Formally, currently there is no memory leak, but if
serial8250_ports[line] is reused with other 8250 driver, then em485
will be already activated and it will cause issues.
Fixes: e490c9144cfa ("tty: Add software emulated RS485 support for 8250")
Signed-off-by: Matwey V. Kornilov <[email protected]>
---
Hi Peter,
There is an issue with my previous patch. Under specific circumstances,
the emulation can be enabled for drivers which don't want (and support) it.
Unfortunately, I can't test this patch because kgdb over 8250_omap console
is the single one option for me on BBB. This patch testing involves
8250_omap module unloading and then inspecting serial8250_ports.
You say that you are preparing some unit-tests. Could you include a test
for this specific issue? And then probably this patch has to be applied.
drivers/tty/serial/8250/8250_core.c | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/drivers/tty/serial/8250/8250_core.c
b/drivers/tty/serial/8250/8250_core.c
index c9720a9..a242881 100644
--- a/drivers/tty/serial/8250/8250_core.c
+++ b/drivers/tty/serial/8250/8250_core.c
@@ -1068,6 +1068,15 @@ void serial8250_unregister_port(int line)
struct uart_8250_port *uart = &serial8250_ports[line];
mutex_lock(&serial_mutex);
+
+ if (uart->em485) {
+ unsigned long flags;
+
+ spin_lock_irqsave(&uart->port.lock, flags);
+ serial8250_em485_destroy(uart);
+ spin_unlock_irqrestore(&uart->port.lock, flags);
+ }
+
uart_remove_one_port(&serial8250_reg, &uart->port);
if (serial8250_isa_devs) {
uart->port.flags &= ~UPF_BOOT_AUTOCONF;
--
2.7.0
