| Message ID | 1649316351-9220-2-git-send-email-quic_vnivarth@quicinc.com (mailing list archive) |
|---|---|
| State | Not Applicable |
| Headers | show |
| Series | [V3] drivers/tty/serial/qcom-geni-serial: Do stop_rx in suspend path for console if console_suspend is disabled | expand |
On 07. 04. 22, 9:25, Vijaya Krishna Nivarthi wrote: > For the case of console_suspend disabled, if back to back suspend/resume > test is executed, at the end of test, sometimes console would appear to > be frozen not responding to input. This would happen because, for > console_suspend disabled, suspend/resume routines only turn resources > off/on but don't do a port close/open. > As a result, during resume, some rx transactions come in before system is > ready, malfunction of rx happens in turn resulting in console appearing > to be stuck. > > Do a stop_rx in suspend sequence to prevent this. start_rx is already > present in resume sequence as part of call to set_termios which does a > stop_rx/start_rx. So why is it OK for every other driver? Should uart_suspend_port() be fixed instead? > Signed-off-by: Vijaya Krishna Nivarthi <quic_vnivarth@quicinc.com> > --- > v3: swapped the order of conditions to be more human readable > v2: restricted patch to contain only stop_rx in suspend sequence > v1: intial patch contained 2 additional unrelated changes in vicinity > --- > drivers/tty/serial/qcom_geni_serial.c | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/drivers/tty/serial/qcom_geni_serial.c b/drivers/tty/serial/qcom_geni_serial.c > index 1543a60..53723d2 100644 > --- a/drivers/tty/serial/qcom_geni_serial.c > +++ b/drivers/tty/serial/qcom_geni_serial.c > @@ -1481,6 +1481,10 @@ static int __maybe_unused qcom_geni_serial_sys_suspend(struct device *dev) > struct uart_port *uport = &port->uport; > struct qcom_geni_private_data *private_data = uport->private_data; > > + /* do a stop_rx here, start_rx is handled in uart_resume_port by call to setermios */ > + if (uart_console(uport) && !console_suspend_enabled) > + uport->ops->stop_rx(uport); > + > /* > * This is done so we can hit the lowest possible state in suspend > * even with no_console_suspend thanks,
On 4/7/2022 1:21 PM, Jiri Slaby wrote: > On 07. 04. 22, 9:25, Vijaya Krishna Nivarthi wrote: >> For the case of console_suspend disabled, if back to back suspend/resume >> test is executed, at the end of test, sometimes console would appear to >> be frozen not responding to input. This would happen because, for >> console_suspend disabled, suspend/resume routines only turn resources >> off/on but don't do a port close/open. >> As a result, during resume, some rx transactions come in before >> system is >> ready, malfunction of rx happens in turn resulting in console appearing >> to be stuck. >> >> Do a stop_rx in suspend sequence to prevent this. start_rx is already >> present in resume sequence as part of call to set_termios which does a >> stop_rx/start_rx. > > So why is it OK for every other driver? Should uart_suspend_port() be > fixed instead? For qcom driver we know that set_termios() call in uart_suspend_port() will recover with a call to start_rx. However that may not be the case with other drivers. We can move stop_rx to uart_suspend_port() and additionally have a start_rx in uart_resume_port() Please let know if such a change would be ok. Thank you.
On Fri, Apr 08, 2022 at 11:45:11AM +0530, Vijaya Krishna Nivarthi wrote: > > On 4/7/2022 1:21 PM, Jiri Slaby wrote: > > On 07. 04. 22, 9:25, Vijaya Krishna Nivarthi wrote: > > > For the case of console_suspend disabled, if back to back suspend/resume > > > test is executed, at the end of test, sometimes console would appear to > > > be frozen not responding to input. This would happen because, for > > > console_suspend disabled, suspend/resume routines only turn resources > > > off/on but don't do a port close/open. > > > As a result, during resume, some rx transactions come in before > > > system is > > > ready, malfunction of rx happens in turn resulting in console appearing > > > to be stuck. > > > > > > Do a stop_rx in suspend sequence to prevent this. start_rx is already > > > present in resume sequence as part of call to set_termios which does a > > > stop_rx/start_rx. > > > > So why is it OK for every other driver? Should uart_suspend_port() be > > fixed instead? > > For qcom driver we know that set_termios() call in uart_suspend_port() will > recover with a call to start_rx. > However that may not be the case with other drivers. > > We can move stop_rx to uart_suspend_port() and additionally have a start_rx > in uart_resume_port() > Please let know if such a change would be ok. This should not be something that each individual driver has to do, please fix it for everyone. thanks, greg k-h
diff --git a/drivers/tty/serial/qcom_geni_serial.c b/drivers/tty/serial/qcom_geni_serial.c index 1543a60..53723d2 100644 --- a/drivers/tty/serial/qcom_geni_serial.c +++ b/drivers/tty/serial/qcom_geni_serial.c @@ -1481,6 +1481,10 @@ static int __maybe_unused qcom_geni_serial_sys_suspend(struct device *dev) struct uart_port *uport = &port->uport; struct qcom_geni_private_data *private_data = uport->private_data; + /* do a stop_rx here, start_rx is handled in uart_resume_port by call to setermios */ + if (uart_console(uport) && !console_suspend_enabled) + uport->ops->stop_rx(uport); + /* * This is done so we can hit the lowest possible state in suspend * even with no_console_suspend
For the case of console_suspend disabled, if back to back suspend/resume test is executed, at the end of test, sometimes console would appear to be frozen not responding to input. This would happen because, for console_suspend disabled, suspend/resume routines only turn resources off/on but don't do a port close/open. As a result, during resume, some rx transactions come in before system is ready, malfunction of rx happens in turn resulting in console appearing to be stuck. Do a stop_rx in suspend sequence to prevent this. start_rx is already present in resume sequence as part of call to set_termios which does a stop_rx/start_rx. Signed-off-by: Vijaya Krishna Nivarthi <quic_vnivarth@quicinc.com> --- v3: swapped the order of conditions to be more human readable v2: restricted patch to contain only stop_rx in suspend sequence v1: intial patch contained 2 additional unrelated changes in vicinity --- drivers/tty/serial/qcom_geni_serial.c | 4 ++++ 1 file changed, 4 insertions(+)