Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Sat, 19 Feb 2022 00:05:55 +0800
From:   Xu Yilun <yilun.xu@intel.com>
To:     Ivan Bornyakov <i.bornyakov@metrotek.ru>
Cc:     mdf@kernel.org, hao.wu@intel.com, trix@redhat.com,
        conor.dooley@microchip.com, linux-kernel@vger.kernel.org,
        linux-fpga@vger.kernel.org, system@metrotek.ru
Subject: Re: [PATCH v4] fpga: microchip-spi: add Microchip FPGA manager
Message-ID: <20220218160555.GA1333893@yilunxu-OptiPlex-7050>
References: <20220214133835.25097-1-i.bornyakov@metrotek.ru>
 <20220217191851.11730-1-i.bornyakov@metrotek.ru>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220217191851.11730-1-i.bornyakov@metrotek.ru>
Precedence: bulk

On Thu, Feb 17, 2022 at 10:18:51PM +0300, Ivan Bornyakov wrote:
> Add support to the FPGA manager for programming Microchip Polarfire
> FPGAs over slave SPI interface.
> 
> Signed-off-by: Ivan Bornyakov <i.bornyakov@metrotek.ru>
> ---
> Changelog:
>   v1 -> v2: fix printk formating
>   v2 -> v3:
>    * replace "microsemi" with "microchip"
>    * replace prefix "microsemi_fpga_" with "mpf_"
>    * more sensible .compatible and .name strings
>    * remove unused defines STATUS_SPI_VIOLATION and STATUS_SPI_ERROR
>   v3 -> v4: fix unused variable warning
>     Put 'mpf_of_ids' definition under conditional compilation, so it
>     would not hang unused if CONFIG_OF is not enabled.
> 
>  drivers/fpga/Kconfig         |   9 +
>  drivers/fpga/Makefile        |   1 +
>  drivers/fpga/microchip-spi.c | 361 +++++++++++++++++++++++++++++++++++
>  3 files changed, 371 insertions(+)
>  create mode 100644 drivers/fpga/microchip-spi.c
> 
> diff --git a/drivers/fpga/Kconfig b/drivers/fpga/Kconfig
> index 26025dbab353..4240c641b100 100644
> --- a/drivers/fpga/Kconfig
> +++ b/drivers/fpga/Kconfig
> @@ -248,4 +248,13 @@ config FPGA_MGR_VERSAL_FPGA
>  	  configure the programmable logic(PL).
>  
>  	  To compile this as a module, choose M here.
> +
> +config FPGA_MGR_MICROCHIP_SPI
> +	tristate "Microchip Polarfire SPI FPGA manager"
> +	depends on SPI
> +	select CRC_CCITT
> +	help
> +	  FPGA manager driver support for Microchip Polarfire FPGAs
> +	  programming over slave SPI interface.
> +
>  endif # FPGA
> diff --git a/drivers/fpga/Makefile b/drivers/fpga/Makefile
> index 4da5273948df..fcb389ca4873 100644
> --- a/drivers/fpga/Makefile
> +++ b/drivers/fpga/Makefile
> @@ -19,6 +19,7 @@ obj-$(CONFIG_FPGA_MGR_XILINX_SPI)	+= xilinx-spi.o
>  obj-$(CONFIG_FPGA_MGR_ZYNQ_FPGA)	+= zynq-fpga.o
>  obj-$(CONFIG_FPGA_MGR_ZYNQMP_FPGA)	+= zynqmp-fpga.o
>  obj-$(CONFIG_FPGA_MGR_VERSAL_FPGA)      += versal-fpga.o
> +obj-$(CONFIG_FPGA_MGR_MICROCHIP_SPI)	+= microchip-spi.o
>  obj-$(CONFIG_ALTERA_PR_IP_CORE)         += altera-pr-ip-core.o
>  obj-$(CONFIG_ALTERA_PR_IP_CORE_PLAT)    += altera-pr-ip-core-plat.o
> 
> diff --git a/drivers/fpga/microchip-spi.c b/drivers/fpga/microchip-spi.c
> new file mode 100644
> index 000000000000..5db25734a27a
> --- /dev/null
> +++ b/drivers/fpga/microchip-spi.c
> @@ -0,0 +1,361 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Microchip Polarfire FPGA programming over slave SPI interface.
> + */
> +
> +#include <linux/module.h>
> +#include <linux/spi/spi.h>
> +#include <linux/of_device.h>
> +#include <linux/fpga/fpga-mgr.h>
> +#include <linux/delay.h>
> +#include <linux/crc-ccitt.h>

Please list them in alphabetical order.

> +
> +#define	SPI_ISC_ENABLE		0x0B
> +#define	SPI_ISC_DISABLE		0x0C
> +#define	SPI_READ_STATUS		0x00
> +#define	SPI_READ_DATA		0x01
> +#define	SPI_FRAME_INIT		0xAE
> +#define	SPI_FRAME		0xEE
> +#define	SPI_PRG_MODE		0x01
> +#define	SPI_RELEASE		0x23
> +
> +#define	SPI_FRAME_SIZE		16
> +
> +#define	HEADER_SIZE_OFFSET	24
> +#define	DATA_SIZE_OFFSET	55
> +
> +#define	LOOKUP_TABLE_RECORD_SIZE	9
> +#define	LOOKUP_TABLE_BLOCK_ID_OFFSET	0
> +#define	LOOKUP_TABLE_BLOCK_START_OFFSET	1
> +
> +#define	COMPONENTS_SIZE_ID	5
> +#define	BITSTREAM_ID		8
> +
> +#define	BITS_PER_COMPONENT_SIZE	22
> +
> +#define	STATUS_POLL_TIMEOUT_MS	1000
> +#define	STATUS_BUSY		BIT(0)
> +#define	STATUS_READY		BIT(1)

Maybe add some prefix for these vendor specific definitions, like
MFP_SPI_ISC_ENABLE.

> +
> +struct mpf_priv {
> +	struct spi_device *spi;
> +	bool program_mode;
> +};
> +
> +static enum fpga_mgr_states mpf_ops_state(struct fpga_manager *mgr)
> +{
> +	struct mpf_priv *priv = mgr->priv;
> +	struct spi_device *spi = priv->spi;
> +	bool program_mode = priv->program_mode;
> +	ssize_t status;
> +
> +	status = spi_w8r8(spi, SPI_READ_STATUS);
> +
> +	if (!program_mode && !status)
> +		return FPGA_MGR_STATE_OPERATING;
> +
> +	return FPGA_MGR_STATE_UNKNOWN;
> +}
> +
> +static int poll_status_not_busy(struct spi_device *spi, u8 mask)
> +{
> +	ssize_t status, timeout = STATUS_POLL_TIMEOUT_MS;
> +
> +	while (timeout--) {
> +		status = spi_w8r8(spi, SPI_READ_STATUS);
> +		if (status < 0)
> +			return status;
> +
> +		if (mask) {
> +			if (!(status & STATUS_BUSY) && (status & mask))
> +				return status;
> +		} else {
> +			if (!(status & STATUS_BUSY))
> +				return status;
> +		}

!(status & STATUS_BUSY) is always checked regardless of mask, so may
move the check out of the if...else statement.

> +
> +		mdelay(1);

busy wait for 1ms is discouraged, maybe usleep_range(). See
Documentation/timers/timers-howto.rst

> +	}

The actual timeout may be much longer than what is defined. To be more
accurate, you may use:

	time_after(jiffies, timeout_jiffies)

> +
> +	return -EBUSY;
> +}
> +
> +static int mpf_spi_write(struct spi_device *spi, const void *buf, size_t buf_size)
> +{
> +	int status = poll_status_not_busy(spi, 0);
> +
> +	if (status < 0)
> +		return status;
> +
> +	return spi_write(spi, buf, buf_size);
> +}
> +
> +static int mpf_spi_write_then_read(struct spi_device *spi,
> +				   const void *txbuf, size_t txbuf_size,
> +				   void *rxbuf, size_t rxbuf_size)
> +{
> +	const u8 read_command[] = { SPI_READ_DATA };
> +	int ret;
> +
> +	ret = mpf_spi_write(spi, txbuf, txbuf_size);
> +	if (ret)
> +		return ret;
> +
> +	ret = poll_status_not_busy(spi, STATUS_READY);
> +	if (ret < 0)
> +		return ret;
> +
> +	return spi_write_then_read(spi, read_command, sizeof(read_command),
> +				   rxbuf, rxbuf_size);
> +}
> +
> +static int mpf_ops_write_init(struct fpga_manager *mgr,
> +			      struct fpga_image_info *info, const char *buf,
> +			      size_t count)
> +{
> +	const u8 isc_en_command[] = { SPI_ISC_ENABLE };
> +	const u8 program_mode[] = { SPI_FRAME_INIT, SPI_PRG_MODE };

Better we follow the reverse xmas tree declaration, so reverse the 2
lines please.

> +	struct mpf_priv *priv = mgr->priv;
> +	struct spi_device *spi = priv->spi;
> +	struct device *dev = &mgr->dev;
> +	u32 isc_ret;
> +	int ret;
> +
> +	if (info->flags & FPGA_MGR_PARTIAL_RECONFIG) {
> +		dev_err(dev, "Partial reconfiguration is not supported\n");
> +

remove the blank line please

> +		return -EOPNOTSUPP;
> +	}
> +
> +	ret = mpf_spi_write_then_read(spi, isc_en_command, sizeof(isc_en_command),
> +				      &isc_ret, sizeof(isc_ret));
> +	if (ret || isc_ret) {
> +		dev_err(dev, "Failed to enable ISC: %d\n", ret ? ret : isc_ret);

							   ret ? : isc_ret

> +

remove the blank line please

> +		return -EFAULT;
> +	}
> +
> +	ret = mpf_spi_write(spi, program_mode, sizeof(program_mode));
> +	if (ret) {
> +		dev_err(dev, "Failed to enter program mode: %d\n", ret);
> +

the same

> +		return ret;
> +	}
> +
> +	priv->program_mode = true;
> +
> +	return 0;
> +}
> +
> +static ssize_t lookup_block_start(int id, const char *buf, size_t buf_size)
> +{
> +	u8 header_size, blocks_num, block_id;
> +	u32 block_start, i;
> +
> +	header_size = *(buf + HEADER_SIZE_OFFSET);
> +
> +	if (header_size > buf_size)
> +		return -EFAULT;
> +
> +	blocks_num = *(buf + header_size - 1);
> +
> +	if (header_size + blocks_num * LOOKUP_TABLE_RECORD_SIZE > buf_size)
> +		return -EFAULT;
> +
> +	for (i = 0; i < blocks_num; i++) {
> +		block_id = *(buf + header_size + LOOKUP_TABLE_RECORD_SIZE * i +
> +			     LOOKUP_TABLE_BLOCK_ID_OFFSET);
> +
> +		if (block_id == id) {
> +			memcpy(&block_start,
> +			       buf + header_size +
> +			       LOOKUP_TABLE_RECORD_SIZE * i +
> +			       LOOKUP_TABLE_BLOCK_START_OFFSET,
> +			       sizeof(block_start));

why a memcpy here, could we just read from the offset?

> +
> +			return le32_to_cpu(block_start);
> +		}
> +	}
> +
> +	return -EFAULT;
> +}
> +
> +static ssize_t parse_bitstream_size(const char *buf, size_t buf_size)
> +{
> +	ssize_t	bitstream_size = 0, components_size_start = 0,
> +		component_size_byte_num, component_size_byte_off, i;
> +	u16 components_num;
> +	u32 component_size;
> +
> +	memcpy(&components_num, buf + DATA_SIZE_OFFSET, sizeof(components_num));

the same, why a memcpy?

> +	components_num = le16_to_cpu(components_num);
> +
> +	components_size_start = lookup_block_start(COMPONENTS_SIZE_ID, buf,
> +						   buf_size);
> +	if (components_size_start < 0)
> +		return components_size_start;
> +
> +	if (components_size_start +
> +	    DIV_ROUND_UP(components_num * BITS_PER_COMPONENT_SIZE,
> +			 BITS_PER_BYTE) > buf_size)
> +		return -EFAULT;
> +
> +	for (i = 0; i < components_num; i++) {
> +		component_size_byte_num =
> +			(i * BITS_PER_COMPONENT_SIZE) / BITS_PER_BYTE;
> +		component_size_byte_off =
> +			(i * BITS_PER_COMPONENT_SIZE) % BITS_PER_BYTE;
> +
> +		memcpy(&component_size,
> +		       buf + components_size_start + component_size_byte_num,
> +		       sizeof(component_size));
> +		component_size = le32_to_cpu(component_size);
> +		component_size >>= component_size_byte_off;
> +		component_size &= GENMASK(BITS_PER_COMPONENT_SIZE - 1, 0);
> +
> +		bitstream_size += component_size;
> +	}
> +
> +	return bitstream_size;
> +}
> +
> +static int mpf_ops_write(struct fpga_manager *mgr, const char *buf, size_t count)
> +{
> +	ssize_t bitstream_start = 0, bitstream_size;
> +	struct mpf_priv *priv = mgr->priv;
> +	struct spi_device *spi = priv->spi;
> +	struct device *dev = &mgr->dev;
> +	u8 tmp_buf[SPI_FRAME_SIZE + 1];
> +	int ret, i;
> +
> +	if (crc_ccitt(0, buf, count)) {
> +		dev_err(dev, "CRC error\n");
> +
> +		return -EINVAL;
> +	}
> +
> +	bitstream_start = lookup_block_start(BITSTREAM_ID, buf, count);
> +	if (bitstream_start < 0) {
> +		dev_err(dev, "Failed to find bitstream start %zd\n",
> +			bitstream_start);
> +
> +		return bitstream_start;
> +	}
> +
> +	bitstream_size = parse_bitstream_size(buf, count);
> +	if (bitstream_size < 0) {
> +		dev_err(dev, "Failed to parse bitstream size %zd\n",
> +			bitstream_size);
> +
> +		return bitstream_size;
> +	}
> +
> +	if (bitstream_start + bitstream_size * SPI_FRAME_SIZE > count) {
> +		dev_err(dev,
> +			"Bitstram outruns firmware. Bitstream start %zd, bitstream size %zd, firmware size %zu\n",

			 Bitstream

> +			bitstream_start, bitstream_size * SPI_FRAME_SIZE, count);
> +
> +		return -EFAULT;
> +	}
> +

If I understand right, this function assumes the users provide the
entire image buffer. But it is possible the image buffer is from a
scatter list and the callback would be called several times.

Maybe the bitstream info at the head of the image could be parsed in
write_init(), and this requires the driver fill the
fpga_manager_ops.initial_header_size

> +	for (i = 0; i < bitstream_size; i++) {
> +		tmp_buf[0] = SPI_FRAME;
> +		memcpy(tmp_buf + 1, buf + bitstream_start + i * SPI_FRAME_SIZE,
> +		       SPI_FRAME_SIZE);
> +
> +		ret = mpf_spi_write(spi, tmp_buf, sizeof(tmp_buf));

Is it possible we use spi_sync_transfer to avoid memcpy the whole
bitstream?

> +		if (ret) {
> +			dev_err(dev,
> +				"Failed to write bitstream frame number %d of %zd\n",
> +				i, bitstream_size);
> +
> +			return ret;
> +		}
> +	}
> +
> +	return 0;
> +}
> +
> +static int mpf_ops_write_complete(struct fpga_manager *mgr,
> +					     struct fpga_image_info *info)

"CHECK: Alignment should match open parenthesis" from checkpatch.pl --strict

> +{
> +	const u8 isc_dis_command[] = { SPI_ISC_DISABLE };
> +	const u8 release_command[] = { SPI_RELEASE };
> +	struct mpf_priv *priv = mgr->priv;
> +	struct spi_device *spi = priv->spi;
> +	struct device *dev = &mgr->dev;
> +	int ret;
> +
> +	ret = mpf_spi_write(spi, isc_dis_command, sizeof(isc_dis_command));
> +	if (ret) {
> +		dev_err(dev, "Failed to disable ISC: %d\n", ret);
> +
> +		return ret;
> +	}
> +
> +	mdelay(1);

Same concern

> +
> +	ret = mpf_spi_write(spi, release_command, sizeof(release_command));
> +	if (ret) {
> +		dev_err(dev, "Failed to exit program mode: %d\n", ret);
> +
> +		return ret;
> +	}
> +
> +	priv->program_mode = false;
> +
> +	return 0;
> +}
> +
> +static const struct fpga_manager_ops mpf_ops = {
> +	.state = mpf_ops_state,
> +	.write_init = mpf_ops_write_init,
> +	.write = mpf_ops_write,
> +	.write_complete = mpf_ops_write_complete,
> +};
> +
> +static int mpf_probe(struct spi_device *spi)
> +{
> +	struct mpf_priv *priv;
> +	struct device *dev = &spi->dev;
> +	struct fpga_manager *mgr;
> +
> +	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
> +	if (!priv)
> +		return -ENOMEM;
> +
> +	priv->spi = spi;
> +
> +	mgr = devm_fpga_mgr_register(dev, "Microchip Polarfire SPI FPGA Manager",
> +				     &mpf_ops, priv);
> +
> +	return PTR_ERR_OR_ZERO(mgr);
> +}
> +
> +static const struct spi_device_id mpf_spi_ids[] = {
> +	{ .name = "mpf-spi-fpga-mgr", },
> +	{},
> +};
> +MODULE_DEVICE_TABLE(spi, mpf_spi_ids);
> +
> +#if IS_ENABLED(CONFIG_OF)
> +static const struct of_device_id mpf_of_ids[] = {
> +	{ .compatible = "microchip,mpf-spi-fpga-mgr" },

From checkpatch.pl

"WARNING: DT compatible string "microchip,mpf-spi-fpga-mgr" appears un-documented -- check ./Documentation/devicetree/bindings/"

Thanks,
Yilun

> +	{},
> +};
> +MODULE_DEVICE_TABLE(of, mpf_of_ids);
> +#endif /* IS_ENABLED(CONFIG_OF) */
> +
> +static struct spi_driver mpf_driver = {
> +	.probe = mpf_probe,
> +	.id_table = mpf_spi_ids,
> +	.driver = {
> +		.name = "microchip_mpf_spi_fpga_mgr",
> +		.of_match_table = of_match_ptr(mpf_of_ids),
> +	},
> +};
> +
> +module_spi_driver(mpf_driver);
> +
> +MODULE_DESCRIPTION("Microchip Polarfire SPI FPGA Manager");
> +MODULE_LICENSE("GPL");
> -- 
> 2.34.1
>