GNU bug report logs - #11508
24.1.50; Off-by-one error in xg_select?

Previous Next

Full log

View this message in rfc822 format

From: help-debbugs <at> gnu.org (GNU bug Tracking System)
To: Ken Brown <kbrown <at> cornell.edu>
Subject: bug#11508: closed (Re: bug#11508: 24.1.50; Off-by-one error in
 xg_select?)
Date: Mon, 21 May 2012 13:41:03 +0000

[Message part 1 (text/plain, inline)]

Your bug report

#11508: 24.1.50; Off-by-one error in xg_select?

which was filed against the emacs package, has been closed.

The explanation is attached below, along with your original report.
If you require more details, please reply to 11508 <at> debbugs.gnu.org.

-- 
11508: http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11508
GNU Bug Tracking System
Contact help-debbugs <at> gnu.org with problems

[Message part 2 (message/rfc822, inline)]

From: Ken Brown <kbrown <at> cornell.edu>
To: Jan Djärv <jan.h.d <at> swipnet.se>
Cc: 11508-done <at> debbugs.gnu.org, Andreas Schwab <schwab <at> linux-m68k.org>
Subject: Re: bug#11508: 24.1.50; Off-by-one error in xg_select?
Date: Mon, 21 May 2012 09:38:10 -0400

Version: 24.2

I've committed a simpler version of the change as bzr revision 108325, 
and I'm closing the bug.

Ken

[Message part 3 (message/rfc822, inline)]

From: Ken Brown <kbrown <at> cornell.edu>
To: bug-gnu-emacs <at> gnu.org
Subject: 24.1.50; Off-by-one error in xg_select?
Date: Fri, 18 May 2012 09:13:41 -0400

I apologize in advance for the noise if I'm misunderstanding something,
but it seems to me that there is an error in the calculation of the
first argument in the call to select in xgselect.c:105. (Line numbers
refer to the current trunk; xgselect.c was last changed in bzr revno
108249.)

The parameter "max_fds" in xg_select, in spite of its name, is initially
1 higher than the maximal file descriptor in the fd_sets in the other
parameters.  If max_fds doesn't get increased in line 78 or 83, then
line 104 does the wrong thing, causing the first argument to select in
line 105 to be 1 higher than it should be.

I think the following patch fixes this.  It also renames "max_fds" to
"fds_lim" to more accurately reflect what it represents.

=== modified file 'src/xgselect.c'
--- src/xgselect.c	2012-05-16 02:22:53 +0000
+++ src/xgselect.c	2012-05-18 12:28:27 +0000
@@ -32,7 +32,7 @@
 static ptrdiff_t gfds_size;
 
 int
-xg_select (int max_fds, SELECT_TYPE *rfds, SELECT_TYPE *wfds, SELECT_TYPE *efds,
+xg_select (int fds_lim, SELECT_TYPE *rfds, SELECT_TYPE *wfds, SELECT_TYPE *efds,
 	   EMACS_TIME *timeout)
 {
   SELECT_TYPE all_rfds, all_wfds;
@@ -41,10 +41,10 @@
   GMainContext *context;
   int have_wfds = wfds != NULL;
   int n_gfds = 0, our_tmo = 0, retval = 0, our_fds = 0;
-  int i, nfds, fds_lim, tmo_in_millisec;
+  int i, nfds, tmo_in_millisec;
 
   if (inhibit_window_system || !display_arg)
-    return select (max_fds, rfds, wfds, efds, timeout);
+    return select (fds_lim, rfds, wfds, efds, timeout);
 
   if (rfds) memcpy (&all_rfds, rfds, sizeof (all_rfds));
   else FD_ZERO (&all_rfds);
@@ -75,12 +75,12 @@
       if (gfds[i].events & G_IO_IN)
         {
           FD_SET (gfds[i].fd, &all_rfds);
-          if (gfds[i].fd > max_fds) max_fds = gfds[i].fd;
+          if (gfds[i].fd >= fds_lim) fds_lim = gfds[i].fd + 1;
         }
       if (gfds[i].events & G_IO_OUT)
         {
           FD_SET (gfds[i].fd, &all_wfds);
-          if (gfds[i].fd > max_fds) max_fds = gfds[i].fd;
+          if (gfds[i].fd >= fds_lim) fds_lim = gfds[i].fd + 1;
           have_wfds = 1;
         }
     }
@@ -101,7 +101,6 @@
       if (our_tmo) tmop = &tmo;
     }
 
-  fds_lim = max_fds + 1;
   nfds = select (fds_lim, &all_rfds, have_wfds ? &all_wfds : NULL, efds, tmop);
 
   if (nfds < 0)

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.