From: Aleksei Mamlin <hidden> Date: 2016-06-06 10:25:08
Add the full description of the Hynix H27UBG8T2BTR-BC NAND chip in the
nand_ids table so that we can later use the NAND ECC infos and ONFI timings
mode in controller drivers.
Signed-off-by: Aleksei Mamlin <redacted>
---
drivers/mtd/nand/nand_ids.c | 4 ++++
1 file changed, 4 insertions(+)
From: Aleksei Mamlin <hidden> Date: 2016-06-06 10:25:29
Enable the NFC and describe the NAND flash connected to this controller.
Signed-off-by: Aleksei Mamlin <redacted>
---
arch/arm/boot/dts/sun7i-a20-wexler-tab7200.dts | 41 ++++++++++++++++++++++++++
1 file changed, 41 insertions(+)
From: Aleksei Mamlin <hidden> Date: 2016-06-06 10:26:52
Enable the NFC and describe the NAND flash connected to this controller.
Signed-off-by: Aleksei Mamlin <redacted>
---
arch/arm/boot/dts/sun4i-a10-marsboard.dts | 41 +++++++++++++++++++++++++++++++
1 file changed, 41 insertions(+)
From: Boris Brezillon <hidden> Date: 2016-06-06 18:55:54
On Mon, 6 Jun 2016 13:24:22 +0300
Aleksei Mamlin [off-list ref] wrote:
Add the full description of the Hynix H27UBG8T2BTR-BC NAND chip in the
nand_ids table so that we can later use the NAND ECC infos and ONFI timings
mode in controller drivers.
Still hoping to get this series [1] merged in 4.8, but if that's
not the case, I'll apply your patch.
BTW, that would be great if you could test it on your platforms.
Regards,
Boris
[1]https://lkml.org/lkml/2016/5/27/264
From: Aleksei Mamlin <hidden> Date: 2016-06-06 19:59:11
On Mon, 6 Jun 2016 20:55:49 +0200
Boris Brezillon [off-list ref] wrote:
On Mon, 6 Jun 2016 13:24:22 +0300
Aleksei Mamlin [off-list ref] wrote:
quoted
Add the full description of the Hynix H27UBG8T2BTR-BC NAND chip in the
nand_ids table so that we can later use the NAND ECC infos and ONFI timings
mode in controller drivers.
Still hoping to get this series [1] merged in 4.8, but if that's
not the case, I'll apply your patch.
BTW, that would be great if you could test it on your platforms.
It seems that Hynix-specific initialization code can't handle H27UBG8T2BTR-BC
chip:
[ 0.886153] nand: Could not find valid ONFI parameter page; aborting
[ 0.892665] nand: device found, Manufacturer ID: 0xad, Chip ID: 0xd7
[ 0.899025] nand: Hynix 1c03000.nand
[ 0.902596] nand: bus width 8 instead 16 bit
[ 0.906858] nand: No NAND device found
[ 0.910620] sunxi_nand 1c03000.nand: failed to init nand chips
[ 0.916528] sunxi_nand: probe of 1c03000.nand failed with error -22
From: Boris Brezillon <hidden> Date: 2016-06-06 20:31:54
On Mon, 6 Jun 2016 22:59:03 +0300
Aleksei Mamlin [off-list ref] wrote:
On Mon, 6 Jun 2016 20:55:49 +0200
Boris Brezillon [off-list ref] wrote:
quoted
On Mon, 6 Jun 2016 13:24:22 +0300
Aleksei Mamlin [off-list ref] wrote:
quoted
Add the full description of the Hynix H27UBG8T2BTR-BC NAND chip in the
nand_ids table so that we can later use the NAND ECC infos and ONFI timings
mode in controller drivers.
Still hoping to get this series [1] merged in 4.8, but if that's
not the case, I'll apply your patch.
BTW, that would be great if you could test it on your platforms.
It seems that Hynix-specific initialization code can't handle H27UBG8T2BTR-BC
chip:
[ 0.886153] nand: Could not find valid ONFI parameter page; aborting
[ 0.892665] nand: device found, Manufacturer ID: 0xad, Chip ID: 0xd7
[ 0.899025] nand: Hynix 1c03000.nand
[ 0.902596] nand: bus width 8 instead 16 bit
[ 0.906858] nand: No NAND device found
[ 0.910620] sunxi_nand 1c03000.nand: failed to init nand chips
[ 0.916528] sunxi_nand: probe of 1c03000.nand failed with error -22
From: Aleksei Mamlin <hidden> Date: 2016-06-06 21:06:52
On Mon, 6 Jun 2016 22:31:38 +0200
Boris Brezillon [off-list ref] wrote:
On Mon, 6 Jun 2016 22:59:03 +0300
Aleksei Mamlin [off-list ref] wrote:
quoted
On Mon, 6 Jun 2016 20:55:49 +0200
Boris Brezillon [off-list ref] wrote:
quoted
On Mon, 6 Jun 2016 13:24:22 +0300
Aleksei Mamlin [off-list ref] wrote:
quoted
Add the full description of the Hynix H27UBG8T2BTR-BC NAND chip in the
nand_ids table so that we can later use the NAND ECC infos and ONFI timings
mode in controller drivers.
Still hoping to get this series [1] merged in 4.8, but if that's
not the case, I'll apply your patch.
BTW, that would be great if you could test it on your platforms.
It seems that Hynix-specific initialization code can't handle H27UBG8T2BTR-BC
chip:
[ 0.886153] nand: Could not find valid ONFI parameter page; aborting
[ 0.892665] nand: device found, Manufacturer ID: 0xad, Chip ID: 0xd7
[ 0.899025] nand: Hynix 1c03000.nand
[ 0.902596] nand: bus width 8 instead 16 bit
[ 0.906858] nand: No NAND device found
[ 0.910620] sunxi_nand 1c03000.nand: failed to init nand chips
[ 0.916528] sunxi_nand: probe of 1c03000.nand failed with error -22
From: Boris Brezillon <hidden> Date: 2016-06-07 05:48:25
On Tue, 7 Jun 2016 00:06:45 +0300
Aleksei Mamlin [off-list ref] wrote:
On Mon, 6 Jun 2016 22:31:38 +0200
Boris Brezillon [off-list ref] wrote:
quoted
On Mon, 6 Jun 2016 22:59:03 +0300
Aleksei Mamlin [off-list ref] wrote:
quoted
On Mon, 6 Jun 2016 20:55:49 +0200
Boris Brezillon [off-list ref] wrote:
quoted
On Mon, 6 Jun 2016 13:24:22 +0300
Aleksei Mamlin [off-list ref] wrote:
quoted
Add the full description of the Hynix H27UBG8T2BTR-BC NAND chip in the
nand_ids table so that we can later use the NAND ECC infos and ONFI timings
mode in controller drivers.
Still hoping to get this series [1] merged in 4.8, but if that's
not the case, I'll apply your patch.
BTW, that would be great if you could test it on your platforms.
It seems that Hynix-specific initialization code can't handle H27UBG8T2BTR-BC
chip:
[ 0.886153] nand: Could not find valid ONFI parameter page; aborting
[ 0.892665] nand: device found, Manufacturer ID: 0xad, Chip ID: 0xd7
[ 0.899025] nand: Hynix 1c03000.nand
[ 0.902596] nand: bus width 8 instead 16 bit
[ 0.906858] nand: No NAND device found
[ 0.910620] sunxi_nand 1c03000.nand: failed to init nand chips
[ 0.916528] sunxi_nand: probe of 1c03000.nand failed with error -22
Yes, it fixes problem. What about ONFI parameter page? Should it be ignored?
It should be ignored: your NAND is not 'ONFI compatible', the datasheet
just says that it supports the' ONFI command set', which is not the
same :).
The trace you're seeing here is just notifying that the core failed to
detect an ONFI NAND, which is expected in your case.
[ 0.886068] nand: Could not find valid ONFI parameter page; aborting
[ 0.892571] nand: device found, Manufacturer ID: 0xad, Chip ID: 0xd7
[ 0.898917] nand: Hynix NAND 4GiB 3,3V 8-bit
[ 0.903198] nand: 4096 MiB, MLC, erase size: 2048 KiB, page size: 8192, OOB size: 640
[ 0.911908] Bad block table found at page 524032, version 0x01
[ 0.918534] Bad block table found at page 523776, version 0x01
[ 0.972112] 5 ofpart partitions found on MTD device 1c03000.nand
[ 0.978116] Creating 5 MTD partitions on "1c03000.nand":
[ 0.983477] 0x000000000000-0x000000200000 : "boot0"
[ 0.988803] 0x000000200000-0x000000400000 : "boot0-rescue"
[ 0.994692] 0x000000400000-0x000000600000 : "uboot"
[ 0.999944] 0x000000600000-0x000000800000 : "uboot-rescue"
[ 1.005763] 0x000000800000-0x000100000000 : "main"
From: Boris Brezillon <hidden> Date: 2016-06-07 05:49:54
On Mon, 6 Jun 2016 22:31:38 +0200
Boris Brezillon [off-list ref] wrote:
On Mon, 6 Jun 2016 22:59:03 +0300
Aleksei Mamlin [off-list ref] wrote:
quoted
On Mon, 6 Jun 2016 20:55:49 +0200
Boris Brezillon [off-list ref] wrote:
quoted
On Mon, 6 Jun 2016 13:24:22 +0300
Aleksei Mamlin [off-list ref] wrote:
quoted
Add the full description of the Hynix H27UBG8T2BTR-BC NAND chip in the
nand_ids table so that we can later use the NAND ECC infos and ONFI timings
mode in controller drivers.
Still hoping to get this series [1] merged in 4.8, but if that's
not the case, I'll apply your patch.
BTW, that would be great if you could test it on your platforms.
It seems that Hynix-specific initialization code can't handle H27UBG8T2BTR-BC
chip:
[ 0.886153] nand: Could not find valid ONFI parameter page; aborting
[ 0.892665] nand: device found, Manufacturer ID: 0xad, Chip ID: 0xd7
[ 0.899025] nand: Hynix 1c03000.nand
[ 0.902596] nand: bus width 8 instead 16 bit
[ 0.906858] nand: No NAND device found
[ 0.910620] sunxi_nand 1c03000.nand: failed to init nand chips
[ 0.916528] sunxi_nand: probe of 1c03000.nand failed with error -22
Can you try this patch? It should fix the problem [1].
Brian, I have a question regarding the extended NAND ids (not full-ids)
defined in the nand_ids table. Are they really valid for all vendors?
If that's the case, why are we extracting the bus width from the id[3]
since we already have this information in the options field?
From: Maxime Ripard <hidden> Date: 2016-06-08 22:03:18
Hi,
On Mon, Jun 06, 2016 at 01:24:24PM +0300, Aleksei Mamlin wrote:
quoted hunk
Enable the NFC and describe the NAND flash connected to this controller.
Signed-off-by: Aleksei Mamlin <redacted>
---
arch/arm/boot/dts/sun7i-a20-wexler-tab7200.dts | 41 ++++++++++++++++++++++++++
1 file changed, 41 insertions(+)
This feels a bit premature. The two boards you're using have an MLC
NAND which is not supported yet. Until there's proper MLC support in
the kernel, I'm not sure we want to enable that for end-users when we
know that things will get wrong.
However, after discussing this with Boris, I appreciate that we don't
have any example available because of this policy for people that want
to opt-in anyway.
What we could do is to still mark the status as disabled in the DTS,
with a big fat warning as comment just before, so that it requires
user action, and that user will have been warned.
Would that work for you?
Maxime
--
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160609/23129dab/attachment.sig>
We usually enable only the pin groups that are actually used by some
board to avoid bloating the DT too much.
And the nodes should be sorted alphabetically.
Thanks!
Maxime
--
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160609/0da21566/attachment-0001.sig>
From: Aleksei Mamlin <hidden> Date: 2016-06-09 08:11:41
On Thu, 9 Jun 2016 00:03:14 +0200
Maxime Ripard [off-list ref] wrote:
Hi,
On Mon, Jun 06, 2016 at 01:24:24PM +0300, Aleksei Mamlin wrote:
quoted
Enable the NFC and describe the NAND flash connected to this controller.
Signed-off-by: Aleksei Mamlin <redacted>
---
arch/arm/boot/dts/sun7i-a20-wexler-tab7200.dts | 41 ++++++++++++++++++++++++++
1 file changed, 41 insertions(+)
This feels a bit premature. The two boards you're using have an MLC
NAND which is not supported yet. Until there's proper MLC support in
the kernel, I'm not sure we want to enable that for end-users when we
know that things will get wrong.
However, after discussing this with Boris, I appreciate that we don't
have any example available because of this policy for people that want
to opt-in anyway.
What we could do is to still mark the status as disabled in the DTS,
with a big fat warning as comment just before, so that it requires
user action, and that user will have been warned.
Would that work for you?
Maxime
Let's drop this two patches with boards dts changes until we get proper MLC
support in kernel.
As for examples - I'll update linux-sunxi wiki with how to enable NAND
controller and describe NAND chips in DTS. People who want to add NAND support
to their boards will be warned that it is risky for now.
--
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
From: Maxime Ripard <hidden> Date: 2016-06-10 08:56:38
On Thu, Jun 09, 2016 at 11:11:00AM +0300, Aleksei Mamlin wrote:
On Thu, 9 Jun 2016 00:03:14 +0200
Maxime Ripard [off-list ref] wrote:
quoted
Hi,
On Mon, Jun 06, 2016 at 01:24:24PM +0300, Aleksei Mamlin wrote:
quoted
Enable the NFC and describe the NAND flash connected to this controller.
Signed-off-by: Aleksei Mamlin <redacted>
---
arch/arm/boot/dts/sun7i-a20-wexler-tab7200.dts | 41 ++++++++++++++++++++++++++
1 file changed, 41 insertions(+)
This feels a bit premature. The two boards you're using have an MLC
NAND which is not supported yet. Until there's proper MLC support in
the kernel, I'm not sure we want to enable that for end-users when we
know that things will get wrong.
However, after discussing this with Boris, I appreciate that we don't
have any example available because of this policy for people that want
to opt-in anyway.
What we could do is to still mark the status as disabled in the DTS,
with a big fat warning as comment just before, so that it requires
user action, and that user will have been warned.
Would that work for you?
Maxime
Let's drop this two patches with boards dts changes until we get proper MLC
support in kernel.
As for examples - I'll update linux-sunxi wiki with how to enable NAND
controller and describe NAND chips in DTS. People who want to add NAND support
to their boards will be warned that it is risky for now.
From: Maxime Ripard <hidden> Date: 2016-06-14 12:33:33
On Tue, Jun 14, 2016 at 02:17:36PM +0300, Aleksei Mamlin wrote:
From: Boris Brezillon <redacted>
Add NAND Flash controller node definition to the A10 SoC.
Signed-off-by: Boris Brezillon <redacted>
Signed-off-by: Aleksei Mamlin <redacted>
From: Maxime Ripard <hidden> Date: 2016-06-14 12:33:56
On Tue, Jun 14, 2016 at 02:17:38PM +0300, Aleksei Mamlin wrote:
From: Boris Brezillon <redacted>
Add NAND Flash controller node definition to the A20 SoC.
Signed-off-by: Boris Brezillon <redacted>
Signed-off-by: Aleksei Mamlin <redacted>
From: Maxime Ripard <hidden> Date: 2016-06-14 12:37:22
Hi,
On Tue, Jun 14, 2016 at 02:17:35PM +0300, Aleksei Mamlin wrote:
From: Boris Brezillon <redacted>
Define the NAND controller pin configs.
Signed-off-by: Boris Brezillon <redacted>
Signed-off-by: Aleksei Mamlin <redacted>
Unfortunately, we don't have any user for it at the moment.
Please re-submit it later when we'll have some :)
Thanks!
Maxime
--
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160614/90e145ec/attachment.sig>