[v2,2/8] serial: sh-sci: Check if TX data was written to device in .tx_empty()

Commit Message

Claudiu Nov. 8, 2024, 10:05 a.m. UTC

From: Claudiu Beznea <[e-mail]>

On the Renesas RZ/G3S, when doing suspend to RAM, the uart_suspend_port()
is called. The uart_suspend_port() calls 3 times the
struct uart_port::ops::tx_empty() before shutting down the port.

According to the documentation, the struct uart_port::ops::tx_empty()
API tests whether the transmitter FIFO and shifter for the port is
empty.

The Renesas RZ/G3S SCIFA IP reports the number of data units stored in the
transmit FIFO through the FDR (FIFO Data Count Register). The data units
in the FIFOs are written in the shift register and transmitted from there.
The TEND bit in the Serial Status Register reports if the data was
transmitted from the shift register.

In the previous code, in the tx_empty() API implemented by the sh-sci
driver, it is considered that the TX is empty if the hardware reports the
TEND bit set and the number of data units in the FIFO is zero.

According to the HW manual, the TEND bit has the following meaning:

0: Transmission is in the waiting state or in progress.
1: Transmission is completed.

It has been noticed that when opening the serial device w/o using it and
then switch to a power saving mode, the tx_empty() call in the
uart_port_suspend() function fails, leading to the "Unable to drain
transmitter" message being printed on the console. This is because the
TEND=0 if nothing has been transmitted and the FIFOs are empty. As the
TEND=0 has double meaning (waiting state, in progress) we can't
determined the scenario described above.

Add a software workaround for this. This sets a variable if any data has
been sent on the serial console (when using PIO) or if the DMA callback has
been called (meaning something has been transmitted).

Fixes: 73a19e4c0301 ("serial: sh-sci: Add DMA support.")
Cc: [e-mail]
Signed-off-by: Claudiu Beznea <[e-mail]>
---

Changes in v2:
- use bool type instead of atomic_t

 drivers/tty/serial/sh-sci.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

Comments

Jiri Slaby Nov. 8, 2024, 10:57 a.m. UTC | #1

On 08. 11. 24, 11:05, Claudiu wrote:
> From: Claudiu Beznea <[e-mail]>
> 
> On the Renesas RZ/G3S, when doing suspend to RAM, the uart_suspend_port()
> is called. The uart_suspend_port() calls 3 times the
> struct uart_port::ops::tx_empty() before shutting down the port.
> 
> According to the documentation, the struct uart_port::ops::tx_empty()
> API tests whether the transmitter FIFO and shifter for the port is
> empty.
> 
> The Renesas RZ/G3S SCIFA IP reports the number of data units stored in the
> transmit FIFO through the FDR (FIFO Data Count Register). The data units
> in the FIFOs are written in the shift register and transmitted from there.
> The TEND bit in the Serial Status Register reports if the data was
> transmitted from the shift register.
> 
> In the previous code, in the tx_empty() API implemented by the sh-sci
> driver, it is considered that the TX is empty if the hardware reports the
> TEND bit set and the number of data units in the FIFO is zero.
> 
> According to the HW manual, the TEND bit has the following meaning:
> 
> 0: Transmission is in the waiting state or in progress.
> 1: Transmission is completed.
> 
> It has been noticed that when opening the serial device w/o using it and
> then switch to a power saving mode, the tx_empty() call in the
> uart_port_suspend() function fails, leading to the "Unable to drain
> transmitter" message being printed on the console. This is because the
> TEND=0 if nothing has been transmitted and the FIFOs are empty. As the
> TEND=0 has double meaning (waiting state, in progress) we can't
> determined the scenario described above.
> 
> Add a software workaround for this. This sets a variable if any data has
> been sent on the serial console (when using PIO) or if the DMA callback has
> been called (meaning something has been transmitted).
> 
> Fixes: 73a19e4c0301 ("serial: sh-sci: Add DMA support.")
> Cc: [e-mail]
> Signed-off-by: Claudiu Beznea <[e-mail]>
> ---
> 
> Changes in v2:
> - use bool type instead of atomic_t
> 
>   drivers/tty/serial/sh-sci.c | 11 +++++++++++
>   1 file changed, 11 insertions(+)
> 
> diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c
> index 136e0c257af1..65514d37bfe2 100644
> --- a/drivers/tty/serial/sh-sci.c
> +++ b/drivers/tty/serial/sh-sci.c
> @@ -157,6 +157,7 @@ struct sci_port {
>   
>   	bool has_rtscts;
>   	bool autorts;
> +	bool first_time_tx;

This is a misnomer. It suggests to be set only during the first TX. What 
about ::did_tx, ::performed_tx, ::transmitted, or alike?

> @@ -885,6 +887,7 @@ static void sci_transmit_chars(struct uart_port *port)
>   		}
>   
>   		sci_serial_out(port, SCxTDR, c);
> +		s->first_time_tx = true;
>   
>   		port->icount.tx++;
>   	} while (--count > 0);
> @@ -1241,6 +1244,8 @@ static void sci_dma_tx_complete(void *arg)
>   	if (kfifo_len(&tport->xmit_fifo) < WAKEUP_CHARS)
>   		uart_write_wakeup(port);
>   
> +	s->first_time_tx = true;

This is too late IMO. The first in-flight dma won't be accounted in 
sci_tx_empty(). From DMA submit up to now.

> @@ -2076,6 +2081,10 @@ static unsigned int sci_tx_empty(struct uart_port *port)
>   {
>   	unsigned short status = sci_serial_in(port, SCxSR);
>   	unsigned short in_tx_fifo = sci_txfill(port);
> +	struct sci_port *s = to_sci_port(port);
> +
> +	if (!s->first_time_tx)
> +		return TIOCSER_TEMT;

So perhaps check if there is a TX DMA running here too?

>   
>   	return (status & SCxSR_TEND(port)) && !in_tx_fifo ? TIOCSER_TEMT : 0;
>   }

thanks,

Claudiu Nov. 8, 2024, 12:19 p.m. UTC | #2

Hi, Jiri,

On 08.11.2024 12:57, Jiri Slaby wrote:
> On 08. 11. 24, 11:05, Claudiu wrote:
>> From: Claudiu Beznea <[e-mail]>
>>
>> On the Renesas RZ/G3S, when doing suspend to RAM, the uart_suspend_port()
>> is called. The uart_suspend_port() calls 3 times the
>> struct uart_port::ops::tx_empty() before shutting down the port.
>>
>> According to the documentation, the struct uart_port::ops::tx_empty()
>> API tests whether the transmitter FIFO and shifter for the port is
>> empty.
>>
>> The Renesas RZ/G3S SCIFA IP reports the number of data units stored in the
>> transmit FIFO through the FDR (FIFO Data Count Register). The data units
>> in the FIFOs are written in the shift register and transmitted from there.
>> The TEND bit in the Serial Status Register reports if the data was
>> transmitted from the shift register.
>>
>> In the previous code, in the tx_empty() API implemented by the sh-sci
>> driver, it is considered that the TX is empty if the hardware reports the
>> TEND bit set and the number of data units in the FIFO is zero.
>>
>> According to the HW manual, the TEND bit has the following meaning:
>>
>> 0: Transmission is in the waiting state or in progress.
>> 1: Transmission is completed.
>>
>> It has been noticed that when opening the serial device w/o using it and
>> then switch to a power saving mode, the tx_empty() call in the
>> uart_port_suspend() function fails, leading to the "Unable to drain
>> transmitter" message being printed on the console. This is because the
>> TEND=0 if nothing has been transmitted and the FIFOs are empty. As the
>> TEND=0 has double meaning (waiting state, in progress) we can't
>> determined the scenario described above.
>>
>> Add a software workaround for this. This sets a variable if any data has
>> been sent on the serial console (when using PIO) or if the DMA callback has
>> been called (meaning something has been transmitted).
>>
>> Fixes: 73a19e4c0301 ("serial: sh-sci: Add DMA support.")
>> Cc: [e-mail]
>> Signed-off-by: Claudiu Beznea <[e-mail]>
>> ---
>>
>> Changes in v2:
>> - use bool type instead of atomic_t
>>
>>   drivers/tty/serial/sh-sci.c | 11 +++++++++++
>>   1 file changed, 11 insertions(+)
>>
>> diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c
>> index 136e0c257af1..65514d37bfe2 100644
>> --- a/drivers/tty/serial/sh-sci.c
>> +++ b/drivers/tty/serial/sh-sci.c
>> @@ -157,6 +157,7 @@ struct sci_port {
>>         bool has_rtscts;
>>       bool autorts;
>> +    bool first_time_tx;
> 
> This is a misnomer. It suggests to be set only during the first TX. 

I chose this naming as this was the scenario I discovered it didn't work.
Reproducible though these steps:

1/ open the serial device (w/o running any TX/RX)
2/ call tx_empty()

What
> about ::did_tx, ::performed_tx, ::transmitted, or alike?

I have nothing against any of these. Can you please let me know if you have
a preferred one?

> 
>> @@ -885,6 +887,7 @@ static void sci_transmit_chars(struct uart_port *port)
>>           }
>>             sci_serial_out(port, SCxTDR, c);
>> +        s->first_time_tx = true;
>>             port->icount.tx++;
>>       } while (--count > 0);
>> @@ -1241,6 +1244,8 @@ static void sci_dma_tx_complete(void *arg)
>>       if (kfifo_len(&tport->xmit_fifo) < WAKEUP_CHARS)
>>           uart_write_wakeup(port);
>>   +    s->first_time_tx = true;
> 
> This is too late IMO. The first in-flight dma won't be accounted in
> sci_tx_empty(). From DMA submit up to now.

If it's in-flight we can't determine it's status anyway with one variable.
We can set this variable later but it wouldn't tell the truth as the TX
might be in progress anyway or may have been finished?

The hardware might help with this though the TEND bit. According to the HW
manual, the TEND bit has the following meaning:

0: Transmission is in the waiting state or in progress.
1: Transmission is completed.

But the problem, from my point of view, is that the 0 has double meaning.

I noticed the tx_empty() is called in kernel multiple times before
declaring TX is empty or not. E.g., uart_suspend_port() call it 3 times,
uart_wait_until_sent() call it in a while () look with a timeout. There is
the uart_ioctl() which calls it though uart_get_lsr_info() only one time
but I presumed the user space might implement the same multiple trials
approach before declaring it empty.

Because of this I considered it wouldn't be harmful for the scenario you
described "The first in-flight dma won't be accounted in sci_tx_empty()"
as the user may try again later to check the status. For this reason I also
chose to have no extra locking around this variable.

Please let me know if you consider otherwise.

Thank you,
Claudiu Beznea

> 
>> @@ -2076,6 +2081,10 @@ static unsigned int sci_tx_empty(struct uart_port
>> *port)
>>   {
>>       unsigned short status = sci_serial_in(port, SCxSR);
>>       unsigned short in_tx_fifo = sci_txfill(port);
>> +    struct sci_port *s = to_sci_port(port);
>> +
>> +    if (!s->first_time_tx)
>> +        return TIOCSER_TEMT;
> 
> So perhaps check if there is a TX DMA running here too?
> 
>>         return (status & SCxSR_TEND(port)) && !in_tx_fifo ? TIOCSER_TEMT
>> : 0;
>>   }
> 
> thanks,

Claudiu Nov. 12, 2024, 8:35 a.m. UTC | #3

Hi, Jiri,

On 08.11.2024 14:19, Claudiu Beznea wrote:
>>> @@ -885,6 +887,7 @@ static void sci_transmit_chars(struct uart_port *port)
>>>           }
>>>             sci_serial_out(port, SCxTDR, c);
>>> +        s->first_time_tx = true;
>>>             port->icount.tx++;
>>>       } while (--count > 0);
>>> @@ -1241,6 +1244,8 @@ static void sci_dma_tx_complete(void *arg)
>>>       if (kfifo_len(&tport->xmit_fifo) < WAKEUP_CHARS)
>>>           uart_write_wakeup(port);
>>>   +    s->first_time_tx = true;
>> This is too late IMO. The first in-flight dma won't be accounted in
>> sci_tx_empty(). From DMA submit up to now.
> If it's in-flight we can't determine it's status anyway with one variable.
> We can set this variable later but it wouldn't tell the truth as the TX
> might be in progress anyway or may have been finished?
> 
> The hardware might help with this though the TEND bit. According to the HW
> manual, the TEND bit has the following meaning:
> 
> 0: Transmission is in the waiting state or in progress.
> 1: Transmission is completed.
> 
> But the problem, from my point of view, is that the 0 has double meaning.
> 
> I noticed the tx_empty() is called in kernel multiple times before
> declaring TX is empty or not. E.g., uart_suspend_port() call it 3 times,
> uart_wait_until_sent() call it in a while () look with a timeout. There is
> the uart_ioctl() which calls it though uart_get_lsr_info() only one time
> but I presumed the user space might implement the same multiple trials
> approach before declaring it empty.
> 
> Because of this I considered it wouldn't be harmful for the scenario you
> described "The first in-flight dma won't be accounted in sci_tx_empty()"
> as the user may try again later to check the status. For this reason I also
> chose to have no extra locking around this variable.
> 
> Please let me know if you consider otherwise.

With the above explanation, can you please let me know if you still
consider I should change the approach for this patch?

Thank you,
Claudiu Beznea

Jiri Slaby Nov. 14, 2024, 6:26 a.m. UTC | #4

Hi,

On 08. 11. 24, 13:19, Claudiu Beznea wrote:
> On 08.11.2024 12:57, Jiri Slaby wrote:
>> On 08. 11. 24, 11:05, Claudiu wrote:
...
>>> --- a/drivers/tty/serial/sh-sci.c
>>> +++ b/drivers/tty/serial/sh-sci.c
>>> @@ -157,6 +157,7 @@ struct sci_port {
>>>          bool has_rtscts;
>>>        bool autorts;
>>> +    bool first_time_tx;
>>
>> This is a misnomer. It suggests to be set only during the first TX.
> 
> I chose this naming as this was the scenario I discovered it didn't work.
> Reproducible though these steps:
> 
> 1/ open the serial device (w/o running any TX/RX)
> 2/ call tx_empty()
> 
> What
>> about ::did_tx, ::performed_tx, ::transmitted, or alike?
> 
> I have nothing against any of these. Can you please let me know if you have
> a preferred one?

No, you choose, or invent even better one :). Or let AI do it for you.

>>> @@ -885,6 +887,7 @@ static void sci_transmit_chars(struct uart_port *port)
>>>            }
>>>              sci_serial_out(port, SCxTDR, c);
>>> +        s->first_time_tx = true;
>>>              port->icount.tx++;
>>>        } while (--count > 0);
>>> @@ -1241,6 +1244,8 @@ static void sci_dma_tx_complete(void *arg)
>>>        if (kfifo_len(&tport->xmit_fifo) < WAKEUP_CHARS)
>>>            uart_write_wakeup(port);
>>>    +    s->first_time_tx = true;
>>
>> This is too late IMO. The first in-flight dma won't be accounted in
>> sci_tx_empty(). From DMA submit up to now.
> 
> If it's in-flight we can't determine it's status anyway with one variable.
> We can set this variable later but it wouldn't tell the truth as the TX
> might be in progress anyway or may have been finished?
> 
> The hardware might help with this though the TEND bit. According to the HW
> manual, the TEND bit has the following meaning:
> 
> 0: Transmission is in the waiting state or in progress.
> 1: Transmission is completed.
> 
> But the problem, from my point of view, is that the 0 has double meaning.
> 
> I noticed the tx_empty() is called in kernel multiple times before
> declaring TX is empty or not. E.g., uart_suspend_port() call it 3 times,
> uart_wait_until_sent() call it in a while () look with a timeout. There is
> the uart_ioctl() which calls it though uart_get_lsr_info() only one time
> but I presumed the user space might implement the same multiple trials
> approach before declaring it empty.
> 
> Because of this I considered it wouldn't be harmful for the scenario you
> described "The first in-flight dma won't be accounted in sci_tx_empty()"
> as the user may try again later to check the status. For this reason I also
> chose to have no extra locking around this variable.

What about the below?

>>> @@ -2076,6 +2081,10 @@ static unsigned int sci_tx_empty(struct uart_port
>>> *port)
>>>    {
>>>        unsigned short status = sci_serial_in(port, SCxSR);
>>>        unsigned short in_tx_fifo = sci_txfill(port);
>>> +    struct sci_port *s = to_sci_port(port);
>>> +
>>> +    if (!s->first_time_tx)
>>> +        return TIOCSER_TEMT;
>>
>> So perhaps check if there is a TX DMA running here too?

This ^^^? Like dmaengine_tx_status()?

>>
>>>          return (status & SCxSR_TEND(port)) && !in_tx_fifo ? TIOCSER_TEMT
>>> : 0;
>>>    }
>>
>> thanks,

Jiri Slaby Nov. 14, 2024, 6:29 a.m. UTC | #5

On 14. 11. 24, 7:26, Jiri Slaby wrote:
>>>> --- a/drivers/tty/serial/sh-sci.c
>>>> +++ b/drivers/tty/serial/sh-sci.c
>>>> @@ -157,6 +157,7 @@ struct sci_port {
>>>>          bool has_rtscts;
>>>>        bool autorts;
>>>> +    bool first_time_tx;
>>>
>>> This is a misnomer. It suggests to be set only during the first TX.
>>
>> I chose this naming as this was the scenario I discovered it didn't work.
>> Reproducible though these steps:
>>
>> 1/ open the serial device (w/o running any TX/RX)
>> 2/ call tx_empty()
>>
>> What
>>> about ::did_tx, ::performed_tx, ::transmitted, or alike?
>>
>> I have nothing against any of these. Can you please let me know if you 
>> have
>> a preferred one?
> 
> No, you choose, or invent even better one :). Or let AI do it for you.

FWIW both gemini and chatgpt answered by "tx_occurred" to my question. 
Which I like the most, perhaps.

Claudiu Nov. 14, 2024, 8:11 a.m. UTC | #6

Hi, Jiri,

On 14.11.2024 08:26, Jiri Slaby wrote:
> Hi,
> 
> On 08. 11. 24, 13:19, Claudiu Beznea wrote:
>> On 08.11.2024 12:57, Jiri Slaby wrote:
>>> On 08. 11. 24, 11:05, Claudiu wrote:
> ...
>>>> --- a/drivers/tty/serial/sh-sci.c
>>>> +++ b/drivers/tty/serial/sh-sci.c
>>>> @@ -157,6 +157,7 @@ struct sci_port {
>>>>          bool has_rtscts;
>>>>        bool autorts;
>>>> +    bool first_time_tx;
>>>
>>> This is a misnomer. It suggests to be set only during the first TX.
>>
>> I chose this naming as this was the scenario I discovered it didn't work.
>> Reproducible though these steps:
>>
>> 1/ open the serial device (w/o running any TX/RX)
>> 2/ call tx_empty()
>>
>> What
>>> about ::did_tx, ::performed_tx, ::transmitted, or alike?
>>
>> I have nothing against any of these. Can you please let me know if you have
>> a preferred one?
> 
> No, you choose, or invent even better one :). Or let AI do it for you.
> 
>>>> @@ -885,6 +887,7 @@ static void sci_transmit_chars(struct uart_port *port)
>>>>            }
>>>>              sci_serial_out(port, SCxTDR, c);
>>>> +        s->first_time_tx = true;
>>>>              port->icount.tx++;
>>>>        } while (--count > 0);
>>>> @@ -1241,6 +1244,8 @@ static void sci_dma_tx_complete(void *arg)
>>>>        if (kfifo_len(&tport->xmit_fifo) < WAKEUP_CHARS)
>>>>            uart_write_wakeup(port);
>>>>    +    s->first_time_tx = true;
>>>
>>> This is too late IMO. The first in-flight dma won't be accounted in
>>> sci_tx_empty(). From DMA submit up to now.
>>
>> If it's in-flight we can't determine it's status anyway with one variable.
>> We can set this variable later but it wouldn't tell the truth as the TX
>> might be in progress anyway or may have been finished?
>>
>> The hardware might help with this though the TEND bit. According to the HW
>> manual, the TEND bit has the following meaning:
>>
>> 0: Transmission is in the waiting state or in progress.
>> 1: Transmission is completed.
>>
>> But the problem, from my point of view, is that the 0 has double meaning.
>>
>> I noticed the tx_empty() is called in kernel multiple times before
>> declaring TX is empty or not. E.g., uart_suspend_port() call it 3 times,
>> uart_wait_until_sent() call it in a while () look with a timeout. There is
>> the uart_ioctl() which calls it though uart_get_lsr_info() only one time
>> but I presumed the user space might implement the same multiple trials
>> approach before declaring it empty.
>>
>> Because of this I considered it wouldn't be harmful for the scenario you
>> described "The first in-flight dma won't be accounted in sci_tx_empty()"
>> as the user may try again later to check the status. For this reason I also
>> chose to have no extra locking around this variable.
> 
> What about the below?
> 
>>>> @@ -2076,6 +2081,10 @@ static unsigned int sci_tx_empty(struct uart_port
>>>> *port)
>>>>    {
>>>>        unsigned short status = sci_serial_in(port, SCxSR);
>>>>        unsigned short in_tx_fifo = sci_txfill(port);
>>>> +    struct sci_port *s = to_sci_port(port);
>>>> +
>>>> +    if (!s->first_time_tx)
>>>> +        return TIOCSER_TEMT;
>>>
>>> So perhaps check if there is a TX DMA running here too?
> 
> This ^^^? Like dmaengine_tx_status()?

I missed that I can use this ^. Thanks for pointing it.

Claudiu

> 
>>>
>>>>          return (status & SCxSR_TEND(port)) && !in_tx_fifo ? TIOCSER_TEMT
>>>> : 0;
>>>>    }
>>>
>>> thanks,

Patch

diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c
index 136e0c257af1..65514d37bfe2 100644
--- a/drivers/tty/serial/sh-sci.c
+++ b/drivers/tty/serial/sh-sci.c
@@ -157,6 +157,7 @@  struct sci_port {
 
 	bool has_rtscts;
 	bool autorts;
+	bool first_time_tx;
 };
 
 #define SCI_NPORTS CONFIG_SERIAL_SH_SCI_NR_UARTS
@@ -850,6 +851,7 @@  static void sci_transmit_chars(struct uart_port *port)
 {
 	struct tty_port *tport = &port->state->port;
 	unsigned int stopped = uart_tx_stopped(port);
+	struct sci_port *s = to_sci_port(port);
 	unsigned short status;
 	unsigned short ctrl;
 	int count;
@@ -885,6 +887,7 @@  static void sci_transmit_chars(struct uart_port *port)
 		}
 
 		sci_serial_out(port, SCxTDR, c);
+		s->first_time_tx = true;
 
 		port->icount.tx++;
 	} while (--count > 0);
@@ -1241,6 +1244,8 @@  static void sci_dma_tx_complete(void *arg)
 	if (kfifo_len(&tport->xmit_fifo) < WAKEUP_CHARS)
 		uart_write_wakeup(port);
 
+	s->first_time_tx = true;
+
 	if (!kfifo_is_empty(&tport->xmit_fifo)) {
 		s->cookie_tx = 0;
 		schedule_work(&s->work_tx);
@@ -2076,6 +2081,10 @@  static unsigned int sci_tx_empty(struct uart_port *port)
 {
 	unsigned short status = sci_serial_in(port, SCxSR);
 	unsigned short in_tx_fifo = sci_txfill(port);
+	struct sci_port *s = to_sci_port(port);
+
+	if (!s->first_time_tx)
+		return TIOCSER_TEMT;
 
 	return (status & SCxSR_TEND(port)) && !in_tx_fifo ? TIOCSER_TEMT : 0;
 }
@@ -2247,6 +2256,7 @@  static int sci_startup(struct uart_port *port)
 
 	dev_dbg(port->dev, "%s(%d)\n", __func__, port->line);
 
+	s->first_time_tx = false;
 	sci_request_dma(port);
 
 	ret = sci_request_irq(s);
@@ -2267,6 +2277,7 @@  static void sci_shutdown(struct uart_port *port)
 	dev_dbg(port->dev, "%s(%d)\n", __func__, port->line);
 
 	s->autorts = false;
+	s->first_time_tx = false;
 	mctrl_gpio_disable_ms(to_sci_port(port)->gpios);
 
 	uart_port_lock_irqsave(port, &flags);