Re: [PATCH v3] spi: Add PPC4xx SPI driver

[Stefan Roese <sr@denx.de>]

On Wednesday 29 October 2008, Josh Boyer wrote:
> Looks pretty good.  Just a few minor comments/questions below.

Thanks. I also added some comments below.

> Also, do you have a patch for a DTS file that gives an example of how to
> instantiate the SPI stuff in the device tree?

OK, I'll add a DTS patch to the next version.

> >diff --git a/drivers/spi/spi_ppc4xx.c b/drivers/spi/spi_ppc4xx.c
> >new file mode 100644
> >index 0000000..6f94acc
> >--- /dev/null
> >+++ b/drivers/spi/spi_ppc4xx.c
> >+#include <linux/module.h>
> >+#include <linux/init.h>
> >+#include <linux/sched.h>
> >+#include <linux/errno.h>
> >+#include <linux/wait.h>
> >+#include <linux/of_platform.h>
> >+#include <linux/of_spi.h>
> >+#include <linux/of_gpio.h>
> >+#include <linux/interrupt.h>
> >+#include <linux/delay.h>
> >+
> >+#include <linux/gpio.h>
> >+#include <linux/spi/spi.h>
> >+#include <linux/spi/spi_bitbang.h>
> >+
> >+#include <asm/io.h>
> >+#include <asm/dcr-native.h>
>
> Should maybe just be <asm/dcr.h>

Fixed.

> >+#include <asm/dcr-regs.h>
> >+
>
> <snip>
>
> >+/* SPI Controller driver's private data. */
> >+struct ppc4xx_spi {
> >+	/* bitbang has to be first */
> >+	struct spi_bitbang bitbang;
> >+	struct completion done;
> >+
> >+	u64 mapbase;
> >+	u64 mapsize;
> >+	int irqnum;
> >+	/* need this to set the SPI clock */
> >+	unsigned int opb_freq;
> >+
> >+	/* for transfers */
> >+	int len;
> >+	int count;
> >+	/* data buffers */
> >+	const unsigned char *tx;
> >+	unsigned char *rx;
> >+
> >+	int *gpios;
> >+	unsigned int num_gpios;
> >+
> >+	volatile struct spi_ppc4xx_regs __iomem *regs; /* pointer to the
> > registers */
>
> volatile?

Fixed.

> >+	struct spi_master *master;
> >+	struct device *dev;
> >+};
>
> <snip>
>
> >+static int spi_ppc4xx_txrx(struct spi_device *spi, struct spi_transfer
> > *t) +{
> >+	struct ppc4xx_spi *hw;
> >+	u8 data;
> >+
> >+	dev_dbg(&spi->dev, "txrx: tx %p, rx %p, len %d\n",
> >+		t->tx_buf, t->rx_buf, t->len);
> >+
> >+	hw = spi_master_get_devdata(spi->master);
> >+
> >+	hw->tx = t->tx_buf;
> >+	hw->rx = t->rx_buf;
> >+	hw->len = t->len;
> >+	hw->count = 0;
> >+
> >+	/* send the first byte */
> >+	data = hw->tx ? hw->tx[0] : 0;
> >+	out_8(&hw->regs->txd, data);
> >+	out_8(&hw->regs->cr, SPI_PPC4XX_CR_STR);
>
> Maybe iowrite8?  Same comment elsewhere.

Why? We use the in_/out_xxx() accessor function for all other 4xx driver as 
well.

> >+	wait_for_completion(&hw->done);
> >+
> >+	return hw->count;
> >+}
> >+
>
> <snip>
>
> >+static struct of_device_id spi_ppc4xx_of_match[] = {
> >+	{ .compatible = "ibm,spi", },
> >+	{},
> >+};
>
> I'm wondering if that is too generic of a match.  In theory,
> IBM could have another SPI controller that isn't for 4xx.
> Maybe "ibm,spi-4xx" ?

Right. I was doing it the same way as already done before, e.g. "ibm,iic". For 
the gpio driver we already switched to "ibm,ppc4xx-gpio". So how 
about "ibm,ppc4xx-spi"?

Thanks.

Best regards,
Stefan