Skip to contentSkip to conversation

mtd: rawnand: marvell: select target before setting up data interface

PATCH In review Daniel Mack 3 participants 10 messages Mar 21, 2019 – Apr 5, 2019

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 mtd: rawnand: marvell: select target before setting up data interface

Now that the nand drivers are responsible for selecting the target prior to hardware access, a call to marvell_nfc_select_target() is necessary from marvell_nfc_setup_data_interface().

This is a regression introduced by commit b25251414f6e ("mtd:
rawnand: marvell: Stop implementing ->select_chip()").

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
Cc Boris Brezillon <[email protected]>
---
 drivers/mtd/nand/raw/marvell_nand.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mtd/nand/raw/marvell_nand.c +2 -0
index 84283c6bb0ff..93abe5be8995 100644
@@ -2325,6 +2325,8 @@ static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr,
 	struct marvell_nfc_timings nfc_tmg;
 	int read_delay;
 
+	marvell_nfc_select_target(chip, chip->cur_cs);
+
 	sdr = nand_get_sdr_timings(conf);
 	if (IS_ERR(sdr))
 		return PTR_ERR(sdr);
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
maintainer Acked-by on v1 1/1
#1 lore raw

Hi Daniel,

+ Richard: this is fixes materials

Daniel Mack <[email protected]> wrote on Thu, 21 Mar 2019 22:20:10 +0100:

Quoted 20 lines

Now that the nand drivers are responsible for selecting the target prior to hardware access, a call to marvell_nfc_select_target() is necessary from marvell_nfc_setup_data_interface().

This is a regression introduced by commit b25251414f6e ("mtd:
rawnand: marvell: Stop implementing ->select_chip()").

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
Cc Boris Brezillon <[email protected]>
---
 drivers/mtd/nand/raw/marvell_nand.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mtd/nand/raw/marvell_nand.c +0 -0
index 84283c6bb0ff..93abe5be8995 100644

drivers/mtd/nand/raw/marvell_nand.c static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr, · line 2332

Quoted 8 lines
drivers/mtd/nand/raw/marvell_nand.c +2 -0
@@ -2325,6 +2325,8 @@ static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr,
 	struct marvell_nfc_timings nfc_tmg;
 	int read_delay;
 
+	marvell_nfc_select_target(chip, chip->cur_cs);
+
 	sdr = nand_get_sdr_timings(conf);
 	if (IS_ERR(sdr))
 		return PTR_ERR(sdr);

I had a short look yesterday and this is very likely to be the root cause of the regression, thank you very much!

Thanks,
Miquèl

#2 lore raw
Quoted 30 lines · On Fri, 22 Mar 2019 09:30:46 +0100 Miquel Raynal <[email protected]>…

Hi Daniel,

+ Richard: this is fixes materials

Daniel Mack <[email protected]> wrote on Thu, 21 Mar 2019 22:20:10 +0100:

Quoted 23 lines

Now that the nand drivers are responsible for selecting the target prior to hardware access, a call to marvell_nfc_select_target() is necessary from marvell_nfc_setup_data_interface().

This is a regression introduced by commit b25251414f6e ("mtd:
rawnand: marvell: Stop implementing ->select_chip()").

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
Cc Boris Brezillon <[email protected]>
---
 drivers/mtd/nand/raw/marvell_nand.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mtd/nand/raw/marvell_nand.c +0 -0
index 84283c6bb0ff..93abe5be8995 100644
@@ -2325,6 +2325,8 @@ static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr,
 	struct marvell_nfc_timings nfc_tmg;
 	int read_delay;
 
Quoted 1 line

drivers/mtd/nand/raw/marvell_nand.c static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr, · line 2328

Quoted 1 line
drivers/mtd/nand/raw/marvell_nand.c +1 -0
+	marvell_nfc_select_target(chip, chip->cur_cs);

Maybe s/chip->cur_cs/chipnr/

Quoted 13 lines

drivers/mtd/nand/raw/marvell_nand.c static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr, · line 2332

Quoted 4 lines
drivers/mtd/nand/raw/marvell_nand.c +1 -0
+
 	sdr = nand_get_sdr_timings(conf);
 	if (IS_ERR(sdr))
 		return PTR_ERR(sdr);  

I had a short look yesterday and this is very likely to be the root cause of the regression, thank you very much!

Thanks,
Miquèl

#3 lore raw
In reply to Miquel Raynal

Hi Miquel,

Quoted 28 lines · Miquel Raynal

+ Richard: this is fixes materials

Daniel Mack <[email protected]> wrote on Thu, 21 Mar 2019 22:20:10 +0100:

Quoted 23 lines

Now that the nand drivers are responsible for selecting the target prior to hardware access, a call to marvell_nfc_select_target() is necessary from marvell_nfc_setup_data_interface().

This is a regression introduced by commit b25251414f6e ("mtd:
rawnand: marvell: Stop implementing ->select_chip()").

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
Cc Boris Brezillon <[email protected]>
---
 drivers/mtd/nand/raw/marvell_nand.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mtd/nand/raw/marvell_nand.c +0 -0
index 84283c6bb0ff..93abe5be8995 100644
@@ -2325,6 +2325,8 @@ static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr,
 	struct marvell_nfc_timings nfc_tmg;
 	int read_delay;
 
Quoted 8 lines

drivers/mtd/nand/raw/marvell_nand.c static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr, · line 2332

Quoted 5 lines
drivers/mtd/nand/raw/marvell_nand.c +2 -0
+	marvell_nfc_select_target(chip, chip->cur_cs);
+
 	sdr = nand_get_sdr_timings(conf);
 	if (IS_ERR(sdr))
 		return PTR_ERR(sdr);

I had a short look yesterday and this is very likely to be the root cause of the regression, thank you very much!

Yes it is, it fixes the regression I have reported.

Do you want me to resend with that nit addressed you mentioned? My version works well, I know that much :)

Thanks,
Daniel

maintainer
#4 lore raw

Hi Daniel,

Daniel Mack <[email protected]> wrote on Fri, 22 Mar 2019 19:48:08 +0100:

Quoted 40 lines

Hi Miquel,

Quoted 28 lines · Miquel Raynal

+ Richard: this is fixes materials

Daniel Mack <[email protected]> wrote on Thu, 21 Mar 2019 22:20:10 +0100:

Quoted 23 lines

Now that the nand drivers are responsible for selecting the target prior to hardware access, a call to marvell_nfc_select_target() is necessary from marvell_nfc_setup_data_interface().

This is a regression introduced by commit b25251414f6e ("mtd:
rawnand: marvell: Stop implementing ->select_chip()").

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
Cc Boris Brezillon <[email protected]>
---
 drivers/mtd/nand/raw/marvell_nand.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mtd/nand/raw/marvell_nand.c +0 -0
index 84283c6bb0ff..93abe5be8995 100644
@@ -2325,6 +2325,8 @@ static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr,
 	struct marvell_nfc_timings nfc_tmg;
 	int read_delay;
 
Quoted 8 lines

drivers/mtd/nand/raw/marvell_nand.c static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr, · line 2332

Quoted 5 lines
drivers/mtd/nand/raw/marvell_nand.c +2 -0
+	marvell_nfc_select_target(chip, chip->cur_cs);
+
 	sdr = nand_get_sdr_timings(conf);
 	if (IS_ERR(sdr))
 		return PTR_ERR(sdr);  

I had a short look yesterday and this is very likely to be the root cause of the regression, thank you very much!

Quoted 4 lines

Yes it is, it fixes the regression I have reported.

Do you want me to resend with that nit addressed you mentioned? My version works well, I know that much :)

Yes please, resent with Boris' comment addressed and we'll be fine.

Thanks,
Miquèl

#5 lore raw

Hi Miquel,

Quoted 43 lines · Miquel Raynal

Daniel Mack <[email protected]> wrote on Fri, 22 Mar 2019 19:48:08 +0100:

Quoted 40 lines

Hi Miquel,

Quoted 36 lines · Miquel Raynal

+ Richard: this is fixes materials

Daniel Mack <[email protected]> wrote on Thu, 21 Mar 2019 22:20:10 +0100:

Quoted 20 lines

Now that the nand drivers are responsible for selecting the target prior to hardware access, a call to marvell_nfc_select_target() is necessary from marvell_nfc_setup_data_interface().

This is a regression introduced by commit b25251414f6e ("mtd:
rawnand: marvell: Stop implementing ->select_chip()").

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
Cc Boris Brezillon <[email protected]>
---
 drivers/mtd/nand/raw/marvell_nand.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mtd/nand/raw/marvell_nand.c +0 -0
index 84283c6bb0ff..93abe5be8995 100644

drivers/mtd/nand/raw/marvell_nand.c static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr, · line 2332

Quoted 8 lines
drivers/mtd/nand/raw/marvell_nand.c +2 -0
@@ -2325,6 +2325,8 @@ static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr,
 	struct marvell_nfc_timings nfc_tmg;
 	int read_delay;
 
+	marvell_nfc_select_target(chip, chip->cur_cs);
+
 	sdr = nand_get_sdr_timings(conf);
 	if (IS_ERR(sdr))
 		return PTR_ERR(sdr);  

I had a short look yesterday and this is very likely to be the root cause of the regression, thank you very much!

Quoted 6 lines
Quoted 4 lines

Yes it is, it fixes the regression I have reported.

Do you want me to resend with that nit addressed you mentioned? My version works well, I know that much :)

Yes please, resent with Boris' comment addressed and we'll be fine.

I tried this now, and with 'chip->cur_cs' instead of 'chipnr', the regression is back. This function is invoked twice during boot, and in both cases, 'chip->cur_cs' is -1, while 'chipnr' is 0. The 2nd time this function is invoked, the 'selected_die' cache value is 0, so marvell_nfc_select_target() bails early, which seems to be a problem.

I can't quite make sense of that yet, all I know is that calling marvell_nfc_select_target() with -1 from marvell_nfc_setup_data_interface() makes the driver functional again. Hmm.

Thanks,
Daniel

#6 lore raw
Quoted 59 lines · On Mon, 25 Mar 2019 17:21:42 +0100 Daniel Mack <[email protected]> wrote:

Hi Miquel,

Quoted 43 lines · Miquel Raynal

Daniel Mack <[email protected]> wrote on Fri, 22 Mar 2019 19:48:08 +0100:

Quoted 40 lines

Hi Miquel,

Quoted 36 lines · Miquel Raynal

+ Richard: this is fixes materials

Daniel Mack <[email protected]> wrote on Thu, 21 Mar 2019 22:20:10 +0100:

Quoted 20 lines

Now that the nand drivers are responsible for selecting the target prior to hardware access, a call to marvell_nfc_select_target() is necessary from marvell_nfc_setup_data_interface().

This is a regression introduced by commit b25251414f6e ("mtd:
rawnand: marvell: Stop implementing ->select_chip()").

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
Cc Boris Brezillon <[email protected]>
---
 drivers/mtd/nand/raw/marvell_nand.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mtd/nand/raw/marvell_nand.c +0 -0
index 84283c6bb0ff..93abe5be8995 100644

drivers/mtd/nand/raw/marvell_nand.c static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr, · line 2332

Quoted 8 lines
drivers/mtd/nand/raw/marvell_nand.c +2 -0
@@ -2325,6 +2325,8 @@ static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr,
 	struct marvell_nfc_timings nfc_tmg;
 	int read_delay;
 
+	marvell_nfc_select_target(chip, chip->cur_cs);
+
 	sdr = nand_get_sdr_timings(conf);
 	if (IS_ERR(sdr))
 		return PTR_ERR(sdr);    

I had a short look yesterday and this is very likely to be the root cause of the regression, thank you very much!

Quoted 6 lines
Quoted 4 lines

Yes it is, it fixes the regression I have reported.

Do you want me to resend with that nit addressed you mentioned? My version works well, I know that much :)

Yes please, resent with Boris' comment addressed and we'll be fine.

I tried this now, and with 'chip->cur_cs' instead of 'chipnr', the regression is back. This function is invoked twice during boot, and in both cases, 'chip->cur_cs' is -1, while 'chipnr' is 0. The 2nd time this function is invoked, the 'selected_die' cache value is 0, so marvell_nfc_select_target() bails early, which seems to be a problem.

Quoted 3 lines

I can't quite make sense of that yet, all I know is that calling marvell_nfc_select_target() with -1 from marvell_nfc_setup_data_interface() makes the driver functional again. Hmm.

I had a quick look at the marvell_nfc_setup_data_interface() function and it doesn't make sense to call marvell_nfc_select_target() there. The fact that calling marvell_nfc_select_target(cur_cs) fixes your problem is probably a side effect, and the bug might come from other changes I made in b25251414f6e (missing call to marvell_nfc_select_target() somewhere else in the driver?).

#7 lore raw
In reply to Daniel Mack
Quoted 59 lines · On Mon, 25 Mar 2019 17:21:42 +0100 Daniel Mack <[email protected]> wrote:

Hi Miquel,

Quoted 43 lines · Miquel Raynal

Daniel Mack <[email protected]> wrote on Fri, 22 Mar 2019 19:48:08 +0100:

Quoted 40 lines

Hi Miquel,

Quoted 36 lines · Miquel Raynal

+ Richard: this is fixes materials

Daniel Mack <[email protected]> wrote on Thu, 21 Mar 2019 22:20:10 +0100:

Quoted 20 lines

Now that the nand drivers are responsible for selecting the target prior to hardware access, a call to marvell_nfc_select_target() is necessary from marvell_nfc_setup_data_interface().

This is a regression introduced by commit b25251414f6e ("mtd:
rawnand: marvell: Stop implementing ->select_chip()").

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
Cc Boris Brezillon <[email protected]>
---
 drivers/mtd/nand/raw/marvell_nand.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mtd/nand/raw/marvell_nand.c +0 -0
index 84283c6bb0ff..93abe5be8995 100644

drivers/mtd/nand/raw/marvell_nand.c static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr, · line 2332

Quoted 8 lines
drivers/mtd/nand/raw/marvell_nand.c +2 -0
@@ -2325,6 +2325,8 @@ static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr,
 	struct marvell_nfc_timings nfc_tmg;
 	int read_delay;
 
+	marvell_nfc_select_target(chip, chip->cur_cs);
+
 	sdr = nand_get_sdr_timings(conf);
 	if (IS_ERR(sdr))
 		return PTR_ERR(sdr);    

I had a short look yesterday and this is very likely to be the root cause of the regression, thank you very much!

Quoted 6 lines
Quoted 4 lines

Yes it is, it fixes the regression I have reported.

Do you want me to resend with that nit addressed you mentioned? My version works well, I know that much :)

Yes please, resent with Boris' comment addressed and we'll be fine.

I tried this now, and with 'chip->cur_cs' instead of 'chipnr', the regression is back. This function is invoked twice during boot, and in both cases, 'chip->cur_cs' is -1, while 'chipnr' is 0. The 2nd time this function is invoked, the 'selected_die' cache value is 0, so marvell_nfc_select_target() bails early, which seems to be a problem.

Quoted 3 lines

I can't quite make sense of that yet, all I know is that calling marvell_nfc_select_target() with -1 from marvell_nfc_setup_data_interface() makes the driver functional again. Hmm.

Can you try with the following diff applied?
--->8---

drivers/mtd/nand/raw/marvell_nand.c +0 -3
index f38e5c1b87e4..e8eccf6afc8f 100644
@@ -722,9 +722,6 @@ static void marvell_nfc_select_target(struct nand_chip *chip,
        struct marvell_nfc *nfc = to_marvell_nfc(chip->controller);
        u32 ndcr_generic;
 
-       if (chip == nfc->selected_chip && die_nr == marvell_nand->selected_die)
-               return;
-
        writel_relaxed(marvell_nand->ndtr0, nfc->regs + NDTR0);
        writel_relaxed(marvell_nand->ndtr1, nfc->regs + NDTR1);
 
author Tested-by
#8 lore raw

Hi Boris,

Quoted 14 lines · Boris Brezillon
Quoted 3 lines · On Mon, 25 Mar 2019 17:21:42 +0100 Daniel Mack <[email protected]> wrote:

I can't quite make sense of that yet, all I know is that calling marvell_nfc_select_target() with -1 from marvell_nfc_setup_data_interface() makes the driver functional again. Hmm.

Can you try with the following diff applied?
--->8---

drivers/mtd/nand/raw/marvell_nand.c +0 -0
index f38e5c1b87e4..e8eccf6afc8f 100644

drivers/mtd/nand/raw/marvell_nand.c static void marvell_nfc_select_target(struct nand_chip *chip, · line 726

Quoted 8 lines
drivers/mtd/nand/raw/marvell_nand.c +0 -3
@@ -722,9 +722,6 @@ static void marvell_nfc_select_target(struct nand_chip *chip,
        struct marvell_nfc *nfc = to_marvell_nfc(chip->controller);
        u32 ndcr_generic;
 
-       if (chip == nfc->selected_chip && die_nr == marvell_nand->selected_die)
-               return;
-
        writel_relaxed(marvell_nand->ndtr0, nfc->regs + NDTR0);
        writel_relaxed(marvell_nand->ndtr1, nfc->regs + NDTR1);

Yes, this works. Thanks a lot!

Do you want to submit a patch for that? You can have my

Best regards,
Daniel

maintainer
#9 lore raw
In reply to Boris Brezillon

Hi Daniel,

Boris Brezillon <[email protected]> wrote on Sat, 30 Mar 2019 19:20:36 +0100:

Quoted 73 lines
Quoted 59 lines · On Mon, 25 Mar 2019 17:21:42 +0100 Daniel Mack <[email protected]> wrote:

Hi Miquel,

Quoted 43 lines · Miquel Raynal

Daniel Mack <[email protected]> wrote on Fri, 22 Mar 2019 19:48:08 +0100:

Quoted 40 lines

Hi Miquel,

Quoted 36 lines · Miquel Raynal

+ Richard: this is fixes materials

Daniel Mack <[email protected]> wrote on Thu, 21 Mar 2019 22:20:10 +0100:

Quoted 20 lines

Now that the nand drivers are responsible for selecting the target prior to hardware access, a call to marvell_nfc_select_target() is necessary from marvell_nfc_setup_data_interface().

This is a regression introduced by commit b25251414f6e ("mtd:
rawnand: marvell: Stop implementing ->select_chip()").

Fixes b25251414f6e ("mtd: rawnand: marvell: Stop implementing ->select_chip()")
Cc Boris Brezillon <[email protected]>
---
 drivers/mtd/nand/raw/marvell_nand.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/mtd/nand/raw/marvell_nand.c +0 -0
index 84283c6bb0ff..93abe5be8995 100644

drivers/mtd/nand/raw/marvell_nand.c static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr, · line 2332

Quoted 8 lines
drivers/mtd/nand/raw/marvell_nand.c +2 -0
@@ -2325,6 +2325,8 @@ static int marvell_nfc_setup_data_interface(struct nand_chip *chip, int chipnr,
 	struct marvell_nfc_timings nfc_tmg;
 	int read_delay;
 
+	marvell_nfc_select_target(chip, chip->cur_cs);
+
 	sdr = nand_get_sdr_timings(conf);
 	if (IS_ERR(sdr))
 		return PTR_ERR(sdr);      

I had a short look yesterday and this is very likely to be the root cause of the regression, thank you very much!

Quoted 6 lines
Quoted 4 lines

Yes it is, it fixes the regression I have reported.

Do you want me to resend with that nit addressed you mentioned? My version works well, I know that much :)

Yes please, resent with Boris' comment addressed and we'll be fine.

I tried this now, and with 'chip->cur_cs' instead of 'chipnr', the regression is back. This function is invoked twice during boot, and in both cases, 'chip->cur_cs' is -1, while 'chipnr' is 0. The 2nd time this function is invoked, the 'selected_die' cache value is 0, so marvell_nfc_select_target() bails early, which seems to be a problem.

Quoted 3 lines

I can't quite make sense of that yet, all I know is that calling marvell_nfc_select_target() with -1 from marvell_nfc_setup_data_interface() makes the driver functional again. Hmm.

Can you try with the following diff applied?
--->8---

drivers/mtd/nand/raw/marvell_nand.c +0 -0
index f38e5c1b87e4..e8eccf6afc8f 100644

drivers/mtd/nand/raw/marvell_nand.c static void marvell_nfc_select_target(struct nand_chip *chip, · line 727

Quoted 9 lines
drivers/mtd/nand/raw/marvell_nand.c +0 -3
@@ -722,9 +722,6 @@ static void marvell_nfc_select_target(struct nand_chip *chip,
        struct marvell_nfc *nfc = to_marvell_nfc(chip->controller);
        u32 ndcr_generic;
 
-       if (chip == nfc->selected_chip && die_nr == marvell_nand->selected_die)
-               return;
-
        writel_relaxed(marvell_nand->ndtr0, nfc->regs + NDTR0);
        writel_relaxed(marvell_nand->ndtr1, nfc->regs + NDTR1);
 

I tested on a pxa board and an armada board and I don't reproduce the bug without this change. What is special with your setup?

I was trying to understand the fix to write a decent commit log but as I cannot reproduce I am stuck. So if you need this fix please send the patch with, hopefully, an explanation of what's wrong.

Thank you!
Miquèl