Re: [PATCH 1/2] mtd: spinand: winbond: add support for W25N04LW
From: Dominique MARTINET
Date: Mon Aug 10 2026 - 04:45:23 EST
Md Sadre Alam wrote on Fri, Aug 07, 2026 at 12:13:38PM +0530:
> Add support for Winbond W25N04LW 4Gbit SPI-NAND.
> It has 8-bit on-die ECC and a 4096+256 byte page
> (4096 byte main area, 128 byte usable OOB once
> on-chip ECC hides the 128 byte parity area),
> organized as 2048 blocks of 64 pages each, unlike
> the existing 2048 byte page W25N04KV/W25N04KW parts.
This is pretty good timing, I was just looking at sending a patch for
W25N04LW myself!
I agree with Miquèl that your patch looks odd, you wrote
> + NAND_MEMORG(1, 4096, 128, 64, 2048, 40, 1, 1, 1),
but the erase blocks are 256K long so it should be
+ NAND_MEMORG(1, 4096, 256, 64, 2048, 40, 1, 1, 1),
did you actually test this device?
If it helps, the datasheet for W25N04LWZExx / W25N04LWTBxx is freely
available from digikey here[1]
[1] https://mm.digikey.com/Volume0/opasdata/d220001/medias/docus/7182/W25N04LWZExx_W25N04LWTBxx_RevD_2025-08-05.pdf
(And it also supports continuous read, so I'm curious to see if it'd
work with the recent continous read support from Miquèl[2] merged in
7.2... I'm not familiar with the subsystem so will need a bit of time to
plug all the hooks in properly)
[2] https://lore.kernel.org/linux-mtd/20260429-winbond-v6-18-rc1-cont-read-v3-0-0f38b3c229ad@xxxxxxxxxxx/
FWIW, I've confirmed it works with the following diff (at least up to
the point of creating an ubi volume and booting from it); happy to send
this as a patch if this has indeed not been tested
------------
diff --git a/drivers/mtd/nand/spi/winbond.c b/drivers/mtd/nand/spi/winbond.c
index 9b78c1e6cbc9..43a4886dc4a5 100644
--- a/drivers/mtd/nand/spi/winbond.c
+++ b/drivers/mtd/nand/spi/winbond.c
@@ -335,6 +335,30 @@ static int w25n02kv_ooblayout_free(struct mtd_info *mtd, int section,
return 0;
}
+static int w25n04lw_ooblayout_ecc(struct mtd_info *mtd, int section,
+ struct mtd_oob_region *region)
+{
+ if (section > 7)
+ return -ERANGE;
+
+ region->offset = 128 + (16 * section);
+ region->length = 13;
+
+ return 0;
+}
+
+static int w25n04lw_ooblayout_free(struct mtd_info *mtd, int section,
+ struct mtd_oob_region *region)
+{
+ if (section > 7)
+ return -ERANGE;
+
+ region->offset = (16 * section) + 2;
+ region->length = 14;
+
+ return 0;
+}
+
static const struct mtd_ooblayout_ops w25n01kv_ooblayout = {
.ecc = w25n01kv_ooblayout_ecc,
.free = w25n02kv_ooblayout_free,
@@ -345,6 +369,11 @@ static const struct mtd_ooblayout_ops w25n02kv_ooblayout = {
.free = w25n02kv_ooblayout_free,
};
+static const struct mtd_ooblayout_ops w25n04lw_ooblayout = {
+ .ecc = w25n04lw_ooblayout_ecc,
+ .free = w25n04lw_ooblayout_free,
+};
+
static int w25n01jw_ooblayout_ecc(struct mtd_info *mtd, int section,
struct mtd_oob_region *region)
{
@@ -768,6 +797,15 @@ static const struct spinand_info winbond_spinand_table[] = {
&update_cache_variants),
0,
SPINAND_ECCINFO(&w25n02kv_ooblayout, w25n02kv_ecc_get_status)),
+ SPINAND_INFO("W25N04LW", /* 1.8V */
+ SPINAND_ID(SPINAND_READID_METHOD_OPCODE_DUMMY, 0xb2, 0x23),
+ NAND_MEMORG(1, 4096, 256, 64, 2048, 40, 1, 1, 1),
+ NAND_ECCREQ(8, 512),
+ SPINAND_INFO_OP_VARIANTS(&read_cache_variants,
+ &write_cache_variants,
+ &update_cache_variants),
+ 0,
+ SPINAND_ECCINFO(&w25n04lw_ooblayout, w25n02kv_ecc_get_status)),
SPINAND_INFO("W35N04JW", /* 1.8V */
SPINAND_ID(SPINAND_READID_METHOD_OPCODE_DUMMY, 0xdf, 0x23),
NAND_MEMORG(1, 4096, 128, 64, 512, 10, 1, 4, 1),
-----------------
Thanks,
--
Dominique Martinet | Asmadeus