linux-serial.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Stefan Agner <stefan@agner.ch>
To: gregkh@linuxfoundation.org
Cc: linux-kernel@lists.infradead.org, stefan@agner.ch,
	yao.yuan@freescale.com, linux-serial@vger.kernel.org,
	jingchang.lu@freescale.com, jslaby@suse.cz,
	linux-arm-kernel@lists.infradead.org
Subject: [PATCH 2/2] serial: fsl_lpuart: calculate DMA burst
Date: Wed,  2 Jul 2014 18:02:57 +0200	[thread overview]
Message-ID: <5d0e45bcd38d9030fcd2b32fb5f09fd26c25f0a6.1404314964.git.stefan@agner.ch> (raw)
In-Reply-To: <54df28cb5dccf71b17523896da17ce2f155855fd.1404314964.git.stefan@agner.ch>

The DMA burst size must match the transmit FIFO depth in order
to make sure all character are transmitted. This patch calculates
DMA burst size by using FIFO depth rather than use the hardcoded
16 bytes. This is required since some UARTs (e.g. UART2 on Vybrid)
have a FIFO depth of 8 bytes.

Signed-off-by: Stefan Agner <stefan@agner.ch>
---
 drivers/tty/serial/fsl_lpuart.c | 27 +++++++++++++--------------
 1 file changed, 13 insertions(+), 14 deletions(-)

diff --git a/drivers/tty/serial/fsl_lpuart.c b/drivers/tty/serial/fsl_lpuart.c
index 6b462a9..f973456 100644
--- a/drivers/tty/serial/fsl_lpuart.c
+++ b/drivers/tty/serial/fsl_lpuart.c
@@ -117,8 +117,6 @@
 #define UARTSFIFO_TXOF		0x02
 #define UARTSFIFO_RXUF		0x01
 
-#define DMA_MAXBURST		16
-#define DMA_MAXBURST_MASK	(DMA_MAXBURST - 1)
 #define FSL_UART_RX_DMA_BUFFER_SIZE	64
 
 #define DRIVER_NAME	"fsl-lpuart"
@@ -240,7 +238,7 @@ static int lpuart_dma_tx(struct lpuart_port *sport, unsigned long count)
 
 	dma_sync_single_for_device(sport->port.dev, sport->dma_tx_buf_bus,
 				UART_XMIT_SIZE, DMA_TO_DEVICE);
-	sport->dma_tx_bytes = count & ~(DMA_MAXBURST_MASK);
+	sport->dma_tx_bytes = count & ~(sport->txfifo_size - 1);
 	tx_bus_addr = sport->dma_tx_buf_bus + xmit->tail;
 	sport->dma_tx_desc = dmaengine_prep_slave_single(sport->dma_tx_chan,
 					tx_bus_addr, sport->dma_tx_bytes,
@@ -269,7 +267,7 @@ static void lpuart_prepare_tx(struct lpuart_port *sport)
 	if (!count)
 		return;
 
-	if (count < DMA_MAXBURST)
+	if (count < sport->txfifo_size)
 		writeb(readb(sport->port.membase + UARTCR5) & ~UARTCR5_TDMAS,
 				sport->port.membase + UARTCR5);
 	else {
@@ -599,15 +597,7 @@ static void lpuart_setup_watermark(struct lpuart_port *sport)
 			UARTCR2_RIE | UARTCR2_RE);
 	writeb(cr2, sport->port.membase + UARTCR2);
 
-	/* determine FIFO size and enable FIFO mode */
 	val = readb(sport->port.membase + UARTPFIFO);
-
-	sport->txfifo_size = 0x1 << (((val >> UARTPFIFO_TXSIZE_OFF) &
-		UARTPFIFO_FIFOSIZE_MASK) + 1);
-
-	sport->rxfifo_size = 0x1 << (((val >> UARTPFIFO_RXSIZE_OFF) &
-		UARTPFIFO_FIFOSIZE_MASK) + 1);
-
 	writeb(val | UARTPFIFO_TXFE | UARTPFIFO_RXFE,
 			sport->port.membase + UARTPFIFO);
 
@@ -652,7 +642,7 @@ static int lpuart_dma_tx_request(struct uart_port *port)
 	dma_buf = sport->port.state->xmit.buf;
 	dma_tx_sconfig.dst_addr = sport->port.mapbase + UARTDR;
 	dma_tx_sconfig.dst_addr_width = DMA_SLAVE_BUSWIDTH_1_BYTE;
-	dma_tx_sconfig.dst_maxburst = DMA_MAXBURST;
+	dma_tx_sconfig.dst_maxburst = sport->txfifo_size;
 	dma_tx_sconfig.direction = DMA_MEM_TO_DEV;
 	ret = dmaengine_slave_config(tx_chan, &dma_tx_sconfig);
 
@@ -765,7 +755,16 @@ static int lpuart_startup(struct uart_port *port)
 	unsigned long flags;
 	unsigned char temp;
 
-	/*whether use dma support by dma request results*/
+	/* determine FIFO size and enable FIFO mode */
+	temp = readb(sport->port.membase + UARTPFIFO);
+
+	sport->txfifo_size = 0x1 << (((temp >> UARTPFIFO_TXSIZE_OFF) &
+		UARTPFIFO_FIFOSIZE_MASK) + 1);
+
+	sport->rxfifo_size = 0x1 << (((temp >> UARTPFIFO_RXSIZE_OFF) &
+		UARTPFIFO_FIFOSIZE_MASK) + 1);
+
+	/* Whether use dma support by dma request results */
 	if (lpuart_dma_tx_request(port) || lpuart_dma_rx_request(port)) {
 		sport->lpuart_dma_use = false;
 	} else {
-- 
2.0.1

      reply	other threads:[~2014-07-02 16:02 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-07-02 16:02 [PATCH 1/2] serial: fsl_lpuart: move DMA RX timeout calculation Stefan Agner
2014-07-02 16:02 ` Stefan Agner [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5d0e45bcd38d9030fcd2b32fb5f09fd26c25f0a6.1404314964.git.stefan@agner.ch \
    --to=stefan@agner.ch \
    --cc=gregkh@linuxfoundation.org \
    --cc=jingchang.lu@freescale.com \
    --cc=jslaby@suse.cz \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@lists.infradead.org \
    --cc=linux-serial@vger.kernel.org \
    --cc=yao.yuan@freescale.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).