On Fri, Apr 27, 2012 at 1:00 PM, Khalid Aziz <[email protected]> wrote:
> Disable Bus Master bit on the device in
> pci_device_shutdown() to ensure PCI devices do not continue
> to DMA data after shutdown. This can cause memory
> corruption in case of a kexec where the current kernel
> shuts down and transfers control to a new kernel while a
> PCI device continues to DMA to memory that does not belong
> to it any more in the new kernel.
>
> I have tested this code on two laptops, two workstations and
> a 16-socket server. kexec worked correctly on all of them.
>
>
> Signed-off-by: Khalid Aziz <[email protected]>
> ---
> ?drivers/pci/pci-driver.c | ? ?6 ++++++
> ?1 files changed, 6 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c
> index 6b54b23..9db5940 100644
> --- a/drivers/pci/pci-driver.c
> +++ b/drivers/pci/pci-driver.c
> @@ -420,6 +420,12 @@ static void pci_device_shutdown(struct device *dev)
> ? ? ? ?pci_msi_shutdown(pci_dev);
> ? ? ? ?pci_msix_shutdown(pci_dev);
>
> + ? ? ? /*
> + ? ? ? ?* Turn off Bus Master bit on the device to tell it to not
> + ? ? ? ?* continue to do DMA
> + ? ? ? ?*/
> + ? ? ? pci_disable_device(pci_dev);
> +
> ? ? ? ?/*
> ? ? ? ? * Devices may be enabled to wake up by runtime PM, but they need not
> ? ? ? ? * be supposed to wake up the system from its "power off" state (e.g.
Any comment on this, Eric? It seems reasonable to me.
On Thu, May 3, 2012 at 5:52 PM, Bjorn Helgaas <[email protected]> wrote:
> On Fri, Apr 27, 2012 at 1:00 PM, Khalid Aziz <[email protected]> wrote:
>> Disable Bus Master bit on the device in
>> pci_device_shutdown() to ensure PCI devices do not continue
>> to DMA data after shutdown. This can cause memory
>> corruption in case of a kexec where the current kernel
>> shuts down and transfers control to a new kernel while a
>> PCI device continues to DMA to memory that does not belong
>> to it any more in the new kernel.
>>
>> I have tested this code on two laptops, two workstations and
>> a 16-socket server. kexec worked correctly on all of them.
>>
>>
>> Signed-off-by: Khalid Aziz <[email protected]>
>> ---
>> ?drivers/pci/pci-driver.c | ? ?6 ++++++
>> ?1 files changed, 6 insertions(+), 0 deletions(-)
>>
>> diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c
>> index 6b54b23..9db5940 100644
>> --- a/drivers/pci/pci-driver.c
>> +++ b/drivers/pci/pci-driver.c
>> @@ -420,6 +420,12 @@ static void pci_device_shutdown(struct device *dev)
>> ? ? ? ?pci_msi_shutdown(pci_dev);
>> ? ? ? ?pci_msix_shutdown(pci_dev);
>>
>> + ? ? ? /*
>> + ? ? ? ?* Turn off Bus Master bit on the device to tell it to not
>> + ? ? ? ?* continue to do DMA
>> + ? ? ? ?*/
>> + ? ? ? pci_disable_device(pci_dev);
>> +
>> ? ? ? ?/*
>> ? ? ? ? * Devices may be enabled to wake up by runtime PM, but they need not
>> ? ? ? ? * be supposed to wake up the system from its "power off" state (e.g.
>
> Any comment on this, Eric? ?It seems reasonable to me.
Applied to my "next" branch, thanks.
Bjorn