| Message ID | 1466101475-8851-1-git-send-email-bjorn.andersson@linaro.org |
|---|---|
| State | New |
| Headers | show |
On Fri, Jun 17, 2016 at 1:02 PM, Nicolas Dechesne <nicolas.dechesne@linaro.org> wrote: > <bjorn.andersson@linaro.org> wrote: >> msm_set_termios() is called whenever the tty is opened. Setting the baud >> rate requires a full reset of the msm serial block, even when the rate >> is unchanged. In the case when the same uart is used as console this >> reset will discard any console output data still being clocked out of >> the TX fifo. >> >> By skipping the rate-change in the case where the baud rate is unchanged >> since last request we can avoid the reset and the discarding of the >> data. >> >> Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> > > > Tested-by: Nicolas Dechesne <nicolas.dechesne@linaro.org> > > I no longer see this type of corruption on serial console > > [ 6.325118] ALSA device �[ 6.333338] Freeing unused kernel > memory: 572K (ffffffc000c10000 - ffffffc000c9f000) > [ 13.800269] remoteproc2: remote processor a20400�[ OK ] Started > Start the WCN core. > [ � Starting Update UTMP about System Runlevel Changes... oops. today i tried on APQ8064 boards , both IFC6410Plus and SD 600eval, and I can no longer 'type' into the serial console, I can get debug messages though. Reverting this change 'fixes' the problem. -- To unsubscribe from this list: send the line "unsubscribe linux-serial" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 21/06/16 10:43, Nicolas Dechesne wrote: > On Fri, Jun 17, 2016 at 1:02 PM, Nicolas Dechesne > <nicolas.dechesne@linaro.org> wrote: >> <bjorn.andersson@linaro.org> wrote: >>> msm_set_termios() is called whenever the tty is opened. Setting the baud >>> rate requires a full reset of the msm serial block, even when the rate >>> is unchanged. In the case when the same uart is used as console this >>> reset will discard any console output data still being clocked out of >>> the TX fifo. >>> >>> By skipping the rate-change in the case where the baud rate is unchanged >>> since last request we can avoid the reset and the discarding of the >>> data. >>> >>> Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> >> >> >> Tested-by: Nicolas Dechesne <nicolas.dechesne@linaro.org> >> >> I no longer see this type of corruption on serial console >> >> [ 6.325118] ALSA device �[ 6.333338] Freeing unused kernel >> memory: 572K (ffffffc000c10000 - ffffffc000c9f000) >> [ 13.800269] remoteproc2: remote processor a20400�[ OK ] Started >> Start the WCN core. >> [ � Starting Update UTMP about System Runlevel Changes... > > > oops. today i tried on APQ8064 boards , both IFC6410Plus and SD > 600eval, and I can no longer 'type' into the serial console, I can get > debug messages though. Reverting this change 'fixes' the problem. Am also hitting the same issue as Nico on IFC6410 with this patch. > -- > To unsubscribe from this list: send the line "unsubscribe linux-arm-msm" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- To unsubscribe from this list: send the line "unsubscribe linux-serial" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/tty/serial/msm_serial.c b/drivers/tty/serial/msm_serial.c index b7d80bd57db9..206149f104fa 100644 --- a/drivers/tty/serial/msm_serial.c +++ b/drivers/tty/serial/msm_serial.c @@ -76,6 +76,7 @@ struct msm_port { bool break_detected; struct msm_dma tx_dma; struct msm_dma rx_dma; + unsigned int last_baud; }; static void msm_handle_tx(struct uart_port *port); @@ -1072,11 +1073,16 @@ static void msm_set_termios(struct uart_port *port, struct ktermios *termios, if (dma->chan) /* Terminate if any */ msm_stop_dma(port, dma); - /* calculate and set baud rate */ + /* calculate and set baud rate, if changed from last request */ baud = uart_get_baud_rate(port, termios, old, 300, 4000000); - baud = msm_set_baud_rate(port, baud, &flags); - if (tty_termios_baud_rate(termios)) - tty_termios_encode_baud_rate(termios, baud, baud); + if (baud != msm_port->last_baud) { + msm_port->last_baud = baud; + + baud = msm_set_baud_rate(port, baud, &flags); + if (tty_termios_baud_rate(termios)) + tty_termios_encode_baud_rate(termios, baud, baud); + uart_update_timeout(port, termios->c_cflag, baud); + } /* calculate parity */ mr = msm_read(port, UART_MR2); @@ -1134,8 +1140,6 @@ static void msm_set_termios(struct uart_port *port, struct ktermios *termios, if (termios->c_iflag & (IGNBRK | BRKINT | PARMRK)) port->read_status_mask |= UART_SR_RX_BREAK; - uart_update_timeout(port, termios->c_cflag, baud); - /* Try to use DMA */ msm_start_rx_dma(msm_port);
msm_set_termios() is called whenever the tty is opened. Setting the baud rate requires a full reset of the msm serial block, even when the rate is unchanged. In the case when the same uart is used as console this reset will discard any console output data still being clocked out of the TX fifo. By skipping the rate-change in the case where the baud rate is unchanged since last request we can avoid the reset and the discarding of the data. Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org> --- drivers/tty/serial/msm_serial.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) -- 2.5.0 -- To unsubscribe from this list: send the line "unsubscribe linux-serial" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html