On Wed, Jul 28, 2010 at 11:46:24AM -0700, Joe Eykholt wrote:
> On 7/28/10 11:42 AM, Chris Leech wrote:
> > If the link state is changing while fipvlan is running, it could result in a
> > packet socket being added to the poll list more than once. When that
> > happened,
> > recvmsg would be called multiple times for a single FIP response.
> >
> > Previously this caused a hang, which the change to MSG_DONTWAIT worked
> > around,
> > but an error message is still printed every time the condition occurs. This
> > change fixes the poll list managment to correctly avoid the error.
> >
> > Signed-off-by: Chris Leech<[email protected]>
> > ---
> >
> > fipvlan.c | 7 +++++++
> > 1 files changed, 7 insertions(+), 0 deletions(-)
> >
> > diff --git a/fipvlan.c b/fipvlan.c
> > index eddffbb..b400427 100644
> > --- a/fipvlan.c
> > +++ b/fipvlan.c
> > @@ -80,6 +80,11 @@ static int pfd_len = 0;
> > void pfd_add(int fd)
> > {
> > struct pollfd *npfd;
> > + int i;
> > +
> > + for (i = 0; i< pfd_len; i++)
> > + if (pfd[i].fd == fd)
> > + return;
> >
> > npfd = realloc(pfd, (pfd_len + 1) * sizeof(struct pollfd));
> > if (!npfd) {
> > @@ -319,6 +324,8 @@ void rtnl_recv_newlink(struct nlmsghdr *nh)
> > iff->running = (ifm->ifi_flags& IFF_RUNNING) == IFF_RUNNING;
> > if (iff->running)
> > pfd_add(iff->ps);
> > + else
> > + pfd_remove(iff->ps);
> > return;
> > }
> >
> >
>
> Is it possible the caller of pfd_add would later want to remove the socket
> from the list and succeed even though it wasn't the last user?
>
> I didn't look into this, but thought I'd mention the issue anyway.
> Maybe it can't happen.
pfd_remove scans for a matching file descriptor, and handles removes in
the middle of the pollfd array correctly I think. I moves up the rest
of the array with memmove, and then shrinks it with realloc.
In my testing I'm pretty sure I saw an fd that wasn't the last one in
the array removed successfully when the link status bounced.
Is that what you were asking about?
- Chris
_______________________________________________
devel mailing list
[email protected]
http://www.open-fcoe.org/mailman/listinfo/devel
