RE: [PATCH] memory: dfl-emif: add the DFL EMIF private feature driver

[Wu, Hao]

> Subject: [PATCH] memory: dfl-emif: add the DFL EMIF private feature driver
> 
> This driver is for the EMIF private feature implemented under FPGA
> Device Feature List (DFL) framework. It is used to expose memory
> interface status information as well as memory clearing control.
> 
> The purpose of memory clearing block is to zero out all private memory
> when FPGA is to be reprogrammed. This gives users a reliable method to
> prevent potential data leakage.
> 
> Signed-off-by: Xu Yilun <yilun.xu@xxxxxxxxx>
> Signed-off-by: Russ Weight <russell.h.weight@xxxxxxxxx>
> ---
>  .../ABI/testing/sysfs-bus-dfl-devices-emif         |  25 +++
>  drivers/memory/Kconfig                             |   9 +
>  drivers/memory/Makefile                            |   2 +
>  drivers/memory/dfl-emif.c                          | 211 +++++++++++++++++++++
>  4 files changed, 247 insertions(+)
>  create mode 100644 Documentation/ABI/testing/sysfs-bus-dfl-devices-emif
>  create mode 100644 drivers/memory/dfl-emif.c
> 
> diff --git a/Documentation/ABI/testing/sysfs-bus-dfl-devices-emif
> b/Documentation/ABI/testing/sysfs-bus-dfl-devices-emif
> new file mode 100644
> index 0000000..33d557e
> --- /dev/null
> +++ b/Documentation/ABI/testing/sysfs-bus-dfl-devices-emif
> @@ -0,0 +1,25 @@
> +What:		/sys/bus/dfl/devices/dfl_dev.X/infX_cal_fail
> +Date:		Sep 2020
> +KernelVersion:	5.10
> +Contact:	Xu Yilun <yilun.xu@xxxxxxxxx>
> +Description:	Read-only. It indicates if the calibration is failed on this
> +		memory interface. "1" for calibration failure, "0" for OK.
> +		Format: %u
> +
> +What:		/sys/bus/dfl/devices/dfl_dev.X/infX_init_done
> +Date:		Sep 2020
> +KernelVersion:	5.10
> +Contact:	Xu Yilun <yilun.xu@xxxxxxxxx>
> +Description:	Read-only. It indicates if the initialization is complete on
> +		this memory interface. "1" for initialization complete, "0"
> +		for not yet.
> +		Format: %u
> +
> +What:		/sys/bus/dfl/devices/dfl_dev.X/infX_clear
> +Date:		Sep 2020
> +KernelVersion:	5.10
> +Contact:	Xu Yilun <yilun.xu@xxxxxxxxx>
> +Description:	Write-only. Writing "1" to this file will zero out all memory
> +		data in this memory interface. Writing other values are
> +		invalid.
> +		Format: %u
> diff --git a/drivers/memory/Kconfig b/drivers/memory/Kconfig
> index 8072204..fb0858f 100644
> --- a/drivers/memory/Kconfig
> +++ b/drivers/memory/Kconfig
> @@ -215,6 +215,15 @@ config STM32_FMC2_EBI
>  	  devices (like SRAM, ethernet adapters, FPGAs, LCD displays, ...) on
>  	  SOCs containing the FMC2 External Bus Interface.
> 
> +config FPGA_DFL_EMIF
> +	tristate "DFL EMIF private feature support"

Yilun,

I see you are submitting several different DFL device driver now, then do you
think if it's better to have some unified style of the naming?

> +	depends on FPGA_DFL && HAS_IOMEM
> +	help
> +	  This driver is for the EMIF private feature implemented under
> +	  FPGA Device Feature List (DFL) framework. It is used to expose
> +	  memory interface status information as well as memory clearing
> +	  control.
> +
>  source "drivers/memory/samsung/Kconfig"
>  source "drivers/memory/tegra/Kconfig"
> 
> diff --git a/drivers/memory/Makefile b/drivers/memory/Makefile
> index e71cf7b..0afbf39 100644
> --- a/drivers/memory/Makefile
> +++ b/drivers/memory/Makefile
> @@ -39,3 +39,5 @@ $(obj)/ti-emif-asm-offsets.h: $(obj)/emif-asm-offsets.s
> FORCE
> 
>  targets += emif-asm-offsets.s
>  clean-files += ti-emif-asm-offsets.h
> +
> +obj-$(CONFIG_FPGA_DFL_EMIF)	+= dfl-emif.o
> diff --git a/drivers/memory/dfl-emif.c b/drivers/memory/dfl-emif.c
> new file mode 100644
> index 0000000..442137f
> --- /dev/null
> +++ b/drivers/memory/dfl-emif.c
> @@ -0,0 +1,211 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Driver for DFL EMIF private feature
> + *
> + * Copyright (C) 2020 Intel Corporation, Inc.
> + *
> + */
> +#include <linux/bitfield.h>
> +#include <linux/errno.h>
> +#include <linux/fpga/dfl-bus.h>
> +#include <linux/io.h>
> +#include <linux/iopoll.h>
> +#include <linux/io-64-nonatomic-lo-hi.h>
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/spinlock.h>
> +#include <linux/stddef.h>
> +#include <linux/types.h>
> +
> +#define EMIF_STAT			0x8
> +#define EMIF_STAT_INIT_DONE_SFT		0
> +#define EMIF_STAT_CALC_FAIL_SFT		8
> +#define EMIF_STAT_CLEAR_BUSY_SFT	16
> +#define EMIF_CTRL			0x10
> +#define EMIF_CTRL_CLEAR_EN_SFT		0
> +#define EMIF_CTRL_CLEAR_EN_MSK		GENMASK_ULL(3, 0)
> +
> +#define EMIF_POLL_INVL			10000 /* us */
> +#define EMIF_POLL_TIMEOUT		5000000 /* us */
> +
> +struct dfl_emif {
> +	struct device *dev;
> +	void __iomem *base;
> +	spinlock_t lock;	/* Serialises access to EMIF_CTRL reg */
> +};
> +
> +struct emif_attr {
> +	struct device_attribute attr;
> +	u32 shift;
> +	u32 index;
> +};
> +
> +#define to_emif_attr(dev_attr) \
> +	container_of(dev_attr, struct emif_attr, attr)
> +
> +static ssize_t emif_state_show(struct device *dev,
> +			       struct device_attribute *attr, char *buf)
> +{
> +	struct emif_attr *eattr = to_emif_attr(attr);
> +	struct dfl_emif *de = dev_get_drvdata(dev);
> +	u64 val;
> +
> +	val = readq(de->base + EMIF_STAT);
> +
> +	return sprintf(buf, "%u\n",
> +		       !!(val & BIT_ULL(eattr->shift + eattr->index)));
> +}
> +
> +static ssize_t emif_clear_store(struct device *dev,
> +				struct device_attribute *attr,
> +				const char *buf, size_t count)
> +{
> +	struct emif_attr *eattr = to_emif_attr(attr);
> +	struct dfl_emif *de = dev_get_drvdata(dev);
> +	u64 clear_busy_msk, clear_en_msk, val;
> +	void __iomem *base = de->base;
> +
> +	if (!sysfs_streq(buf, "1"))
> +		return -EINVAL;
> +
> +	clear_busy_msk = BIT_ULL(EMIF_STAT_CLEAR_BUSY_SFT + eattr-
> >index);
> +	clear_en_msk = BIT_ULL(EMIF_CTRL_CLEAR_EN_SFT + eattr->index);
> +
> +	spin_lock(&de->lock);
> +	/* The CLEAR_EN field is WO, but other fields are RW */
> +	val = readq(base + EMIF_CTRL);
> +	val &= ~EMIF_CTRL_CLEAR_EN_MSK;
> +	val |= clear_en_msk;
> +	writeq(val, base + EMIF_CTRL);
> +	spin_unlock(&de->lock);

Should this lock protect below status polling? Is it possible another clear
happening right after this spin unlock and cause a different value to 
EMIF_CTRL before below status polling?

Thanks
Hao

> +
> +	if (readq_poll_timeout(base + EMIF_STAT, val,
> +			       !(val & clear_busy_msk),
> +			       EMIF_POLL_INVL, EMIF_POLL_TIMEOUT)) {
> +		dev_err(de->dev, "timeout, fail to clear\n");
> +		return -ETIMEDOUT;
> +	}
> +
> +	return count;
> +}
> +
> +#define emif_state_attr(_name, _shift, _index)
> 	\
> +	struct emif_attr emif_attr_##inf##_index##_##_name =
> 	\
> +		{ .attr = __ATTR(inf##_index##_##_name, 0444,
> 	\
> +				 emif_state_show, NULL),		\
> +		  .shift = (_shift), .index = (_index) }
> +
> +#define emif_clear_attr(_index)
> 	\
> +	struct emif_attr emif_attr_##inf##_index##_clear =		\
> +		{ .attr = __ATTR(inf##_index##_clear, 0200,		\
> +				 NULL, emif_clear_store),		\
> +		  .index = (_index) }
> +
> +emif_state_attr(init_done, EMIF_STAT_INIT_DONE_SFT, 0);
> +emif_state_attr(init_done, EMIF_STAT_INIT_DONE_SFT, 1);
> +emif_state_attr(init_done, EMIF_STAT_INIT_DONE_SFT, 2);
> +emif_state_attr(init_done, EMIF_STAT_INIT_DONE_SFT, 3);
> +
> +emif_state_attr(cal_fail, EMIF_STAT_CALC_FAIL_SFT, 0);
> +emif_state_attr(cal_fail, EMIF_STAT_CALC_FAIL_SFT, 1);
> +emif_state_attr(cal_fail, EMIF_STAT_CALC_FAIL_SFT, 2);
> +emif_state_attr(cal_fail, EMIF_STAT_CALC_FAIL_SFT, 3);
> +
> +emif_clear_attr(0);
> +emif_clear_attr(1);
> +emif_clear_attr(2);
> +emif_clear_attr(3);
> +
> +static struct attribute *dfl_emif_attrs[] = {
> +	&emif_attr_inf0_init_done.attr.attr,
> +	&emif_attr_inf0_cal_fail.attr.attr,
> +	&emif_attr_inf0_clear.attr.attr,
> +
> +	&emif_attr_inf1_init_done.attr.attr,
> +	&emif_attr_inf1_cal_fail.attr.attr,
> +	&emif_attr_inf1_clear.attr.attr,
> +
> +	&emif_attr_inf2_init_done.attr.attr,
> +	&emif_attr_inf2_cal_fail.attr.attr,
> +	&emif_attr_inf2_clear.attr.attr,
> +
> +	&emif_attr_inf3_init_done.attr.attr,
> +	&emif_attr_inf3_cal_fail.attr.attr,
> +	&emif_attr_inf3_clear.attr.attr,
> +
> +	NULL,
> +};
> +
> +static umode_t dfl_emif_visible(struct kobject *kobj,
> +				struct attribute *attr, int n)
> +{
> +	struct dfl_emif *de = dev_get_drvdata(kobj_to_dev(kobj));
> +	struct emif_attr *eattr = container_of(attr, struct emif_attr,
> +					       attr.attr);
> +	u64 val;
> +
> +	/*
> +	 * This device supports upto 4 memory interfaces, but not all
> +	 * interfaces are used on different platforms. The read out value of
> +	 * CLEAN_EN field (which is a bitmap) could tell how many interfaces
> +	 * are available.
> +	 */
> +	val = FIELD_GET(EMIF_CTRL_CLEAR_EN_MSK, readq(de->base +
> EMIF_CTRL));
> +
> +	return (val & BIT_ULL(eattr->index)) ? attr->mode : 0;
> +}
> +
> +static const struct attribute_group dfl_emif_group = {
> +	.is_visible = dfl_emif_visible,
> +	.attrs = dfl_emif_attrs,
> +};
> +
> +static const struct attribute_group *dfl_emif_groups[] = {
> +	&dfl_emif_group,
> +	NULL,
> +};
> +
> +static int dfl_emif_probe(struct dfl_device *dfl_dev)
> +{
> +	struct device *dev = &dfl_dev->dev;
> +	struct dfl_emif *de;
> +
> +	de = devm_kzalloc(dev, sizeof(*de), GFP_KERNEL);
> +	if (!de)
> +		return -ENOMEM;
> +
> +	de->base = devm_ioremap_resource(&dfl_dev->dev, &dfl_dev-
> >mmio_res);
> +	if (IS_ERR(de->base)) {
> +		dev_err(dev, "get mem resource fail!\n");
> +		return PTR_ERR(de->base);
> +	}
> +
> +	de->dev = dev;
> +	spin_lock_init(&de->lock);
> +	dev_set_drvdata(dev, de);
> +
> +	return 0;
> +}
> +
> +#define FME_FEATURE_ID_EMIF	0x9
> +
> +static const struct dfl_device_id dfl_emif_ids[] = {
> +	{ FME_ID, FME_FEATURE_ID_EMIF },
> +	{ }
> +};
> +
> +static struct dfl_driver dfl_emif_driver = {
> +	.drv	= {
> +		.name       = "dfl-emif",
> +		.dev_groups = dfl_emif_groups,
> +	},
> +	.id_table = dfl_emif_ids,
> +	.probe   = dfl_emif_probe,
> +};
> +
> +module_dfl_driver(dfl_emif_driver);
> +
> +MODULE_DEVICE_TABLE(dfl, dfl_emif_ids);
> +MODULE_DESCRIPTION("DFL EMIF driver");
> +MODULE_AUTHOR("Intel Corporation");
> +MODULE_LICENSE("GPL v2");
> --
> 2.7.4