I see absolutely nothing security-related about this, but it is a
potential bug report, so we should consider it.
At 10:53 AM -0600 1/29/02, Gilles Detillieux wrote:
> > status = ::connect(sock, (struct sockaddr *)&server,
>sizeof(server));
>> + if (status < 0 && errno == EAGAIN)
>> + {
>> + sleep(1);
>> + continue;
>> + }
>> if (status < 0 && errno == EINTR && !allow_EINTR)
>> {
>> ::close(sock);
>> open();
> > continue;
The problem I have with this is the errno == EAGAIN. While it seems
to compile OK on Mac OS X and Solaris (which don't mention it in
their manpages), I'm not so sure how portable that is. And I'm
hesitant to mess around with portability problems right before a
release. In contrast, here's the code currently in 3.1.6
status = ::connect(sock, (struct sockaddr *)&server, sizeof(server));
//
// Disable alarm and restore previous policy if any
//
alarm(0);
sigaction(SIGALRM, &old_action, 0);
if (status == 0 || errno == EALREADY || errno == EISCONN)
{
connected = 1;
return OK;
}
//
// Only loop if timed out. Other errors are fatal.
//
if (status < 0 && errno != EINTR)
break;
::close(sock);
open();
sleep(wait_time);
In this case, you're saying that Linux is going to set errno to
EAGAIN != EINTR, so it will break and fail immediately, rather than
timeout and retry. (Interestingly, it seems like the allow_EINTR has
disappeared, which could be a problem.)
I'd be glad to revise this *if* someone can confirm that EAGAIN is
portable and/or in some spec like POSIX.
Otherwise, I worry about releasing 3.1.6 and finding this new
revision in Connection.cc prevents compilation on some platforms.
(Remember that this section of the code is the most heavily
autoconf'ed.)
--
--
-Geoff Hutchison
Williams Students Online
http://wso.williams.edu/
_______________________________________________
htdig-general mailing list <[EMAIL PROTECTED]>
To unsubscribe, send a message to <[EMAIL PROTECTED]> with a
subject of unsubscribe
FAQ: http://htdig.sourceforge.net/FAQ.html
