signed-off-by: Yinghai Lu <yinghai@xxxxxxxxxx>
> Cc: Randy Dunlap <rdunlap@xxxxxxxxxxxx>
> Cc: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
> Cc: linux-doc@xxxxxxxxxxxxxxx
> ---
>  Documentation/ABI/testing/sysfs-bus-pci |    9 +++++++++
>  drivers/pci/pci-sysfs.c                 |   21 +++++++++++++++++++++
>  2 files changed, 30 insertions(+), 0 deletions(-)
>
> diff --git a/Documentation/ABI/testing/sysfs-bus-pci
> b/Documentation/ABI/testing/sysfs-bus-pci
> index 22392de..eb826bd 100644
> --- a/Documentation/ABI/testing/sysfs-bus-pci
> +++ b/Documentation/ABI/testing/sysfs-bus-pci
> @@ -66,6 +66,15 @@ Description:
>                re-discover previously removed devices.
>                Depends on CONFIG_HOTPLUG.
>
> +What:          /sys/bus/pci/rescan_root
> +Date:          March 2012
> +Contact:       Linux PCI developers <linux-pci@xxxxxxxxxxxxxxx>
> +Description:
> +               Writing a non-zero value to this attribute will
> +               cause a rescan of all PCI root buses in the system, and
> +               re-discover previously removed PCI root buses.

This seems sort of backwards to me. Architecture-specific code
enumerated the host bridges/root buses in the first place, not the PCI
core. Now we're asking the PCI core (/sys/bus/pci) to re-enumerate
them.

But maybe it's still the right answer, I dunno.

> +               Depends on CONFIG_HOTPLUG.
> +
>  What:          /sys/bus/pci/devices/.../msi_irqs/
>  Date:          September, 2011
>  Contact:       Neil Horman <nhorman@xxxxxxxxxxxxx>
> diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c
> index f855b4b..d6221e5 100644
> --- a/drivers/pci/pci-sysfs.c
> +++ b/drivers/pci/pci-sysfs.c
> @@ -302,8 +302,29 @@ static ssize_t bus_rescan_store(struct bus_type *bus,
> const char *buf,
>        return count;
>  }
>
> +void __weak acpi_pci_root_rescan(void) { }
> +void __weak pcibios_root_rescan(void) { }
> +
> +static ssize_t bus_rescan_root_store(struct bus_type *bus, const char
> *buf,
> +                               size_t count)
> +{
> +       unsigned long val;
> +
> +       if (strict_strtoul(buf, 0, &val) < 0)
> +               return -EINVAL;
> +
> +       if (val) {
> +               mutex_lock(&pci_remove_rescan_mutex);
> +               acpi_pci_root_rescan();

This architecture-dependent function name should not be here. How
will we make this work for powerpc, sparc, etc.?

> +               pcibios_root_rescan();
> +               mutex_unlock(&pci_remove_rescan_mutex);
> +       }
> +       return count;
> +}
> +
>  struct bus_attribute pci_bus_attrs[] = {
>        __ATTR(rescan, (S_IWUSR|S_IWGRP), NULL, bus_rescan_store),
> +       __ATTR(rescan_root, (S_IWUSR|S_IWGRP), NULL,
> bus_rescan_root_store),
>        __ATTR_NULL
>  };
>
> --
> 1.7.7
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-pci" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/