Re: [BUG] mtd: cfi_cmdset_0002: write regression since v4.17-rc1

[Tokunori Ikegami <ikegami.t@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 3339 bytes --]

Hi Ahmad-san,

On 2022/03/08 18:44, Ahmad Fatoum wrote:
> Hello Tokunori,
>
> On 06.03.22 16:49, Tokunori Ikegami wrote:
>> Hi,
>>
>> On 2022/03/04 20:11, Ahmad Fatoum wrote:
>>> Hello Tokunori-san,
>>>
>>> On 20.02.22 13:22, Tokunori Ikegami wrote:
>>>> Hi Ahmad-san,
>>>>
>>>> Could you please try the version 2 patch attached for the error case?
>>>> This version is to check the DQ true data 0xFF by chip_good().
>>> I had a similar patch locally as well at first. I just tested yours
>>> and I can't reproduce the issue.
>> Thanks for your support.
>> Sorry if possible could you please retest the attached the patch again since this fixed the version 1 patch maintainer review comments?
> Works good.
>
> Tested-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
Thank you so much for your test.
>
>>>> But I am not sure if this works or not since the error is possible to be caused by Hi-Z 0xff on floating bus or etc.
>>> That it works for me could be because of Hi-Z 0xff, which is why
>>> decided against it.
>> I see.
>>>>>>>> What seems to work for me is checking if chip_good or chip_ready
>>>>>>>> and map_word is equal to 0xFF. I can't justify why this is ok though.
>>>>>>>> (Worst case bus is floating at this point of time and Hi-Z is read
>>>>>>>> as 0xff on CPU data lines...)
>>>>>>> Sorry I am not sure about this.
>>>>>>> I thought the chip_ready() itself is correct as implemented as the data sheet in the past.
>>>>>>> But it did not work correctly so changed to use chip_good() instead as it is also correct.
>>>>>> What exactly in the datasheet makes you believe chip_good is not appropriate?
>>>>> I just mentioned about the actual issue behaviors as not worked chip_good() on S29GL964N and not worked chip_ready() on MX29GL512FHT2I-11G before etc.
>>>>> Anyway let me recheck the data sheet details as just checked it again quickly but needed more investigation to understand.
>>>> As far as I checked still both chip_good() and chip_ready() seem correct but still the root cause is unknown.
>>>> If as you mentioned the issue was cased by the DQ true data 0xFF I am not sure why the read work without any error after the write operation.
>>>> Also if the error was caused by the Hi-Z 0xff on floating bus as mentioned I am not sure why the read work without any error after the write operation with chip_ready().
>>>> Sorry anyway the root cause is also unknown when the write operation was changed to use chip_good() instead of chip_ready().
>>> I've be ok with v1 then. Restores working behavior for me and shouldn't break others.
>> Noted but still I am thinking the version 2 patch to check 0xff seems better than to use chip_ready() so let me consider this again later.
> The original version has less room for surprise as it restores previously
> working behavior. Assuming 0xFF to be good without backing from documentation
> is more risky IMO.
The change to check 0xFF can be limited for the S29GL064N chip do you 
have any comment about this?
Just attached the patch changed as so and thinking to send the patch as 
version 3 to the maintainer if you are okay.

Regards,
Ikegami

>
> Thanks for your continued support,
> Ahmad
>
>> Regards,
>> Ikegami
>>
>>> Cheers and thanks again,
>>> Ahmad
>>>
>>>> Regards,
>>>> Ikegami
>>>>
>>>>> Regards,
>>>>> Ikegami
>>>>>
>>>>>> Cheers,
>>>>>> Ahmad
>>>>>>
>>>>>>
>

[-- Attachment #2: v3-0001-mtd-cfi_cmdset_0002-Change-chip_good-to-check-DQ-.patch --]
[-- Type: text/plain, Size: 5504 bytes --]

From f4e767b4c9b2d5139387175f0c57afd81f0b62de Mon Sep 17 00:00:00 2001
From: Tokunori Ikegami <ikegami.t@gmail.com>
Date: Sat, 19 Feb 2022 19:39:32 +0900
Subject: [PATCH v3] mtd: cfi_cmdset_0002: Change chip_good() to check DQ true
 data 0xFF on S29GL064N

The regression issue has been caused on S29GL064N and reported it.
The change mentioned for regression is to use chip_good() for buffered write.
Also it seems that the 0xFF value is read on the error case.
It is possible to be caused by DQ true data described by S29GL064N datasheet.
So change chip_good() to check DQ true data 0xFF additionally for the error.

Fixes: dfeae1073583("mtd: cfi_cmdset_0002: Change write buffer to check correct value")
Signed-off-by: Tokunori Ikegami <ikegami.t@gmail.com>
Cc: Miquel Raynal <miquel.raynal@bootlin.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Vignesh Raghavendra <vigneshr@ti.com>
Cc: linux-mtd@lists.infradead.org
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/b687c259-6413-26c9-d4c9-b3afa69ea124@pengutronix.de/
---
 drivers/mtd/chips/cfi_cmdset_0002.c | 36 ++++++++++++++++++++---------
 1 file changed, 25 insertions(+), 11 deletions(-)

diff --git a/drivers/mtd/chips/cfi_cmdset_0002.c b/drivers/mtd/chips/cfi_cmdset_0002.c
index a761134fd3be..99c1c6741b69 100644
--- a/drivers/mtd/chips/cfi_cmdset_0002.c
+++ b/drivers/mtd/chips/cfi_cmdset_0002.c
@@ -48,6 +48,7 @@
 #define SST49LF040B		0x0050
 #define SST49LF008A		0x005a
 #define AT49BV6416		0x00d6
+#define S29GL064N_MN12		0x0c01
 
 /*
  * Status Register bit description. Used by flash devices that don't
@@ -462,7 +463,7 @@ static struct cfi_fixup cfi_fixup_table[] = {
 	{ CFI_MFR_AMD, 0x0056, fixup_use_secsi },
 	{ CFI_MFR_AMD, 0x005C, fixup_use_secsi },
 	{ CFI_MFR_AMD, 0x005F, fixup_use_secsi },
-	{ CFI_MFR_AMD, 0x0c01, fixup_s29gl064n_sectors },
+	{ CFI_MFR_AMD, S29GL064N_MN12, fixup_s29gl064n_sectors },
 	{ CFI_MFR_AMD, 0x1301, fixup_s29gl064n_sectors },
 	{ CFI_MFR_AMD, 0x1a00, fixup_s29gl032n_sectors },
 	{ CFI_MFR_AMD, 0x1a01, fixup_s29gl032n_sectors },
@@ -837,6 +838,11 @@ static int __xipram chip_ready(struct map_info *map, struct flchip *chip,
 	return map_word_equal(map, d, t);
 }
 
+static bool __xipram cfi_use_dq_true_data(struct cfi_private *cfi)
+{
+	return cfi->mfr == CFI_MFR_AMD && cfi->id == S29GL064N_MN12;
+}
+
 /*
  * Return true if the chip is ready and has the correct value.
  *
@@ -853,7 +859,7 @@ static int __xipram chip_ready(struct map_info *map, struct flchip *chip,
  *
  */
 static int __xipram chip_good(struct map_info *map, struct flchip *chip,
-			      unsigned long addr, map_word expected)
+			      unsigned long addr, map_word *expected)
 {
 	struct cfi_private *cfi = map->fldrv_priv;
 	map_word oldd, curd;
@@ -875,8 +881,16 @@ static int __xipram chip_good(struct map_info *map, struct flchip *chip,
 	oldd = map_read(map, addr);
 	curd = map_read(map, addr);
 
-	return	map_word_equal(map, oldd, curd) &&
-		map_word_equal(map, curd, expected);
+	if (!map_word_equal(map, oldd, curd))
+		return 0;
+
+	if (expected && map_word_equal(map, curd, *expected))
+		return 1;
+
+	if (cfi_use_dq_true_data(cfi))
+		return map_word_equal(map, oldd, map_word_ff(map));
+
+	return 0;
 }
 
 static int get_chip(struct map_info *map, struct flchip *chip, unsigned long adr, int mode)
@@ -1699,7 +1713,7 @@ static int __xipram do_write_oneword_once(struct map_info *map,
 		 * "chip_good" to avoid the failure due to scheduling.
 		 */
 		if (time_after(jiffies, timeo) &&
-		    !chip_good(map, chip, adr, datum)) {
+		    !chip_good(map, chip, adr, &datum)) {
 			xip_enable(map, chip, adr);
 			printk(KERN_WARNING "MTD %s(): software timeout\n", __func__);
 			xip_disable(map, chip, adr);
@@ -1707,7 +1721,7 @@ static int __xipram do_write_oneword_once(struct map_info *map,
 			break;
 		}
 
-		if (chip_good(map, chip, adr, datum)) {
+		if (chip_good(map, chip, adr, &datum)) {
 			if (cfi_check_err_status(map, chip, adr))
 				ret = -EIO;
 			break;
@@ -1979,14 +1993,14 @@ static int __xipram do_write_buffer_wait(struct map_info *map,
 		 * "chip_good" to avoid the failure due to scheduling.
 		 */
 		if (time_after(jiffies, timeo) &&
-		    !chip_good(map, chip, adr, datum)) {
+		    !chip_good(map, chip, adr, &datum)) {
 			pr_err("MTD %s(): software timeout, address:0x%.8lx.\n",
 			       __func__, adr);
 			ret = -EIO;
 			break;
 		}
 
-		if (chip_good(map, chip, adr, datum)) {
+		if (chip_good(map, chip, adr, &datum)) {
 			if (cfi_check_err_status(map, chip, adr))
 				ret = -EIO;
 			break;
@@ -2282,7 +2296,7 @@ static int do_panic_write_oneword(struct map_info *map, struct flchip *chip,
 		udelay(1);
 	}
 
-	if (!chip_good(map, chip, adr, datum) ||
+	if (!chip_good(map, chip, adr, &datum) ||
 	    cfi_check_err_status(map, chip, adr)) {
 		/* reset on all failures. */
 		map_write(map, CMD(0xF0), chip->start);
@@ -2478,7 +2492,7 @@ static int __xipram do_erase_chip(struct map_info *map, struct flchip *chip)
 			chip->erase_suspended = 0;
 		}
 
-		if (chip_good(map, chip, adr, map_word_ff(map))) {
+		if (chip_good(map, chip, adr, NULL)) {
 			if (cfi_check_err_status(map, chip, adr))
 				ret = -EIO;
 			break;
@@ -2577,7 +2591,7 @@ static int __xipram do_erase_oneblock(struct map_info *map, struct flchip *chip,
 			chip->erase_suspended = 0;
 		}
 
-		if (chip_good(map, chip, adr, map_word_ff(map))) {
+		if (chip_good(map, chip, adr, NULL)) {
 			if (cfi_check_err_status(map, chip, adr))
 				ret = -EIO;
 			break;
-- 
2.32.0