| Message ID | 20210604095704.756190-1-yoshihiro.shimoda.uh@renesas.com |
|---|---|
| State | New |
| Headers | show |
| Series | [v2] serial: sh-sci: Stop dmaengine transfer in sci_stop_tx() | expand |
Hi Jiri, Thank you for your review! > From: Jiri Slaby, Sent: Friday, June 4, 2021 7:22 PM > > On 04. 06. 21, 11:57, Yoshihiro Shimoda wrote: > > Stop dmaengine transfer in sci_stop_tx(). Otherwise, the following > > message is possible output when system enters suspend and while > > transferring data, because clearing TIE bit in SCSCR is not able to > > stop any dmaengine transfer. > > > > sh-sci e6550000.serial: ttySC1: Unable to drain transmitter > > > > Notes: > > - sci_dma_stop_tx() is a macro in the .h file because struct sci_port > > is declared in the .c file and #ifdef should be in the .h file. > > - This patch uses dmaengine_terminate_async() so that we can apply > > this into longterm kernel v4.9.x or later. > > > > Fixes: 73a19e4c0301 ("serial: sh-sci: Add DMA support.") > > Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> > > --- <snip> > > diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c > > index 4baf1316ea72..3793cf9f352c 100644 > > --- a/drivers/tty/serial/sh-sci.c > > +++ b/drivers/tty/serial/sh-sci.c > > @@ -610,6 +610,8 @@ static void sci_stop_tx(struct uart_port *port) > > ctrl &= ~SCSCR_TIE; > > > > serial_port_out(port, SCSCR, ctrl); > > + > > + sci_dma_stop_tx(port); > > } > > > > static void sci_start_rx(struct uart_port *port) > > diff --git a/drivers/tty/serial/sh-sci.h b/drivers/tty/serial/sh-sci.h > > index c0dfe4382898..435c674af1ce 100644 > > --- a/drivers/tty/serial/sh-sci.h > > +++ b/drivers/tty/serial/sh-sci.h > > @@ -174,3 +174,19 @@ enum { > > (((port)->type == PORT_SCI) ? SCI_TDxE_CLEAR : SCIF_TDxE_CLEAR) > > #define SCxSR_BREAK_CLEAR(port) \ > > (((port)->type == PORT_SCI) ? SCI_BREAK_CLEAR : SCIF_BREAK_CLEAR) > > + > > +#ifdef CONFIG_SERIAL_SH_SCI_DMA > > +#define sci_dma_stop_tx(port) \ > > Ouch. First, why this needs to be in a header? The v1 patch [1] putted #ifdef in .c file, so that I got feedback from Greg like below: "Please do not put #ifdef in .c files, this should be possible without that." [1] https://lore.kernel.org/linux-renesas-soc/TY2PR01MB3692C545672195370487FA60D83C9@TY2PR01MB3692.jpnprd01.prod.outlook.com/T/#md5b58540e26a28ef1acddfc326c62bfde73e8e8a > Second, please don't do this by a macro. I got it. I'll try to fix this issue somehow... Best regards, Yoshihiro Shimoda
diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c index 4baf1316ea72..3793cf9f352c 100644 --- a/drivers/tty/serial/sh-sci.c +++ b/drivers/tty/serial/sh-sci.c @@ -610,6 +610,8 @@ static void sci_stop_tx(struct uart_port *port) ctrl &= ~SCSCR_TIE; serial_port_out(port, SCSCR, ctrl); + + sci_dma_stop_tx(port); } static void sci_start_rx(struct uart_port *port) diff --git a/drivers/tty/serial/sh-sci.h b/drivers/tty/serial/sh-sci.h index c0dfe4382898..435c674af1ce 100644 --- a/drivers/tty/serial/sh-sci.h +++ b/drivers/tty/serial/sh-sci.h @@ -174,3 +174,19 @@ enum { (((port)->type == PORT_SCI) ? SCI_TDxE_CLEAR : SCIF_TDxE_CLEAR) #define SCxSR_BREAK_CLEAR(port) \ (((port)->type == PORT_SCI) ? SCI_BREAK_CLEAR : SCIF_BREAK_CLEAR) + +#ifdef CONFIG_SERIAL_SH_SCI_DMA +#define sci_dma_stop_tx(port) \ +{ \ + struct sci_port *s = to_sci_port(port); \ + \ + if (s->chan_tx && !dma_submit_error(s->cookie_tx)) { \ + dmaengine_terminate_async(s->chan_tx); \ + s->cookie_tx = -EINVAL; \ + } \ +} +#else +static inline void sci_dma_stop_tx(struct uart_port *port) +{ +} +#endif /* CONFIG_SERIAL_SH_SCI_DMA */
Stop dmaengine transfer in sci_stop_tx(). Otherwise, the following message is possible output when system enters suspend and while transferring data, because clearing TIE bit in SCSCR is not able to stop any dmaengine transfer. sh-sci e6550000.serial: ttySC1: Unable to drain transmitter Notes: - sci_dma_stop_tx() is a macro in the .h file because struct sci_port is declared in the .c file and #ifdef should be in the .h file. - This patch uses dmaengine_terminate_async() so that we can apply this into longterm kernel v4.9.x or later. Fixes: 73a19e4c0301 ("serial: sh-sci: Add DMA support.") Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> --- Changes from v1: - Don't put #ifdef in the .c file. - Update the commit description. https://lore.kernel.org/linux-renesas-soc/TY2PR01MB3692C545672195370487FA60D83C9@TY2PR01MB3692.jpnprd01.prod.outlook.com/T/#t drivers/tty/serial/sh-sci.c | 2 ++ drivers/tty/serial/sh-sci.h | 16 ++++++++++++++++ 2 files changed, 18 insertions(+)