On 2012/9/19 14:18, Huang Ying wrote:
> On Wed, 2012-09-19 at 14:09 +0800, Yijing Wang wrote:
>> On 2012/9/19 13:57, Huang Ying wrote:
>>> On Wed, 2012-09-19 at 10:40 +0800, Yijing Wang wrote:
>>>> Since hot plug for pci devices while doing aer inject, some newly created child buses'
>>>> pci_ops will be assigned to pci_ops_aer. Aer_inject does not track these pci_ops_aer(not
>>>> list in pci_bus_ops_list),we should clean all of these when rmmod aer_inject module.
>>>>
>>>> Signed-off-by: Yijing Wang <wangyijing@xxxxxxxxxx>
>>>> ---
>>>> drivers/pci/pcie/aer/aer_inject.c | 35 +++++++++++++++++++++++++++++++++++
>>>> 1 files changed, 35 insertions(+), 0 deletions(-)
>>>>
>>>> diff --git a/drivers/pci/pcie/aer/aer_inject.c b/drivers/pci/pcie/aer/aer_inject.c
>>>> index 442147b..e2400b5 100644
>>>> --- a/drivers/pci/pcie/aer/aer_inject.c
>>>> +++ b/drivers/pci/pcie/aer/aer_inject.c
>>>> @@ -299,6 +299,32 @@ static void pci_bus_ops_init(struct pci_bus_ops *bus_ops,
>>>> bus_ops->ops = ops;
>>>> }
>>>>
>>>> +static void pci_clean_child_aer_ops(struct pci_bus *bus)
>>>> +{
>>>> + struct pci_bus *child;
>>>> +
>>>> + list_for_each_entry(child, &bus->children, node) {
>>>> + if (child->ops == &pci_ops_aer)
>>>> + pci_bus_set_ops(child, bus->ops);
>>>> + pci_clean_child_aer_ops(child);
>>>> + }
>>>
>>> Why not use pci_walk_bus?
>>>
>>
>> I think here walk all child buses is enough, pci_walk_bus will walk all child devices under this bus.
>>
>>>> +}
>>>> +
>>>> +/* find pci_ops_aer from root bus, and replace it by parent bus's pci_ops.
>>>> + * pci_ops of root bus won't be pci_ops_aer here*/
>>>> +static void clean_untracked_pci_ops_aer(void)
>>>> +{
>>>> + struct pci_bus_ops *bus_ops;
>>>> +
>>>> + list_for_each_entry(bus_ops, &pci_bus_ops_list, list) {
>>>> + /* the bus with untracked pci_ops_aer always the
>>>> + * child bus of root bus tracked in pci_bus_ops_list
>>>> + */
>>>> + if (pci_is_root_bus(bus_ops->bus))
>>>> + pci_clean_child_aer_ops(bus_ops->bus);
>>>> + }
>>>> +}
>>>> +
>>>> static int pci_bus_set_aer_ops(struct pci_bus *bus)
>>>> {
>>>> struct pci_ops *ops;
>>>> @@ -558,9 +584,18 @@ static void __exit aer_inject_exit(void)
>>>>
>>>> misc_deregister(&aer_inject_device);
>>>>
>>>> + /* clean pci_bus_ops tracked in pci_bus_ops_list */
>>>> while ((bus_ops = pci_bus_ops_get(bus_ops)))
>>>> pci_bus_set_ops(bus_ops->bus, bus_ops->ops);
>>>
>>> Why not do clean here inside loop?
>>>
>>
>> Because I also need these bus_ops to scan untracked pci_ops_aer.
>
> I mean why not call pci_clean_child_aer_ops() inside loop.
>
My idea is first clean all pci_ops_aer tracked in pci_bus_ops_list, second clean all untracked pci_ops_aer.
If put pci_clean_child_aer_ops inside loop, when I clean tracked pci_ops_aer for root bus 0, next I will use
bus 0 's pci_ops to clean bus 1 's pci_ops_aer. Actually maybe bus 1 have its own pci_ops_aer tracked in pci_bus_ops_list.
So I think use every bus's own bus_ops to clean pci_ops_aer first is better.
\-[0000:00]-+-00.0 Intel Corporation 5500 I/O Hub to ESI Port
+-01.0-[01]--+-00.0 Broadcom Corporation NetXtreme II BCM5709 Gigabit Ethernet
| \-00.1 Broadcom Corporation NetXtreme II BCM5709 Gigabit Ethernet
> Best Regards,
> Huang Ying
>
>>>
>>>>
>>>> + /* Inject aer errors and hotplug the same pcie device
>>>> + * maybe assign some newly created buses' pci_ops pci_ops_aer.
>>>> + * Since these pci_ops_aer are not tracked in pci_bus_ops_list,
>>>> + * we need to find and clean untracked pci_ops_aer before aer_inject
>>>> + * module exit
>>>> + */
>>>> + clean_untracked_pci_ops_aer();
>>>> +
>>>> while ((bus_ops = pci_bus_ops_pop()))
>>>> kfree(bus_ops);
>>>>
>>>
>>>
>>>
>>> .
>>>
>>
>>
>
>
>
> .
>
--
Thanks!
Yijing
--
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
