PATCH: tioccons() br0ken in 2.2 and 2.3

Miquel van Smoorenburg (miquels@cistron.nl)
Tue, 31 Aug 1999 18:15:56 +0200

In 2.2.x and 2.3.x the tioccons() function in tty_io.c tests if you
are using it on the real console, which means that the console redirection
should be turned off.

However it doesn't work since

1. the test is wrong; the wrong tty->driver.type is tested, and
tty->driver.subtype should be tested as well
2. it won't work anyway as those fields don't come from the new /dev/console
device but from the actual device behind it (/dev/tty1 or /dev/ttyS0)

So a better way is to test inode->i_rdev to check for /dev/console,
just like it's done in tty_write()

Patch against 2.3.15 follows, also applies against 2.2.12 (should go
into 2.2.13 as well I think)

diff -ruN linux-2.3.15/drivers/char/tty_io.c linux-2.3.15.new/drivers/char/tty_io.c
--- linux-2.3.15/drivers/char/tty_io.c Tue Aug 31 16:04:40 1999
+++ linux-2.3.15.new/drivers/char/tty_io.c Tue Aug 31 17:46:13 1999
@@ -1499,10 +1499,11 @@
return 0;
}

-static int tioccons(struct tty_struct *tty, struct tty_struct *real_tty)
+static int tioccons(struct inode *inode,
+ struct tty_struct *tty, struct tty_struct *real_tty)
{
- if (tty->driver.type == TTY_DRIVER_TYPE_CONSOLE ||
- tty->driver.type == TTY_DRIVER_TYPE_SYSCONS) {
+ if (inode->i_rdev == SYSCONS_DEV ||
+ inode->i_rdev == CONSOLE_DEV) {
if (!suser())
return -EPERM;
redirect = NULL;
@@ -1712,7 +1713,7 @@
case TIOCSWINSZ:
return tiocswinsz(tty, real_tty, (struct winsize *) arg);
case TIOCCONS:
- return tioccons(tty, real_tty);
+ return tioccons(inode, tty, real_tty);
case FIONBIO:
return fionbio(file, (int *) arg);
case TIOCEXCL:

Mike.

-- 
... somehow I have a feeling the hurting hasn't even begun yet
	-- Bill, "The Terrible Thunderlizards"


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/