Skip to contentSkip to conversation

backlight: aw99706: DT parsing and blank state fixes

PATCH Applied Junjie Cao 4 participants 11 messages Aug 4 – Aug 12

Subsystems (5)
lore
Share & tools
lore

Junjie Cao

author
#0 lore raw

v3 cover backlight: aw99706: DT parsing and blank state fixes

The driver reads its tuning properties under names the binding does not define (patch 1), the validation paths accept values the binding forbids (patch 2), and update_status() ignores the blank state, leaving bl_power inoperative (patch 3).

v3:
- patch 2: reject 0 for awinic,sw-freq-hz (reserved-slot marker collided with it); add Fixes tag
- patch 3: add stable tag
v2: only patch 1/3 reached the lists (outbound mail failure):

    https://lore.kernel.org/all/[email protected]/

v1: https://lore.kernel.org/all/[email protected]/

Junjie Cao (3):
backlight: aw99706: Fix DT property names to match binding
backlight: aw99706: Validate all DT property values consistently
backlight: aw99706: Honor the core blank state in update_status()

 drivers/video/backlight/aw99706.c | 25 +++++++++++++++----------
 1 file changed, 15 insertions(+), 10 deletions(-)
Signature · 1 line
2.43.0
author
#1 lore raw

PATCH v3 1/3 backlight: aw99706: Fix DT property names to match binding

From: Junjie Cao <[email protected]>

The driver reads four tuning properties without the unit suffixes that the binding mandates: "awinic,sw-freq" instead of "awinic,sw-freq-hz", "awinic,sw-ilmt" instead of "awinic,sw-ilmt-microamp", "awinic,iled-max" instead of "awinic,iled-max-microamp", and "awinic,uvlo-thres" instead of "awinic,uvlo-thres-microvolt".

As a result, device_property_read_u32() never finds these properties in a binding-conformant device tree and silently falls back to the compiled-in defaults for switching frequency, switching current limit, max LED current, and UVLO threshold.

Fix by aligning the property name strings in aw99706_dt_props[] with the binding. No value/range changes are needed since both sides already use the same units and enumerations.

Fixes 147b38a5ad06 ("backlight: aw99706: Add support for Awinic AW99706 backlight")
---
 drivers/video/backlight/aw99706.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/video/backlight/aw99706.c +4 -4
index 18299faf06ad..e130f164303a 100644
@@ -130,23 +130,23 @@ static const struct aw99706_dt_prop aw99706_dt_props[] = {
 		AW99706_CFG0_REG, AW99706_DIM_MODE_MASK, 1,
 	},
 	{
-		"awinic,sw-freq", aw99706_dt_property_lookup,
+		"awinic,sw-freq-hz", aw99706_dt_property_lookup,
 		aw99706_sw_freq_tbl, ARRAY_SIZE(aw99706_sw_freq_tbl),
 		AW99706_CFG1_REG, AW99706_SW_FREQ_MASK, 750000,
 	},
 	{
-		"awinic,sw-ilmt", aw99706_dt_property_lookup,
+		"awinic,sw-ilmt-microamp", aw99706_dt_property_lookup,
 		aw99706_sw_ilmt_tbl, ARRAY_SIZE(aw99706_sw_ilmt_tbl),
 		AW99706_CFG1_REG, AW99706_SW_ILMT_MASK, 3000000,
 	},
 	{
-		"awinic,iled-max", aw99706_dt_property_iled_max_convert,
+		"awinic,iled-max-microamp", aw99706_dt_property_iled_max_convert,
 		NULL, 0,
 		AW99706_CFG2_REG, AW99706_ILED_MAX_MASK, 20000,
 
 	},
 	{
-		"awinic,uvlo-thres", aw99706_dt_property_lookup,
+		"awinic,uvlo-thres-microvolt", aw99706_dt_property_lookup,
 		aw99706_ulvo_thres_tbl, ARRAY_SIZE(aw99706_ulvo_thres_tbl),
 		AW99706_CFG2_REG, AW99706_UVLOSEL_MASK, 2200000,
 	},
Signature · 1 line
2.43.0
#2 lore raw

Re: [PATCH v3 1/3] backlight: aw99706: Fix DT property names to match binding

Thank you for your contribution! Sashiko AI review found 3 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 3 potential issue(s) to consider:

Pre-existing issues:
- [High] The driver attempts to read the I2C CHIPID register before configuring and asserting the hardware enable GPIO.
- [High] The update_status callback directly uses bl->props.brightness and fails to account for backlight blanking or suspend power states.
- [High] The initial default-brightness parsed from the device tree is never applied to the hardware.
--

--- Patch [1]: [PATCH v3 1/3] backlight: aw99706: Fix DT property names to match binding --- commit 0780fe78c9307963271e82c84a46a86db0bcde59
Author: Junjie Cao <[email protected]>

backlight: aw99706: Fix DT property names to match binding

The driver reads four tuning properties without the unit suffixes that the binding mandates. Fix by aligning the property name strings in aw99706_dt_props[] with the binding.

drivers/video/backlight/aw99706.c static const struct aw99706_dt_prop aw99706_dt_props[] = { · line 133

Quoted 10 lines
drivers/video/backlight/aw99706.c +1 -1
index 18299faf06ad4..e130f164303a2 100644
@@ -130,23 +130,23 @@ static const struct aw99706_dt_prop aw99706_dt_props[] = {
 		AW99706_CFG0_REG, AW99706_DIM_MODE_MASK, 1,
 	},
 	{
-		"awinic,sw-freq", aw99706_dt_property_lookup,
+		"awinic,sw-freq-hz", aw99706_dt_property_lookup,

[ ... ]

[Severity: High]
This is a pre-existing issue, but does aw99706_bl_update_status() handle backlight blanking or suspend power states?

drivers/video/backlight/aw99706.c:aw99706_bl_update_status() {
	struct aw99706_device *aw = bl_get_data(bl);

	return aw99706_update_brightness(aw, bl->props.brightness);
}

By directly accessing bl->props.brightness instead of using backlight_get_brightness(bl), the driver ignores BL_CORE_FBBLANK and BL_CORE_SUSPENDED flags, keeping the screen illuminated when it should be blanked.

[Severity: High]
This is a pre-existing issue, but can the device probe successfully if the I2C read occurs before the hardware enable GPIO is asserted?

drivers/video/backlight/aw99706.c:aw99706_probe() {
	...
	ret = aw99706_chip_id_read(aw);
	if (ret != AW99706_ID)
		return dev_err_probe(dev, -ENODEV,
				     "Unknown chip id 0x%02x\n", ret);

	aw99706_dt_parse(aw, &props);

	aw->hwen_gpio = devm_gpiod_get(aw->dev, "enable", GPIOD_OUT_LOW);
	...
}

If the bootloader does not leave the enable GPIO asserted, the device will be powered off or in reset, rendering I2C communication impossible.

[Severity: High]
This is a pre-existing issue, but is the initial default-brightness applied to the hardware?

drivers/video/backlight/aw99706.c:aw99706_probe() {
	...
	bl_dev = devm_backlight_device_register(dev, "aw99706-backlight", dev,
						aw, &aw99706_bl_ops, &props);
	if (IS_ERR(bl_dev))
		return dev_err_probe(dev, PTR_ERR(bl_dev),
				     "Failed to register backlight!\n");

	aw->bl_dev = bl_dev;

	return 0;
}

The probe routine successfully parses default-brightness and registers the backlight device, but doesn't call backlight_update_status(bl_dev) to sync the parsed initial brightness to the hardware registers.

Without an explicit initial brightness update to overwrite the hardware reset state, the display remains off upon boot until userspace manually sets the brightness.

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
maintainer Reviewed-by on v3 1/3
#3 lore raw

Re: [PATCH v3 1/3] backlight: aw99706: Fix DT property names to match binding

In reply to Junjie Cao
Quoted 13 lines · Junjie Cao

From: Junjie Cao <[email protected]>

The driver reads four tuning properties without the unit suffixes that the binding mandates: "awinic,sw-freq" instead of "awinic,sw-freq-hz", "awinic,sw-ilmt" instead of "awinic,sw-ilmt-microamp", "awinic,iled-max" instead of "awinic,iled-max-microamp", and "awinic,uvlo-thres" instead of "awinic,uvlo-thres-microvolt".

As a result, device_property_read_u32() never finds these properties in a binding-conformant device tree and silently falls back to the compiled-in defaults for switching frequency, switching current limit, max LED current, and UVLO threshold.

Quoted 7 lines

Fix by aligning the property name strings in aw99706_dt_props[] with the binding. No value/range changes are needed since both sides already use the same units and enumerations.

Fixes 147b38a5ad06 ("backlight: aw99706: Add support for Awinic AW99706 backlight")

There don't seem to be any usage of these properties in-tree so I think it is probably OK to change the driver behaviour without fallbacks for DTs with version skew!

Daniel.

author
#4 lore raw

PATCH v3 2/3 backlight: aw99706: Validate all DT property values consistently

In reply to Junjie Cao

From: Junjie Cao <[email protected]>

The lookup helpers for dim-mode and ramp-ctl take a shortcut when lookup_tbl is NULL: they accept any u32 value without range-checking and return success unconditionally. Out-of-range values get silently truncated by regmap_update_bits instead of triggering the dev_warn + default-fallback path that the other properties use.

Add a field-width check for the NULL-table case so that values exceeding the register field maximum are rejected the same way a table-lookup miss is.

The switching frequency table has a second hole: reserved slots use 0 as their marker, so "awinic,sw-freq-hz = <0>" matches slot 0 and programs a reserved encoding. Make the reserved marker U32_MAX and skip such slots during lookup.

While here, also switch the error returns to -EINVAL for consistency.

Fixes 147b38a5ad06 ("backlight: aw99706: Add support for Awinic AW99706 backlight")
---
 drivers/video/backlight/aw99706.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)
drivers/video/backlight/aw99706.c +10 -5
index e130f164303a..6ec49b6cb14c 100644
@@ -60,7 +60,7 @@
 #define AW99706_MTPLDOSEL_REG			0x1E
 #define AW99706_MTPRUN_REG			0x1F
 
-#define RESV	0
+#define RESV	U32_MAX
 
 /* Boost switching frequency table, in Hz */
 static const u32 aw99706_sw_freq_tbl[] = {
@@ -94,17 +94,19 @@ static int aw99706_dt_property_lookup(const struct aw99706_dt_prop *prop,
 	int i;
 
 	if (!prop->lookup_tbl) {
+		if (dt_val > (prop->mask >> __ffs(prop->mask)))
+			return -EINVAL;
 		*val = dt_val;
 		return 0;
 	}
 
 	for (i = 0; i < prop->tbl_size; i++)
-		if (prop->lookup_tbl[i] == dt_val)
+		if (prop->lookup_tbl[i] != RESV && prop->lookup_tbl[i] == dt_val)
 			break;
 
 	*val = i;
 
-	return i == prop->tbl_size ? -1 : 0;
+	return i == prop->tbl_size ? -EINVAL : 0;
 }
 
 #define MIN_ILED_MAX	5000
@@ -116,11 +118,14 @@ aw99706_dt_property_iled_max_convert(const struct aw99706_dt_prop *prop,
 				     u32 dt_val, u8 *val)
 {
 	if (dt_val > MAX_ILED_MAX || dt_val < MIN_ILED_MAX)
-		return -1;
+		return -EINVAL;
+
+	if ((dt_val - MIN_ILED_MAX) % STEP_ILED_MAX)
+		return -EINVAL;
 
 	*val = (dt_val - MIN_ILED_MAX) / STEP_ILED_MAX;
 
-	return (dt_val - MIN_ILED_MAX) % STEP_ILED_MAX;
+	return 0;
 }
 
 static const struct aw99706_dt_prop aw99706_dt_props[] = {
Signature · 1 line
2.43.0
#5 lore raw

Re: [PATCH v3 2/3] backlight: aw99706: Validate all DT property values consistently

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:

Pre-existing issues:
- [High] I2C chip ID read is attempted before the hardware enable GPIO is asserted.
- [High] Initial backlight brightness state is not synchronized to the hardware during probe. --

--- Patch [2]: [PATCH v3 2/3] backlight: aw99706: Validate all DT property values consistently --- commit 7ac792a99fb6a9cdd91d97a6eb34756415f65058
Author: Junjie Cao <[email protected]>

backlight: aw99706: Validate all DT property values consistently

The lookup helpers for dim-mode and ramp-ctl take a shortcut when lookup_tbl is NULL, accepting any u32 value without range-checking. This commit adds a field-width check for the NULL-table case so that values exceeding the register field maximum are rejected. It also updates the switching frequency table reserved marker to U32_MAX and changes error returns to -EINVAL.

drivers/video/backlight/aw99706.c

Quoted 4 lines
drivers/video/backlight/aw99706.c +0 -0
index e130f164303a2..6ec49b6cb14c6 100644

[ ... ]

[Severity: High]
This isn't a bug introduced by this patch, but does aw99706_probe() attempt an I2C read before the hardware enable GPIO is asserted?

In aw99706_probe(), the driver calls aw99706_chip_id_read() before fetching hwen_gpio and calling aw99706_hw_init() to assert the enable pin:

drivers/video/backlight/aw99706.c:aw99706_probe() {
    ...
	ret = aw99706_chip_id_read(aw);
	if (ret != AW99706_ID)
		return dev_err_probe(dev, -ENODEV,
				     "Unknown chip id 0x%02x\n", ret);

	aw99706_dt_parse(aw, &props);

	aw->hwen_gpio = devm_gpiod_get(aw->dev, "enable", GPIOD_OUT_LOW);
    ...
}

Could this cause the probe to fail with an I2C NACK or timeout if the bootloader hasn't already left the GPIO asserted?

[Severity: High]
This is a pre-existing issue, but is the initial backlight brightness state synchronized to the hardware during probe?

aw99706_probe() parses the default brightness and registers the backlight device, but returns without flushing the initial state:

drivers/video/backlight/aw99706.c:aw99706_probe() {
    ...
	bl_dev = devm_backlight_device_register(dev, "aw99706-backlight", dev,
						aw, &aw99706_bl_ops, &props);
	if (IS_ERR(bl_dev))
		return dev_err_probe(dev, PTR_ERR(bl_dev),
				     "Failed to register backlight!\n");

	aw->bl_dev = bl_dev;

	return 0;
}

Since backlight_update_status() or aw99706_update_brightness() is never called here, does this mean the parsed brightness and the backlight enable bit aren't written to the hardware, potentially leaving a black screen during boot?

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2
maintainer Reviewed-by on v3 2/3
#6 lore raw

Re: [PATCH v3 2/3] backlight: aw99706: Validate all DT property values consistently

In reply to Junjie Cao
Quoted 17 lines · Junjie Cao

From: Junjie Cao <[email protected]>

The lookup helpers for dim-mode and ramp-ctl take a shortcut when lookup_tbl is NULL: they accept any u32 value without range-checking and return success unconditionally. Out-of-range values get silently truncated by regmap_update_bits instead of triggering the dev_warn + default-fallback path that the other properties use.

Add a field-width check for the NULL-table case so that values exceeding the register field maximum are rejected the same way a table-lookup miss is.

The switching frequency table has a second hole: reserved slots use 0 as their marker, so "awinic,sw-freq-hz = <0>" matches slot 0 and programs a reserved encoding. Make the reserved marker U32_MAX and skip such slots during lookup.

Quoted 4 lines

While here, also switch the error returns to -EINVAL for consistency.

Fixes 147b38a5ad06 ("backlight: aw99706: Add support for Awinic AW99706 backlight")

Daniel.

author
#7 lore raw

PATCH v3 3/3 backlight: aw99706: Honor the core blank state in update_status()

In reply to Junjie Cao

From: Junjie Cao <[email protected]>

update_status() passes props.brightness straight to the hardware and ignores the power/blank state tracked by the core. Writing 4 to the bl_power sysfs attribute or blanking the framebuffer therefore leaves the backlight lit.

Use backlight_get_brightness(), which returns 0 while the device is blanked.

Fixes 147b38a5ad06 ("backlight: aw99706: Add support for Awinic AW99706 backlight")
---
 drivers/video/backlight/aw99706.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/video/backlight/aw99706.c +1 -1
index 6ec49b6cb14c..c4a66b59ecf1 100644
@@ -340,7 +340,7 @@ static int aw99706_bl_update_status(struct backlight_device *bl)
 {
 	struct aw99706_device *aw = bl_get_data(bl);
 
-	return aw99706_update_brightness(aw, bl->props.brightness);
+	return aw99706_update_brightness(aw, backlight_get_brightness(bl));
 }
 
 static const struct backlight_ops aw99706_bl_ops = {
Signature · 1 line
2.43.0
#8 lore raw

Re: [PATCH v3 3/3] backlight: aw99706: Honor the core blank state in update_status()

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:

Pre-existing issues:
- [High] I2C read is performed before the hardware is powered up or enabled.
- [High] The initial backlight state parsed from device tree is never applied to the hardware during probe. --

--- Patch [3]: [PATCH v3 3/3] backlight: aw99706: Honor the core blank state in update_status() --- commit 4fd7fd023a2e220d60fcbcb9c519d9b49b430467
Author: Junjie Cao <[email protected]>

backlight: aw99706: Honor the core blank state in update_status()

update_status() passes props.brightness straight to the hardware and ignores the power/blank state tracked by the core. Writing 4 to the bl_power sysfs attribute or blanking the framebuffer therefore leaves the backlight lit.

drivers/video/backlight/aw99706.c static int aw99706_bl_update_status(struct backlight_device *bl) · line 346

Quoted 13 lines
drivers/video/backlight/aw99706.c +1 -1
index 6ec49b6cb14c6..c4a66b59ecf13 100644
@@ -340,7 +340,7 @@ static int aw99706_bl_update_status(struct backlight_device *bl)
 {
 	struct aw99706_device *aw = bl_get_data(bl);
 
-	return aw99706_update_brightness(aw, bl->props.brightness);
+	return aw99706_update_brightness(aw, backlight_get_brightness(bl));
 }
 
 static const struct backlight_ops aw99706_bl_ops = {

[Severity: High]
This is a pre-existing issue, but I noticed an I2C read is performed before the hardware is powered up or enabled.

In drivers/video/backlight/aw99706.c:aw99706_probe():

	ret = aw99706_chip_id_read(aw);
	if (ret != AW99706_ID)
		return dev_err_probe(dev, -ENODEV,
				     "Unknown chip id 0x%02x\n", ret);

	aw99706_dt_parse(aw, &props);

	aw->hwen_gpio = devm_gpiod_get(aw->dev, "enable", GPIOD_OUT_LOW);

	ret = aw99706_hw_init(aw);

Will the driver probe fail with -ENODEV here if the hardware is disabled by default, since the hardware enable GPIO (aw->hwen_gpio) is not requested and asserted until later via devm_gpiod_get() and aw99706_hw_init()?

[Severity: High]
This is also a pre-existing issue, but the initial backlight state parsed from device tree is never applied to the hardware during probe.

In drivers/video/backlight/aw99706.c:aw99706_probe():

	bl_dev = devm_backlight_device_register(dev, "aw99706-backlight", dev,
						aw, &aw99706_bl_ops, &props);
	if (IS_ERR(bl_dev))
		return dev_err_probe(dev, PTR_ERR(bl_dev),
				     "Failed to register backlight!\n");

	aw->bl_dev = bl_dev;

	return 0;

Does the backlight remain off during the entire boot process until userspace writes to the sysfs brightness attribute?

The probe function completes without calling backlight_update_status() to synchronize the parsed device tree brightness properties to the hardware, and the backlight core does not automatically sync the initial software state upon registration.

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=3
maintainer Reviewed-by on v3 3/3
#9 lore raw

Re: [PATCH v3 3/3] backlight: aw99706: Honor the core blank state in update_status()

In reply to Junjie Cao
Quoted 7 lines · Junjie Cao

From: Junjie Cao <[email protected]>

update_status() passes props.brightness straight to the hardware and ignores the power/blank state tracked by the core. Writing 4 to the bl_power sysfs attribute or blanking the framebuffer therefore leaves the backlight lit.

Quoted 6 lines

Use backlight_get_brightness(), which returns 0 while the device is blanked.

Fixes 147b38a5ad06 ("backlight: aw99706: Add support for Awinic AW99706 backlight")

Daniel.

maintainer Applied
#10 lore raw
In reply to Junjie Cao
Quoted 10 lines · Junjie Cao

The driver reads its tuning properties under names the binding does not define (patch 1), the validation paths accept values the binding forbids (patch 2), and update_status() ignores the blank state, leaving bl_power inoperative (patch 3).

v3:
- patch 2: reject 0 for awinic,sw-freq-hz (reserved-slot marker collided with it); add Fixes tag
- patch 3: add stable tag
v2: only patch 1/3 reached the lists (outbound mail failure):

Quoted 4 lines
    https://lore.kernel.org/all/[email protected]/

v1: https://lore.kernel.org/all/[email protected]/

[...]

Applied, thanks!

[1/3] backlight: aw99706: Fix DT property names to match binding

      commit: daf120e464386f750b3870dc04d5bc6a786f0000
[2/3] backlight: aw99706: Validate all DT property values consistently
      commit: 3d142b19b09ea62d6e9a33d46d1b858b3ad05707
[3/3] backlight: aw99706: Honor the core blank state in update_status()
      commit: fbf8b5d251e8b4e1b01c591be313c63cb9e1526a

--
Lee Jones [李琼斯]