[GIT PULL 0/3] Renesas SH SCI updates for v3.14

[GIT PULL 0/3] Renesas SH SCI updates for v3.14

[Simon Horman]

Hi Kevin, Hi Olof, Hi Arnd,

please consider these Renesas SH SCI updates for v3.14.

They have been acked by Greg Kroah-Hartman to be taken
through my tree. The reason for this is that they are dependencies
for subsequent changes to use the common clock framework.


The following changes since commit 6ce4eac1f600b34f2f7f58f9cd8f0503d79e42ae:

  Linux 3.13-rc1 (2013-11-22 11:30:55 -0800)

are available in the git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas.git tags/renesas-sh-sci-for-v3.14

for you to fetch changes up to b016b646e8676858f39ea9be760494b04b9ee0af:

  serial: sh-sci: Convert to clk_prepare/unprepare (2013-12-14 09:59:31 +0900)

----------------------------------------------------------------
Renesas SH SCI updates for v3.14

Updates for the SH SCI serial driver

* Convert to clk_prepare/unprepare
* Fix warnings due to improper casts and printk formats

----------------------------------------------------------------
Laurent Pinchart (3):
      serial: sh-sci: Fix warnings due to improper casts and printk formats
      serial: sh-sci: Don't enable/disable port from within break timer
      serial: sh-sci: Convert to clk_prepare/unprepare

 drivers/tty/serial/sh-sci.c | 39 ++++++++++++++++++++++-----------------
 1 file changed, 22 insertions(+), 17 deletions(-)

[PATCH 1/3] serial: sh-sci: Fix warnings due to improper casts and printk formats

[Simon Horman]

From: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>

Use the %zu and %pad printk specifiers to print size_t and dma_addr_t
variables, and cast pointers to uintptr_t instead of unsigned int where
applicable. This fixes warnings on platforms where pointers and/or
dma_addr_t have a different size than int.

Cc: linux-serial at vger.kernel.org
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
---
 drivers/tty/serial/sh-sci.c | 19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)

diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c
index 7d8103c..6e5ce62 100644
--- a/drivers/tty/serial/sh-sci.c
+++ b/drivers/tty/serial/sh-sci.c
@@ -557,7 +557,7 @@ static inline int sci_rxd_in(struct uart_port *port)
 		return 1;
 
 	/* Cast for ARM damage */
-	return !!__raw_readb((void __iomem *)s->cfg->port_reg);
+	return !!__raw_readb((void __iomem *)(uintptr_t)s->cfg->port_reg);
 }
 
 /* ********************************************************************** *
@@ -1309,7 +1309,7 @@ static int sci_dma_rx_push(struct sci_port *s, size_t count)
 	}
 
 	if (room < count)
-		dev_warn(port->dev, "Rx overrun: dropping %u bytes\n",
+		dev_warn(port->dev, "Rx overrun: dropping %zu bytes\n",
 			 count - room);
 	if (!room)
 		return room;
@@ -1442,7 +1442,7 @@ static void work_fn_rx(struct work_struct *work)
 		int count;
 
 		chan->device->device_control(chan, DMA_TERMINATE_ALL, 0);
-		dev_dbg(port->dev, "Read %u bytes with cookie %d\n",
+		dev_dbg(port->dev, "Read %zu bytes with cookie %d\n",
 			sh_desc->partial, sh_desc->cookie);
 
 		spin_lock_irqsave(&port->lock, flags);
@@ -1691,16 +1691,17 @@ static void sci_request_dma(struct uart_port *port)
 		s->chan_tx = chan;
 		sg_init_table(&s->sg_tx, 1);
 		/* UART circular tx buffer is an aligned page. */
-		BUG_ON((int)port->state->xmit.buf & ~PAGE_MASK);
+		BUG_ON((uintptr_t)port->state->xmit.buf & ~PAGE_MASK);
 		sg_set_page(&s->sg_tx, virt_to_page(port->state->xmit.buf),
-			    UART_XMIT_SIZE, (int)port->state->xmit.buf & ~PAGE_MASK);
+			    UART_XMIT_SIZE,
+			    (uintptr_t)port->state->xmit.buf & ~PAGE_MASK);
 		nent = dma_map_sg(port->dev, &s->sg_tx, 1, DMA_TO_DEVICE);
 		if (!nent)
 			sci_tx_dma_release(s, false);
 		else
-			dev_dbg(port->dev, "%s: mapped %d@%p to %x\n", __func__,
-				sg_dma_len(&s->sg_tx),
-				port->state->xmit.buf, sg_dma_address(&s->sg_tx));
+			dev_dbg(port->dev, "%s: mapped %d@%p to %pad\n", __func__,
+				sg_dma_len(&s->sg_tx), port->state->xmit.buf,
+				&sg_dma_address(&s->sg_tx));
 
 		s->sg_len_tx = nent;
 
@@ -1740,7 +1741,7 @@ static void sci_request_dma(struct uart_port *port)
 
 			sg_init_table(sg, 1);
 			sg_set_page(sg, virt_to_page(buf[i]), s->buf_len_rx,
-				    (int)buf[i] & ~PAGE_MASK);
+				    (uintptr_t)buf[i] & ~PAGE_MASK);
 			sg_dma_address(sg) = dma[i];
 		}
 
-- 
1.8.4

[PATCH 2/3] serial: sh-sci: Don't enable/disable port from within break timer

[Simon Horman]

From: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>

The break timer accesses hardware registers and thus requires the port
to be enabled. It currently ensures this by enabling the port at the
beginning of the timer handler, and disabling it at the end. However,
the enable/disable operations call the runtime PM sync functions, which
are not allowed in atomic context. The current situation is thus broken.

This change relies on non-atomic code to enable/disable the port. The
break timer will only be started from the IRQ handler, which already
runs with the port enabled. We just need to ensure that the port won't
be disabled with the timer running, and that's easily done by just
cancelling the timer in the port disable function.

Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Acked-by: Magnus Damm <damm@opensource.se>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
---
 drivers/tty/serial/sh-sci.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c
index 6e5ce62..1ebac3e 100644
--- a/drivers/tty/serial/sh-sci.c
+++ b/drivers/tty/serial/sh-sci.c
@@ -431,6 +431,14 @@ static void sci_port_disable(struct sci_port *sci_port)
 	if (!sci_port->port.dev)
 		return;
 
+	/* Cancel the break timer to ensure that the timer handler will not try
+	 * to access the hardware with clocks and power disabled. Reset the
+	 * break flag to make the break debouncing state machine ready for the
+	 * next break.
+	 */
+	del_timer_sync(&sci_port->break_timer);
+	sci_port->break_flag = 0;
+
 	clk_disable(sci_port->fclk);
 	clk_disable(sci_port->iclk);
 
@@ -733,8 +741,6 @@ static void sci_break_timer(unsigned long data)
 {
 	struct sci_port *port = (struct sci_port *)data;
 
-	sci_port_enable(port);
-
 	if (sci_rxd_in(&port->port) == 0) {
 		port->break_flag = 1;
 		sci_schedule_break_timer(port);
@@ -744,8 +750,6 @@ static void sci_break_timer(unsigned long data)
 		sci_schedule_break_timer(port);
 	} else
 		port->break_flag = 0;
-
-	sci_port_disable(port);
 }
 
 static int sci_handle_errors(struct uart_port *port)
-- 
1.8.4

[PATCH 3/3] serial: sh-sci: Convert to clk_prepare/unprepare

[Simon Horman]

From: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>

Turn clk_enable() and clk_disable() calls into clk_prepare_enable() and
clk_disable_unprepare() to get ready for the migration to the common
clock framework.

Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Acked-by: Magnus Damm <damm@opensource.se>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Simon Horman <horms+renesas@verge.net.au>
---
 drivers/tty/serial/sh-sci.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c
index 1ebac3e..1a3fc7a 100644
--- a/drivers/tty/serial/sh-sci.c
+++ b/drivers/tty/serial/sh-sci.c
@@ -421,9 +421,9 @@ static void sci_port_enable(struct sci_port *sci_port)
 
 	pm_runtime_get_sync(sci_port->port.dev);
 
-	clk_enable(sci_port->iclk);
+	clk_prepare_enable(sci_port->iclk);
 	sci_port->port.uartclk = clk_get_rate(sci_port->iclk);
-	clk_enable(sci_port->fclk);
+	clk_prepare_enable(sci_port->fclk);
 }
 
 static void sci_port_disable(struct sci_port *sci_port)
@@ -439,8 +439,8 @@ static void sci_port_disable(struct sci_port *sci_port)
 	del_timer_sync(&sci_port->break_timer);
 	sci_port->break_flag = 0;
 
-	clk_disable(sci_port->fclk);
-	clk_disable(sci_port->iclk);
+	clk_disable_unprepare(sci_port->fclk);
+	clk_disable_unprepare(sci_port->iclk);
 
 	pm_runtime_put_sync(sci_port->port.dev);
 }
-- 
1.8.4

[GIT PULL 0/3] Renesas SH SCI updates for v3.14

[Olof Johansson]

On Mon, Dec 23, 2013 at 11:08:41AM +0900, Simon Horman wrote:
> Hi Kevin, Hi Olof, Hi Arnd,
> 
> please consider these Renesas SH SCI updates for v3.14.
> 
> They have been acked by Greg Kroah-Hartman to be taken
> through my tree. The reason for this is that they are dependencies
> for subsequent changes to use the common clock framework.
> 
> 
> The following changes since commit 6ce4eac1f600b34f2f7f58f9cd8f0503d79e42ae:
> 
>   Linux 3.13-rc1 (2013-11-22 11:30:55 -0800)
> 
> are available in the git repository at:
> 
>   git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas.git tags/renesas-sh-sci-for-v3.14
> 
> for you to fetch changes up to b016b646e8676858f39ea9be760494b04b9ee0af:
> 
>   serial: sh-sci: Convert to clk_prepare/unprepare (2013-12-14 09:59:31 +0900)

Merged into next/cleanup, since they're mostly bugfixes and they're a prereq
for other code so the dependency flow works better that way.


-Olof