On Fri, Oct 05, 2007 at 04:06:51AM -0400, Mike Frysinger wrote:
> looks like the fix for the race condition may have included some unintended
> side effects ... if you execute say:
> xterm -e script
> and then resize the xterm, script will terminate because the read() in
> doinput() returned -1 (with errno being set to EINTR) ... i tried just making
> the logic (read() > 0 || errno == EINTR), but that didnt seem to work very
> well ;)
Thanks! Fixed.
Karel
commit 2b8bbb5fb0f024ea3917fedbbbaf0fab1c1f6555
Author: Karel Zak <kzak@xxxxxxxxxx>
Date: Fri Oct 5 12:22:13 2007 +0200
script: dies on SIGWINCH
The "doinput" process doesn't make a difference between SIGWINCH and
SIGCHILD. This process also sends unnecessary SIGWINCH to child (the
signal is ignored by child). Fixed.
Signed-off-by: Karel Zak <kzak@xxxxxxxxxx>
diff --git a/misc-utils/script.c b/misc-utils/script.c
index d3272df..3b957d8 100644
--- a/misc-utils/script.c
+++ b/misc-utils/script.c
@@ -99,6 +99,7 @@ int tflg = 0;
static char *progname;
int die;
+int resized;
static void
die_if_link(char *fn) {
@@ -235,8 +236,14 @@ doinput() {
if (die == 0 && child && kill(child, 0) == -1 && errno == ESRCH)
die = 1;
- while (die == 0 && (cc = read(0, ibuf, BUFSIZ)) > 0)
- (void) write(master, ibuf, cc);
+ while (die == 0) {
+ if ((cc = read(0, ibuf, BUFSIZ)) > 0)
+ (void) write(master, ibuf, cc);
+ else if (cc == -1 && errno == EINTR && resized)
+ resized = 0;
+ else
+ break;
+ }
done();
}
@@ -255,11 +262,10 @@ finish(int dummy) {
void
resize(int dummy) {
+ resized = 1;
/* transmit window change information to the child */
(void) ioctl(0, TIOCGWINSZ, (char *)&win);
(void) ioctl(slave, TIOCSWINSZ, (char *)&win);
-
- kill(child, SIGWINCH);
}
/*
-
To unsubscribe from this list: send the line "unsubscribe util-linux-ng" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
