Skip to contentSkip to conversation

Revert "core: device: use dev_power_domain_on"

PATCH New Igor Opaniuk 6 participants 8 messages Dec 17, 2019 – Dec 30, 2019

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 Revert "core: device: use dev_power_domain_on"

From: Igor Opaniuk <[email protected]>

This reverts commit f0cc4eae9a1702a768817ea25d9f23cece69d021

This was previously reported that f0cc4eae9a ("core: device:
use dev_power_domain_on") breaks initial boot on Colibri iMX8X and IMX8 QM ROM 7720a1 board. Revert it until the problem is properly fixed.

---

 drivers/core/device.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/core/device.c +3 -3
index 4e037083a6..c5b232c259 100644
@@ -313,6 +313,7 @@ static void *alloc_priv(int size, uint flags)
 
 int device_probe(struct udevice *dev)
 {
+	struct power_domain pd;
 	const struct driver *drv;
 	int size = 0;
 	int ret;
@@ -396,9 +397,8 @@ int device_probe(struct udevice *dev)
 	if (CONFIG_IS_ENABLED(POWER_DOMAIN) && dev->parent &&
 	    (device_get_uclass_id(dev) != UCLASS_POWER_DOMAIN) &&
 	    !(drv->flags & DM_FLAG_DEFAULT_PD_CTRL_OFF)) {
-		ret = dev_power_domain_on(dev);
-		if (ret)
-			goto fail;
+		if (!power_domain_get(dev, &pd))
+			power_domain_on(&pd);
 	}
 
 	ret = uclass_pre_probe_device(dev);
Signature · 1 line
2.17.1
#1 lore raw

PATCH 1/1 Revert "core: device: use dev_power_domain_on"

Quoted 10 lines · Igor Opaniuk

From: Igor Opaniuk <[email protected]>

This reverts commit f0cc4eae9a1702a768817ea25d9f23cece69d021

This was previously reported that f0cc4eae9a ("core: device:
use dev_power_domain_on") breaks initial boot on Colibri iMX8X and IMX8 QM ROM 7720a1 board. Revert it until the problem is properly fixed.

Quoted 30 lines

---

 drivers/core/device.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/core/device.c +3 -4
index 4e037083a6..c5b232c259 100644
@@ -313,6 +313,7 @@ static void *alloc_priv(int size, uint flags)
 
 int device_probe(struct udevice *dev)
 {
+       struct power_domain pd;
        const struct driver *drv;
        int size = 0;
        int ret;
@@ -396,9 +397,8 @@ int device_probe(struct udevice *dev)
        if (CONFIG_IS_ENABLED(POWER_DOMAIN) && dev->parent &&
            (device_get_uclass_id(dev) != UCLASS_POWER_DOMAIN) &&
            !(drv->flags & DM_FLAG_DEFAULT_PD_CTRL_OFF)) {
-               ret = dev_power_domain_on(dev);
-               if (ret)
-                       goto fail;
+               if (!power_domain_get(dev, &pd))
+                       power_domain_on(&pd);
        }
 
        ret = uclass_pre_probe_device(dev);
--
Quoted 2 lines

2.17.1

--
Best regards
Oleksandr Suvorov

Toradex AG
Altsagenstrasse 5 | 6048 Horw/Luzern | Switzerland | T: +41 41 500 4800 (main line)

#2 lore raw

PATCH 1/1 Revert "core: device: use dev_power_domain_on"

In reply to Igor Opaniuk
Quoted 10 lines · Igor Opaniuk

From: Igor Opaniuk <[email protected]>

This reverts commit f0cc4eae9a1702a768817ea25d9f23cece69d021

This was previously reported that f0cc4eae9a ("core: device:
use dev_power_domain_on") breaks initial boot on Colibri iMX8X and IMX8 QM ROM 7720a1 board. Revert it until the problem is properly fixed.

Being so close from the 2020.01, I think this is the safest approach for now:

#3 lore raw

PATCH 1/1 Revert "core: device: use dev_power_domain_on"

In reply to Igor Opaniuk
Quoted 9 lines

Subject: [PATCH v1] Revert "core: device: use dev_power_domain_on"

From: Igor Opaniuk <[email protected]>

This reverts commit f0cc4eae9a1702a768817ea25d9f23cece69d021

This was previously reported that f0cc4eae9a ("core: device:
use dev_power_domain_on") breaks initial boot on Colibri iMX8X and IMX8 QM ROM 7720a1 board. Revert it until the problem is properly fixed.

This is not the root cause. Please check your driver power domain.

Thanks
Peng.

Quoted 33 lines

---

 drivers/core/device.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/core/device.c index renamed from drivers/core/device.c +0 -0

4e037083a6..c5b232c259 100644

drivers/core/device.c +3 -4
@@ -313,6 +313,7 @@ static void *alloc_priv(int size, uint flags)
 
 int device_probe(struct udevice *dev)
 {
+	struct power_domain pd;
 	const struct driver *drv;
 	int size = 0;
 	int ret;
@@ -396,9 +397,8 @@ int device_probe(struct udevice *dev)
 	if (CONFIG_IS_ENABLED(POWER_DOMAIN) && dev->parent &&
 	    (device_get_uclass_id(dev) != UCLASS_POWER_DOMAIN) &&
 	    !(drv->flags & DM_FLAG_DEFAULT_PD_CTRL_OFF)) {
-		ret = dev_power_domain_on(dev);
-		if (ret)
-			goto fail;
+		if (!power_domain_get(dev, &pd))
+			power_domain_on(&pd);
 	}
 
 	ret = uclass_pre_probe_device(dev);
--

2.17.1

#4 lore raw

PATCH 1/1 Revert "core: device: use dev_power_domain_on"

Hi Peng,

Quoted 5 lines · Peng Fan
Quoted 5 lines

Subject: [PATCH v1] Revert "core: device: use dev_power_domain_on"

From: Igor Opaniuk <[email protected]>

Quoted 7 lines
Quoted 5 lines

This reverts commit f0cc4eae9a1702a768817ea25d9f23cece69d021

This was previously reported that f0cc4eae9a ("core: device:
use dev_power_domain_on") breaks initial boot on Colibri iMX8X and IMX8 QM ROM 7720a1 board. Revert it until the problem is properly fixed.

This is not the root cause. Please check your driver power domain.

Could this error be dependant on the SCFW version?

BTW, are you able to boot mx8qxp mek with mainline?

It is hanging for me as posted at
https://lists.denx.de/pipermail/u-boot/2019-December/394054.html

#5 lore raw

PATCH 1/1 Revert "core: device: use dev_power_domain_on"

In reply to Peng Fan

Hi Peng,

Quoted 5 lines · Peng Fan
Quoted 5 lines

Subject: [PATCH v1] Revert "core: device: use dev_power_domain_on"

From: Igor Opaniuk <[email protected]>

Quoted 7 lines
Quoted 5 lines

This reverts commit f0cc4eae9a1702a768817ea25d9f23cece69d021

This was previously reported that f0cc4eae9a ("core: device:
use dev_power_domain_on") breaks initial boot on Colibri iMX8X and IMX8 QM ROM 7720a1 board. Revert it until the problem is properly fixed.

This is not the root cause. Please check your driver power domain.

There is no any statement anywhere that this is the "root cause". Me and Oliver pointed to this commit as it was the first one when two MX8-based boards from different vendors just stoped booting.

That is pretty obvious that there were updates also on SCFW/TF-A side, but unfortunately I haven't found any related information (nor in the commit message, neither any
updates in board/freescale/imx8*/README), just some notice in one of ML threads, where
Oliver actually reported this issue.

Please provide more info about what version of SECO, SCFW and TF-A using.

Thanks

Quoted 38 lines

Thanks
Peng.

Quoted 33 lines

---

 drivers/core/device.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/core/device.c index renamed from drivers/core/device.c +0 -0

4e037083a6..c5b232c259 100644

drivers/core/device.c +3 -4
@@ -313,6 +313,7 @@ static void *alloc_priv(int size, uint flags)
 
 int device_probe(struct udevice *dev)
 {
+     struct power_domain pd;
      const struct driver *drv;
      int size = 0;
      int ret;
@@ -396,9 +397,8 @@ int device_probe(struct udevice *dev)
      if (CONFIG_IS_ENABLED(POWER_DOMAIN) && dev->parent &&
          (device_get_uclass_id(dev) != UCLASS_POWER_DOMAIN) &&
          !(drv->flags & DM_FLAG_DEFAULT_PD_CTRL_OFF)) {
-             ret = dev_power_domain_on(dev);
-             if (ret)
-                     goto fail;
+             if (!power_domain_get(dev, &pd))
+                     power_domain_on(&pd);
      }
 
      ret = uclass_pre_probe_device(dev);
--

2.17.1

Signature · 8 lines
Best regards - Freundliche Grüsse - Meilleures salutations

Igor Opaniuk

mailto: igor.opaniuk at gmail.com
skype: igor.opanyuk
+380 (93) 836 40 67
http://ua.linkedin.com/in/iopaniuk
#6 lore raw

PATCH 1/1 Revert "core: device: use dev_power_domain_on"

In reply to Igor Opaniuk
Quoted 9 lines · Igor Opaniuk

From: Igor Opaniuk <[email protected]>

This reverts commit f0cc4eae9a1702a768817ea25d9f23cece69d021

This was previously reported that f0cc4eae9a ("core: device:
use dev_power_domain_on") breaks initial boot on Colibri iMX8X and IMX8 QM ROM 7720a1 board. Revert it until the problem is properly fixed.

NAK. This breaks other drivers with multiple power domains that rely on core framework to enable.

Quoted 18 lines

---

 drivers/core/device.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/core/device.c +1 -0
index 4e037083a6..c5b232c259 100644
@@ -313,6 +313,7 @@ static void *alloc_priv(int size, uint flags)
 
 int device_probe(struct udevice *dev)
 {
+	struct power_domain pd;
 	const struct driver *drv;
 	int size = 0;
 	int ret;

drivers/core/device.c int device_probe(struct udevice *dev) · line 401

Quoted 6 lines
drivers/core/device.c +0 -3
@@ -396,9 +397,8 @@ int device_probe(struct udevice *dev)
 	if (CONFIG_IS_ENABLED(POWER_DOMAIN) && dev->parent &&
 	    (device_get_uclass_id(dev) != UCLASS_POWER_DOMAIN) &&
 	    !(drv->flags & DM_FLAG_DEFAULT_PD_CTRL_OFF)) {
-		ret = dev_power_domain_on(dev);
-		if (ret)
-			goto fail;

Can you check by not returning on failure here? If yes then check the power-domain/driver that is failing. If any driver doesn't expect core to enable power-domain then enable DM_FLAG_DEFAULT_PD_CTRL_OFF in the respective driver.

Thanks and regards,
Lokesh

Quoted 6 lines
drivers/core/device.c +2 -0
+		if (!power_domain_get(dev, &pd))
+			power_domain_on(&pd);
 	}
 
 	ret = uclass_pre_probe_device(dev);
#7 lore raw

PATCH 1/1 Revert "core: device: use dev_power_domain_on"

Quoted 36 lines · Lokesh Vutla
Quoted 9 lines · Igor Opaniuk

From: Igor Opaniuk <[email protected]>

This reverts commit f0cc4eae9a1702a768817ea25d9f23cece69d021

This was previously reported that f0cc4eae9a ("core: device:
use dev_power_domain_on") breaks initial boot on Colibri iMX8X and IMX8 QM ROM 7720a1 board. Revert it until the problem is properly fixed.

NAK. This breaks other drivers with multiple power domains that rely on core framework to enable.

Quoted 20 lines

---

 drivers/core/device.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/core/device.c +1 -0
index 4e037083a6..c5b232c259 100644
@@ -313,6 +313,7 @@ static void *alloc_priv(int size, uint flags)
 
 int device_probe(struct udevice *dev)
 {
+	struct power_domain pd;
 	const struct driver *drv;
 	int size = 0;
 	int ret;
@@ -396,9 +397,8 @@ int device_probe(struct udevice *dev)
 	if (CONFIG_IS_ENABLED(POWER_DOMAIN) && dev->parent &&
 	    (device_get_uclass_id(dev) != UCLASS_POWER_DOMAIN) &&
Quoted 8 lines

drivers/core/device.c int device_probe(struct udevice *dev) · line 401

Quoted 4 lines
drivers/core/device.c +0 -3
 	    !(drv->flags & DM_FLAG_DEFAULT_PD_CTRL_OFF)) {
-		ret = dev_power_domain_on(dev);
-		if (ret)
-			goto fail;

Can you check by not returning on failure here? If yes then check the power-domain/driver that is failing. If any driver doesn't expect core to enable power-domain then enable DM_FLAG_DEFAULT_PD_CTRL_OFF in the respective driver.

I'm adding a few other people to this part of the thread as I don't think breaking some boards to fix other boards is a viable choice, even at this stage in the release. Thanks all.

Signature · 8 lines
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.denx.de/pipermail/u-boot/attachments/20191230/8d0f4557/attachment.sig>