[Click] click Digest, Vol 92, Issue 4
Joonwoo Park
joonwpark81 at gmail.com
Wed Feb 2 14:04:11 EST 2011
Hi Bobby,
Thanks for having look at this.
Please see inlined below.
On Wed, Feb 2, 2011 at 10:24 AM, Bobby Longpocket
<bobbylongpocket at yahoo.com> wrote:
> Hi Joonwoo,
>
> In ToUserDevice: It looks like the existing IOCTL isn't doing anything that requires locking, so you should be able to leave out the mutex.
I don't think so... filp->private_data is not atomic on every platform.
>
> In FromDevice:click_rx_handler(), the call to _real_rx_handler could presumably eat the sk_buff, so the next line needs to be conditional upon (and use) the sk_buff returned from that call rather than the original skb?
Nice catch.
But setting _real_rx_handler_data after calling _real_rx_handler() is
necessary regardless skb was eaten or not.
I'll fix with a different way that save dev before calling
_real_rx_handler() and set it _real_rx_handler_data.
Thanks,
Joonwoo
>
> Regards,
> BBL
>
>
>
>> Hi All,
>>
>> I just sent pull request of patches below to Eddie and rest
>> of collaborators.
>> Any input will be greatly appreciated.
>>
>>
>> Joonwoo
>
>
>
>
> _______________________________________________
> click mailing list
> click at amsterdam.lcs.mit.edu
> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
>
More information about the click
mailing list