Skip to contentSkip to conversation

mtd: spi-nor: OTP support

PATCH Superseded Michael Walle 2 participants 6 messages Feb 16, 2021 – Mar 1, 2021

Subsystems (2)

v3 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

v3 cover mtd: spi-nor: OTP support

The first version of this patchset was posted over a year ago and since then, there was not a single comment. I have to admit, this is rather frustrating. I've now stripped down the patchset to only two patches and moved the code into an own module otp.c. If you like, I can also add myself to a reviewer for the OTP code.

This patchset implements the MTD OTP functions to allow access to the SPI OTP data. Specific support is added for Winbond flash chips.

In the past there was already an attempt by Rahul Bedarkar to add this, but there was no response. These patches are slightly based on his work.

https://lore.kernel.org/linux-mtd/[email protected]/

Changes since v2:
- improved commit messages
- add buffer size check in spi_nor_mtd_otp_info(). just to be sure, the buffer is hardcoded to 4k by the mtd subsys
- moved all code to otp.c
- dropped the patches introduced in v2

Changes since v1:
- added methods for Macronix and similar flashes
- added patch to cleanup/consolidate code in core.c

Michael Walle (2):
mtd: spi-nor: add OTP support
mtd: spi-nor: implement OTP support for Winbond and similar flashes

 drivers/mtd/spi-nor/Makefile  |   1 +
 drivers/mtd/spi-nor/core.c    |  10 +-
 drivers/mtd/spi-nor/core.h    |  56 ++++++
 drivers/mtd/spi-nor/otp.c     | 313 ++++++++++++++++++++++++++++++++++
 drivers/mtd/spi-nor/winbond.c |  17 +-
 include/linux/mtd/spi-nor.h   |   9 +
 6 files changed, 403 insertions(+), 3 deletions(-)
 create mode 100644 drivers/mtd/spi-nor/otp.c
Signature · 1 line
2.20.1
#1 lore raw

PATCH v3 1/2 mtd: spi-nor: add OTP support

SPI flashes sometimes have a special OTP area, which can (and is) used to store immutable properties like board serial number or vendor assigned network hardware addresses.

The MTD subsystem already supports accessing such areas and some (non SPI-NOR) flashes already implement support for it. It differentiates between user and factory areas. User areas can be written by the user and factory ones are pre-programmed and locked down by the vendor, usually containing an "electrical serial number". This patch will only add support for the user areas.

Lay the foundation and implement the MTD callbacks for the SPI-NOR and add necessary parameters to the flash_info structure. If a flash supports OTP it can be added by the convenience macro OTP_INFO(). Sometimes there are individual regions, which might have individual offsets. Therefore, it is possible to specify the starting address of the first regions as well as the distance between two regions (e.g. Winbond devices uses this method).

Additionally, the regions might be locked down. Once locked, no further write access is possible.

For SPI-NOR flashes the OTP area is accessed like the normal memory, e.g. by offset addressing; except that you either have to use special read/write commands (Winbond) or you have to enter (and exit) a specific OTP mode (Macronix, Micron).

Thus we introduce four operations to which the MTD callbacks will be
mapped: .read(), .write(), .lock() and .is_locked(). The read and the write ops will be given an address offset to operate on while the locking ops use regions because locking always affects a whole region. It is up to the flash driver to implement these ops.

---
 drivers/mtd/spi-nor/Makefile |   1 +
 drivers/mtd/spi-nor/core.c   |   8 ++
 drivers/mtd/spi-nor/core.h   |  50 +++++++++++
 drivers/mtd/spi-nor/otp.c    | 157 +++++++++++++++++++++++++++++++++++
 4 files changed, 216 insertions(+)
 create mode 100644 drivers/mtd/spi-nor/otp.c
drivers/mtd/spi-nor/Makefile +1 -0
index 653923896205..2ed2e76ce4f1 100644
@@ -12,6 +12,7 @@ spi-nor-objs			+= intel.o
 spi-nor-objs			+= issi.o
 spi-nor-objs			+= macronix.o
 spi-nor-objs			+= micron-st.o
+spi-nor-objs			+= otp.o
 spi-nor-objs			+= spansion.o
 spi-nor-objs			+= sst.o
 spi-nor-objs			+= winbond.o
drivers/mtd/spi-nor/core.c +8 -0
index 0522304f52fa..af9d7f194f01 100644
@@ -3012,6 +3012,12 @@ static void spi_nor_info_init_params(struct spi_nor *nor)
 	spi_nor_set_erase_type(&map->erase_type[i], info->sector_size,
 			       SPINOR_OP_SE);
 	spi_nor_init_uniform_erase_map(map, erase_mask, params->size);
+
+	/* OTP parameters */
+	nor->params->otp_info.otp_size = info->otp_size;
+	nor->params->otp_info.n_otps = info->n_otps;
+	nor->params->otp_info.otp_start_addr = info->otp_start_addr;
+	nor->params->otp_info.otp_addr_offset = info->otp_addr_offset;
 }
 
 /**
@@ -3502,6 +3508,8 @@ int spi_nor_scan(struct spi_nor *nor, const char *name,
 		mtd->_is_locked = spi_nor_is_locked;
 	}
 
+	spi_nor_otp_init(nor);
+
 	if (info->flags & USE_FSR)
 		nor->flags |= SNOR_F_USE_FSR;
 	if (info->flags & SPI_NOR_HAS_TB) {
drivers/mtd/spi-nor/core.h +50 -0

104 lines not on this page · raw message

drivers/mtd/spi-nor/otp.c new file +157 -0

157 lines not on this page · raw message

Signature · 6 lines
2.20.1


______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/
#2 lore raw

Re: [PATCH v3 1/2] mtd: spi-nor: add OTP support

Hi, Michael,

Quoted 28 lines · Michael Walle

EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe

SPI flashes sometimes have a special OTP area, which can (and is) used to store immutable properties like board serial number or vendor assigned network hardware addresses.

The MTD subsystem already supports accessing such areas and some (non SPI-NOR) flashes already implement support for it. It differentiates between user and factory areas. User areas can be written by the user and factory ones are pre-programmed and locked down by the vendor, usually containing an "electrical serial number". This patch will only add support for the user areas.

Lay the foundation and implement the MTD callbacks for the SPI-NOR and add necessary parameters to the flash_info structure. If a flash supports OTP it can be added by the convenience macro OTP_INFO(). Sometimes there are individual regions, which might have individual offsets. Therefore, it is possible to specify the starting address of the first regions as well as the distance between two regions (e.g. Winbond devices uses this method).

Additionally, the regions might be locked down. Once locked, no further write access is possible.

For SPI-NOR flashes the OTP area is accessed like the normal memory, e.g. by offset addressing; except that you either have to use special read/write commands (Winbond) or you have to enter (and exit) a specific OTP mode (Macronix, Micron).

Quoted 6 lines

Thus we introduce four operations to which the MTD callbacks will be
mapped: .read(), .write(), .lock() and .is_locked(). The read and the write ops will be given an address offset to operate on while the locking ops use regions because locking always affects a whole region. It is up to the flash driver to implement these ops.

SPI NORs can support some OTP-like behaviour, meaning that in addition to the tipical OTP ops (read, write, lock), the SPI NORs can also erase the OTP-like memory before the permanent lock. I find the erase useful, in case one writes something wrong from the start, in case of errors where what was written differs from what is read back, or simply at development stage, to check the implementation. So I think we should add support for that too. If not now, maybe later.

Michael, the overall implementation looks good, and I think we can have a version of it merged in the following merge window. Some suggestions and comments below.

cut

Quoted 4 lines
drivers/mtd/spi-nor/core.c +0 -0
index 0522304f52fa..af9d7f194f01 100644

cut

drivers/mtd/spi-nor/core.c int spi_nor_scan(struct spi_nor *nor, const char *name, · line 3511

Quoted 5 lines
drivers/mtd/spi-nor/core.c +1 -0
@@ -3502,6 +3508,8 @@ int spi_nor_scan(struct spi_nor *nor, const char *name,
                mtd->_is_locked = spi_nor_is_locked;
        }
 
+       spi_nor_otp_init(nor);

since this returns void, we can do it the last thing in the spi_nor_scan(), so that we don't gratuitously init fields in case of errors.

Quoted 16 lines
drivers/mtd/spi-nor/core.c +1 -0
+
        if (info->flags & USE_FSR)
                nor->flags |= SNOR_F_USE_FSR;
        if (info->flags & SPI_NOR_HAS_TB) {
drivers/mtd/spi-nor/core.h +4 -0
index 4a3f7f150b5d..5fb54ae08c5b 100644
@@ -175,6 +175,21 @@ struct spi_nor_erase_map {
        u8                              uniform_erase_type;
 };
 
+/**
+ * struct spi_nor_otp_info - Structure to describe the SPI NOR OTP region
+ * @otp_size:          size of one OTP region in bytes.
+ * @n_otps:            number of individual OTP regions.

drivers/mtd/spi-nor/core.h struct spi_nor_erase_map { · line 191

Quoted 10 lines
drivers/mtd/spi-nor/core.h +10 -0
+ * @otp_start_addr:    start address of the OTP area.
+ * @otp_addr_offset:   offset between consecutive OTP regions if there are
+ *                     more than one.
+ */
+struct spi_nor_otp_info {
+       u32 otp_size;
+       int n_otps;
+       u32 otp_start_addr;
+       u32 otp_addr_offset;
+};

How about the following:

struct spi_nor_otp_memory_organization {
	loff_t base;
	loff_t offset;
	size_t len;
	unsigned int n_regions;
};
Quoted 7 lines
drivers/mtd/spi-nor/core.h +1 -0
+
 /**
  * struct spi_nor_locking_ops - SPI NOR locking methods
  * @lock:      lock a region of the SPI NOR.
@@ -187,6 +202,20 @@ struct spi_nor_locking_ops {
        int (*is_locked)(struct spi_nor *nor, loff_t ofs, uint64_t len);
 };

drivers/mtd/spi-nor/core.h struct spi_nor_locking_ops { · line 213

Quoted 10 lines
drivers/mtd/spi-nor/core.h +9 -0
 
+/**
+ * struct spi_nor_otp_ops - SPI NOR OTP methods
+ * @read:      read from the SPI NOR OTP area.
+ * @write:     write to the SPI NOR OTP area.
+ * @lock:      lock an OTP region.
+ * @is_locked: check if an OTP region of the SPI NOR is locked.
+ */
+struct spi_nor_otp_ops {
+       int (*read)(struct spi_nor *nor, loff_t ofs, uint64_t len, u8 *buf);

int (*read)(struct spi_nor *nor, loff_t offset, size_t len, u8 *buf);

drivers/mtd/spi-nor/core.h struct spi_nor_locking_ops { · line 214

Quoted 1 line
drivers/mtd/spi-nor/core.h +1 -0
+       int (*write)(struct spi_nor *nor, loff_t ofs, uint64_t len, u8 *buf);

same here

Quoted 32 lines
drivers/mtd/spi-nor/core.h +7 -0
+       int (*lock)(struct spi_nor *nor, unsigned int region);
+       int (*is_locked)(struct spi_nor *nor, unsigned int region);
+};
+
 /**
  * struct spi_nor_flash_parameter - SPI NOR flash parameters and settings.
  * Includes legacy flash parameters and settings that can be overwritten
@@ -208,6 +237,7 @@ struct spi_nor_locking_ops {
  *                      higher index in the array, the higher priority.
  * @erase_map:         the erase map parsed from the SFDP Sector Map Parameter
  *                      Table.
+ * @otp_info:          describes the OTP regions.
  * @octal_dtr_enable:  enables SPI NOR octal DTR mode.
  * @quad_enable:       enables SPI NOR quad mode.
  * @set_4byte_addr_mode: puts the SPI NOR in 4 byte addressing mode.
@@ -219,6 +249,7 @@ struct spi_nor_locking_ops {
  *                      e.g. different opcodes, specific address calculation,
  *                      page size, etc.
  * @locking_ops:       SPI NOR locking methods.
+ * @otp_ops:           SPI NOR OTP methods.
  */
 struct spi_nor_flash_parameter {
        u64                             size;
@@ -232,6 +263,7 @@ struct spi_nor_flash_parameter {
        struct spi_nor_pp_command       page_programs[SNOR_CMD_PP_MAX];
 
        struct spi_nor_erase_map        erase_map;
+       struct spi_nor_otp_info         otp_info;
 
        int (*octal_dtr_enable)(struct spi_nor *nor, bool enable);
        int (*quad_enable)(struct spi_nor *nor);

drivers/mtd/spi-nor/core.h struct spi_nor_flash_parameter { · line 276

Quoted 5 lines
drivers/mtd/spi-nor/core.h +1 -0
@@ -240,6 +272,7 @@ struct spi_nor_flash_parameter {
        int (*setup)(struct spi_nor *nor, const struct spi_nor_hwcaps *hwcaps);
 
        const struct spi_nor_locking_ops *locking_ops;
+       const struct spi_nor_otp_ops *otp_ops;
 };

Let's group these altogether:

struct spi_nor_otp {
	struct spi_nor_otp_memory_organization memorg;
	const struct spi_nor_otp_ops *ops;
};

and then in:

struct spi_nor_flash_parameter {
	...
	struct spi_nor_erase_map        erase_map;
	struct spi_nor_otp		otp;
	...
}

So that we use:
nor->params->otp->memorg and nor->params->otp->ops

Quoted 5 lines
drivers/mtd/spi-nor/core.h +0 -0
 /**
@@ -341,6 +374,15 @@ struct flash_info {
 
        /* Part specific fixup hooks. */

drivers/mtd/spi-nor/core.h struct flash_info { · line 385

Quoted 10 lines
drivers/mtd/spi-nor/core.h +9 -0
        const struct spi_nor_fixups *fixups;
+
+       /* OTP size in bytes */
+       u16 otp_size;
+       /* Number of OTP banks */
+       u16 n_otps;
+       /* Start address of OTP area */
+       u32 otp_start_addr;
+       /* Offset between consecutive OTP banks if there are more than one */
+       u32 otp_addr_offset;

Let's use the structure that we have already defined:
struct spi_nor_otp_memory_organization otp_memorg;

cut

Quoted 9 lines
drivers/mtd/spi-nor/otp.c new file +3 -0
new file mode 100644
index 000000000000..59bd1a3f450d
@@ -0,0 +1,157 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * OTP support for SPI-NOR flashes

drivers/mtd/spi-nor/otp.c line 13

Quoted 10 lines
drivers/mtd/spi-nor/otp.c new file +10 -0
+ *
+ * Copyright (C) 2021 Michael Walle <[email protected]>
+ */
+
+#include <linux/mtd/mtd.h>
+#include <linux/mtd/spi-nor.h>
+
+#include "core.h"
+
+static loff_t spi_nor_otp_region_start(struct spi_nor *nor, int region)

const struct spi_nor *nor, unsigned int region

drivers/mtd/spi-nor/otp.c line 20

Quoted 7 lines
drivers/mtd/spi-nor/otp.c new file +7 -0
+{
+       struct spi_nor_otp_info *info = &nor->params->otp_info;
+
+       return info->otp_start_addr + region * info->otp_addr_offset;
+}
+
+static loff_t spi_nor_otp_region_end(struct spi_nor *nor, int region)

same

drivers/mtd/spi-nor/otp.c line 30

Quoted 10 lines
drivers/mtd/spi-nor/otp.c new file +10 -0
+{
+       struct spi_nor_otp_info *info = &nor->params->otp_info;
+
+       return spi_nor_otp_region_start(nor, region) + info->otp_size - 1;
+}
+
+static int spi_nor_mtd_otp_info(struct mtd_info *mtd, size_t len,
+                               size_t *retlen, struct otp_info *buf)
+{
+       struct spi_nor *nor = mtd_to_spi_nor(mtd);

const struct spi_nor *nor

drivers/mtd/spi-nor/otp.c line 32

Quoted 2 lines
drivers/mtd/spi-nor/otp.c new file +2 -0
+       int n_otps = nor->params->otp_info.n_otps;
+       int locked, i;

unsigned int i;

Quoted 13 lines
drivers/mtd/spi-nor/otp.c new file +13 -0
+
+       if (len < n_otps * sizeof(*buf))
+               return -ENOSPC;
+
+       for (i = 0; i < n_otps; i++) {
+               buf[i].start = spi_nor_otp_region_start(nor, i);
+               buf[i].length = nor->params->otp_info.otp_size;
+
+               locked = nor->params->otp_ops->is_locked(nor, i);
+               if (locked < 0)
+                       return locked;
+
+               buf[i].locked = !!locked;

drivers/mtd/spi-nor/otp.c line 55

Quoted 10 lines
drivers/mtd/spi-nor/otp.c new file +10 -0
+       }
+
+       *retlen = n_otps * sizeof(*buf);
+
+       return 0;
+}
+
+static int spi_nor_otp_addr_to_region(struct spi_nor *nor, loff_t addr)
+{
+       int i;

unsigned int i;

Quoted 4 lines
drivers/mtd/spi-nor/otp.c new file +4 -0
+
+       for (i = 0; i < nor->params->otp_info.n_otps; i++)
+               if (addr >= spi_nor_otp_region_start(nor, i) &&
+                   addr <= spi_nor_otp_region_end(nor, i))

drivers/mtd/spi-nor/otp.c line 69

Quoted 10 lines
drivers/mtd/spi-nor/otp.c new file +10 -0
+                       return i;
+
+       return -EINVAL;
+}
+
+static int spi_nor_mtd_otp_read_write(struct mtd_info *mtd, loff_t ofs,
+                                     size_t len, size_t *retlen, u_char *buf,
+                                     bool is_write)
+{
+       struct spi_nor *nor = mtd_to_spi_nor(mtd);

const

Quoted 6 lines
drivers/mtd/spi-nor/otp.c new file +6 -0
+       int region;
+       int ret;
+
+       *retlen = 0;
+
+       region = spi_nor_otp_addr_to_region(nor, ofs);

drivers/mtd/spi-nor/otp.c line 85

Quoted 10 lines
drivers/mtd/spi-nor/otp.c new file +10 -0
+       if (region < 0)
+               return 0;
+
+       if (ofs < spi_nor_otp_region_start(nor, region))
+               return 0;
+
+       if ((ofs + len - 1) > spi_nor_otp_region_end(nor, region))
+               return 0;
+
+       ret = spi_nor_lock_and_prep(nor);

please check the ret value

Quoted 20 lines
drivers/mtd/spi-nor/otp.c new file +20 -0
+
+       if (is_write)
+               ret = nor->params->otp_ops->write(nor, ofs, len, buf);
+       else
+               ret = nor->params->otp_ops->read(nor, ofs, len, buf);
+
+       spi_nor_unlock_and_unprep(nor);
+
+       if (ret < 0)
+               return ret;
+
+       *retlen = len;
+       return 0;
+}
+
+static int spi_nor_mtd_otp_read(struct mtd_info *mtd, loff_t from, size_t len,
+                               size_t *retlen, u_char *buf)
+{
+       return spi_nor_mtd_otp_read_write(mtd, from, len, retlen, buf, false);
+}

drivers/mtd/spi-nor/otp.c line 115

Quoted 10 lines
drivers/mtd/spi-nor/otp.c new file +10 -0
+
+static int spi_nor_mtd_otp_write(struct mtd_info *mtd, loff_t to, size_t len,
+                                size_t *retlen, u_char *buf)
+{
+       return spi_nor_mtd_otp_read_write(mtd, to, len, retlen, buf, true);
+}
+
+static int spi_nor_mtd_otp_lock(struct mtd_info *mtd, loff_t from, size_t len)
+{
+       struct spi_nor *nor = mtd_to_spi_nor(mtd);

const

drivers/mtd/spi-nor/otp.c line 124

Quoted 9 lines
drivers/mtd/spi-nor/otp.c new file +9 -0
+       int region;
+       int ret;
+
+       region = spi_nor_otp_addr_to_region(nor, from);
+       if (region < 0)
+               return -EINVAL;
+
+       if (len != nor->params->otp_info.otp_size)
+               return -EINVAL;

Does the otp memory organization matter for the end user?
Can't we lock/read/write past region size, for example 2 or 3 regions in a row, depending on length?

Cheers,
ta

Quoted 34 lines
drivers/mtd/spi-nor/otp.c new file +33 -1
+
+       ret = spi_nor_lock_and_prep(nor);
+       if (ret)
+               return ret;
+
+       ret = nor->params->otp_ops->lock(nor, region);
+
+       spi_nor_unlock_and_unprep(nor);
+
+       return ret;
+}
+
+void spi_nor_otp_init(struct spi_nor *nor)
+{
+       struct mtd_info *mtd = &nor->mtd;
+
+       if (!nor->params->otp_ops)
+               return;
+
+       /*
+        * We only support user_prot callbacks (yet).
+        *
+        * Some SPI-NOR flashes like Macronix ones can be ordered in two
+        * different variants. One with a factory locked OTP area and one where
+        * it is left to the user to write to it. The factory locked OTP is
+        * usually preprogrammed with an "electrical serial number". We don't
+        * support these for now.
+        */
+       mtd->_get_user_prot_info = spi_nor_mtd_otp_info;
+       mtd->_read_user_prot_reg = spi_nor_mtd_otp_read;
+       mtd->_write_user_prot_reg = spi_nor_mtd_otp_write;
+       mtd->_lock_user_prot_reg = spi_nor_mtd_otp_lock;
+}
--
Quoted 2 lines

2.20.1

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

#3 lore raw

Re: [PATCH v3 1/2] mtd: spi-nor: add OTP support

Hi Tudor,

thanks for the review.

Am 2021-02-28 13:00, schrieb [email protected]:

Quoted 36 lines · Michael Walle
Quoted 28 lines

EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe

SPI flashes sometimes have a special OTP area, which can (and is) used to store immutable properties like board serial number or vendor assigned network hardware addresses.

The MTD subsystem already supports accessing such areas and some (non SPI-NOR) flashes already implement support for it. It differentiates between user and factory areas. User areas can be written by the user and factory ones are pre-programmed and locked down by the vendor, usually containing an "electrical serial number". This patch will only add support for the user areas.

Lay the foundation and implement the MTD callbacks for the SPI-NOR and add necessary parameters to the flash_info structure. If a flash supports OTP it can be added by the convenience macro OTP_INFO(). Sometimes there are individual regions, which might have individual offsets. Therefore, it is possible to specify the starting address of the first regions as well as the distance between two regions (e.g. Winbond devices uses this method).

Additionally, the regions might be locked down. Once locked, no further write access is possible.

For SPI-NOR flashes the OTP area is accessed like the normal memory, e.g. by offset addressing; except that you either have to use special read/write commands (Winbond) or you have to enter (and exit) a specific OTP mode (Macronix, Micron).

Quoted 6 lines

Thus we introduce four operations to which the MTD callbacks will be
mapped: .read(), .write(), .lock() and .is_locked(). The read and the write ops will be given an address offset to operate on while the locking ops use regions because locking always affects a whole region. It is up to the flash driver to implement these ops.

Quoted 7 lines

SPI NORs can support some OTP-like behaviour, meaning that in addition to the tipical OTP ops (read, write, lock), the SPI NORs can also erase the OTP-like memory before the permanent lock. I find the erase useful, in case one writes something wrong from the start, in case of errors where what was written differs from what is read back, or simply at development stage, to check the implementation. So I think we should add support for that too. If not now, maybe later.

As mentioned on IRC, this should be a seperate ioctl() thus the mtd-utils need to be changed, too. I wouldn't like to change the semantics of the current (more or less well-established) ioctl interface around MTD otp operations.

Thus maybe there might be OTPERASE ioctl, similar to OTPLOCK. I'm open to this, but in a seperate patch series.

Quoted 11 lines
Quoted 4 lines
drivers/mtd/spi-nor/core.c +0 -0
index 0522304f52fa..af9d7f194f01 100644

cut

Quoted 5 lines
drivers/mtd/spi-nor/core.c +0 -0
@@ -3502,6 +3508,8 @@ int spi_nor_scan(struct spi_nor *nor, const char 
 *name,
                mtd->_is_locked = spi_nor_is_locked;
        }
 
Quoted 4 lines

drivers/mtd/spi-nor/core.c int spi_nor_scan(struct spi_nor *nor, const char *name, · line 3511

Quoted 1 line
drivers/mtd/spi-nor/core.c +1 -0
+       spi_nor_otp_init(nor);

since this returns void, we can do it the last thing in the spi_nor_scan(), so that we don't gratuitously init fields in case of errors.

Mh, it used to be at this point because it sets the mtd ops. In my previous patch version this was still inline. So sure I can move it, but I'd move it right before spi_nor_init() because that will actually interface with the SPI flash.

Quoted 31 lines
Quoted 22 lines
drivers/mtd/spi-nor/core.c +1 -0
+
        if (info->flags & USE_FSR)
                nor->flags |= SNOR_F_USE_FSR;
        if (info->flags & SPI_NOR_HAS_TB) {
drivers/mtd/spi-nor/core.h +2 -0
index 4a3f7f150b5d..5fb54ae08c5b 100644
@@ -175,6 +175,21 @@ struct spi_nor_erase_map {
        u8                              uniform_erase_type;
 };
 
+/**
+ * struct spi_nor_otp_info - Structure to describe the SPI NOR OTP 

region

+ * @otp_size:          size of one OTP region in bytes.
+ * @n_otps:            number of individual OTP regions.
+ * @otp_start_addr:    start address of the OTP area.
+ * @otp_addr_offset:   offset between consecutive OTP regions if 

there are
+ * more than one.
+ */

drivers/mtd/spi-nor/core.h struct spi_nor_erase_map { · line 191

Quoted 6 lines
drivers/mtd/spi-nor/core.h +6 -0
+struct spi_nor_otp_info {
+       u32 otp_size;
+       int n_otps;
+       u32 otp_start_addr;
+       u32 otp_addr_offset;
+};

How about the following:

Quoted 1 line

struct spi_nor_otp_memory_organization {

I've used spi_nor_otp_organization, first because its a shorter name and second it might or might not have something to do with memory.

Quoted 5 lines
	loff_t base;
	loff_t offset;
	size_t len;
	unsigned int n_regions;
};

otherwise I've put the len argument as the first property, because other flashes usually contain just one region, thus base and offset are normally unused (or zero).

[..]

Quoted 17 lines
Quoted 12 lines
drivers/mtd/spi-nor/core.h +8 -0
@@ -341,6 +374,15 @@ struct flash_info {
 
        /* Part specific fixup hooks. */
        const struct spi_nor_fixups *fixups;
+
+       /* OTP size in bytes */
+       u16 otp_size;
+       /* Number of OTP banks */
+       u16 n_otps;
+       /* Start address of OTP area */
+       u32 otp_start_addr;
+       /* Offset between consecutive OTP banks if there are more than 
Quoted 2 lines

one */
+ u32 otp_addr_offset;

Let's use the structure that we have already defined:
struct spi_nor_otp_memory_organization otp_memorg;

Ok, this was due to the convenience macro OTP_INFO(). Which now looks like:

#define OTP_INFO(_len, _n_regions, _base, _offset)			\
		.otp_org = (struct spi_nor_otp_organization) {		\
			.len = (_len),				\
			.base = (_base),				\
			.offset = (_offset),				\
			.n_regions = (_n_regions),			\
		},

btw there is also a shorter OTP_INFO1(_len) which I used for the macronix flashes (which for now was dropped in this series).

Quoted 33 lines
Quoted 19 lines
drivers/mtd/spi-nor/otp.c new file +13 -0
new file mode 100644
index 000000000000..59bd1a3f450d
@@ -0,0 +1,157 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * OTP support for SPI-NOR flashes
+ *
+ * Copyright (C) 2021 Michael Walle <[email protected]>
+ */
+
+#include <linux/mtd/mtd.h>
+#include <linux/mtd/spi-nor.h>
+
+#include "core.h"
+
+static loff_t spi_nor_otp_region_start(struct spi_nor *nor, int 
Quoted 1 line

region)

const struct spi_nor *nor, unsigned int region

drivers/mtd/spi-nor/otp.c line 20

Quoted 7 lines
drivers/mtd/spi-nor/otp.c new file +7 -0
+{
+       struct spi_nor_otp_info *info = &nor->params->otp_info;
+
+       return info->otp_start_addr + region * info->otp_addr_offset;
+}
+
+static loff_t spi_nor_otp_region_end(struct spi_nor *nor, int region)

same

Quoted 13 lines

drivers/mtd/spi-nor/otp.c line 30

Quoted 11 lines
drivers/mtd/spi-nor/otp.c new file +10 -1
+{
+       struct spi_nor_otp_info *info = &nor->params->otp_info;
+
+       return spi_nor_otp_region_start(nor, region) + info->otp_size 
- 1;
+}
+
+static int spi_nor_mtd_otp_info(struct mtd_info *mtd, size_t len,
+                               size_t *retlen, struct otp_info *buf)
+{
+       struct spi_nor *nor = mtd_to_spi_nor(mtd);

const struct spi_nor *nor

Isn't possible, because is_locked() is called which in turn calls spi_nor_read_xx() which takes "struct spi_nor".

Quoted 42 lines

drivers/mtd/spi-nor/otp.c line 32

Quoted 2 lines
drivers/mtd/spi-nor/otp.c new file +2 -0
+       int n_otps = nor->params->otp_info.n_otps;
+       int locked, i;

unsigned int i;

Quoted 22 lines
drivers/mtd/spi-nor/otp.c new file +21 -0
+
+       if (len < n_otps * sizeof(*buf))
+               return -ENOSPC;
+
+       for (i = 0; i < n_otps; i++) {
+               buf[i].start = spi_nor_otp_region_start(nor, i);
+               buf[i].length = nor->params->otp_info.otp_size;
+
+               locked = nor->params->otp_ops->is_locked(nor, i);
+               if (locked < 0)
+                       return locked;
+
+               buf[i].locked = !!locked;
+       }
+
+       *retlen = n_otps * sizeof(*buf);
+
+       return 0;
+}
+
+static int spi_nor_otp_addr_to_region(struct spi_nor *nor, loff_t 

addr)

drivers/mtd/spi-nor/otp.c

Quoted 2 lines
drivers/mtd/spi-nor/otp.c new file +2 -0
+{
+       int i;

unsigned int i;

Quoted 10 lines
drivers/mtd/spi-nor/otp.c new file +10 -0
+
+       for (i = 0; i < nor->params->otp_info.n_otps; i++)
+               if (addr >= spi_nor_otp_region_start(nor, i) &&
+                   addr <= spi_nor_otp_region_end(nor, i))
+                       return i;
+
+       return -EINVAL;
+}
+
+static int spi_nor_mtd_otp_read_write(struct mtd_info *mtd, loff_t 
Quoted 8 lines

drivers/mtd/spi-nor/otp.c line 69

Quoted 6 lines

ofs,
+ size_t len, size_t *retlen,
u_char *buf,

drivers/mtd/spi-nor/otp.c new file +3 -0
+                                     bool is_write)
+{
+       struct spi_nor *nor = mtd_to_spi_nor(mtd);

const

likewise

Quoted 11 lines
Quoted 11 lines
drivers/mtd/spi-nor/otp.c new file +10 -0
+       int region;
+       int ret;
+
+       *retlen = 0;
+
+       region = spi_nor_otp_addr_to_region(nor, ofs);
+       if (region < 0)
+               return 0;
+
+       if (ofs < spi_nor_otp_region_start(nor, region))
Quoted 8 lines

drivers/mtd/spi-nor/otp.c line 85

Quoted 6 lines
drivers/mtd/spi-nor/otp.c new file +6 -0
+               return 0;
+
+       if ((ofs + len - 1) > spi_nor_otp_region_end(nor, region))
+               return 0;
+
+       ret = spi_nor_lock_and_prep(nor);

please check the ret value

good catch.

Quoted 30 lines
Quoted 30 lines
drivers/mtd/spi-nor/otp.c new file +16 -0
+
+       if (is_write)
+               ret = nor->params->otp_ops->write(nor, ofs, len, buf);
+       else
+               ret = nor->params->otp_ops->read(nor, ofs, len, buf);
+
+       spi_nor_unlock_and_unprep(nor);
+
+       if (ret < 0)
+               return ret;
+
+       *retlen = len;
+       return 0;
+}
+
+static int spi_nor_mtd_otp_read(struct mtd_info *mtd, loff_t from, 

size_t len,

drivers/mtd/spi-nor/otp.c new file +3 -0
+                               size_t *retlen, u_char *buf)
+{
+       return spi_nor_mtd_otp_read_write(mtd, from, len, retlen, buf, 

false);

drivers/mtd/spi-nor/otp.c new file +3 -0
+}
+
+static int spi_nor_mtd_otp_write(struct mtd_info *mtd, loff_t to, 

size_t len,

drivers/mtd/spi-nor/otp.c new file +3 -0
+                                size_t *retlen, u_char *buf)
+{
+       return spi_nor_mtd_otp_read_write(mtd, to, len, retlen, buf, 

true);

Quoted 8 lines

drivers/mtd/spi-nor/otp.c

Quoted 6 lines
drivers/mtd/spi-nor/otp.c new file +3 -0
+}
+
+static int spi_nor_mtd_otp_lock(struct mtd_info *mtd, loff_t from, 

size_t len)

drivers/mtd/spi-nor/otp.c new file +2 -0
+{
+       struct spi_nor *nor = mtd_to_spi_nor(mtd);

const

likewise

Quoted 14 lines

drivers/mtd/spi-nor/otp.c line 124

Quoted 10 lines
drivers/mtd/spi-nor/otp.c new file +9 -0
+       int region;
+       int ret;
+
+       region = spi_nor_otp_addr_to_region(nor, from);
+       if (region < 0)
+               return -EINVAL;
+
+       if (len != nor->params->otp_info.otp_size)
+               return -EINVAL;

Does the otp memory organization matter for the end user?
Can't we lock/read/write past region size, for example 2 or 3 regions in a row, depending on length?

Mhh tough one. I guess the question really is: Do we want to remap the 0x1000, 0x2000, 0x3000 offsets?
- 0x1000 -> 0
- 0x2000 -> 1 * region_size
- 0x3000 -> 2 * region_size

This is just an example, some devices may us other offsets.

I'd see this as a prerequsite for handling multiple regions in one write, because otherwise you'll have to handle the holes which makes it impossible I guess. For example what would happen with (given an otp size of 0x100):
(1) lock(0, 0x100)
(2) lock(0x100, 0xf00)
(3) lock(0, 0x1000)

(1) will work, (2) should return -EINVAL; but what will (3) return. -EINVAL too, I guess. But then, ops spanning multiple regions doesn't make sense at all, because they will always return -EINVAL.

Unfortunately, I don't know how userspace might access it.

This is how it looks like at the moment:

# flash_otp_info -u /dev/mtd1
Number of OTP user blocks on /dev/mtd1: 3
block  0:  offset = 0x1000  size = 256 bytes  [unlocked]
block  1:  offset = 0x2000  size = 256 bytes  [unlocked]
block  2:  offset = 0x3000  size = 256 bytes  [unlocked]

-michael

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

#4 lore raw

Re: [PATCH v3 1/2] mtd: spi-nor: add OTP support

Am 2021-03-01 18:32, schrieb Michael Walle:

Quoted 29 lines

Am 2021-02-28 13:00, schrieb [email protected]:

Quoted 4 lines · Michael Walle

Does the otp memory organization matter for the end user?
Can't we lock/read/write past region size, for example 2 or 3 regions in a row, depending on length?

Mhh tough one. I guess the question really is: Do we want to remap the 0x1000, 0x2000, 0x3000 offsets?
- 0x1000 -> 0
- 0x2000 -> 1 * region_size
- 0x3000 -> 2 * region_size

This is just an example, some devices may us other offsets.

I'd see this as a prerequsite for handling multiple regions in one write, because otherwise you'll have to handle the holes which makes it impossible I guess. For example what would happen with (given an otp size of 0x100):
(1) lock(0, 0x100)
(2) lock(0x100, 0xf00)
(3) lock(0, 0x1000)

(1) will work, (2) should return -EINVAL; but what will (3) return. -EINVAL too, I guess. But then, ops spanning multiple regions doesn't make sense at all, because they will always return -EINVAL.

Unfortunately, I don't know how userspace might access it.

Quoted 7 lines

This is how it looks like at the moment:

# flash_otp_info -u /dev/mtd1
Number of OTP user blocks on /dev/mtd1: 3
block  0:  offset = 0x1000  size = 256 bytes  [unlocked]
block  1:  offset = 0x2000  size = 256 bytes  [unlocked]
block  2:  offset = 0x3000  size = 256 bytes  [unlocked]

Hm, I just found the following in the mtd-utils [1]:
"offset and size must match on OTP region boundaries".

[1] http://git.infradead.org/mtd-utils.git/blob/HEAD:/misc-utils/flash_otp_lock.c#l25

-michael

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

#5 lore raw

PATCH v3 2/2 mtd: spi-nor: implement OTP support for Winbond and similar flashes

In reply to Michael Walle

Use the new OTP ops to implement OTP access on Winbond flashes. Most Winbond flashes provides up to four different OTP regions ("Security Registers").

Winbond devices use a special opcode to read and write to the OTP regions, just like the RDSFDP opcode. In fact, it seems that the (undocumented) first OTP area of the newer flashes is the actual SFDP table.

On a side note, Winbond devices also allow erasing the OTP regions as long as the area isn't locked down.

This was tested on a Winbond W25Q32JW as well as on a W25Q32FW.

---
 drivers/mtd/spi-nor/core.c    |   2 +-
 drivers/mtd/spi-nor/core.h    |   6 ++
 drivers/mtd/spi-nor/otp.c     | 163 ++++++++++++++++++++++++++++++++++
 drivers/mtd/spi-nor/winbond.c |  17 +++-
 include/linux/mtd/spi-nor.h   |   9 ++
 5 files changed, 194 insertions(+), 3 deletions(-)
drivers/mtd/spi-nor/core.c +1 -1
index af9d7f194f01..1115046e2680 100644
@@ -1034,7 +1034,7 @@ static int spi_nor_write_16bit_sr_and_check(struct spi_nor *nor, u8 sr1)
  *
  * Return: 0 on success, -errno otherwise.
  */
-static int spi_nor_write_16bit_cr_and_check(struct spi_nor *nor, u8 cr)
+int spi_nor_write_16bit_cr_and_check(struct spi_nor *nor, u8 cr)
 {
 	int ret;
 	u8 *sr_cr = nor->bouncebuf;
drivers/mtd/spi-nor/core.h +6 -0
index 5fb54ae08c5b..d755f2bf77e0 100644
@@ -492,6 +492,7 @@ int spi_nor_read_sr(struct spi_nor *nor, u8 *sr);
 int spi_nor_read_cr(struct spi_nor *nor, u8 *cr);
 int spi_nor_write_sr(struct spi_nor *nor, const u8 *sr, size_t len);
 int spi_nor_write_sr_and_check(struct spi_nor *nor, u8 sr1);
+int spi_nor_write_16bit_cr_and_check(struct spi_nor *nor, u8 cr);
 
 int spi_nor_xread_sr(struct spi_nor *nor, u8 *sr);
 ssize_t spi_nor_read_data(struct spi_nor *nor, loff_t from, size_t len,
@@ -499,6 +500,11 @@ ssize_t spi_nor_read_data(struct spi_nor *nor, loff_t from, size_t len,
 ssize_t spi_nor_write_data(struct spi_nor *nor, loff_t to, size_t len,
 			   const u8 *buf);
 
+int spi_nor_otp_read_secr(struct spi_nor *nor, loff_t addr, uint64_t len, u8 *buf);
+int spi_nor_otp_write_secr(struct spi_nor *nor, loff_t addr, uint64_t len, u8 *buf);
+int spi_nor_otp_lock_sr2(struct spi_nor *nor, unsigned int region);
+int spi_nor_otp_is_locked_sr2(struct spi_nor *nor, unsigned int region);
+
 int spi_nor_hwcaps_read2cmd(u32 hwcaps);
 u8 spi_nor_convert_3to4_read(u8 opcode);
 void spi_nor_set_read_settings(struct spi_nor_read_command *read,
drivers/mtd/spi-nor/otp.c +163 -0

169 lines not on this page · raw message

drivers/mtd/spi-nor/winbond.c +15 -2
index e5dfa786f190..d3bbd2df7500 100644
@@ -55,14 +55,18 @@ static const struct flash_info winbond_parts[] = {
 	{ "w25q32", INFO(0xef4016, 0, 64 * 1024,  64, SECT_4K) },
 	{ "w25q32dw", INFO(0xef6016, 0, 64 * 1024,  64,
 			   SECT_4K | SPI_NOR_DUAL_READ | SPI_NOR_QUAD_READ |
-			   SPI_NOR_HAS_LOCK | SPI_NOR_HAS_TB) },
+			   SPI_NOR_HAS_LOCK | SPI_NOR_HAS_TB)
+			   OTP_INFO(256, 3, 0x1000, 0x1000)
+	},
+
 	{ "w25q32jv", INFO(0xef7016, 0, 64 * 1024,  64,
 			   SECT_4K | SPI_NOR_DUAL_READ | SPI_NOR_QUAD_READ |
 			   SPI_NOR_HAS_LOCK | SPI_NOR_HAS_TB)
 	},
 	{ "w25q32jwm", INFO(0xef8016, 0, 64 * 1024,  64,
 			    SECT_4K | SPI_NOR_DUAL_READ | SPI_NOR_QUAD_READ |
-			    SPI_NOR_HAS_LOCK | SPI_NOR_HAS_TB) },
+			    SPI_NOR_HAS_LOCK | SPI_NOR_HAS_TB)
+			    OTP_INFO(256, 3, 0x1000, 0x1000) },
 	{ "w25q64jwm", INFO(0xef8017, 0, 64 * 1024, 128,
 			    SECT_4K | SPI_NOR_DUAL_READ | SPI_NOR_QUAD_READ |
 			    SPI_NOR_HAS_LOCK | SPI_NOR_HAS_TB) },
@@ -131,9 +135,18 @@ static int winbond_set_4byte_addr_mode(struct spi_nor *nor, bool enable)
 	return spi_nor_write_disable(nor);
 }
 
+static const struct spi_nor_otp_ops winbond_otp_ops = {
+	.read = spi_nor_otp_read_secr,
+	.write = spi_nor_otp_write_secr,
+	.lock = spi_nor_otp_lock_sr2,
+	.is_locked = spi_nor_otp_is_locked_sr2,
+};
+
 static void winbond_default_init(struct spi_nor *nor)
 {
 	nor->params->set_4byte_addr_mode = winbond_set_4byte_addr_mode;
+	if (nor->params->otp_info.n_otps)
+		nor->params->otp_ops = &winbond_otp_ops;
 }
 
 static const struct spi_nor_fixups winbond_fixups = {
include/linux/mtd/spi-nor.h +9 -0
index a0d572855444..6d1956049e90 100644
@@ -107,6 +107,11 @@
 #define SPINOR_OP_RD_EVCR      0x65    /* Read EVCR register */
 #define SPINOR_OP_WD_EVCR      0x61    /* Write EVCR register */
 
+/* Used for GigaDevices and Winbond flashes. */
+#define SPINOR_OP_ESECR		0x44	/* Erase Security registers */
+#define SPINOR_OP_PSECR		0x42	/* Program Security registers */
+#define SPINOR_OP_RSECR		0x48	/* Read Security registers */
+
 /* Status Register bits. */
 #define SR_WIP			BIT(0)	/* Write in progress */
 #define SR_WEL			BIT(1)	/* Write enable latch */
@@ -138,8 +143,12 @@
 
 /* Status Register 2 bits. */
 #define SR2_QUAD_EN_BIT1	BIT(1)
+#define SR2_LB1			BIT(3)	/* Security Register Lock Bit 1 */
+#define SR2_LB2			BIT(4)	/* Security Register Lock Bit 2 */
+#define SR2_LB3			BIT(5)	/* Security Register Lock Bit 3 */
 #define SR2_QUAD_EN_BIT7	BIT(7)
 
+
 /* Supported SPI protocols */
 #define SNOR_PROTO_INST_MASK	GENMASK(23, 16)
 #define SNOR_PROTO_INST_SHIFT	16
Signature · 6 lines
2.20.1


______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/