[<prev] [next>] [day] [month] [year] [list]
Message-ID: <CAF8QhUhwTbJCXnJr4662AdR_fbgEQ6=Sn0HBMyaqW-mj1=wx8Q@mail.gmail.com>
Date: Wed, 16 Jan 2019 15:00:47 +0800
From: fei phung <feiphung27@...il.com>
To: mst@...hat.com, netdev@...r.kernel.org, feiphung@...mail.com
Subject: Re: Question on ptr_ring linux header
Hi,
> https://gist.github.com/promach/65e9331d55a43a2815239430a28e29c6#file-circ_ring-c-L62
> racy if there are multiple consumers.
> just call ptr_ring_consume_any.
If I modify pop_circ_queue() below to directly use
ptr_ring_consume_any() without
ptr_ring_empty_any() , I am afraid I am running into system crash due
to NULL pointer
deferencing for *item_pop:
*item_pop = *((struct item *)ptr_ring_consume_any(buffer));
Just one silly beginner question, why will ptr_ring_empty_any() lead
to data race problem ?
inline int pop_circ_queue(struct ptr_ring * buffer, struct item * item_pop)
{
DEBUG_MSG(KERN_INFO "Before pop, head = %u , tail = %u\n",
buffer->consumer_head, buffer->consumer_tail);
/* extract one item struct containing two unsigned integers
from the buffer */
*item_pop = *((struct item *)ptr_ring_consume_any(buffer));
if((item_pop != NULL) && (item_pop->val1 > 0))
{
// val1 will never be zero since the event number starts from
1 (so, val1 in push_circ_queue() will not be zero, same case after
pop_circ_queue()), and 0 is only possible during initialization, not
during pop_circ_queue()
DEBUG_MSG(KERN_INFO "val1 = %u , val2 = %u\n",
item_pop->val1, item_pop->val2);
DEBUG_MSG(KERN_INFO "After pop, head = %u , tail =
%u\n", buffer->consumer_head, buffer->consumer_tail);
return 0;
}
else {
DEBUG_MSG(KERN_INFO "empty, nothing to pop from the ring\n");
return 1;
}
}
Regards,
Phung
Powered by blists - more mailing lists