atmel_spi: fix dmachain oops with DEBUG enabled

In atmel_spi_next_xfer, xfer can be NULL because the next transfer may
already have been submitted to the PDC (using DMA chaining).  This can
cause an oops, since the debug message assumed it was never null.  The
fix changes how those debug messages are issued, ensuring that one is
issued each time a transfer is started instead of once per call.

Also, properly indent the "can this transfer be chained" test so it's
not hidden as if it were non-conditional code.
Signed-off-by: NHaavard Skinnemoen <hskinnemoen@atmel.com>
Signed-off-by: NDavid Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

drivers/spi/atmel_spi.c

@@ -198,6 +198,11 @@ static void atmel_spi_next_xfer(struct spi_master *master,
 			len >>= 1;
 		spi_writel(as, RCR, len);
 		spi_writel(as, TCR, len);
+
+		dev_dbg(&msg->spi->dev,
+			"  start xfer %p: len %u tx %p/%08x rx %p/%08x\n",
+			xfer, xfer->len, xfer->tx_buf, xfer->tx_dma,
+			xfer->rx_buf, xfer->rx_dma);
 	} else {
 		xfer = as->next_transfer;
 		remaining = as->next_remaining_bytes;
@@ -208,8 +213,8 @@ static void atmel_spi_next_xfer(struct spi_master *master,
 
 	if (remaining > 0)
 		len = remaining;
-	else if (!atmel_spi_xfer_is_last(msg, xfer) &&
-		atmel_spi_xfer_can_be_chained(xfer)) {
+	else if (!atmel_spi_xfer_is_last(msg, xfer)
+			&& atmel_spi_xfer_can_be_chained(xfer)) {
 		xfer = list_entry(xfer->transfer_list.next,
 				struct spi_transfer, transfer_list);
 		len = xfer->len;
@@ -230,6 +235,11 @@ static void atmel_spi_next_xfer(struct spi_master *master,
 			len >>= 1;
 		spi_writel(as, RNCR, len);
 		spi_writel(as, TNCR, len);
+
+		dev_dbg(&msg->spi->dev,
+			"  next xfer %p: len %u tx %p/%08x rx %p/%08x\n",
+			xfer, xfer->len, xfer->tx_buf, xfer->tx_dma,
+			xfer->rx_buf, xfer->rx_dma);
 	} else {
 		spi_writel(as, RNCR, 0);
 		spi_writel(as, TNCR, 0);
@@ -246,12 +256,6 @@ static void atmel_spi_next_xfer(struct spi_master *master,
 	 * It should be doable, though. Just not now...
 	 */
 	spi_writel(as, IER, SPI_BIT(ENDRX) | SPI_BIT(OVRES));
-
-	dev_dbg(&msg->spi->dev,
-		"  start xfer %p: len %u tx %p/%08x rx %p/%08x imr %03x\n",
-		xfer, xfer->len, xfer->tx_buf, xfer->tx_dma,
-		xfer->rx_buf, xfer->rx_dma, spi_readl(as, IMR));
-
 	spi_writel(as, PTCR, SPI_BIT(TXTEN) | SPI_BIT(RXTEN));
 }