mail archive of the barebox mailing list
 help / color / mirror / Atom feed
From: Marc Kleine-Budde <mkl@pengutronix.de>
To: Ahmad Fatoum <a.fatoum@pengutronix.de>
Cc: barebox@lists.infradead.org
Subject: Re: [PATCH 2/3] common: board: phytec: import SoM detection for imx8m based SoM from u-boot
Date: Tue, 6 Jun 2023 16:32:53 +0200	[thread overview]
Message-ID: <20230606-gratitude-chemist-598b553aab88-mkl@pengutronix.de> (raw)
In-Reply-To: <4430e8b6-3044-f7e0-0cee-a71fdf03838f@pengutronix.de>

[-- Attachment #1: Type: text/plain, Size: 16844 bytes --]

On 06.06.2023 14:51:05, Ahmad Fatoum wrote:
> On 06.06.23 12:50, Marc Kleine-Budde wrote:
> > This patch imports and cleans up the SoM detection for imx8n based SoM
> > from u-boot.
> > 
> > Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
> > ---
> >  common/boards/Kconfig                             |   7 +
> >  common/boards/Makefile                            |   1 +
> >  common/boards/phytec/Makefile                     |   4 +
> >  common/boards/phytec/phytec-som-detection.c       | 209 ++++++++++++++++++++++
> >  common/boards/phytec/phytec-som-imx8m-detection.c | 151 ++++++++++++++++
> >  include/phytec-som-detection.h                    |  69 +++++++
> >  include/phytec-som-imx8m-detection.h              |  19 ++
> >  7 files changed, 460 insertions(+)
> > 
> > diff --git a/common/boards/Kconfig b/common/boards/Kconfig
> > index e27273b7671d..b240548b484d 100644
> > --- a/common/boards/Kconfig
> > +++ b/common/boards/Kconfig
> > @@ -2,3 +2,10 @@
> >  
> >  config BOARD_QEMU_VIRT
> >  	bool
> > +
> > +config BOARD_PHYTEC_SOM_DETECTION
> > +	bool
> > +
> > +config BOARD_PHYTEC_SOM_IMX8M_DETECTION
> > +	bool
> > +	select BOARD_PHYTEC_SOM_DETECTION
> > diff --git a/common/boards/Makefile b/common/boards/Makefile
> > index 5b4e429c13e9..2a96ce6aec5c 100644
> > --- a/common/boards/Makefile
> > +++ b/common/boards/Makefile
> > @@ -1,3 +1,4 @@
> >  # SPDX-License-Identifier: GPL-2.0-only
> >  
> >  obj-$(CONFIG_BOARD_QEMU_VIRT)	+= qemu-virt/
> > +obj-y += phytec/
> > diff --git a/common/boards/phytec/Makefile b/common/boards/phytec/Makefile
> > new file mode 100644
> > index 000000000000..741a0e2eb704
> > --- /dev/null
> > +++ b/common/boards/phytec/Makefile
> > @@ -0,0 +1,4 @@
> > +# SPDX-License-Identifier: GPL-2.0-only
> > +
> > +lwl-$(CONFIG_BOARD_PHYTEC_SOM_DETECTION) += phytec-som-detection.o
> > +lwl-$(CONFIG_BOARD_PHYTEC_SOM_IMX8M_DETECTION) += phytec-som-imx8m-detection.o
> > diff --git a/common/boards/phytec/phytec-som-detection.c b/common/boards/phytec/phytec-som-detection.c
> > new file mode 100644
> > index 000000000000..d9479f8ced69
> > --- /dev/null
> > +++ b/common/boards/phytec/phytec-som-detection.c
> > @@ -0,0 +1,209 @@
> > +// SPDX-License-Identifier: GPL-2.0+
> > +/*
> > + * Copyright (C) 2020 PHYTEC Messtechnik GmbH
> > + * Author: Teresa Remmet <t.remmet@phytec.de>
> > + */
> > +
> > +#include <common.h>
> > +#include <pbl/eeprom.h>
> > +#include <phytec-som-imx8m-detection.h>
> > +
> > +struct phytec_eeprom_data eeprom_data;
> > +
> > +#define POLY (0x1070U << 3)
> > +
> > +static u8 _crc8(u16 data)
> > +{
> > +	int i;
> > +
> > +	for (i = 0; i < 8; i++) {
> > +		if (data & 0x8000)
> > +			data = data ^ POLY;
> > +		data = data << 1;
> > +	}
> > +
> > +	return data >> 8;
> > +}
> > +
> > +static unsigned int crc8(unsigned int crc, const u8 *vptr, int len)
> > +{
> > +	int i;
> > +
> > +	for (i = 0; i < len; i++)
> > +		crc = _crc8((crc ^ vptr[i]) << 8);
> > +
> > +	return crc;
> > +}
> 
> There's already a crc8 implementation. Why can't you reuse it?

It's a lookup table based approach and not pbl ready.

> > +
> > +char *phytec_get_opt(struct phytec_eeprom_data *data)
> 
> const return and const argument?

fixed

> > +{
> > +	char *opt;
> > +
> > +	if (!data)
> > +		data = &eeprom_data;
> > +
> > +	switch (data->api_rev) {
> > +	case PHYTEC_API_REV0:
> > +	case PHYTEC_API_REV1:
> > +		opt = data->data.data_api0.opt;
> > +		break;
> > +	case PHYTEC_API_REV2:
> > +		opt = data->data.data_api2.opt;
> > +		break;
> > +	default:
> > +		opt = NULL;
> > +		break;
> > +	};
> > +
> > +	return opt;
> > +}
> > +
> > +static int phytec_eeprom_data_init(struct pbl_i2c *i2c,
> > +				   struct phytec_eeprom_data *data,
> > +				   int addr, u8 phytec_som_type)
> > +{
> > +	unsigned int crc;
> > +	char *opt;
> > +	int *ptr;
> > +	int ret = -1, i;
> > +	u8 som;
> > +
> > +	if (!data)
> > +		data = &eeprom_data;
> > +
> > +	eeprom_read(i2c, addr, I2C_ADDR_16_BIT, data, sizeof(struct phytec_eeprom_data));
> > +
> > +	if (data->api_rev == 0xff) {
> > +		pr_err("%s: EEPROM is not flashed. Prototype?\n", __func__);
> > +		return -EINVAL;
> > +	}
> > +
> > +	for (i = 0, ptr = (int *)data;
> > +	     i < sizeof(struct phytec_eeprom_data);
> > +	     i += sizeof(ptr), ptr++)
> > +		if (*ptr != 0x0)
> > +			break;
> > +
> > +	if (i == sizeof(struct phytec_eeprom_data)) {
> > +		pr_err("%s: EEPROM data is all zero. Erased?\n", __func__);
> > +		return -EINVAL;
> > +	}
> > +
> > +	if (data->api_rev > PHYTEC_API_REV2) {
> > +		pr_err("%s: EEPROM API revision %u not supported\n",
> > +		       __func__, data->api_rev);
> > +		return -EINVAL;
> > +	}
> > +
> > +	/* We are done here for early revisions */
> > +	if (data->api_rev <= PHYTEC_API_REV1)
> > +		return 0;
> > +
> > +	crc = crc8(0, (const unsigned char *)data,
> > +		   sizeof(struct phytec_eeprom_data));
> > +	pr_debug("%s: crc: %x\n", __func__, crc);
> > +
> > +	if (crc) {
> > +		pr_err("%s: CRC mismatch. EEPROM data is not usable\n", __func__);
> > +		return -EINVAL;
> > +	}
> > +
> > +	som = data->data.data_api2.som_no;
> > +	pr_debug("%s: som id: %u\n", __func__, som);
> > +	opt = phytec_get_opt(data);
> > +	if (!opt)
> > +		return -EINVAL;
> > +
> > +	if (IS_ENABLED(CONFIG_BOARD_PHYTEC_SOM_IMX8M_DETECTION))
> 
> Why is this behind CONFIG_ symbol?

The original u-boot code offers the possibility to switch it off, too.

> > +		ret = phytec_imx8m_detect(som, opt, phytec_som_type);
> > +
> > +	if (ret) {
> > +		pr_err("%s: SoM ID does not match. Wrong EEPROM data?\n", __func__);
> > +		return -EINVAL;
> > +	}
> > +
> > +	return 0;
> > +}
> > +
> > +void phytec_print_som_info(struct phytec_eeprom_data *data)
> 
> const?

fixed

> 
> > +{
> > +	struct phytec_api2_data *api2;
> > +	char pcb_sub_rev;
> > +	unsigned int ksp_no, sub_som_type1 = -1, sub_som_type2 = -1;
> > +
> > +	if (!data)
> > +		data = &eeprom_data;
> > +
> > +	if (data->api_rev < PHYTEC_API_REV2)
> > +		return;
> > +
> > +	api2 = &data->data.data_api2;
> > +
> > +	/* Calculate PCB subrevision */
> > +	pcb_sub_rev = api2->pcb_sub_opt_rev & 0x0f;
> > +	pcb_sub_rev = pcb_sub_rev ? ((pcb_sub_rev - 1) + 'a') : ' ';
> > +
> > +	/* print standard product string */
> > +	if (api2->som_type <= 1) {
> > +		pr_info("SoM: %s-%03u-%s.%s PCB rev: %u%c\n",
> > +			phytec_som_type_str[api2->som_type], api2->som_no,
> > +			api2->opt, api2->bom_rev, api2->pcb_rev, pcb_sub_rev);
> > +		return;
> > +	}
> > +	/* print KSP/KSM string */
> > +	if (api2->som_type <= 3) {
> > +		ksp_no = (api2->ksp_no << 8) | api2->som_no;
> > +		pr_info("SoM: %s-%u ",
> > +			phytec_som_type_str[api2->som_type], ksp_no);
> > +		/* print standard product based KSP/KSM strings */
> > +	} else {
> > +		switch (api2->som_type) {
> > +		case 4:
> > +			sub_som_type1 = 0;
> > +			sub_som_type2 = 3;
> > +			break;
> > +		case 5:
> > +			sub_som_type1 = 0;
> > +			sub_som_type2 = 2;
> > +			break;
> > +		case 6:
> > +			sub_som_type1 = 1;
> > +			sub_som_type2 = 3;
> > +			break;
> > +		case 7:
> > +			sub_som_type1 = 1;
> > +			sub_som_type2 = 2;
> > +			break;
> > +		default:
> > +			break;
> > +		};
> > +
> > +		pr_info("SoM: %s-%03u-%s-%03u ",
> > +			phytec_som_type_str[sub_som_type1],
> > +			api2->som_no, phytec_som_type_str[sub_som_type2],
> > +			api2->ksp_no);
> > +	}
> > +
> > +	pr_cont("Option: %s BOM rev: %s PCB rev: %u%c\n", api2->opt,
> > +		api2->bom_rev, api2->pcb_rev, pcb_sub_rev);
> 
> pr_cont doesn't work as you'd expect in barebox if you output to log.
> Just do a separate print.

fixed

> 
> > +}
> > +
> > +int phytec_eeprom_data_setup(struct pbl_i2c *i2c, struct phytec_eeprom_data *data,
> > +			     int addr, int addr_fallback, u8 cpu_type)
> > +{
> > +	int ret;
> > +
> > +	ret = phytec_eeprom_data_init(i2c, data, addr, cpu_type);
> > +	if (ret) {
> > +		pr_err("%s: init failed. Trying fall back address 0x%x\n",
> > +		       __func__, addr_fallback);
> > +		ret = phytec_eeprom_data_init(i2c, data, addr_fallback, cpu_type);
> > +	}
> > +
> > +	if (ret)
> > +		pr_err("%s: EEPROM data init failed\n", __func__);
> > +	else
> > +		pr_debug("%s: init successful\n", __func__);
> > +
> > +	return ret;
> > +}
> > diff --git a/common/boards/phytec/phytec-som-imx8m-detection.c b/common/boards/phytec/phytec-som-imx8m-detection.c
> > new file mode 100644
> > index 000000000000..cc7fb6971548
> > --- /dev/null
> > +++ b/common/boards/phytec/phytec-som-imx8m-detection.c
> > @@ -0,0 +1,151 @@
> > +// SPDX-License-Identifier: GPL-2.0+
> > +/*
> > + * Copyright (C) 2020 PHYTEC Messtechnik GmbH
> > + * Author: Teresa Remmet <t.remmet@phytec.de>
> > + */
> > +
> > +#include <common.h>
> > +#include <mach/imx/generic.h>
> > +#include <phytec-som-imx8m-detection.h>
> > +
> > +extern struct phytec_eeprom_data eeprom_data;
> > +
> > +/* Check if the SoM is actually one of the following products:
> > + * - i.MX8MM
> > + * - i.MX8MN
> > + * - i.MX8MP
> > + * - i.MX8MQ
> > + *
> > + * Returns 0 in case it's a known SoM. Otherwise, returns -errno.
> > + */
> > +u8 phytec_imx8m_detect(u8 som, char *opt, u8 cpu_type)
> 
> int. Even moreso as you're returning -EINVAL in the error case.

fixed

> 
> > +{
> > +	if (som == PHYTEC_IMX8MP_SOM && cpu_type == IMX_CPU_IMX8MP)
> > +		return 0;
> > +
> > +	if (som == PHYTEC_IMX8MM_SOM) {
> > +		if (((opt[0] - '0') != 0) &&
> > +		    ((opt[1] - '0') == 0) && cpu_type == IMX_CPU_IMX8MM)
> > +			return 0;
> > +		else if (((opt[0] - '0') == 0) &&
> > +			 ((opt[1] - '0') != 0) && cpu_type == IMX_CPU_IMX8MN)
> > +			return 0;
> > +		return -EINVAL;
> > +	}
> > +
> > +	if (som == PHYTEC_IMX8MQ_SOM && cpu_type == IMX_CPU_IMX8MQ)
> > +		return 0;
> > +
> > +	return -EINVAL;
> > +}
> > +
> > +/*
> > + * So far all PHYTEC i.MX8M boards have RAM size definition at the
> > + * same location.
> > + */
> > +u8 phytec_get_imx8m_ddr_size(struct phytec_eeprom_data *data)
> > +{
> > +	char *opt;
> > +	u8 ddr_id;
> > +
> > +	if (!data)
> > +		data = &eeprom_data;
> > +
> > +	opt = phytec_get_opt(data);
> > +	if (opt)
> > +		ddr_id = opt[2] - '0';
> > +	else
> > +		ddr_id = PHYTEC_EEPROM_INVAL;
> > +
> > +	pr_debug("%s: ddr id: %u\n", __func__, ddr_id);
> > +
> > +	return ddr_id;
> > +}
> > +
> > +/*
> > + * Filter SPI-NOR flash information. All i.MX8M boards have this at
> > + * the same location.
> > + * returns: 0x0 if no SPI is poulated. Otherwise a board depended
> 
> populated.

fixed

> 
> > + * code for the size. PHYTEC_EEPROM_INVAL when the data is invalid.
> > + */
> > +u8 phytec_get_imx8m_spi(struct phytec_eeprom_data *data)
> > +{
> > +	char *opt;
> > +	u8 spi;
> > +
> > +	if (!data)
> > +		data = &eeprom_data;
> > +
> > +	if (data->api_rev < PHYTEC_API_REV2)
> > +		return PHYTEC_EEPROM_INVAL;
> > +
> > +	opt = phytec_get_opt(data);
> > +	if (opt)
> > +		spi = opt[4] - '0';
> > +	else
> > +		spi = PHYTEC_EEPROM_INVAL;
> 
> Why not return -EINVAL? If you want to handle 0xff
> specially, just check for it and return -EINVAL.
> 
> > +
> > +	pr_debug("%s: spi: %u\n", __func__, spi);
> 
> Do we really need this debug print? I think the debug print should rather
> go into the future caller.

This comes from the original u-boot code and it's not used here. Should
I remove it completely?

> 
> > +
> > +	return spi;
> > +}
> > +
> > +/*
> > + * Filter ethernet phy information. All i.MX8M boards have this at
> > + * the same location.
> > + * returns: 0x0 if no ethernet phy is poulated. 0x1 if it is populated.
> > + * PHYTEC_EEPROM_INVAL when the data is invalid.
> > + */
> > +u8 phytec_get_imx8m_eth(struct phytec_eeprom_data *data)
> > +{
> > +	char *opt;
> > +	u8 eth;
> > +
> > +	if (!data)
> > +		data = &eeprom_data;
> > +
> > +	if (data->api_rev < PHYTEC_API_REV2)
> > +		return PHYTEC_EEPROM_INVAL;
> > +
> > +	opt = phytec_get_opt(data);
> > +	if (opt) {
> > +		eth = opt[5] - '0';
> > +		eth &= 0x1;
> > +	} else {
> > +		eth = PHYTEC_EEPROM_INVAL;
> > +	}
> > +
> > +	pr_debug("%s: eth: %u\n", __func__, eth);
> > +
> > +	return eth;
> > +}
> > +
> > +/*
> > + * Filter RTC information.
> > + * returns: 0 if no RTC is poulated. 1 if it is populated.
> > + * PHYTEC_EEPROM_INVAL when the data is invalid.
> > + */
> > +u8 phytec_get_imx8mp_rtc(struct phytec_eeprom_data *data)
> > +{
> > +	char *opt;
> > +	u8 rtc;
> > +
> > +	if (!data)
> > +		data = &eeprom_data;
> > +
> > +	if (data->api_rev < PHYTEC_API_REV2)
> > +		return PHYTEC_EEPROM_INVAL;
> > +
> > +	opt = phytec_get_opt(data);
> > +	if (opt) {
> > +		rtc = opt[5] - '0';
> > +		rtc &= 0x4;
> > +		rtc = !(rtc >> 2);
> > +	} else {
> > +		rtc = PHYTEC_EEPROM_INVAL;
> > +	}
> > +
> > +	pr_debug("%s: rtc: %u\n", __func__, rtc);
> > +
> > +	return rtc;
> > +}
> > diff --git a/include/phytec-som-detection.h b/include/phytec-som-detection.h
> > new file mode 100644
> > index 000000000000..65c7cb561e1d
> > --- /dev/null
> > +++ b/include/phytec-som-detection.h
> > @@ -0,0 +1,69 @@
> > +/* SPDX-License-Identifier: GPL-2.0-or-later */
> > +/*
> > + * Copyright (C) 2020 PHYTEC Messtechnik GmbH
> > + * Author: Teresa Remmet <t.remmet@phytec.de>
> > + */
> > +
> > +#ifndef _PHYTEC_SOM_DETECTION_H
> > +#define _PHYTEC_SOM_DETECTION_H
> > +
> > +#include <common.h>
> > +#include <pbl/i2c.h>
> > +
> > +#define PHYTEC_MAX_OPTIONS	17
> > +#define PHYTEC_IMX8MQ_SOM	66
> > +#define PHYTEC_IMX8MM_SOM	69
> > +#define PHYTEC_IMX8MP_SOM	70
> > +
> > +#define PHYTEC_EEPROM_INVAL	0xff
> > +
> > +enum {
> > +	PHYTEC_API_REV0 = 0,
> > +	PHYTEC_API_REV1,
> > +	PHYTEC_API_REV2,
> > +};
> > +
> > +static const char * const phytec_som_type_str[] = {
> > +	"PCM",
> > +	"PCL",
> > +	"KSM",
> > +	"KSP",
> > +};
> > +
> > +struct phytec_api0_data {
> > +	u8 pcb_rev;		/* PCB revision of SoM */
> > +	u8 som_type;		/* SoM type */
> > +	u8 ksp_no;		/* KSP no */
> > +	char opt[16];		/* SoM options */
> > +	u8 mac[6];		/* MAC address (optional) */
> > +	u8 pad[5];		/* padding */
> 
> __pad

fixed

> > +	u8 cksum;		/* checksum */
> > +} __attribute__ ((__packed__));
> > +
> > +struct phytec_api2_data {
> > +	u8 pcb_rev;		/* PCB revision of SoM */
> > +	u8 pcb_sub_opt_rev;	/* PCB subrevision and opt revision */
> > +	u8 som_type;		/* SoM type */
> > +	u8 som_no;		/* SoM number */
> > +	u8 ksp_no;		/* KSP information */
> > +	char opt[PHYTEC_MAX_OPTIONS]; /* SoM options */
> > +	char bom_rev[2];	/* BOM revision */
> > +	u8 mac[6];		/* MAC address (optional) */
> > +	u8 crc8;		/* checksum */
> > +} __attribute__ ((__packed__));
> > +
> > +struct phytec_eeprom_data {
> > +	u8 api_rev;
> > +	union {
> > +		struct phytec_api0_data data_api0;
> > +		struct phytec_api2_data data_api2;
> > +	} data;
> > +} __attribute__ ((__packed__));
> > +
> > +char *phytec_get_opt(struct phytec_eeprom_data *data);
> > +int phytec_eeprom_data_setup(struct pbl_i2c *i2c, struct phytec_eeprom_data *data,
> > +			     int addr, int addr_fallback, u8 cpu_type);
> > +
> > +void phytec_print_som_info(struct phytec_eeprom_data *data);
> > +
> > +#endif /* _PHYTEC_SOM_DETECTION_H */
> > diff --git a/include/phytec-som-imx8m-detection.h b/include/phytec-som-imx8m-detection.h
> > new file mode 100644
> > index 000000000000..108d40bb4030
> > --- /dev/null
> > +++ b/include/phytec-som-imx8m-detection.h
> 
> include/boards/phytec/som-imx8m-detection.h or similar.

good idea!

> 
> > @@ -0,0 +1,19 @@
> > +/* SPDX-License-Identifier: GPL-2.0-or-later */
> > +/*
> > + * Copyright (C) 2020 PHYTEC Messtechnik GmbH
> > + * Author: Teresa Remmet <t.remmet@phytec.de>
> > + */
> > +
> > +#ifndef _PHYTEC_SOM_IMX8M_DETECTION_H
> > +#define _PHYTEC_SOM_IMX8M_DETECTION_H
> > +
> > +#include <common.h>
> > +#include <phytec-som-detection.h>
> > +
> > +u8 phytec_imx8m_detect(u8 som, char *opt, u8 cpu_type);
> > +u8 phytec_get_imx8m_ddr_size(struct phytec_eeprom_data *data);
> > +u8 phytec_get_imx8mp_rtc(struct phytec_eeprom_data *data);
> > +u8 phytec_get_imx8m_spi(struct phytec_eeprom_data *data);
> > +u8 phytec_get_imx8m_eth(struct phytec_eeprom_data *data);
> > +
> > +#endif /* _PHYTEC_SOM_IMX8M_DETECTION_H */
> > 

Marc

-- 
Pengutronix e.K.                 | Marc Kleine-Budde          |
Embedded Linux                   | https://www.pengutronix.de |
Vertretung Nürnberg              | Phone: +49-5121-206917-129 |
Amtsgericht Hildesheim, HRA 2686 | Fax:   +49-5121-206917-9   |

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

  reply	other threads:[~2023-06-06 14:34 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-06-06 10:50 [PATCH 0/3] ARM: i.MX8MM: add Phytec i.MX8 SoM support Marc Kleine-Budde
2023-06-06 10:50 ` [PATCH 1/3] i2c: add <pbl/eeprom.h> for PBL use Marc Kleine-Budde
2023-06-06 11:21   ` Sascha Hauer
2023-06-06 11:33     ` Marc Kleine-Budde
2023-06-06 10:50 ` [PATCH 2/3] common: board: phytec: import SoM detection for imx8m based SoM from u-boot Marc Kleine-Budde
2023-06-06 12:51   ` Ahmad Fatoum
2023-06-06 14:32     ` Marc Kleine-Budde [this message]
2023-06-06 10:50 ` [PATCH 3/3] ARM: i.MX8MM: add Phytec i.MX8 SoM support Marc Kleine-Budde

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20230606-gratitude-chemist-598b553aab88-mkl@pengutronix.de \
    --to=mkl@pengutronix.de \
    --cc=a.fatoum@pengutronix.de \
    --cc=barebox@lists.infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox