Skip to contentSkip to conversation

mtd/cfi: add support for SST 4KB sector granularity

PATCH New Angelo Dureghello 2 participants 2 messages Dec 1, 2012 – Jan 10, 2013 Memory technology devices (MTD)

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 mtd/cfi: add support for SST 4KB sector granularity

Add support for SST 4KB sector granularity.

Many recent SST flashes, i.e. SST39VF3201B and similar of this family are declared CFI-conformant from SST. They support CFI query, but implement 2 different sector sizes in the same memory: a 64KB sector (they call it "block", std AMD erase cmd=0x30), and a 4KB sector (they call it "sector", erase cmd=0x50). Also, CFI query on these chips, reading from address 0x2dh of cfi query struct, detects a number of secotrs for the 4KB granularity (flinfo shows it).

For all other aspects, they are CFI compliant, so, as Linux do, i think it's a good idea to handle these chips in the CFI driver, with a fixup to allow 4KB granularity, as should be expected, instead of 64KB.

---
 drivers/mtd/cfi_flash.c |   25 ++++++++++++++++++++++++-
 include/flash.h         |    1 +
 2 files changed, 25 insertions(+), 1 deletion(-)
drivers/mtd/cfi_flash.c +24 -1
index b2dfc53..b46f1ca 100644
@@ -1128,7 +1128,7 @@ int flash_erase (flash_info_t * info, int s_first, int s_last)
 						AMD_CMD_ERASE_START);
 				flash_unlock_seq (info, sect);
 				flash_write_cmd (info, sect, 0,
-						 AMD_CMD_ERASE_SECTOR);
+						 info->cmd_erase_sector);
 				break;
 #ifdef CONFIG_FLASH_CFI_LEGACY
 			case CFI_CMDSET_AMD_LEGACY:
@@ -1733,6 +1733,7 @@ static void cmdset_amd_read_jedec_ids(flash_info_t *info)
 static int cmdset_amd_init(flash_info_t *info, struct cfi_qry *qry)
 {
 	info->cmd_reset = AMD_CMD_RESET;
+	info->cmd_erase_sector = AMD_CMD_ERASE_SECTOR;
 
 	cmdset_amd_read_jedec_ids(info);
 	flash_write_cmd(info, 0, info->cfi_offset, FLASH_CMD_CFI);
@@ -2003,6 +2004,25 @@ static void flash_fixup_stm(flash_info_t *info, struct cfi_qry *qry)
 	}
 }
 
+static void flash_fixup_sst(flash_info_t *info, struct cfi_qry *qry)
+{
+	/*
+	 * SST, for many recent nor parallel flashes, says they are
+	 * CFI-conformant. This is not true, since qry struct. 
+	 * reports a std. AMD command set (0x0002), while SST allows to
+	 * erase two different sector sizes for the same memory.
+	 * 64KB sector (SST call it block)  needs 0x30 to be erased.
+	 * 4KB  sector (SST call it sector) needs 0x50 to be erased.
+	 * Since CFI query detect the 4KB number of sectors, users expects 
+	 * a sector granularity of 4KB, and it is here set.
+	 */ 
+	if (info->device_id == 0x5D23 || /* SST39VF3201B */
+	    info->device_id == 0x5C23) { /* SST39VF3202B */
+		/* set sector granularity to 4KB */	
+		info->cmd_erase_sector=0x50;
+	}
+}
+
 /*
  * The following code cannot be run from FLASH!
  *
@@ -2081,6 +2101,9 @@ ulong flash_get_size (phys_addr_t base, int banknum)
 		case 0x0020:
 			flash_fixup_stm(info, &qry);
 			break;
+		case 0x00bf: /* SST */
+			flash_fixup_sst(info, &qry);
+			break;
 		}
 
 		debug ("manufacturer is %d\n", info->vendor);
include/flash.h +1 -0
index a04ce90..0ba2d33 100644
@@ -44,6 +44,7 @@ typedef struct {
 	ulong	buffer_write_tout;	/* maximum buffer write timeout		*/
 	ushort	vendor;			/* the primary vendor id		*/
 	ushort	cmd_reset;		/* vendor specific reset command	*/
+	uchar   cmd_erase_sector;	/* vendor specific erase sect. command	*/
 	ushort	interface;		/* used for x8/x16 adjustments		*/
 	ushort	legacy_unlock;		/* support Intel legacy (un)locking	*/
 	ushort	manufacturer_id;	/* manufacturer id			*/
#1 lore raw

PATCH 1/1 mtd/cfi: add support for SST 4KB sector granularity

Quoted 10 lines · Angelo Dureghello

Add support for SST 4KB sector granularity.

Many recent SST flashes, i.e. SST39VF3201B and similar of this family are declared CFI-conformant from SST. They support CFI query, but implement 2 different sector sizes in the same memory: a 64KB sector (they call it "block", std AMD erase cmd=0x30), and a 4KB sector (they call it "sector", erase cmd=0x50). Also, CFI query on these chips, reading from address 0x2dh of cfi query struct, detects a number of secotrs for the 4KB granularity (flinfo shows it).

Quoted 3 lines

For all other aspects, they are CFI compliant, so, as Linux do, i think it's a good idea to handle these chips in the CFI driver, with a fixup to allow 4KB granularity, as should be expected, instead of 64KB.

Applied to u-boot-cfi-flash/master. Thanks.

Thanks,
Stefan