MTD: pxa3xx_nand: fix nand detection issue

PATCH New Lei Wen 5 participants 11 messages Jun 7, 2011 – Jul 22, 2011

Subsystems (2)

Memory technology devices (MTD) NAND flash

lore

Share & tools

lore

Lei Wen

author

2011-06-07 10:01 UTC #0 lore raw

PATCH 1/2 MTD: pxa3xx_nand: fix nand detection issue

When keep_config is set, the detection would goes different routine. That the driver would read out the setting which is set previously by bootloader. While most bootloader keep the irq mask as off, and current driver need all irq default open, keep_config behavior would lead to no irq at all.

Signed-off-by Lei Wen <[email protected]>

Tested-by Daniel Mack <[email protected]>

Cc stable at kernel.org

---
 drivers/mtd/nand/pxa3xx_nand.c |   12 +++++++-----
 1 files changed, 7 insertions(+), 5 deletions(-)

drivers/mtd/nand/pxa3xx_nand.c +7 -5

index 1fb3b3a..faa0edd 100644
@@ -813,7 +813,7 @@ static int pxa3xx_nand_detect_config(struct pxa3xx_nand_info *info)
 	info->page_size = ndcr & NDCR_PAGE_SZ ? 2048 : 512;
 	/* set info fields needed to read id */
 	info->read_id_bytes = (info->page_size == 2048) ? 4 : 2;
-	info->reg_ndcr = ndcr;
+	info->reg_ndcr = ndcr & ~NDCR_INT_MASK;
 	info->cmdset = &default_cmdset;
 
 	info->ndtr0cs0 = nand_readl(info, NDTR0CS0);

@@ -882,7 +882,7 @@ static int pxa3xx_nand_scan(struct mtd_info *mtd)
 	struct pxa3xx_nand_info *info = mtd->priv;
 	struct platform_device *pdev = info->pdev;
 	struct pxa3xx_nand_platform_data *pdata = pdev->dev.platform_data;
-	struct nand_flash_dev pxa3xx_flash_ids[2] = { {NULL,}, {NULL,} };
+	struct nand_flash_dev pxa3xx_flash_ids[2], *def = NULL;
 	const struct pxa3xx_nand_flash *f = NULL;
 	struct nand_chip *chip = mtd->priv;
 	uint32_t id = -1;

@@ -942,8 +942,10 @@ static int pxa3xx_nand_scan(struct mtd_info *mtd)
 	pxa3xx_flash_ids[0].erasesize = f->page_size * f->page_per_block;
 	if (f->flash_width == 16)
 		pxa3xx_flash_ids[0].options = NAND_BUSWIDTH_16;
+	pxa3xx_flash_ids[1].name = NULL;
+	def = pxa3xx_flash_ids;
 KEEP_CONFIG:
-	if (nand_scan_ident(mtd, 1, pxa3xx_flash_ids))
+	if (nand_scan_ident(mtd, 1, def))
 		return -ENODEV;
 	/* calculate addressing information */
 	info->col_addr_cycles = (mtd->writesize >= 2048) ? 2 : 1;

@@ -954,9 +956,9 @@ KEEP_CONFIG:
 		info->row_addr_cycles = 2;
 	mtd->name = mtd_names[0];
 	chip->ecc.mode = NAND_ECC_HW;
-	chip->ecc.size = f->page_size;
+	chip->ecc.size = info->page_size;
 
-	chip->options = (f->flash_width == 16) ? NAND_BUSWIDTH_16 : 0;
+	chip->options = (info->reg_ndcr & NDCR_DWIDTH_M) ? NAND_BUSWIDTH_16 : 0;
 	chip->options |= NAND_NO_AUTOINCR;
 	chip->options |= NAND_NO_READRDY;

Signature · 1 line

1.7.0.4

Lei Wen

author

2011-06-07 10:01 UTC #1 lore raw

PATCH 2/2 MTD: pxa3xx_nand: Fix blank page ECC mismatch

From: Daniel Mack <[email protected]>

This bug was introduced in f8155a40 ("mtd: pxa3xx_nand: rework irq logic") and causes the PXA3xx NAND controller fail to operate with NAND flash that has empty pages. According to the comment in this block, the hardware controller will report a double-bit error for empty pages, which can and must be ignored.

This patch restores the original behaviour of the driver.

Signed-off-by Daniel Mack <[email protected]>

Acked-by Lei Wen <[email protected]>

Cc Haojian Zhuang <[email protected]>

Cc David Woodhouse <[email protected]>

Cc stable at kernel.org

---
 drivers/mtd/nand/pxa3xx_nand.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

drivers/mtd/nand/pxa3xx_nand.c +2 -0

index faa0edd..30689cc 100644
@@ -685,6 +685,8 @@ static int pxa3xx_nand_read_page_hwecc(struct mtd_info *mtd,
 		 * OOB, ignore such double bit errors
 		 */
 		if (is_buf_blank(buf, mtd->writesize))
+			info->retcode = ERR_NONE;
+		else
 			mtd->ecc_stats.failed++;
 	}

Signature · 1 line

1.7.0.4

Artem Bityutskiy

2011-06-07 11:17 UTC #2 lore raw

PATCH 2/2 MTD: pxa3xx_nand: Fix blank page ECC mismatch

Quoted 8 lines · Lei Wen

From: Daniel Mack <[email protected]>
This bug was introduced in f8155a40 ("mtd: pxa3xx_nand: rework irq logic") and causes the PXA3xx NAND controller fail to operate with NAND flash that has empty pages. According to the comment in this block, the hardware controller will report a double-bit error for empty pages, which can and must be ignored.

Quoted 7 lines

This patch restores the original behaviour of the driver.
Signed-off-by Daniel Mack <[email protected]>
Acked-by Lei Wen <[email protected]>
Cc Haojian Zhuang <[email protected]>
Cc David Woodhouse <[email protected]>
Cc stable at kernel.org

This was introduced in 2.6.38, so let's make it:

Cc stable at kernel.org [2.6.38+]

Signature · 2 lines

Best Regards,
Artem Bityutskiy (????? ????????)

Lei Wen

2011-06-07 12:28 UTC #3 lore raw

PATCH 2/2 MTD: pxa3xx_nand: Fix blank page ECC mismatch

Quoted 17 lines · Artem Bityutskiy

Quoted 8 lines · Lei Wen
From: Daniel Mack <[email protected]>
This bug was introduced in f8155a40 ("mtd: pxa3xx_nand: rework irq logic") and causes the PXA3xx NAND controller fail to operate with NAND flash that has empty pages. According to the comment in this block, the hardware controller will report a double-bit error for empty pages, which can and must be ignored.
Quoted 7 lines
This patch restores the original behaviour of the driver.
Signed-off-by Daniel Mack <[email protected]>
Acked-by Lei Wen <[email protected]>
Cc Haojian Zhuang <[email protected]>
Cc David Woodhouse <[email protected]>
Cc stable at kernel.org

Quoted 4 lines

This was introduced in 2.6.38, so let's make it:
Cc stable at kernel.org [2.6.38+]

Thanks for correcting this.

Best regards,
Lei

Daniel Mack

2011-06-25 11:48 UTC #4 lore raw

PATCH 2/2 MTD: pxa3xx_nand: Fix blank page ECC mismatch

Quoted 18 lines · Lei Wen

Quoted 17 lines · Artem Bityutskiy
Quoted 15 lines · Lei Wen
From: Daniel Mack <[email protected]>
This bug was introduced in f8155a40 ("mtd: pxa3xx_nand: rework irq logic") and causes the PXA3xx NAND controller fail to operate with NAND flash that has empty pages. According to the comment in this block, the hardware controller will report a double-bit error for empty pages, which can and must be ignored.
This patch restores the original behaviour of the driver.
Signed-off-by Daniel Mack <[email protected]>
Acked-by Lei Wen <[email protected]>
Cc Haojian Zhuang <[email protected]>
Cc David Woodhouse <[email protected]>
Cc stable at kernel.org

Quoted 5 lines

Quoted 4 lines
This was introduced in 2.6.38, so let's make it:
Cc stable at kernel.org [2.6.38+]
Thanks for correcting this.

Has anyone queued this for 3.0 yet? This fixes a serious regression, so we need this merged, along with Lei's other patch titled "MTD:
pxa3xx_nand: fix nand detection issue".

Thanks,
Daniel

Brian Norris

2011-07-21 18:55 UTC #5 lore raw

PATCH 2/2 MTD: pxa3xx_nand: Fix blank page ECC mismatch

Quoted 3 lines · Daniel Mack

Has anyone queued this for 3.0 yet? This fixes a serious regression, so we need this merged, along with Lei's other patch titled "MTD:
pxa3xx_nand: fix nand detection issue".

Hmm, I'm a little late on this one...I was just sorting through a few e-mails and saw this one.

It looks like it made it as far as Artem's l2-mtd-2.6.git:
http://git.infradead.org/users/dedekind/l2-mtd-2.6.git/commit/e119967aad8df3b87834e28249feca696c650647

but things have been sitting there for months now. I believe every so often, dwmw sorts through Artem's l2-mtd-2.6 repo, pulls patches into mtd-2.6, and gets them merged upstream. I think the last time this happened was early June (between v3.0-rc1 and v3.0-rc2). Once such patches hit mainline, then the stable team is alerted and the appropriately-labeled patches are sent to the stable repos. Don't know when this will happen though. I'm kinda wondering myself...

Brian

Artem Bityutskiy

2011-07-22 08:35 UTC #6 lore raw

PATCH 2/2 MTD: pxa3xx_nand: Fix blank page ECC mismatch

Quoted 11 lines · Brian Norris

Quoted 3 lines · Daniel Mack
Has anyone queued this for 3.0 yet? This fixes a serious regression, so we need this merged, along with Lei's other patch titled "MTD:
pxa3xx_nand: fix nand detection issue".
Hmm, I'm a little late on this one...I was just sorting through a few e-mails and saw this one.
It looks like it made it as far as Artem's l2-mtd-2.6.git:
http://git.infradead.org/users/dedekind/l2-mtd-2.6.git/commit/e119967aad8df3b87834e28249feca696c650647

Quoted 7 lines

but things have been sitting there for months now. I believe every so often, dwmw sorts through Artem's l2-mtd-2.6 repo, pulls patches into mtd-2.6, and gets them merged upstream. I think the last time this happened was early June (between v3.0-rc1 and v3.0-rc2). Once such patches hit mainline, then the stable team is alerted and the appropriately-labeled patches are sent to the stable repos. Don't know when this will happen though. I'm kinda wondering myself...

My guess is that this patch will reach upstream only when the linux-3.1 merge window opens.

Signature · 2 lines

Best Regards,
Artem Bityutskiy

Artem Bityutskiy

2011-07-22 09:37 UTC #7 lore raw

PATCH 2/2 MTD: pxa3xx_nand: Fix blank page ECC mismatch

Quoted 2 lines · Artem Bityutskiy

My guess is that this patch will reach upstream only when the linux-3.1 merge window opens.

And it is open now :-)

Signature · 2 lines

Best Regards,
Artem Bityutskiy

Artem Bityutskiy

2011-06-07 11:16 UTC #8 lore raw

PATCH 1/2 MTD: pxa3xx_nand: fix nand detection issue

In reply to Lei Wen

Quoted 9 lines · Lei Wen

When keep_config is set, the detection would goes different routine. That the driver would read out the setting which is set previously by bootloader. While most bootloader keep the irq mask as off, and current driver need all irq default open, keep_config behavior would lead to no irq at all.
Signed-off-by Lei Wen <[email protected]>
Tested-by Daniel Mack <[email protected]>
Cc stable at kernel.org

When the regression was introduced? Commit id?

Signature · 2 lines

Best Regards,
Artem Bityutskiy (????? ????????)

Lei Wen

2011-06-07 12:27 UTC #9 lore raw

PATCH 1/2 MTD: pxa3xx_nand: fix nand detection issue

Hi Artem,

Quoted 7 lines · Artem Bityutskiy

Quoted 6 lines · Lei Wen
When keep_config is set, the detection would goes different routine. That the driver would read out the setting which is set previously by bootloader. While most bootloader keep the irq mask as off, and current driver need all irq default open, keep_config behavior would lead to no irq at all.

Quoted 5 lines

Quoted 3 lines
Signed-off-by Lei Wen <[email protected]>
Tested-by Daniel Mack <[email protected]>
Cc stable at kernel.org
When the regression was introduced? Commit id?

This regression should be introduced by below commit:

commit f8155a404db95656f1519b28fdb96cb68f8b2364

Author: Lei Wen <[email protected]>
Date:   Mon Feb 28 10:32:11 2011 +0800

mtd: pxa3xx_nand: rework irq logic

So its should be:
Cc: stable at kernel.org [2.6.38+]

Best regards,
Lei

Artem Bityutskiy

2011-06-07 12:33 UTC #10 lore raw

PATCH 1/2 MTD: pxa3xx_nand: fix nand detection issue

In reply to Lei Wen

Quoted 9 lines · Lei Wen

When keep_config is set, the detection would goes different routine. That the driver would read out the setting which is set previously by bootloader. While most bootloader keep the irq mask as off, and current driver need all irq default open, keep_config behavior would lead to no irq at all.
Signed-off-by Lei Wen <[email protected]>
Tested-by Daniel Mack <[email protected]>
Cc stable at kernel.org

Pushed both patches to l2-mtd-2.6.git, thanks.

Signature · 2 lines

Best Regards,
Artem Bityutskiy (????? ????????)