[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: ssh_select possible improvement? SOLVED, patch attached.
[Thread Prev] | [Thread Next]
- Subject: Re: ssh_select possible improvement? SOLVED, patch attached.
- From: Norbert Kiesel <nkiesel@xxxxxxxxxxxxxxx>
- Reply-to: libssh@xxxxxxxxxx
- Date: Mon, 20 Sep 2010 11:45:33 -0700
- To: libssh@xxxxxxxxxx
On Mon, 2010-09-20 at 19:27 +0200, Andreas Schneider wrote:
> On Monday, September 20, 2010 18:37:57 you wrote:
> > Hi Andreas,
>
> Hi Vic,
>
> > This is a really trivial fix and a tiny patch. Please review.
>
> thanks for your patch, I've pushed a modified version to master and v0-4.
>
>
> -- andreas
>
>
Just nitpicking, but the code now reads:
void ssh_socket_fd_set(struct socket *s, fd_set *set, socket_t *max_fd)
{
if (s->fd == SSH_INVALID_SOCKET)
return;
FD_SET(s->fd,set);
if (s->fd >= 0 &&
s->fd >= *max_fd &&
s->fd != SSH_INVALID_SOCKET) {
*max_fd = s->fd + 1;
}
}
I'd say that the "s->fd != SSH_INVALID_SOCKET" can be dropped because of
the test 6 lines above and that the "s->fd >= 0" is of dubious value: it
used to guard against -1 in s->fd but that got replaced with the test
against SSH_INVALID_SOCKET, so a negative but valid fd (of course not
possible in Linux/POSIX) should update max_fd. So I'd propose to change
it to
if (s->fd >= *max_fd) {
*max_fd = s->fd + 1;
}
| ssh_select possible improvement? | Vic Lee <llyzs@xxxxxxx> |
| Re: ssh_select possible improvement? | Vic Lee <llyzs@xxxxxxx> |
| Re: ssh_select possible improvement? SOLVED, patch attached. | Vic Lee <llyzs@xxxxxxx> |
| Re: ssh_select possible improvement? SOLVED, patch attached. | Andreas Schneider <asn@xxxxxxxxxxxx> |