[PATCH v2 1/2] mtd: nand: spi: Only one dummy byte in QUADIO

[PATCH v2 1/2] mtd: nand: spi: Only one dummy byte in QUADIO

[Reto Schneider]

From: Hauke Mehrtens <hauke@hauke-m.de>

The datasheet only lists one dummy byte in the 0xEB operation for the
following chips:
* GD5F1GQ4xExxG
* GD5F1GQ4xFxxG
* GD5F1GQ4UAYIG
* GD5F4GQ4UAYIG

Reto Schneider:
- Linux patch ported to U-Boot
- Checked for compatibility with GD5F1GQ4xBxxG
- Fixed operation code in original commit message (0xEH -> 0xEB)

Signed-off-by: Reto Schneider <reto.schneider@husqvarnagroup.com>
Reviewed-by: Stefan Roese <sr@denx.de>
---
Changes in v2:
 - Adjust authorship to original author, declare own work separately

 drivers/mtd/nand/spi/gigadevice.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/mtd/nand/spi/gigadevice.c b/drivers/mtd/nand/spi/gigadevice.c
index 0b228dcb5b..5de0ebbb7b 100644
--- a/drivers/mtd/nand/spi/gigadevice.c
+++ b/drivers/mtd/nand/spi/gigadevice.c
@@ -20,7 +20,7 @@
 #define GD5FXGQ4XEXXG_REG_STATUS2		0xf0
 
 static SPINAND_OP_VARIANTS(read_cache_variants,
-		SPINAND_PAGE_READ_FROM_CACHE_QUADIO_OP(0, 2, NULL, 0),
+		SPINAND_PAGE_READ_FROM_CACHE_QUADIO_OP(0, 1, NULL, 0),
 		SPINAND_PAGE_READ_FROM_CACHE_X4_OP(0, 1, NULL, 0),
 		SPINAND_PAGE_READ_FROM_CACHE_DUALIO_OP(0, 1, NULL, 0),
 		SPINAND_PAGE_READ_FROM_CACHE_X2_OP(0, 1, NULL, 0),
-- 
2.29.2

[PATCH v2 2/2] mtd: nand: spi: Support GigaDevice GD5F1GQ5UExxG

[Reto Schneider]

From: Reto Schneider <reto.schneider@husqvarnagroup.com>

The relevant changes to the already existing GD5F1GQ4UExxG support has
been determined by consulting the GigaDevice product change notice
AN-0392-10, version 1.0 from November 30, 2020.

As the overlaps are huge, variable names have been generalized
accordingly.

Apart form the lowered ECC strength (4 instead of 8 bits per 512 bytes),
the new device ID, and the extra quad IO dummy byte, no changes had to
be taken into account.

New hardware features are not supported, namely:
 - Power on reset
 - Unique ID
 - Double transfer rate (DTR)
 - Parameter page
 - Random data quad IO

The inverted semantic of the "driver strength" register bits, defaulting
to 100% instead of 50% for the Q5 devices, got ignored as the driver has
never touched them anyway.

The no longer supported "read from cache during block erase"
functionality is not reflected as the current SPI NAND core does not
support it anyway.

Implementation has been tested on MediaTek MT7688 based GARDENA smart
Gateways using both, GigaDevice GD5F1GQ5UEYIG and GD5F1GQ4UBYIG.

Signed-off-by: Reto Schneider <reto.schneider@husqvarnagroup.com>
Reviewed-by: Stefan Roese <sr@denx.de>
---
Changes in v2:
 - Clarify effect of no longer supported feature

 drivers/mtd/nand/spi/gigadevice.c | 79 +++++++++++++++++++++++++++----
 1 file changed, 69 insertions(+), 10 deletions(-)

diff --git a/drivers/mtd/nand/spi/gigadevice.c b/drivers/mtd/nand/spi/gigadevice.c
index 5de0ebbb7b..a2c93486f4 100644
--- a/drivers/mtd/nand/spi/gigadevice.c
+++ b/drivers/mtd/nand/spi/gigadevice.c
@@ -17,9 +17,13 @@
 #define GD5FXGQ4XA_STATUS_ECC_1_7_BITFLIPS	(1 << 4)
 #define GD5FXGQ4XA_STATUS_ECC_8_BITFLIPS	(3 << 4)
 
-#define GD5FXGQ4XEXXG_REG_STATUS2		0xf0
+#define GD5FXGQ5XE_STATUS_ECC_1_4_BITFLIPS	(1 << 4)
+#define GD5FXGQ5XE_STATUS_ECC_4_BITFLIPS	(3 << 4)
 
-static SPINAND_OP_VARIANTS(read_cache_variants,
+#define GD5FXGQXXEXXG_REG_STATUS2		0xf0
+
+/* Q4 devices, QUADIO: Dummy bytes valid for 1 and 2 GBit variants */
+static SPINAND_OP_VARIANTS(gd5fxgq4_read_cache_variants,
 		SPINAND_PAGE_READ_FROM_CACHE_QUADIO_OP(0, 1, NULL, 0),
 		SPINAND_PAGE_READ_FROM_CACHE_X4_OP(0, 1, NULL, 0),
 		SPINAND_PAGE_READ_FROM_CACHE_DUALIO_OP(0, 1, NULL, 0),
@@ -27,6 +31,15 @@ static SPINAND_OP_VARIANTS(read_cache_variants,
 		SPINAND_PAGE_READ_FROM_CACHE_OP(true, 0, 1, NULL, 0),
 		SPINAND_PAGE_READ_FROM_CACHE_OP(false, 0, 1, NULL, 0));
 
+/* Q5 devices, QUADIO: Dummy bytes only valid for 1 GBit variants */
+static SPINAND_OP_VARIANTS(gd5f1gq5_read_cache_variants,
+		SPINAND_PAGE_READ_FROM_CACHE_QUADIO_OP(0, 2, NULL, 0),
+		SPINAND_PAGE_READ_FROM_CACHE_X4_OP(0, 1, NULL, 0),
+		SPINAND_PAGE_READ_FROM_CACHE_DUALIO_OP(0, 1, NULL, 0),
+		SPINAND_PAGE_READ_FROM_CACHE_X2_OP(0, 1, NULL, 0),
+		SPINAND_PAGE_READ_FROM_CACHE_OP(true, 0, 1, NULL, 0),
+		SPINAND_PAGE_READ_FROM_CACHE_OP(false, 0, 1, NULL, 0));
+
 static SPINAND_OP_VARIANTS(write_cache_variants,
 		SPINAND_PROG_LOAD_X4(true, 0, NULL, 0),
 		SPINAND_PROG_LOAD(true, 0, NULL, 0));
@@ -35,7 +48,7 @@ static SPINAND_OP_VARIANTS(update_cache_variants,
 		SPINAND_PROG_LOAD_X4(false, 0, NULL, 0),
 		SPINAND_PROG_LOAD(false, 0, NULL, 0));
 
-static int gd5fxgq4xexxg_ooblayout_ecc(struct mtd_info *mtd, int section,
+static int gd5fxgqxxexxg_ooblayout_ecc(struct mtd_info *mtd, int section,
 				       struct mtd_oob_region *region)
 {
 	if (section)
@@ -47,7 +60,7 @@ static int gd5fxgq4xexxg_ooblayout_ecc(struct mtd_info *mtd, int section,
 	return 0;
 }
 
-static int gd5fxgq4xexxg_ooblayout_free(struct mtd_info *mtd, int section,
+static int gd5fxgqxxexxg_ooblayout_free(struct mtd_info *mtd, int section,
 					struct mtd_oob_region *region)
 {
 	if (section)
@@ -64,7 +77,7 @@ static int gd5fxgq4xexxg_ecc_get_status(struct spinand_device *spinand,
 					u8 status)
 {
 	u8 status2;
-	struct spi_mem_op op = SPINAND_GET_FEATURE_OP(GD5FXGQ4XEXXG_REG_STATUS2,
+	struct spi_mem_op op = SPINAND_GET_FEATURE_OP(GD5FXGQXXEXXG_REG_STATUS2,
 						      &status2);
 	int ret;
 
@@ -102,21 +115,67 @@ static int gd5fxgq4xexxg_ecc_get_status(struct spinand_device *spinand,
 	return -EINVAL;
 }
 
-static const struct mtd_ooblayout_ops gd5fxgq4xexxg_ooblayout = {
-	.ecc = gd5fxgq4xexxg_ooblayout_ecc,
-	.rfree = gd5fxgq4xexxg_ooblayout_free,
+static int gd5fxgq5xexxg_ecc_get_status(struct spinand_device *spinand,
+					u8 status)
+{
+	u8 status2;
+	struct spi_mem_op op = SPINAND_GET_FEATURE_OP(GD5FXGQXXEXXG_REG_STATUS2,
+						      &status2);
+	int ret;
+
+	switch (status & STATUS_ECC_MASK) {
+	case STATUS_ECC_NO_BITFLIPS:
+		return 0;
+
+	case GD5FXGQ5XE_STATUS_ECC_1_4_BITFLIPS:
+		/*
+		 * Read status2 register to determine a more fine grained
+		 * bit error status
+		 */
+		ret = spi_mem_exec_op(spinand->slave, &op);
+		if (ret)
+			return ret;
+
+		/*
+		 * 1 ... 4 bits are flipped (and corrected)
+		 */
+		/* bits sorted this way (1...0): ECCSE1, ECCSE0 */
+		return ((status2 & STATUS_ECC_MASK) >> 4) + 1;
+
+	case STATUS_ECC_UNCOR_ERROR:
+		return -EBADMSG;
+
+	default:
+		break;
+	}
+
+	return -EINVAL;
+}
+
+static const struct mtd_ooblayout_ops gd5fxgqxxexxg_ooblayout = {
+	.ecc = gd5fxgqxxexxg_ooblayout_ecc,
+	.rfree = gd5fxgqxxexxg_ooblayout_free,
 };
 
 static const struct spinand_info gigadevice_spinand_table[] = {
 	SPINAND_INFO("GD5F1GQ4UExxG", 0xd1,
 		     NAND_MEMORG(1, 2048, 128, 64, 1024, 1, 1, 1),
 		     NAND_ECCREQ(8, 512),
-		     SPINAND_INFO_OP_VARIANTS(&read_cache_variants,
+		     SPINAND_INFO_OP_VARIANTS(&gd5fxgq4_read_cache_variants,
 					      &write_cache_variants,
 					      &update_cache_variants),
 		     0,
-		     SPINAND_ECCINFO(&gd5fxgq4xexxg_ooblayout,
+		     SPINAND_ECCINFO(&gd5fxgqxxexxg_ooblayout,
 				     gd5fxgq4xexxg_ecc_get_status)),
+	SPINAND_INFO("GD5F1GQ5UExxG", 0x51,
+		     NAND_MEMORG(1, 2048, 128, 64, 1024, 1, 1, 1),
+		     NAND_ECCREQ(4, 512),
+		     SPINAND_INFO_OP_VARIANTS(&gd5f1gq5_read_cache_variants,
+					      &write_cache_variants,
+					      &update_cache_variants),
+		     0,
+		     SPINAND_ECCINFO(&gd5fxgqxxexxg_ooblayout,
+				     gd5fxgq5xexxg_ecc_get_status)),
 };
 
 static int gigadevice_spinand_detect(struct spinand_device *spinand)
-- 
2.29.2