linux-stable.git/drivers/tty/tty_ioctl.c, branch v6.7 Linux kernel stable tree tty: convert THROTTLE constants into enum 2023-10-03T12:31:16+00:00 Jiri Slaby (SUSE) jirislaby@kernel.org 2023-09-19T08:51:53+00:00 e5d0424ac31154e47bfce857ba908bbe861c7a92 And make an explicit constant for zero too. This allows for easier type checking of the parameter. Note: tty_struct::flow_change is kept as int because include/tty.h (tty_struct) doesn't see tty/tty.h (this enum). Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org> Link: https://lore.kernel.org/r/20230919085156.1578-13-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
And make an explicit constant for zero too. This allows for easier type
checking of the parameter.

Note: tty_struct::flow_change is kept as int because include/tty.h
(tty_struct) doesn't see tty/tty.h (this enum).

Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20230919085156.1578-13-jirislaby@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tty: fix up and plug in tty_ioctl kernel-doc 2023-10-03T12:31:15+00:00 Jiri Slaby (SUSE) jirislaby@kernel.org 2023-09-19T08:51:49+00:00 3b1a696bb96882d4f85c7cca07383d94cb7c2de3 The ioctl helpers are well documented, except they are not plugged in the Documentation. So fix up the minor issues in the kernel-doc and plug it in. The minor issues include: * bad \t on every line (sphinx misinterprets the description otherwise) * missing colon after Return * superfluous \n after the comment * make some struct members and constants a hyperlink * and so on Perhaps better to use --word-diff if one wants to see the "real" changes. Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org> Link: https://lore.kernel.org/r/20230919085156.1578-9-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
The ioctl helpers are well documented, except they are not plugged in
the Documentation. So fix up the minor issues in the kernel-doc and plug
it in.

The minor issues include:
* bad \t on every line (sphinx misinterprets the description otherwise)
* missing colon after Return
* superfluous \n after the comment
* make some struct members and constants a hyperlink
* and so on

Perhaps better to use --word-diff if one wants to see the "real"
changes.

Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20230919085156.1578-9-jirislaby@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tty: invert return values of tty_{,un}throttle_safe() 2023-10-03T12:31:15+00:00 Jiri Slaby (SUSE) jirislaby@kernel.org 2023-09-19T08:51:48+00:00 5b4f9cf3cc339565be1ecd64d608b6ffdfba7c1e If tty_{,un}throttle_safe() returned true on success (similar to *_trylock()), it would make the conditions in callers more obvious. So perform the switch to these inverted values (and fix the callers). Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org> Link: https://lore.kernel.org/r/20230919085156.1578-8-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
If tty_{,un}throttle_safe() returned true on success (similar to
*_trylock()), it would make the conditions in callers more obvious. So
perform the switch to these inverted values (and fix the callers).

Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20230919085156.1578-8-jirislaby@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tty: switch tty_{,un}throttle_safe() to return a bool 2023-10-03T12:31:15+00:00 Jiri Slaby (SUSE) jirislaby@kernel.org 2023-09-19T08:51:47+00:00 c2a36609dab3b7c937ab95bfb8b98e72391f772e They return 0 or 1 -- a boolean value, so make it clear than noone should expect negative or other values. Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org> Link: https://lore.kernel.org/r/20230919085156.1578-7-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
They return 0 or 1 -- a boolean value, so make it clear than noone
should expect negative or other values.

Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20230919085156.1578-7-jirislaby@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tty: make tty_change_softcar() more understandable 2023-08-11T19:12:44+00:00 Jiri Slaby (SUSE) jirislaby@kernel.org 2023-08-10T09:14:42+00:00 77b425e4efe5a40cab602bb6538d8d27ba4d9948 * rename 'arg' to 'enable' as that is what it means. * make 'bit' a tcflag_t, not int, as that is what cflags are. Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org> Link: https://lore.kernel.org/r/20230810091510.13006-9-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
* rename 'arg' to 'enable' as that is what it means.
* make 'bit' a tcflag_t, not int, as that is what cflags are.

Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20230810091510.13006-9-jirislaby@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tty: drop tty_debug_wait_until_sent() 2023-08-11T19:12:44+00:00 Jiri Slaby (SUSE) jirislaby@kernel.org 2023-08-10T09:14:41+00:00 d1150d29906cdfddf8a43aaf5a4cafa4f22474fa It's a nop for everyone as TTY_DEBUG_WAIT_UNTIL_SENT is never set. Provided, we have better debugging/printout mechanisms nowadays, remove this mechanism. Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org> Link: https://lore.kernel.org/r/20230810091510.13006-8-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
It's a nop for everyone as TTY_DEBUG_WAIT_UNTIL_SENT is never set.
Provided, we have better debugging/printout mechanisms nowadays, remove
this mechanism.

Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20230810091510.13006-8-jirislaby@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tty: change tty_write_lock()'s ndelay parameter to bool 2023-08-11T19:12:44+00:00 Jiri Slaby (SUSE) jirislaby@kernel.org 2023-08-10T09:14:39+00:00 af815336556df28f800669c58ab3bdad7d786b98 It's a yes-no parameter, so convert it to bool to be obvious. Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org> Link: https://lore.kernel.org/r/20230810091510.13006-6-jirislaby@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
It's a yes-no parameter, so convert it to bool to be obvious.

Signed-off-by: "Jiri Slaby (SUSE)" <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20230810091510.13006-6-jirislaby@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tty: Prevent writing chars during tcsetattr TCSADRAIN/FLUSH 2023-03-29T08:55:53+00:00 Ilpo Järvinen ilpo.jarvinen@linux.intel.com 2023-03-17T11:33:17+00:00 094fb49a2d0d6827c86d2e0840873e6db0c491d2 If userspace races tcsetattr() with a write, the drained condition might not be guaranteed by the kernel. There is a race window after checking Tx is empty before tty_set_termios() takes termios_rwsem for write. During that race window, more characters can be queued by a racing writer. Any ongoing transmission might produce garbage during HW's ->set_termios() call. The intent of TCSADRAIN/FLUSH seems to be preventing such a character corruption. If those flags are set, take tty's write lock to stop any writer before performing the lower layer Tx empty check and wait for the pending characters to be sent (if any). The initial wait for all-writers-done must be placed outside of tty's write lock to avoid deadlock which makes it impossible to use tty_wait_until_sent(). The write lock is retried if a racing write is detected. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Link: https://lore.kernel.org/r/20230317113318.31327-2-ilpo.jarvinen@linux.intel.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
If userspace races tcsetattr() with a write, the drained condition
might not be guaranteed by the kernel. There is a race window after
checking Tx is empty before tty_set_termios() takes termios_rwsem for
write. During that race window, more characters can be queued by a
racing writer.

Any ongoing transmission might produce garbage during HW's
->set_termios() call. The intent of TCSADRAIN/FLUSH seems to be
preventing such a character corruption. If those flags are set, take
tty's write lock to stop any writer before performing the lower layer
Tx empty check and wait for the pending characters to be sent (if any).

The initial wait for all-writers-done must be placed outside of tty's
write lock to avoid deadlock which makes it impossible to use
tty_wait_until_sent(). The write lock is retried if a racing write is
detected.

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://lore.kernel.org/r/20230317113318.31327-2-ilpo.jarvinen@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tty_ioctl: Use BIT() for internal flags 2023-03-17T14:01:08+00:00 Ilpo Järvinen ilpo.jarvinen@linux.intel.com 2023-03-09T08:20:29+00:00 0b05223bdf6058b96a596609c50a18c7b010224d Convert internal flags to use BIT(). Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Link: https://lore.kernel.org/r/20230309082035.14880-3-ilpo.jarvinen@linux.intel.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
Convert internal flags to use BIT().

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://lore.kernel.org/r/20230309082035.14880-3-ilpo.jarvinen@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tty: Return bool from tty_termios_hw_change() 2023-01-19T15:04:35+00:00 Ilpo Järvinen ilpo.jarvinen@linux.intel.com 2023-01-17T09:03:56+00:00 87f22db4c251ff92d588c2cc710031a59d5e4ec0 Change tty_termios_hw_change() to return bool. Reviewed-by: Jiri Slaby <jirislaby@kernel.org> Reviewed-by: Johan Hovold <johan@kernel.org> Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Link: https://lore.kernel.org/r/20230117090358.4796-11-ilpo.jarvinen@linux.intel.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
Change tty_termios_hw_change() to return bool.

Reviewed-by: Jiri Slaby <jirislaby@kernel.org>
Reviewed-by: Johan Hovold <johan@kernel.org>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://lore.kernel.org/r/20230117090358.4796-11-ilpo.jarvinen@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>