Hi Vladimir,

On Tue, Jun 28, 2022 at 04:08:10PM +0000, Vladimir Oltean wrote:

On Tue, Jun 28, 2022 at 01:17:01AM -0700, Colin Foster wrote:

quoted

diff --git a/include/linux/mfd/ocelot.h b/include/linux/mfd/ocelot.h
new file mode 100644
index 000000000000..5c95e4ee38a6
--- /dev/null
+++ b/include/linux/mfd/ocelot.h

@@ -0,0 +1,27 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
+/* Copyright 2022 Innovative Advantage Inc. */
+
+#include <linux/err.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+#include <linux/types.h>
+
+struct resource;
+
+static inline struct regmap *
+ocelot_platform_init_regmap_from_resource(struct platform_device *pdev,
+					  unsigned int index,
+					  const struct regmap_config *config)

I think this function name is too long (especially if you're going to
also introduce ocelot_platform_init_regmap_from_resource_optional),
and I have the impression that the "platform_init_" part of the name
doesn't bring too much value. How about ocelot_regmap_from_resource()?

I thought the same thing after your first email. My thought was "how do
I indent that?" :-)

quoted

+{
+	struct resource *res;
+	u32 __iomem *regs;
+
+	regs = devm_platform_get_and_ioremap_resource(pdev, index, &res);
+
+	if (!res)
+		return ERR_PTR(-ENOENT);
+	else if (IS_ERR(regs))
+		return ERR_CAST(regs);
+	else
+		return devm_regmap_init_mmio(&pdev->dev, regs, config);
+}
-- 
2.25.1

To illustrate what I'm trying to say, these would be the shim
definitions:

static inline struct regmap *
ocelot_regmap_from_resource(struct platform_device *pdev,
			    unsigned int index,
			    const struct regmap_config *config)
{
	struct resource *res;
	void __iomem *regs;

	regs = devm_platform_get_and_ioremap_resource(pdev, index, &res);
	if (IS_ERR(regs))
		return regs;

	return devm_regmap_init_mmio(&pdev->dev, regs, config);
}

static inline struct regmap *
ocelot_regmap_from_resource_optional(struct platform_device *pdev,
				     unsigned int index,
				     const struct regmap_config *config)
{
	struct resource *res;
	void __iomem *regs;

	res = platform_get_resource(pdev, IORESOURCE_MEM, index);
	if (!res)
		return NULL;

	regs = devm_ioremap_resource(&pdev->dev, r);
	if (IS_ERR(regs))
		return regs;

	return devm_regmap_init_mmio(&pdev->dev, regs, config);
}

and these would be the full versions:

static struct regmap *
ocelot_regmap_from_mem_resource(struct device *dev, struct resource *res,
				const struct regmap_config *config)
{
	void __iomem *regs;

	regs = devm_ioremap_resource(dev, r);
	if (IS_ERR(regs))
		return regs;

	return devm_regmap_init_mmio(dev, regs, config);
}

static struct regmap *
ocelot_regmap_from_reg_resource(struct device *dev, struct resource *res,
				const struct regmap_config *config)
{
	/* Open question: how to differentiate SPI from I2C resources? */

My expectation is to set something up in drivers/mfd/ocelot-{spi,i2c}.c
and have an if/else / switch. PCIe might actually be our first hardware
spin.

	return ocelot_spi_init_regmap(dev->parent, dev, res);
}

struct regmap *
ocelot_regmap_from_resource_optional(struct platform_device *pdev,
				     unsigned int index,
				     const struct regmap_config *config)
{
	struct device *dev = &pdev->dev;
	struct resource *res;

	res = platform_get_resource(pdev, IORESOURCE_MEM, index);
	if (res)
		return ocelot_regmap_from_mem_resource(dev, res, config);

	/*
	 * Fall back to using IORESOURCE_REG, which is possible in an
	 * MFD configuration
	 */
	res = platform_get_resource(pdev, IORESOURCE_REG, index);
	if (res)
		return ocelot_regmap_from_reg_resource(dev, res, config);

	return NULL;
}

struct regmap *
ocelot_regmap_from_resource(struct platform_device *pdev,
			    unsigned int index,
			    const struct regmap_config *config)
{
	struct regmap *map;

	map = ocelot_regmap_from_resource_optional(pdev, index, config);
	return map ? : ERR_PTR(-ENOENT);
}

I hope I didn't get something wrong, this is all code written within the
email client, so it is obviously not compiled/tested....

Yep - I definitely get the point. And thanks for the review.

The other (bigger?) issue is around how this MFD can be loaded as a
module. Right now it is pretty straightforward to say
#if IS_ENABLED(CONFIG_MFD_OCELOT). Theres a level of nuance if
CONFIG_MFD_OCELOT=m while the child devices are compiled in
(CONFIG_PINCTRL_MICROCHIP_SGPIO=y for example). It still feels like this
code belongs somewhere in platform / resource / device / mfd...?

It might be perfectly valid to have multiple SGPIO controllers - one
local and one remote / SPI. But without the CONFIG_MFD_OCELOT module
loaded, I don't think the SGPIO module would work.

This patch set deals with the issue by setting MFD_OCELOT to a boolean -
but in the long run I think a module makes sense. I admittedly haven't
spent enough time researching (bashing my head against the wall) this,
but this seems like a good opportunity to at least express that I'm
expecting to have to deal with this issue soon. I met with Alexandre at
ELC this past week, and he said Arnd (both added to CC) might be a good
resource - but again I'd like to do a little more searching before
throwing it over the wall.