[PATCH 3/4] mtd: nand: support Micron READ RETRY
Brian Norris
computersforpeace at gmail.com
Thu Dec 5 15:58:38 EST 2013
Caught a simple bug when testing with a non-READ_RETRY NAND. See below.
On Thu, Dec 05, 2013 at 12:19:59PM -0800, Brian Norris wrote:
...
> One thing I'm not sure about: do all relevant (i.e., ONFI-capable) NAND drivers
> support SET_FEATURES properly? If not, then it's possible that this could break
> nand_do_read_ops for such drivers. Not sure what the best method of handling
> that would be.
...
> --- a/drivers/mtd/nand/nand_base.c
> +++ b/drivers/mtd/nand/nand_base.c
...
> @@ -1512,8 +1535,27 @@ static int nand_do_read_ops(struct mtd_info *mtd, loff_t from,
> nand_wait_ready(mtd);
> }
>
> - if (mtd->ecc_stats.failed - ecc_failures)
> - ecc_fail = true;
> + if (mtd->ecc_stats.failed - ecc_failures) {
> + retry_mode++;
This increment shouldn't be done for all ECC failures, but only for
failures where there are remaining read-retry modes (i.e., flash that
support read retry).
> + if (retry_mode < chip->read_retries) {
> + pr_debug("ECC error; performing READ RETRY %d\n",
> + retry_mode);
> +
> + ret = nand_set_read_retry(mtd,
> + retry_mode);
> + if (ret < 0)
> + break;
> +
> + /* Reset failures */
> + mtd->ecc_stats.failed = ecc_failures;
> + continue;
> + } else {
> + /* No more retry modes; real failure */
> + ecc_fail = true;
> + }
> + }
> +
> + buf += bytes;
> } else {
> memcpy(buf, chip->buffers->databuf + col, bytes);
> buf += bytes;
I'll send a v2 with the following extra diff:
diff --git a/drivers/mtd/nand/nand_base.c b/drivers/mtd/nand/nand_base.c
index 82efd819f31b..aeafcdce6cc0 100644
--- a/drivers/mtd/nand/nand_base.c
+++ b/drivers/mtd/nand/nand_base.c
@@ -1536,8 +1536,8 @@ static int nand_do_read_ops(struct mtd_info *mtd, loff_t from,
}
if (mtd->ecc_stats.failed - ecc_failures) {
- retry_mode++;
- if (retry_mode < chip->read_retries) {
+ if (retry_mode + 1 < chip->read_retries) {
+ retry_mode++;
pr_debug("ECC error; performing READ RETRY %d\n",
retry_mode);
---
Brian
More information about the linux-mtd
mailing list