Skip to contentSkip to conversation

Add some second source legacy flash chips 256x8

PATCH New Niklaus Giger 3 participants 3 messages Jul 22, 2009 Memory technology devices (MTD)

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 Add some second source legacy flash chips 256x8

---

We tested some boot flashes as possible second source.
As some had ids above 0x7f, we had to implement JEDEC JEP106Z in cfi_flash.c. I now that I just missed the merge window, but the patches are not urgent and I was away for a three week

 drivers/mtd/cfi_flash.c   |   16 +++++++++--
 drivers/mtd/jedec_flash.c |   67 +++++++++++++++++++++++++++++++++++++++++++++
 include/flash.h           |   10 ++++++-
 3 files changed, 89 insertions(+), 4 deletions(-)
drivers/mtd/cfi_flash.c +13 -3
index 81ac5d3..4cd0116 100644
@@ -106,6 +106,8 @@
 #define ATM_CMD_SOFTLOCK_START		0x80
 #define ATM_CMD_LOCK_SECT		0x40
 
+#define FLASH_CONTINUATION_CODE         0x7F
+
 #define FLASH_OFFSET_MANUFACTURER_ID	0x00
 #define FLASH_OFFSET_DEVICE_ID		0x01
 #define FLASH_OFFSET_DEVICE_ID2		0x0E
@@ -1541,13 +1543,21 @@ static int cmdset_intel_init(flash_info_t *info, struct cfi_qry *qry)
 
 static void cmdset_amd_read_jedec_ids(flash_info_t *info)
 {
-	flash_write_cmd(info, 0, 0, AMD_CMD_RESET);
+	ushort bankId = 0;
+	uchar  manuId;
+
+        flash_write_cmd(info, 0, 0, AMD_CMD_RESET);
 	flash_unlock_seq(info, 0);
 	flash_write_cmd(info, 0, info->addr_unlock1, FLASH_CMD_READ_ID);
 	udelay(1000); /* some flash are slow to respond */
 
-	info->manufacturer_id = flash_read_uchar (info,
-					FLASH_OFFSET_MANUFACTURER_ID);
+	manuId = flash_read_uchar (info, FLASH_OFFSET_MANUFACTURER_ID);
+	/* JEDEC JEP106Z specifies ID codes up to bank 7 */
+	while(manuId == FLASH_CONTINUATION_CODE && bankId < 0x800) {
+		bankId += 0x100;
+		manuId = flash_read_uchar (info, bankId | FLASH_OFFSET_MANUFACTURER_ID);
+	}
+	info->manufacturer_id = manuId;
 
 	switch (info->chipwidth){
 	case FLASH_CFI_8BIT:
drivers/mtd/jedec_flash.c +67 -0

79 lines not on this page · raw message

include/flash.h +9 -1
index b016162..8feca1b 100644
@@ -46,7 +46,7 @@ typedef struct {
 	ushort	cmd_reset;		/* vendor specific reset command	*/
 	ushort	interface;		/* used for x8/x16 adjustments		*/
 	ushort	legacy_unlock;		/* support Intel legacy (un)locking	*/
-	uchar	manufacturer_id;	/* manufacturer id			*/
+	ushort	manufacturer_id;	/* manufacturer id			*/
 	ushort	device_id;		/* device id				*/
 	ushort	device_id2;		/* extended device id			*/
 	ushort	ext_addr;		/* extended query table address		*/
@@ -154,6 +154,7 @@ extern flash_info_t *flash_get_info(ulong base);
  * Device IDs
  */
 
+/* Manufacturers inside bank 0 have ids like 0x00xx00xx */
 #define AMD_MANUFACT	0x00010001	/* AMD	   manuf. ID in D23..D16, D7..D0 */
 #define FUJ_MANUFACT	0x00040004	/* FUJITSU manuf. ID in D23..D16, D7..D0 */
 #define ATM_MANUFACT	0x001F001F	/* ATMEL */
@@ -166,6 +167,13 @@ extern flash_info_t *flash_get_info(ulong base);
 #define TOSH_MANUFACT	0x00980098	/* TOSHIBA manuf. ID in D23..D16, D7..D0 */
 #define MT2_MANUFACT	0x002C002C	/* alternate MICRON manufacturer ID*/
 #define EXCEL_MANUFACT	0x004A004A	/* Excel Semiconductor			*/
+#define AMIC_MANUFACT	0x00370037	/* AMIC    manuf. ID in D23..D16, D7..D0 */
+#define WINB_MANUFACT	0x00DA00DA	/* Winbond manuf. ID in D23..D16, D7..D0 */
+
+/* Manufacturers inside bank 1 have ids like 0x01xx01xx */
+#define EON_MANUFACT	0x011C011C	/* EON     manuf. ID in D23..D16, D7..D0 */
+
+/* Manufacturers inside bank 2 have ids like 0x02xx02xx */
 
 					/* Micron Technologies (INTEL compat.)	*/
 #define MT_ID_28F400_T	0x44704470	/* 28F400B3 ID ( 4 M, top boot sector)	*/

--
1.6.3.3

#1 lore raw

PATCH 1/1 Add some second source legacy flash chips 256x8

Dear Niklaus Giger,

Quoted 5 lines · In message <[email protected]> you…

---

We tested some boot flashes as possible second source.
As some had ids above 0x7f, we had to implement JEDEC JEP106Z in cfi_flash.c.

Please provide at least this statement as part of the commit message.

Quoted 2 lines

I now that I just missed the merge window, but the patches are not urgent and I was away for a three week

Stefan is probably willin gto consider this as a bug fix (at least I would not reprimand him if he did :-)

drivers/mtd/cfi_flash.c static int cmdset_intel_init(flash_info_t *info, struct cfi_qry *qry) · line 1549

Quoted 7 lines

static void cmdset_amd_read_jedec_ids(flash_info_t *info) {

drivers/mtd/cfi_flash.c +4 -1
-	flash_write_cmd(info, 0, 0, AMD_CMD_RESET);
+	ushort bankId = 0;
+	uchar  manuId;
+
+        flash_write_cmd(info, 0, 0, AMD_CMD_RESET);

Please use TAB for indentation.

Best regards,

Wolfgang Denk

Signature · 4 lines
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
Use C++ to confuse your enemies; use C to produce stable code.
#2 lore raw

PATCH 1/1 Add some second source legacy flash chips 256x8

In reply to Niklaus Giger
Quoted 1 line · On Wednesday 22 July 2009 08:41:49 Niklaus Giger wrote:

Apart from Wolfgang's comments, please find some minor coding style comments below.

Quoted 33 lines

We tested some boot flashes as possible second source.
As some had ids above 0x7f, we had to implement JEDEC JEP106Z in cfi_flash.c. I now that I just missed the merge window, but the patches are not urgent and I was away for a three week

 drivers/mtd/cfi_flash.c   |   16 +++++++++--
 drivers/mtd/jedec_flash.c |   67
+++++++++++++++++++++++++++++++++++++++++++++ include/flash.h           |  
10 ++++++-
 3 files changed, 89 insertions(+), 4 deletions(-)
drivers/mtd/cfi_flash.c +5 -1
index 81ac5d3..4cd0116 100644
@@ -106,6 +106,8 @@
 #define ATM_CMD_SOFTLOCK_START		0x80
 #define ATM_CMD_LOCK_SECT		0x40
 
+#define FLASH_CONTINUATION_CODE         0x7F
+
 #define FLASH_OFFSET_MANUFACTURER_ID	0x00
 #define FLASH_OFFSET_DEVICE_ID		0x01
 #define FLASH_OFFSET_DEVICE_ID2		0x0E
@@ -1541,13 +1543,21 @@ static int cmdset_intel_init(flash_info_t *info,
 struct cfi_qry *qry)
 
 static void cmdset_amd_read_jedec_ids(flash_info_t *info)
 {
-	flash_write_cmd(info, 0, 0, AMD_CMD_RESET);
+	ushort bankId = 0;
+	uchar  manuId;
+

drivers/mtd/cfi_flash.c static int cmdset_intel_init(flash_info_t *info, · line 1557

Quoted 10 lines
drivers/mtd/cfi_flash.c +4 -2
+        flash_write_cmd(info, 0, 0, AMD_CMD_RESET);
 	flash_unlock_seq(info, 0);
 	flash_write_cmd(info, 0, info->addr_unlock1, FLASH_CMD_READ_ID);
 	udelay(1000); /* some flash are slow to respond */
 
-	info->manufacturer_id = flash_read_uchar (info,
-					FLASH_OFFSET_MANUFACTURER_ID);
+	manuId = flash_read_uchar (info, FLASH_OFFSET_MANUFACTURER_ID);
+	/* JEDEC JEP106Z specifies ID codes up to bank 7 */
+	while(manuId == FLASH_CONTINUATION_CODE && bankId < 0x800) {

Space before "(".

Thanks.

Best regards,
Stefan

=====================================================================
DENX Software Engineering GmbH, MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-0 Fax: +49-8142-66989-80 Email: [email protected]
=====================================================================