[Bug 1090] Increase MAX_SESSIONS?

bugzilla-daemon at mindrot.org bugzilla-daemon at mindrot.org
Mon Oct 3 13:37:35 EST 2005


http://bugzilla.mindrot.org/show_bug.cgi?id=1090





------- Additional Comments From dtucker at zip.com.au  2005-10-03 13:37 -------
(From update of attachment 963)
Personally, I have no objection to this in principle.

Some comments on the patch:

>+		num_sessions=1;

I'd be tempted to allocate them in blocks (eg of 8) to avoid excessive reallocs
but that's not critical.

>+		sessions=calloc(num_sessions,sizeof(sessions[0]));

Since sessions is initialized as NULL, could you use xrealloc (which is
guaranteed to be happy with xrealloc(NULL, size)) rather than calloc/realloc to
simplify this?

>+		Session *n=realloc(sessions,++num_sessions*sizeof(Session));
>+		if (!n)
>+		    return NULL;

If the realloc fails you will have already incremented num_sessions, so the
next new channel will overflow the array bounds.

Along similar lines: it's unlikely but what's to prevent num_sessions exceeding
INT_MAX and wrapping when MaxSessions=0 and you have gobs of memory?

>+		sessions=n;
>+		s=sessions+num_sessions-1;

The rest of the function uses array syntax, it's probably easier to follow if
you stick to that.  "s = sessions[i]" would be right, no?

>+.It Cm MaxSession

Should be "MaxSessions"?.

There were a few style nits too but they're not a big deal.




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.




More information about the openssh-bugs mailing list