NetBSD Problem Report #18849
Received: (qmail 3809 invoked by uid 605); 29 Oct 2002 08:39:08 -0000
Message-Id: <200210290839.g9T8d1L01780@loop.home>
Date: Tue, 29 Oct 2002 01:39:01 -0700 (MST)
From: swp@alumni.rice.edu
Sender: gnats-bugs-owner@netbsd.org
Reply-To: swp@alumni.rice.edu
To: gnats-bugs@gnats.netbsd.org
Subject: open of dcm modem port sometimes hangs
X-Send-Pr-Version: 3.95
>Number: 18849
>Category: port-hp300
>Synopsis: open of dcm modem port sometimes hangs
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: port-hp300-maintainer
>State: closed
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Tue Oct 29 08:40:00 +0000 2002
>Closed-Date: Sun May 04 02:10:42 +0000 2003
>Last-Modified: Sun May 04 02:11:00 +0000 2003
>Originator: Steve Peurifoy
>Release: NetBSD 1.5.2
>Organization:
>Environment:
System: NetBSD loop.home 1.5.2 NetBSD 1.5.2 (loop) #2: Fri Oct 25 23:49:35 MDT 2002 root@loop.home:/usr/obj/sys/arch/hp300/compile/loop hp300
>Description:
Occasionally a blocking open of dcm port 0 will hang in ttyopen()
waiting for carrier detect to be asserted when it's already
present.
>How-To-Repeat:
I've encountered this at random times wth a couple of different
modems.
>Fix:
The problem is that dcmmint() will return before updating the
value of sc->sc_mcndlast if the device is not open. If carrier
detect drops after the device is closed, sc->sc_mcndlast will
no longer correctly represent the "previous" state of the
signals when the next transition of carrier detect occurs.
A subsequent blocking open will wait forever for an event
which has in fact taken place.
Index: dcm.c
===================================================================
RCS file: /cvs/src/sys/arch/hp300/dev/dcm.c,v
retrieving revision 1.2
diff -c -r1.2 dcm.c
*** dcm.c 2002/10/21 08:36:57 1.2
--- dcm.c 2002/10/26 05:46:09
***************
*** 908,917 ****
struct tty *tp;
struct dcmdevice *dcm = sc->sc_dcm;
- tp = sc->sc_tty[port];
- if (tp == NULL || (tp->t_state & TS_ISOPEN) == 0)
- return;
-
#ifdef DEBUG
if (dcmdebug & DDB_MODEM)
printf("%s port %d: dcmmint: mcnd %x mcndlast %x\n",
--- 908,913 ----
***************
*** 919,924 ****
--- 915,924 ----
#endif
delta = mcnd ^ sc->sc_mcndlast[port];
sc->sc_mcndlast[port] = mcnd;
+ tp = sc->sc_tty[port];
+ if (tp == NULL || (tp->t_state & TS_ISOPEN) == 0)
+ return;
+
if ((delta & MI_CTS) && (tp->t_state & TS_ISOPEN) &&
(tp->t_cflag & CCTS_OFLOW)) {
if (mcnd & MI_CTS) {
>Release-Note:
>Audit-Trail:
State-Changed-From-To: open->closed
State-Changed-By: gmcgarry
State-Changed-When: Sun May 4 02:10:24 UTC 2003
State-Changed-Why:
patch applied.
From: Gregory McGarry <gmcgarry@netbsd.org>
To: gnats-bugs@gnats.netbsd.org
Cc:
Subject: pr/18849 CVS commit: src/sys/arch/hp300/dev
Date: Sun, 4 May 2003 02:10:09 +0000 (UTC)
Module Name: src
Committed By: gmcgarry
Date: Sun May 4 02:10:08 UTC 2003
Modified Files:
src/sys/arch/hp300/dev: dcm.c
Log Message:
Fix from Steve Peurifoy in PR#18849:
"Occasionally a blocking open of dcm port 0 will hang in ttyopen()
waiting for carrier detect to be asserted when it's already
present.
The problem is that dcmmint() will return before updating the
value of sc->sc_mcndlast if the device is not open. If carrier
detect drops after the device is closed, sc->sc_mcndlast will
no longer correctly represent the "previous" state of the
signals when the next transition of carrier detect occurs.
A subsequent blocking open will wait forever for an event
which has in fact taken place."
To generate a diff of this commit:
cvs rdiff -r1.57 -r1.58 src/sys/arch/hp300/dev/dcm.c
Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.
>Unformatted:
(Contact us)
$NetBSD: query-full-pr,v 1.51 2026/08/10 02:28:17 riastradh Exp $
$NetBSD: gnats_config.sh,v 1.10 2026/05/13 22:00:09 riastradh Exp $
Copyright © 1994-2026
The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.