diff mbox

[v1,2/5] nand/denali: change read_status function method

Message ID 20100812022314.GB25657@intel.com
State Accepted
Commit f0bc0c778fe12e43658193fae975fc5762e98773
Headers show

Commit Message

Dong, Chuanxiao Aug. 12, 2010, 2:23 a.m. UTC 
From 8b09365e8e58ea0d18fe0ce03b20531c98d43a7a Mon Sep 17 00:00:00 2001
From: Chuanxiao Dong <chuanxiao.dong@intel.com>
Date: Wed, 11 Aug 2010 17:14:59 +0800
Subject: [PATCH 2/5] nand/denali: change read_status function method

In mtd->write, Denali controller will use MODE_11 mode to read
NAND flash status, then return back to MODE_1O mode to do page
write.
Here comes a bug for this kind of using, sometimes controller will
not write data to NAND and just return a good interrupt to tell
driver writing work is done. The data in this page is all 0xff and
this page can not be written again. The reason is unknow.

So read Denali controller register WRITE_PROTECT to get NAND status
instead.

Signed-off-by: Chuanxiao Dong <chuanxiao.dong@intel.com>
---
 drivers/mtd/nand/denali.c |   12 +++++-------
 1 files changed, 5 insertions(+), 7 deletions(-)
diff mbox

Patch

diff --git a/drivers/mtd/nand/denali.c b/drivers/mtd/nand/denali.c
index 3ba8952..975a893 100644
--- a/drivers/mtd/nand/denali.c
+++ b/drivers/mtd/nand/denali.c
@@ -169,13 +169,11 @@  static void read_status(struct denali_nand_info *denali)
 	/* initialize the data buffer to store status */
 	reset_buf(denali);
 
-	/* initiate a device status read */
-	cmd = MODE_11 | BANK(denali->flash_bank);
-	index_addr(denali, cmd | COMMAND_CYCLE, 0x70);
-	iowrite32(cmd | STATUS_CYCLE, denali->flash_mem);
-
-	/* update buffer with status value */
-	write_byte_to_buf(denali, ioread32(denali->flash_mem + 0x10));
+	cmd = ioread32(denali->flash_reg + WRITE_PROTECT);
+	if (cmd)
+		write_byte_to_buf(denali, NAND_STATUS_WP);
+	else
+		write_byte_to_buf(denali, 0);
 }
 
 /* resets a specific device connected to the core */