| Message ID | 20240804221535.291923-1-mwalle@kernel.org |
|---|---|
| State | Accepted |
| Headers | show |
| Series | mtd: spi-nor: winbond: add Zetta ZD25Q128C support | expand |
On Mon, Aug 05 2024, Michael Walle wrote: > Zetta normally uses BAh as its vendor ID. But for the ZD25Q128C they > took the one from Winbond and messed up the size parameters in SFDP. > Most functions seem compatible with the W25Q128, we just have to fix up > the size. > > Link: http://www.zettadevice.com/upload/file/20150821/DS_Zetta_25Q128_RevA.pdf > Link: https://www.lcsc.com/datasheet/lcsc_datasheet_2312081757_Zetta-ZD25Q128CSIGT_C19626875.pdf > Signed-off-by: Michael Walle <mwalle@kernel.org> > --- [...] > diff --git a/drivers/mtd/spi-nor/winbond.c b/drivers/mtd/spi-nor/winbond.c > index e065e4fd42a3..9f7ce5763e71 100644 > --- a/drivers/mtd/spi-nor/winbond.c > +++ b/drivers/mtd/spi-nor/winbond.c > @@ -17,6 +17,31 @@ > SPI_MEM_OP_NO_DUMMY, \ > SPI_MEM_OP_DATA_OUT(1, buf, 0)) > > +static int > +w25q128_post_bfpt_fixups(struct spi_nor *nor, > + const struct sfdp_parameter_header *bfpt_header, > + const struct sfdp_bfpt *bfpt) > +{ > + /* > + * Zetta ZD25Q128C is a clone of the Winbond device. But the encoded > + * size is really wrong. It seems that they confused Mbit with MiB. > + * Thus the flash is discovered as a 2MiB device. > + */ > + if (bfpt_header->major == SFDP_JESD216_MAJOR && > + bfpt_header->minor == SFDP_JESD216_MINOR && > + nor->params->size == SZ_2M && > + nor->params->erase_map.regions[0].size == SZ_2M) { > + nor->params->size = SZ_16M; > + nor->params->erase_map.regions[0].size = SZ_16M; > + } Since the size is 16 MiB for both Zetta and Winbond, why do you have these conditions here? Why not just do it unconditionally? What situation do you want to protect against? > + > + return 0; > +} > + > +static const struct spi_nor_fixups w25q128_fixups = { > + .post_bfpt = w25q128_post_bfpt_fixups, > +}; > + > static int > w25q256_post_bfpt_fixups(struct spi_nor *nor, > const struct sfdp_parameter_header *bfpt_header, > @@ -108,6 +133,7 @@ static const struct flash_info winbond_nor_parts[] = { > .size = SZ_16M, > .flags = SPI_NOR_HAS_LOCK | SPI_NOR_HAS_TB, > .no_sfdp_flags = SECT_4K | SPI_NOR_DUAL_READ | SPI_NOR_QUAD_READ, > + .fixups = &w25q128_fixups, > }, { > .id = SNOR_ID(0xef, 0x40, 0x19), > .name = "w25q256",
Hi, > > Zetta normally uses BAh as its vendor ID. But for the ZD25Q128C they > > took the one from Winbond and messed up the size parameters in SFDP. > > Most functions seem compatible with the W25Q128, we just have to fix up > > the size. > > > > Link: http://www.zettadevice.com/upload/file/20150821/DS_Zetta_25Q128_RevA.pdf > > Link: https://www.lcsc.com/datasheet/lcsc_datasheet_2312081757_Zetta-ZD25Q128CSIGT_C19626875.pdf > > Signed-off-by: Michael Walle <mwalle@kernel.org> > > --- > [...] > > diff --git a/drivers/mtd/spi-nor/winbond.c b/drivers/mtd/spi-nor/winbond.c > > index e065e4fd42a3..9f7ce5763e71 100644 > > --- a/drivers/mtd/spi-nor/winbond.c > > +++ b/drivers/mtd/spi-nor/winbond.c > > @@ -17,6 +17,31 @@ > > SPI_MEM_OP_NO_DUMMY, \ > > SPI_MEM_OP_DATA_OUT(1, buf, 0)) > > > > +static int > > +w25q128_post_bfpt_fixups(struct spi_nor *nor, > > + const struct sfdp_parameter_header *bfpt_header, > > + const struct sfdp_bfpt *bfpt) > > +{ > > + /* > > + * Zetta ZD25Q128C is a clone of the Winbond device. But the encoded > > + * size is really wrong. It seems that they confused Mbit with MiB. > > + * Thus the flash is discovered as a 2MiB device. > > + */ > > + if (bfpt_header->major == SFDP_JESD216_MAJOR && > > + bfpt_header->minor == SFDP_JESD216_MINOR && > > + nor->params->size == SZ_2M && > > + nor->params->erase_map.regions[0].size == SZ_2M) { > > + nor->params->size = SZ_16M; > > + nor->params->erase_map.regions[0].size = SZ_16M; > > + } > > Since the size is 16 MiB for both Zetta and Winbond, why do you have > these conditions here? Why not just do it unconditionally? What > situation do you want to protect against? Two things, I want to make sure, the values I'll overwrite (which were set indirectly by the SFDP data I check here) are really the "before" value I expect them to be. And secondly, I want to fingerprint the SFDP as accurately as possible. I mean you wouldn't need the SFDP version either. Think of a preparation for a possible .match() callback in zetta.c. That would probably look like this, i.e. check the flash id and these four conditions. What are your concerns? -michael
On Mon, Aug 12 2024, Michael Walle wrote: > Hi, > >> > Zetta normally uses BAh as its vendor ID. But for the ZD25Q128C they >> > took the one from Winbond and messed up the size parameters in SFDP. >> > Most functions seem compatible with the W25Q128, we just have to fix up >> > the size. >> > >> > Link: http://www.zettadevice.com/upload/file/20150821/DS_Zetta_25Q128_RevA.pdf >> > Link: https://www.lcsc.com/datasheet/lcsc_datasheet_2312081757_Zetta-ZD25Q128CSIGT_C19626875.pdf >> > Signed-off-by: Michael Walle <mwalle@kernel.org> >> > --- >> [...] >> > diff --git a/drivers/mtd/spi-nor/winbond.c b/drivers/mtd/spi-nor/winbond.c >> > index e065e4fd42a3..9f7ce5763e71 100644 >> > --- a/drivers/mtd/spi-nor/winbond.c >> > +++ b/drivers/mtd/spi-nor/winbond.c >> > @@ -17,6 +17,31 @@ >> > SPI_MEM_OP_NO_DUMMY, \ >> > SPI_MEM_OP_DATA_OUT(1, buf, 0)) >> > >> > +static int >> > +w25q128_post_bfpt_fixups(struct spi_nor *nor, >> > + const struct sfdp_parameter_header *bfpt_header, >> > + const struct sfdp_bfpt *bfpt) >> > +{ >> > + /* >> > + * Zetta ZD25Q128C is a clone of the Winbond device. But the encoded >> > + * size is really wrong. It seems that they confused Mbit with MiB. >> > + * Thus the flash is discovered as a 2MiB device. >> > + */ >> > + if (bfpt_header->major == SFDP_JESD216_MAJOR && >> > + bfpt_header->minor == SFDP_JESD216_MINOR && >> > + nor->params->size == SZ_2M && >> > + nor->params->erase_map.regions[0].size == SZ_2M) { >> > + nor->params->size = SZ_16M; >> > + nor->params->erase_map.regions[0].size = SZ_16M; >> > + } >> >> Since the size is 16 MiB for both Zetta and Winbond, why do you have >> these conditions here? Why not just do it unconditionally? What >> situation do you want to protect against? > > Two things, I want to make sure, the values I'll overwrite (which > were set indirectly by the SFDP data I check here) are really the > "before" value I expect them to be. And secondly, I want to > fingerprint the SFDP as accurately as possible. I mean you wouldn't > need the SFDP version either. Think of a preparation for a possible > .match() callback in zetta.c. That would probably look like this, > i.e. check the flash id and these four conditions. Hmm, I see. I personally find doing it unconditional a bit easier to grok but this is fine too I think. Reviewed-by: Pratyush Yadav <pratyush@kernel.org> > What are your concerns? None. I was mostly curious why you did it this way.
On Mon, Aug 05 2024, Michael Walle wrote: > Zetta normally uses BAh as its vendor ID. But for the ZD25Q128C they > took the one from Winbond and messed up the size parameters in SFDP. > Most functions seem compatible with the W25Q128, we just have to fix up > the size. > > Link: http://www.zettadevice.com/upload/file/20150821/DS_Zetta_25Q128_RevA.pdf > Link: https://www.lcsc.com/datasheet/lcsc_datasheet_2312081757_Zetta-ZD25Q128CSIGT_C19626875.pdf > Signed-off-by: Michael Walle <mwalle@kernel.org> Applied to spi-nor/next. Thanks!
diff --git a/drivers/mtd/spi-nor/winbond.c b/drivers/mtd/spi-nor/winbond.c index e065e4fd42a3..9f7ce5763e71 100644 --- a/drivers/mtd/spi-nor/winbond.c +++ b/drivers/mtd/spi-nor/winbond.c @@ -17,6 +17,31 @@ SPI_MEM_OP_NO_DUMMY, \ SPI_MEM_OP_DATA_OUT(1, buf, 0)) +static int +w25q128_post_bfpt_fixups(struct spi_nor *nor, + const struct sfdp_parameter_header *bfpt_header, + const struct sfdp_bfpt *bfpt) +{ + /* + * Zetta ZD25Q128C is a clone of the Winbond device. But the encoded + * size is really wrong. It seems that they confused Mbit with MiB. + * Thus the flash is discovered as a 2MiB device. + */ + if (bfpt_header->major == SFDP_JESD216_MAJOR && + bfpt_header->minor == SFDP_JESD216_MINOR && + nor->params->size == SZ_2M && + nor->params->erase_map.regions[0].size == SZ_2M) { + nor->params->size = SZ_16M; + nor->params->erase_map.regions[0].size = SZ_16M; + } + + return 0; +} + +static const struct spi_nor_fixups w25q128_fixups = { + .post_bfpt = w25q128_post_bfpt_fixups, +}; + static int w25q256_post_bfpt_fixups(struct spi_nor *nor, const struct sfdp_parameter_header *bfpt_header, @@ -108,6 +133,7 @@ static const struct flash_info winbond_nor_parts[] = { .size = SZ_16M, .flags = SPI_NOR_HAS_LOCK | SPI_NOR_HAS_TB, .no_sfdp_flags = SECT_4K | SPI_NOR_DUAL_READ | SPI_NOR_QUAD_READ, + .fixups = &w25q128_fixups, }, { .id = SNOR_ID(0xef, 0x40, 0x19), .name = "w25q256",
Zetta normally uses BAh as its vendor ID. But for the ZD25Q128C they took the one from Winbond and messed up the size parameters in SFDP. Most functions seem compatible with the W25Q128, we just have to fix up the size. Link: http://www.zettadevice.com/upload/file/20150821/DS_Zetta_25Q128_RevA.pdf Link: https://www.lcsc.com/datasheet/lcsc_datasheet_2312081757_Zetta-ZD25Q128CSIGT_C19626875.pdf Signed-off-by: Michael Walle <mwalle@kernel.org> --- $ cat /sys/class/spi_master/spi0/spi0.0/spi-nor/jedec_id ef4018 $ md5sum /sys/class/spi_master/spi0/spi0.0/spi-nor/sfdp f19e214c6709a259511bac4bd9c87407 /sys/class/spi_master/spi0/spi0.0/spi-nor/sfdp $ xxd -p /sys/class/spi_master/spi0/spi0.0/spi-nor/sfdp 53464450000101ff00000109300000ffb3000103600000ffffffffffffff ffffffffffffffffffffffffffffffffffffe520f9ffffffff0044eb086b 083b80bbeeffffffffff00ffffff00ff0c200f5210d80881ffffffffffff ffffffffffff003600279ff97764fccbffff $ cat /sys/kernel/debug/spi-nor/spi0.0/capabilities Supported read modes by the flash 1S-1S-1S opcode 0x03 mode cycles 0 dummy cycles 0 1S-1S-1S (fast read) opcode 0x0b mode cycles 0 dummy cycles 8 1S-1S-2S opcode 0x3b mode cycles 0 dummy cycles 8 1S-2S-2S opcode 0xbb mode cycles 4 dummy cycles 0 1S-1S-4S opcode 0x6b mode cycles 0 dummy cycles 8 1S-4S-4S opcode 0xeb mode cycles 2 dummy cycles 4 Supported page program modes by the flash 1S-1S-1S opcode 0x02 $ cat /sys/kernel/debug/spi-nor/spi0.0/params name w25q128 id ef 40 18 ef 40 18 size 16.0 MiB write size 1 page size 256 address nbytes 3 flags HAS_SR_TB | HAS_LOCK | HAS_16BIT_SR opcodes read 0x0b dummy cycles 8 erase 0x20 program 0x02 8D extension none protocols read 1S-1S-1S write 1S-1S-1S register 1S-1S-1S erase commands 81 (256 B) [0] 20 (4.00 KiB) [1] 52 (32.0 KiB) [2] d8 (64.0 KiB) [3] c7 (16.0 MiB) sector map region (in hex) | erase mask | overlaid ------------------+------------+---------- 00000000-00ffffff | [ 1 ] | no $ dd if=/dev/urandom of=spi_test bs=1M count=2 2+0 records in 2+0 records out $ mtd_debug erase /dev/mtd0 0 2097152 Erased 2097152 bytes from address 0x00000000 in flash $ mtd_debug read /dev/mtd0 0 2097152 spi_read Copied 2097152 bytes from address 0x00000000 in flash to spi_read $ hexdump spi_read 0000000 ffff ffff ffff ffff ffff ffff ffff ffff * 0200000 $ sha256sum spi_read 4bda3a28f4ffe603c0ec1258c0034d65a1a0d35ab7bd523a834608adabf03cc5 spi_read $ mtd_debug write /dev/mtd0 0 2097152 spi_test Copied 2097152 bytes from spi_test to address 0x00000000 in flash $ mtd_debug read /dev/mtd0 0 2097152 spi_read Copied 2097152 bytes from address 0x00000000 in flash to spi_read $ sha256sum spi* 90145978caa36e84ecc07734d193c97a54ae601884a5ff4e0b90ff8c1f540a03 spi_read 90145978caa36e84ecc07734d193c97a54ae601884a5ff4e0b90ff8c1f540a03 spi_test $ mtd_debug erase /dev/mtd0 0 2097152 Erased 2097152 bytes from address 0x00000000 in flash $ mtd_debug read /dev/mtd0 0 2097152 spi_read Copied 2097152 bytes from address 0x00000000 in flash to spi_read $ sha256sum spi* 4bda3a28f4ffe603c0ec1258c0034d65a1a0d35ab7bd523a834608adabf03cc5 spi_read 90145978caa36e84ecc07734d193c97a54ae601884a5ff4e0b90ff8c1f540a03 spi_test $ mtd_debug info /dev/mtd0 mtd.type = MTD_NORFLASH mtd.flags = MTD_CAP_NORFLASH mtd.size = 16777216 (16M) mtd.erasesize = 4096 (4K) mtd.writesize = 1 mtd.oobsize = 0 regions = 0 --- drivers/mtd/spi-nor/winbond.c | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+)