Thanks all for your comments.
On Mon, Mar 12, 2012 at 10:11 PM, Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote:
> On Mon, 12 Mar 2012, Michal Nazarewicz wrote:
>
>> On Mon, 12 Mar 2012 14:00:19 +0100, Huajun Li <huajun.li.lee@xxxxxxxxx> wrote:
>>
>> > Non-hub device has no child, and even a real USB hub has ports far
>> > less than USB_MAXCHILDREN, so there is no need using a fix array for
>> > child devices, just allocate it dynamically according real port
>> > number.
>> >
>> > Signed-off-by: Huajun Li <huajun.li.lee@xxxxxxxxx>
>> > ---
>> > drivers/usb/core/hub.c | 10 +++++++---
>> > include/linux/usb.h | 2 +-
>> > 2 files changed, 8 insertions(+), 4 deletions(-)
>> >
>> > diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
>> > index 72c51cd..28664eb 100644
>> > --- a/drivers/usb/core/hub.c
>> > +++ b/drivers/usb/core/hub.c
>> > @@ -1047,8 +1047,10 @@ static int hub_configure(struct usb_hub *hub,
>> > dev_info (hub_dev, "%d port%s detected\n", hdev->maxchild,
>> > (hdev->maxchild == 1) ? "" : "s");
>> >
>> > + hdev->children = kzalloc(hdev->maxchild *
>> > + sizeof(struct usb_device *), GFP_KERNEL);
>>
>> kcalloc()
Hi Michal,
Thanks for your reminder, will update it.
>>
>> > hub->port_owners = kzalloc(hdev->maxchild * sizeof(void *), GFP_KERNEL);
>> > - if (!hub->port_owners) {
>> > + if (!hdev->children || !hub->port_owners) {
>>
>> Memory leak if hdev->children != NULL. Just add:
>>
>> kfree(hdev->children);
>>
>> if you don't want to add another “if” or a label.
>
> No, this is correct as it stands. There is no memory leak.
>
There is no memory leak as Alan said.
You know, once hub_configure() fails, hub_disconnect() will be called
to release resources.
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
