Skip to contentSkip to conversation

ARM: dts: am437x-idk: Fix QSPI compatible string

PATCH New Vignesh R 3 participants 9 messages Oct 13, 2016 – Nov 28, 2016 Arm (32-bit)

lore
Share & tools
lore

Vignesh R

author
#0 lore raw

PATCH 1/1 ARM: dts: am437x-idk: Fix QSPI compatible string

Unlike Linux kernel, U-Boot depends on "spi-flash" compatible to probe m25p80 spi-nor devices. Hence, add "spi-flash" compatible string to m25p80 node. Without this patch, flash device DT data is not parsed and QSPI operates in unsupported mode leading to data corruption.

---
 arch/arm/dts/am437x-idk-evm.dts | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/dts/am437x-idk-evm.dts +1 -1
index 478f0a62cb1c..e45464716549 100644
@@ -341,7 +341,7 @@
 
 	spi-max-frequency = <48000000>;
 	m25p80 at 0 {
-		compatible = "mx66l51235l";
+		compatible = "mx66l51235l", "spi-flash";
 		spi-max-frequency = <48000000>;
 		reg = <0>;
 		spi-cpol;
Signature · 1 line
2.10.1
#1 lore raw

PATCH 1/1 ARM: dts: am437x-idk: Fix QSPI compatible string

Quoted 7 lines · Vignesh R

Unlike Linux kernel, U-Boot depends on "spi-flash" compatible to probe m25p80 spi-nor devices. Hence, add "spi-flash" compatible string to m25p80 node. Without this patch, flash device DT data is not parsed and QSPI operates in unsupported mode leading to data corruption.

---

Applied to u-boot-spi/master

thanks!

Signature · 4 lines
Jagan Teki
Free Software Engineer | www.openedev.com
U-Boot, Linux | Upstream Maintainer
Hyderabad, India.
#2 lore raw

PATCH 1/1 ARM: dts: am437x-idk: Fix QSPI compatible string

Quoted 9 lines · Jagan Teki
Quoted 6 lines · Vignesh R

Unlike Linux kernel, U-Boot depends on "spi-flash" compatible to probe m25p80 spi-nor devices. Hence, add "spi-flash" compatible string to m25p80 node. Without this patch, flash device DT data is not parsed and QSPI operates in unsupported mode leading to data corruption.

Applied to u-boot-spi/master

... I don't like that we need a non u-boot prefixed string here for the binding to work as that will lead to harder re-syncs later on the dt files. Why aren't we matching on the existing part? Thanks!

Signature · 8 lines
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20161013/3ec62020/attachment.sig>
#3 lore raw

PATCH 1/1 ARM: dts: am437x-idk: Fix QSPI compatible string

Quoted 7 lines · Tom Rini
Quoted 6 lines · Jagan Teki
Quoted 5 lines · Vignesh R

Unlike Linux kernel, U-Boot depends on "spi-flash" compatible to probe m25p80 spi-nor devices. Hence, add "spi-flash" compatible string to m25p80 node. Without this patch, flash device DT data is not parsed and QSPI operates in unsupported mode leading to data corruption.

Quoted 7 lines
Quoted 3 lines
Quoted 1 line

Applied to u-boot-spi/master

... I don't like that we need a non u-boot prefixed string here for the binding to work as that will lead to harder re-syncs later on the dt files. Why aren't we matching on the existing part? Thanks!

Most of these compatibles from Linux, and we're unable to support as of now. and moment we moved to spi-nor the we will maintain the compatible of all these like Linux.

Too speed-up, can you please look at the last approach was submitted on [1]

[1] [PATCH RFC v8 00/16] SPI-NOR/MTD addition

thanks!

Signature · 4 lines
Jagan Teki
Free Software Engineer | www.openedev.com
U-Boot, Linux | Upstream Maintainer
Hyderabad, India.
author
#4 lore raw

PATCH 1/1 ARM: dts: am437x-idk: Fix QSPI compatible string

In reply to Tom Rini

Hi Jagan,

Quoted 7 lines · Tom Rini
Quoted 6 lines · Jagan Teki
Quoted 5 lines · Vignesh R

Unlike Linux kernel, U-Boot depends on "spi-flash" compatible to probe m25p80 spi-nor devices. Hence, add "spi-flash" compatible string to m25p80 node. Without this patch, flash device DT data is not parsed and QSPI operates in unsupported mode leading to data corruption.

Quoted 8 lines
Quoted 3 lines
Quoted 1 line

Applied to u-boot-spi/master

... I don't like that we need a non u-boot prefixed string here for the binding to work as that will lead to harder re-syncs later on the dt files. Why aren't we matching on the existing part? Thanks!

I don't see this patch in u-boot master yet. If this patch was dropped due to Tom's comment above, then could you suggest how to address the issue?

Signature · 2 lines
Regards
Vignesh
#5 lore raw

PATCH 1/1 ARM: dts: am437x-idk: Fix QSPI compatible string

Quoted 14 lines · Vignesh R

Hi Jagan,

Quoted 11 lines · Tom Rini
Quoted 9 lines · Jagan Teki
Quoted 6 lines · Vignesh R

Unlike Linux kernel, U-Boot depends on "spi-flash" compatible to probe m25p80 spi-nor devices. Hence, add "spi-flash" compatible string to m25p80 node. Without this patch, flash device DT data is not parsed and QSPI operates in unsupported mode leading to data corruption.

Applied to u-boot-spi/master

Quoted 7 lines
Quoted 4 lines

... I don't like that we need a non u-boot prefixed string here for the binding to work as that will lead to harder re-syncs later on the dt files. Why aren't we matching on the existing part? Thanks!

I don't see this patch in u-boot master yet. If this patch was dropped due to Tom's comment above, then could you suggest how to address the issue?

Some how missed this, but do you still unable to probe the flash w/o "spi-flash"? I think there is a device_probe when !device_active(dev) that will detect the flash chip. Please try once.

thanks!

Signature · 4 lines
Jagan Teki
Free Software Engineer | www.openedev.com
U-Boot, Linux | Upstream Maintainer
Hyderabad, India.
author
#6 lore raw

PATCH 1/1 ARM: dts: am437x-idk: Fix QSPI compatible string

Quoted 20 lines · Jagan Teki
Quoted 19 lines · Vignesh R

Hi Jagan,

Quoted 15 lines · Tom Rini
Quoted 9 lines · Jagan Teki
Quoted 6 lines · Vignesh R

Unlike Linux kernel, U-Boot depends on "spi-flash" compatible to probe m25p80 spi-nor devices. Hence, add "spi-flash" compatible string to m25p80 node. Without this patch, flash device DT data is not parsed and QSPI operates in unsupported mode leading to data corruption.

Applied to u-boot-spi/master

... I don't like that we need a non u-boot prefixed string here for the binding to work as that will lead to harder re-syncs later on the dt files. Why aren't we matching on the existing part? Thanks!

Quoted 7 lines
Quoted 2 lines

I don't see this patch in u-boot master yet. If this patch was dropped due to Tom's comment above, then could you suggest how to address the issue?

Some how missed this, but do you still unable to probe the flash w/o "spi-flash"? I think there is a device_probe when !device_active(dev) that will detect the flash chip. Please try once.

Yes, the flash is detected. But the device is not bound due to missing "spi-flash" compatible and hence DT properties of flash node are not available to the driver. So, the QSPI controller tries to operate in mode 0 instead of mode 3 and default frequency instead of 48MHz as specified by spi-max-frequency property leading to failure of flashing operations.

Signature · 2 lines
Regards
Vignesh
author
#7 lore raw

PATCH 1/1 ARM: dts: am437x-idk: Fix QSPI compatible string

Quoted 31 lines · Vignesh R
Quoted 20 lines · Jagan Teki
Quoted 19 lines · Vignesh R

Hi Jagan,

Quoted 15 lines · Tom Rini
Quoted 9 lines · Jagan Teki
Quoted 6 lines · Vignesh R

Unlike Linux kernel, U-Boot depends on "spi-flash" compatible to probe m25p80 spi-nor devices. Hence, add "spi-flash" compatible string to m25p80 node. Without this patch, flash device DT data is not parsed and QSPI operates in unsupported mode leading to data corruption.

Applied to u-boot-spi/master

... I don't like that we need a non u-boot prefixed string here for the binding to work as that will lead to harder re-syncs later on the dt files. Why aren't we matching on the existing part? Thanks!

Quoted 7 lines
Quoted 2 lines

I don't see this patch in u-boot master yet. If this patch was dropped due to Tom's comment above, then could you suggest how to address the issue?

Some how missed this, but do you still unable to probe the flash w/o "spi-flash"? I think there is a device_probe when !device_active(dev) that will detect the flash chip. Please try once.

Quoted 7 lines

Yes, the flash is detected. But the device is not bound due to missing "spi-flash" compatible and hence DT properties of flash node are not available to the driver. So, the QSPI controller tries to operate in mode 0 instead of mode 3 and default frequency instead of 48MHz as specified by spi-max-frequency property leading to failure of flashing operations.

In case its not clear from above explanation, this patch is still _needed_ in order for DT properties to be picked up. Without this QSPI write operations fail on am437x. Can this patch be picked up for this rc?

Signature · 2 lines
Regards
Vignesh
#8 lore raw

PATCH 1/1 ARM: dts: am437x-idk: Fix QSPI compatible string

Quoted 42 lines · Vignesh R
Quoted 31 lines · Vignesh R
Quoted 20 lines · Jagan Teki
Quoted 19 lines · Vignesh R

Hi Jagan,

Quoted 15 lines · Tom Rini
Quoted 9 lines · Jagan Teki
Quoted 6 lines · Vignesh R

Unlike Linux kernel, U-Boot depends on "spi-flash" compatible to probe m25p80 spi-nor devices. Hence, add "spi-flash" compatible string to m25p80 node. Without this patch, flash device DT data is not parsed and QSPI operates in unsupported mode leading to data corruption.

Applied to u-boot-spi/master

... I don't like that we need a non u-boot prefixed string here for the binding to work as that will lead to harder re-syncs later on the dt files. Why aren't we matching on the existing part? Thanks!

Quoted 7 lines
Quoted 2 lines

I don't see this patch in u-boot master yet. If this patch was dropped due to Tom's comment above, then could you suggest how to address the issue?

Some how missed this, but do you still unable to probe the flash w/o "spi-flash"? I think there is a device_probe when !device_active(dev) that will detect the flash chip. Please try once.

Quoted 7 lines

Yes, the flash is detected. But the device is not bound due to missing "spi-flash" compatible and hence DT properties of flash node are not available to the driver. So, the QSPI controller tries to operate in mode 0 instead of mode 3 and default frequency instead of 48MHz as specified by spi-max-frequency property leading to failure of flashing operations.

Quoted 3 lines

In case its not clear from above explanation, this patch is still _needed_ in order for DT properties to be picked up. Without this QSPI write operations fail on am437x. Can this patch be picked up for this rc?

Will pick.

thanks!

Signature · 4 lines
Jagan Teki
Free Software Engineer | www.openedev.com
U-Boot, Linux | Upstream Maintainer
Hyderabad, India.