Make sure UART is powered up when dumping MCTRL status
RFC PATCH 1/1 Make sure UART is powered up when dumping MCTRL status
Greetings,
Since serial devices are powered down when not in use and some of those devices cannot be accessed when powered down, we need to enable power around calls to get_mcrtl() when dumping port state via uart_line_info(). This resolves hangs observed on some machines while reading serial device registers when a port is powered off.
---
drivers/serial/serial_core.c | 12 ++++++++++++ 1 files changed, 12 insertions(+)
drivers/serial/serial_core.c +12 -0
@@ -1652,6 +1652,7 @@ static const char *uart_type(struct uartstatic int uart_line_info (char *buf,struct uart_driver *drv,int i) {struct uart_state *state = drv->state + i; +int pm_state;struct uart_port *port = state->port;char stat_buf[32 ];unsigned int status;
@@ -1674,9 +1675,16 @@ static int uart_line_info(char *buf, strif (capable (CAP_SYS_ADMIN )) { +mutex_lock (&state->mutex); + pm_state = state->pm_state; +if (pm_state) +uart_change_pm (state,0 );spin_lock_irq (&port->lock); status = port->ops->get_mctrl (port);spin_unlock_irq (&port->lock); +if (pm_state) +uart_change_pm (state, pm_state); +mutex_unlock (&state->mutex); ret +=sprintf (buf + ret," tx:%d rx:%d" , port->icount.tx, port->icount.rx);
@@ -2068,6 +2076,10 @@ uart_configure_port(struct uart_driver *uart_report_port (drv, port); +/* Power up port for set_mctrl() */ +if (!uart_console (port)) +uart_change_pm (state,0 ); +/* * Ensure that the modem control lines are de-activated. * We probably don't need a spinlock around this, but
Comments appreciated. TIA!
--
Regards,
George
drivers/serial/serial_core.c uart_configure_port(struct uart_driver * · line 2082
Quoted 8 lines · George G. Davis
drivers/serial/serial_core.c +4
-0@@ -2068,6 +2076,10 @@ uart_configure_port(struct uart_driver *uart_report_port (drv, port); +/* Power up port for set_mctrl() */ +if (!uart_console (port)) +uart_change_pm (state,0 ); +
If it's possible for uarts to be powered down here, wouldn't it be a good idea to ensure that the console is also powered up?
Signature · 3 lines
Russell King Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/ maintainer of: 2.6 Serial core
Quoted 5 lines · Russell King
Quoted 4 lines · George G. Davis
drivers/serial/serial_core.c +0
-0@@ -2068,6 +2076,10 @@ uart_configure_port(struct uart_driver *uart_report_port (drv, port);
Quoted 7 lines
drivers/serial/serial_core.c uart_configure_port(struct uart_driver * · line 2082
Quoted 4 lines
drivers/serial/serial_core.c +4
-0+/* Power up port for set_mctrl() */ +if (!uart_console (port)) +uart_change_pm (state,0 ); +If it's possible for uarts to be powered down here, wouldn't it be a good idea to ensure that the console is also powered up?
Hrm, yes. The above !uart_console(port) merely assumed that the device is already powered up if it's the selected console. In that way we only call uart_change_pm() to power it up once, earlier but I haven't verified that it was indeed done properly. That check can be removed I guess.
Thanks!
--
Regards,
George
Quoted 5 lines · Russell King
Quoted 4 lines · George G. Davis
(quoted diff) +0
-0@@ -2068,6 +2076,10 @@ uart_configure_port(struct uart_driver * uart_report_port(drv, port);
Quoted 7 lines
Review uart_configure_port(struct uart_driver * · line 2082
Quoted 4 lines
(quoted diff) +4
-0+ /* Power up port for set_mctrl() */ + if (!uart_console(port)) + uart_change_pm(state, 0); +If it's possible for uarts to be powered down here, wouldn't it be a good idea to ensure that the console is also powered up?
Apologies for the delay... Here's an update which insures all serial ports are powered up prior to accessing device registers.
Since serial devices are powered down when not in use and some of those devices cannot be accessed when powered down, we need to enable power around calls to get_mcrtl() when dumping port state via uart_line_info(). This resolves hangs observed on some machines while reading serial device registers when a port is powered off.
drivers/serial/serial_core.c | 11 +++++++++++ 1 files changed, 11 insertions(+)
drivers/serial/serial_core.c +11 -0
@@ -1652,6 +1652,7 @@ static const char *uart_type(struct uartstatic int uart_line_info (char *buf,struct uart_driver *drv,int i) {struct uart_state *state = drv->state + i; +int pm_state;struct uart_port *port = state->port;char stat_buf[32 ];unsigned int status;
@@ -1674,9 +1675,16 @@ static int uart_line_info(char *buf, strif (capable (CAP_SYS_ADMIN )) { +mutex_lock (&state->mutex); + pm_state = state->pm_state; +if (pm_state) +uart_change_pm (state,0 );spin_lock_irq (&port->lock); status = port->ops->get_mctrl (port);spin_unlock_irq (&port->lock); +if (pm_state) +uart_change_pm (state, pm_state); +mutex_unlock (&state->mutex); ret +=sprintf (buf + ret," tx:%d rx:%d" , port->icount.tx, port->icount.rx);
@@ -2071,6 +2079,9 @@ uart_configure_port(struct uart_driver *uart_report_port (drv, port); +/* Power up port for set_mctrl() */ +uart_change_pm (state,0 ); +/* * Ensure that the modem control lines are de-activated. * We probably don't need a spinlock around this, but
--
Regards,
George
Ping,
Quoted 17 lines · George G. Davis
Quoted 5 lines · Russell King
Quoted 4 lines · George G. Davis
drivers/serial/serial_core.c +0
-0@@ -2068,6 +2076,10 @@ uart_configure_port(struct uart_driver *uart_report_port (drv, port);Quoted 7 lines
drivers/serial/serial_core.c uart_configure_port(struct uart_driver * · line 2084
Quoted 4 lines
drivers/serial/serial_core.c +4
-0+/* Power up port for set_mctrl() */ +if (!uart_console (port)) +uart_change_pm (state,0 ); +If it's possible for uarts to be powered down here, wouldn't it be a good idea to ensure that the console is also powered up?
Apologies for the delay... Here's an update which insures all serial ports are powered up prior to accessing device registers.
Quoted 7 lines
Since serial devices are powered down when not in use and some of those devices cannot be accessed when powered down, we need to enable power around calls to
get_mcrtl()when dumping port state viauart_line_info(). This resolves hangs observed on some machines while reading serial device registers when a port is powered off.Signed-off-by George G. Davis <[email protected]>
Any chance of this getting committed? The Freescale i.MX31 shuts off clocks when ports are not in use and the kernel hangs without this change. It's likely other targets may start managing clocks in this way too. So this may be required for others at some point. In the meantime, it should be harmless for those targets which don't manage device clocks, no?
TIA!
--
Regards,
George