[PATCH v2 0/2] Add dev_err_probe() to i2c designware platform probing

[PATCH v2 0/2] Add dev_err_probe() to i2c designware platform probing

[Benoît Monin]

I recently spend some time debugging a case where the i2c controller
never showed up. In the end it was caused by a missing reset controller
due to a typo in the device tree.

While this has nothing to do with the i2c designware driver, not having
any hint about why the device stays in deferred probe state does not
help.

The first patch convert dev_err() to dev_err_probe() on request IRQ
failure, this simplifies error handling.

The second patch add dev_err_probe() in the error paths that can return
-EPROBE_DEFER in dw_i2c_plat_probe() to aid in debugging deferred probe
errors.

Signed-off-by: Benoît Monin <benoit.monin@bootlin.com>
---
Changes in v2:
- Split the patch in two: one to convert dev_err() to dev_err_probe()
  on request IRQ failure and the other adding dev_err_probe() to error
  paths in dw_i2c_plat_probe()
- Link to v1: https://lore.kernel.org/r/20250904-i2c-dw-dev-err-probe-v1-1-acca6ffd122e@bootlin.com

---
Benoît Monin (2):
      i2c: designware: convert to dev_err_probe() on request IRQ error
      i2c: designware: use dev_err_probe() when probing platform device

 drivers/i2c/busses/i2c-designware-master.c  |  9 ++++-----
 drivers/i2c/busses/i2c-designware-platdrv.c | 10 ++++++----
 drivers/i2c/busses/i2c-designware-slave.c   |  9 ++++-----
 3 files changed, 14 insertions(+), 14 deletions(-)
---
base-commit: b320789d6883cc00ac78ce83bccbfe7ed58afcf0
change-id: 20250822-i2c-dw-dev-err-probe-eaf9bce3ef4c

Best regards,
-- 
Benoît Monin, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

[PATCH v2 1/2] i2c: designware: convert to dev_err_probe() on request IRQ error

[Benoît Monin]

Simplify the error handling of devm_request_irq() in
i2c_dw_probe_master() and i2c_dw_probe_slave() by converting to:

    return dev_err_probe();

instead of calling:

    dev_err();
    return ret;

This also handle deferred probe error without spamming the log.

Signed-off-by: Benoît Monin <benoit.monin@bootlin.com>
---
 drivers/i2c/busses/i2c-designware-master.c | 9 ++++-----
 drivers/i2c/busses/i2c-designware-slave.c  | 9 ++++-----
 2 files changed, 8 insertions(+), 10 deletions(-)

diff --git a/drivers/i2c/busses/i2c-designware-master.c b/drivers/i2c/busses/i2c-designware-master.c
index cbd88ffa561010ff2d29086836d9119da5b8885c..c7a72c28786c2b59a249a768d43a7954119bc018 100644
--- a/drivers/i2c/busses/i2c-designware-master.c
+++ b/drivers/i2c/busses/i2c-designware-master.c
@@ -1068,11 +1068,10 @@ int i2c_dw_probe_master(struct dw_i2c_dev *dev)
 	if (!(dev->flags & ACCESS_POLLING)) {
 		ret = devm_request_irq(dev->dev, dev->irq, i2c_dw_isr,
 				       irq_flags, dev_name(dev->dev), dev);
-		if (ret) {
-			dev_err(dev->dev, "failure requesting irq %i: %d\n",
-				dev->irq, ret);
-			return ret;
-		}
+		if (ret)
+			return dev_err_probe(dev->dev, ret,
+					     "failure requesting irq %i: %d\n",
+					     dev->irq, ret);
 	}
 
 	ret = i2c_dw_init_recovery_info(dev);
diff --git a/drivers/i2c/busses/i2c-designware-slave.c b/drivers/i2c/busses/i2c-designware-slave.c
index b936a240db0a9308f005148cdf4c4f9fd512be05..6eb16b7d75a6d059c7abcead609258f9d514d012 100644
--- a/drivers/i2c/busses/i2c-designware-slave.c
+++ b/drivers/i2c/busses/i2c-designware-slave.c
@@ -266,11 +266,10 @@ int i2c_dw_probe_slave(struct dw_i2c_dev *dev)
 
 	ret = devm_request_irq(dev->dev, dev->irq, i2c_dw_isr_slave,
 			       IRQF_SHARED, dev_name(dev->dev), dev);
-	if (ret) {
-		dev_err(dev->dev, "failure requesting IRQ %i: %d\n",
-			dev->irq, ret);
-		return ret;
-	}
+	if (ret)
+		return dev_err_probe(dev->dev, ret,
+				     "failure requesting IRQ %i: %d\n",
+				     dev->irq, ret);
 
 	ret = i2c_add_numbered_adapter(adap);
 	if (ret)

-- 
2.51.0

Re: [PATCH v2 1/2] i2c: designware: convert to dev_err_probe() on request IRQ error

[Andy Shevchenko]

On Thu, Sep 04, 2025 at 04:31:06PM +0200, Benoît Monin wrote:
> Simplify the error handling of devm_request_irq() in
> i2c_dw_probe_master() and i2c_dw_probe_slave() by converting to:
> 
>     return dev_err_probe();
> 
> instead of calling:
> 
>     dev_err();
>     return ret;
> 
> This also handle deferred probe error without spamming the log.

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v2 1/2] i2c: designware: convert to dev_err_probe() on request IRQ error

[Jarkko Nikula]

On 9/4/25 6:02 PM, Andy Shevchenko wrote:
> On Thu, Sep 04, 2025 at 04:31:06PM +0200, Benoît Monin wrote:
>> Simplify the error handling of devm_request_irq() in
>> i2c_dw_probe_master() and i2c_dw_probe_slave() by converting to:
>>
>>      return dev_err_probe();
>>
>> instead of calling:
>>
>>      dev_err();
>>      return ret;
>>
>> This also handle deferred probe error without spamming the log.
> 
> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
> 
Acked-by: Jarkko Nikula <jarkko.nikula@linux.intel.com>

[PATCH v2 2/2] i2c: designware: use dev_err_probe() when probing platform device

[Benoît Monin]

Add calls to dev_err_probe() on error paths that can return
-EPROBE_DEFER when probing platform device. Namely when requesting the
reset controller, when probing for lock support and when requesting the
clocks.

PCI device probing already use dev_err_probe().

Signed-off-by: Benoît Monin <benoit.monin@bootlin.com>
---
 drivers/i2c/busses/i2c-designware-platdrv.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/i2c/busses/i2c-designware-platdrv.c b/drivers/i2c/busses/i2c-designware-platdrv.c
index a35e4c64a1d46f43aa2d37c0d20fbbd4bc1ff600..be2f9533cef94e3d07cb52401528fe5c1985198b 100644
--- a/drivers/i2c/busses/i2c-designware-platdrv.c
+++ b/drivers/i2c/busses/i2c-designware-platdrv.c
@@ -238,7 +238,7 @@ static int dw_i2c_plat_probe(struct platform_device *pdev)
 
 	dev->rst = devm_reset_control_get_optional_exclusive(device, NULL);
 	if (IS_ERR(dev->rst))
-		return PTR_ERR(dev->rst);
+		return dev_err_probe(device, PTR_ERR(dev->rst), "failed to acquire reset\n");
 
 	reset_control_deassert(dev->rst);
 
@@ -247,21 +247,23 @@ static int dw_i2c_plat_probe(struct platform_device *pdev)
 		goto exit_reset;
 
 	ret = i2c_dw_probe_lock_support(dev);
-	if (ret)
+	if (ret) {
+		ret = dev_err_probe(device, ret, "failed to probe lock support\n");
 		goto exit_reset;
+	}
 
 	i2c_dw_configure(dev);
 
 	/* Optional interface clock */
 	dev->pclk = devm_clk_get_optional(device, "pclk");
 	if (IS_ERR(dev->pclk)) {
-		ret = PTR_ERR(dev->pclk);
+		ret = dev_err_probe(device, PTR_ERR(dev->pclk), "failed to acquire pclk\n");
 		goto exit_reset;
 	}
 
 	dev->clk = devm_clk_get_optional(device, NULL);
 	if (IS_ERR(dev->clk)) {
-		ret = PTR_ERR(dev->clk);
+		ret = dev_err_probe(device, PTR_ERR(dev->clk), "failed to acquire clock\n");
 		goto exit_reset;
 	}
 

-- 
2.51.0

Re: [PATCH v2 2/2] i2c: designware: use dev_err_probe() when probing platform device

[Andy Shevchenko]

On Thu, Sep 04, 2025 at 04:31:07PM +0200, Benoît Monin wrote:
> Add calls to dev_err_probe() on error paths that can return
> -EPROBE_DEFER when probing platform device. Namely when requesting the
> reset controller, when probing for lock support and when requesting the
> clocks.
> 
> PCI device probing already use dev_err_probe().

Makes sense by at least two aspects:
1) less log spamming for deferred probes;
2) easier to debug an issue (I assume it's your case).

Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v2 2/2] i2c: designware: use dev_err_probe() when probing platform device

[Jarkko Nikula]

On 9/4/25 6:04 PM, Andy Shevchenko wrote:
> On Thu, Sep 04, 2025 at 04:31:07PM +0200, Benoît Monin wrote:
>> Add calls to dev_err_probe() on error paths that can return
>> -EPROBE_DEFER when probing platform device. Namely when requesting the
>> reset controller, when probing for lock support and when requesting the
>> clocks.
>>
>> PCI device probing already use dev_err_probe().
> 
> Makes sense by at least two aspects:
> 1) less log spamming for deferred probes;
> 2) easier to debug an issue (I assume it's your case).
> 
> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
> 
Acked-by: Jarkko Nikula <jarkko.nikula@linux.intel.com>