Skip to contentSkip to conversation

mtd: spi-nor: Fix non-uniform erase map issues

PATCH Applied [email protected] 2 participants 8 messages Oct 2, 2020 – Feb 5, 2021

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

cover mtd: spi-nor: Fix non-uniform erase map issues

From: Takahiro Kuwano <[email protected]>

This patchset fixes the issues in handling non-uniform erase map.

Takahiro Kuwano (4):
mtd: spi-nor: sfdp: Fix wrong erase type bitmask for overlaid region
mtd: spi-nor: sfdp: Fix wrong index of last erase region
mtd: spi-nor: core: Fix erase type discovery for overlaid region
mtd: spi-nor: core: Fix erase command list setup for

 drivers/mtd/spi-nor/core.c | 10 ++++++----
 drivers/mtd/spi-nor/sfdp.c |  5 ++---
 2 files changed, 8 insertions(+), 7 deletions(-)
Signature · 6 lines
2.25.1


______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/
#1 lore raw

PATCH 1/4 mtd: spi-nor: sfdp: Fix wrong erase type bitmask for overlaid region

From: Takahiro Kuwano <[email protected]>

At the time spi_nor_region_check_overlay() is called, the erase types are sorted in ascending order of erase size. The 'erase_type' should be masked with 'BIT(erase[i].idx)' instead of 'BIT(i)'.

---
 drivers/mtd/spi-nor/sfdp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mtd/spi-nor/sfdp.c +1 -1
index e2a43d39eb5f..5332345b3cb8 100644
@@ -760,7 +760,7 @@ spi_nor_region_check_overlay(struct spi_nor_erase_region *region,
 	int i;
 
 	for (i = 0; i < SNOR_ERASE_TYPE_MAX; i++) {
-		if (!(erase_type & BIT(i)))
+		if (!(erase[i].size && erase_type & BIT(erase[i].idx)))
 			continue;
 		if (region->size & erase[i].size_mask) {
 			spi_nor_region_mark_overlay(region);
Signature · 6 lines
2.25.1


______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/
#2 lore raw

PATCH 2/4 mtd: spi-nor: sfdp: Fix last erase region marking

In reply to [email protected]

From: Takahiro Kuwano <[email protected]>

The place of spi_nor_region_mark_end() must be moved, because 'i' is re-used for the index of erase[].

---
 drivers/mtd/spi-nor/sfdp.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/mtd/spi-nor/sfdp.c +1 -2
index 5332345b3cb8..08de2a2b4452 100644
@@ -830,6 +830,7 @@ spi_nor_init_non_uniform_erase_map(struct spi_nor *nor,
 		offset = (region[i].offset & ~SNOR_ERASE_FLAGS_MASK) +
 			 region[i].size;
 	}
+	spi_nor_region_mark_end(&region[i - 1]);
 
 	save_uniform_erase_type = map->uniform_erase_type;
 	map->uniform_erase_type = spi_nor_sort_erase_mask(map,
@@ -853,8 +854,6 @@ spi_nor_init_non_uniform_erase_map(struct spi_nor *nor,
 		if (!(regions_erase_type & BIT(erase[i].idx)))
 			spi_nor_set_erase_type(&erase[i], 0, 0xFF);
 
-	spi_nor_region_mark_end(&region[i - 1]);
-
 	return 0;
 }
 
Signature · 6 lines
2.25.1


______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/
#3 lore raw

PATCH 3/4 mtd: spi-nor: core: Fix erase type discovery for overlaid region

In reply to [email protected]

From: Takahiro Kuwano <[email protected]>

For overlaid regions, 'region->size' is assined to 'cmd->size' later in spi_nor_init_erase_cmd() so 'erase->size' can be greater than 'len'.

---
 drivers/mtd/spi-nor/core.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
drivers/mtd/spi-nor/core.c +5 -4
index 0369d98b2d12..c7d5c86905fa 100644
@@ -1212,14 +1212,15 @@ spi_nor_find_best_erase_type(const struct spi_nor_erase_map *map,
 
 		erase = &map->erase_type[i];
 
+		/* Alignment is not mandatory for overlaid regions */
+		if (region->offset & SNOR_OVERLAID_REGION &&
+		    region->size <= len)
+			return erase;
+
 		/* Don't erase more than what the user has asked for. */
 		if (erase->size > len)
 			continue;
 
-		/* Alignment is not mandatory for overlaid regions */
-		if (region->offset & SNOR_OVERLAID_REGION)
-			return erase;
-
 		spi_nor_div_by_erase_size(erase, addr, &rem);
 		if (rem)
 			continue;
Signature · 6 lines
2.25.1


______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/
#4 lore raw

PATCH 4/4 mtd: spi-nor: core: Add erase size check for erase command initialization

In reply to [email protected]

From: Takahiro Kuwano <[email protected]>

Even if erase type is same as previous region, erase size can be different if the previous region is overlaid region. Since 'region->size' is assigned to 'cmd->size' for overlaid region, comparing 'erase->size' and 'cmd->size' can detect previous overlaid region.

---
 drivers/mtd/spi-nor/core.c | 1 +
 1 file changed, 1 insertion(+)
drivers/mtd/spi-nor/core.c +1 -0
index c7d5c86905fa..a560e8ce13cd 100644
@@ -1364,6 +1364,7 @@ static int spi_nor_init_erase_cmd_list(struct spi_nor *nor,
 			goto destroy_erase_cmd_list;
 
 		if (prev_erase != erase ||
+		    erase->size != cmd->size ||
 		    region->offset & SNOR_OVERLAID_REGION) {
 			cmd = spi_nor_init_erase_cmd(region, erase);
 			if (IS_ERR(cmd)) {
Signature · 6 lines
2.25.1


______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/
#5 lore raw
In reply to [email protected]

Hi, Takahiro,

Are these theoretical fixes or do they fix the non uniform erase on some flash?

Cheers,
ta

Quoted 12 lines · On 10/2/20 8:17 AM, [email protected] wrote:

EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe

From: Takahiro Kuwano <[email protected]>

This patchset fixes the issues in handling non-uniform erase map.

Takahiro Kuwano (4):
mtd: spi-nor: sfdp: Fix wrong erase type bitmask for overlaid region
mtd: spi-nor: sfdp: Fix wrong index of last erase region
mtd: spi-nor: core: Fix erase type discovery for overlaid region
mtd: spi-nor: core: Fix erase command list setup for

Quoted 7 lines
 drivers/mtd/spi-nor/core.c | 10 ++++++----
 drivers/mtd/spi-nor/sfdp.c |  5 ++---
 2 files changed, 8 insertions(+), 7 deletions(-)

--
2.25.1

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

#6 lore raw

Hi Tudor,

Quoted 4 lines · On 1/24/2021 7:34 PM, [email protected] wrote:

Hi, Takahiro,

Are these theoretical fixes or do they fix the non uniform erase on some flash?

They fix non uniform erase on the S25HL-T/S25HS-T that added by:
https://patchwork.ozlabs.org/project/linux-mtd/patch/[email protected]/

And theoretically, they should fix on S25FS-S family that also has non uniform erase option.

Thank you for looking into this.

Best Regards,
Takahiro

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

#7 lore raw
In reply to [email protected]
Quoted 9 lines · On Fri, 2 Oct 2020 14:17:58 +0900, [email protected] wrote:

This patchset fixes the issues in handling non-uniform erase map.

Takahiro Kuwano (4):
mtd: spi-nor: sfdp: Fix wrong erase type bitmask for overlaid region
mtd: spi-nor: sfdp: Fix wrong index of last erase region
mtd: spi-nor: core: Fix erase type discovery for overlaid region
mtd: spi-nor: core: Fix erase command list setup for

[...]

Updated commit description in 3/4, and added Fixes tags and Cc to stable for all.

Applied to spi-nor/next, thanks!

[1/4] mtd: spi-nor: sfdp: Fix wrong erase type bitmask for overlaid region

      https://git.kernel.org/mtd/c/abdf5a5ef965
[2/4] mtd: spi-nor: sfdp: Fix last erase region marking
      https://git.kernel.org/mtd/c/9166f4af32db
[3/4] mtd: spi-nor: core: Fix erase type discovery for overlaid region
      https://git.kernel.org/mtd/c/969b276718de
[4/4] mtd: spi-nor: core: Add erase size check for erase command initialization
      https://git.kernel.org/mtd/c/58fa22f68fca

Best regards,

Signature · 5 lines
Tudor Ambarus <[email protected]>

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/