Re: [PATCH 24/26] dmaengine: ti-edma: Remove redundant dev_err()/dev_err_probe()

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] Unhandled probe deferral from `platform_get_irq_byname` causes the driver to silently probe without interrupts.
--

--- Patch [24]: [PATCH 24/26] dmaengine: ti-edma: Remove redundant dev_err()/dev_err_probe() ---
commit 3f067543fc6aa46c8b00f2557e8220b1465e86e3
Author: Pan Chuang <panchuang@vivo.com>

dmaengine: ti-edma: Remove redundant dev_err()/dev_err_probe()

The commit removes redundant dev_err() logging around devm_request_irq()
calls. The devm_request_irq() function now automatically logs detailed error
messages on failure, eliminating the need for driver-specific messages.

> diff --git a/drivers/dma/ti/edma.c b/drivers/dma/ti/edma.c
> index d97db5af35555..e023ca9640dec 100644
> --- a/drivers/dma/ti/edma.c
> +++ b/drivers/dma/ti/edma.c
> @@ -2415,10 +2415,8 @@ static int edma_probe(struct platform_device *pdev)
>  
>  	irq = platform_get_irq_byname(pdev, "edma3_ccint");
>  	if (irq < 0 && node)
>  		irq = irq_of_parse_and_map(node, 0);
>  
>  	if (irq > 0) {
>  		irq_name = devm_kasprintf(dev, GFP_KERNEL, "%s_ccint",
>  					  dev_name(dev));
>  		if (!irq_name) {
>  			ret = -ENOMEM;
>  			goto err_disable_pm;
>  		}
>  
>  		ret = devm_request_irq(dev, irq, dma_irq_handler, 0, irq_name,
>  				       ecc);
> -		if (ret) {
> -			dev_err(dev, "CCINT (%d) failed --> %d\n", irq, ret);
> +		if (ret)
>  			goto err_disable_pm;
> -		}
>  		ecc->ccint = irq;
>  	}

[Severity: High]
This is a pre-existing issue, but does this logic handle probe deferrals
correctly?

If platform_get_irq_byname() returns -EPROBE_DEFER, the irq value is
either overwritten to 0 by irq_of_parse_and_map(), or remains -EPROBE_DEFER.
In both scenarios, the if (irq > 0) check fails, causing the driver to
silently skip interrupt setup and successfully complete the probe without
interrupts.

Could this lead to dependent devices hanging or experiencing timeouts
because DMA completions are never signaled?

[ ... ]

> @@ -2436,10 +2434,8 @@ static int edma_probe(struct platform_device *pdev)
>  
>  	irq = platform_get_irq_byname(pdev, "edma3_ccerrint");
>  	if (irq < 0 && node)
>  		irq = irq_of_parse_and_map(node, 2);
>  
>  	if (irq > 0) {
>  		irq_name = devm_kasprintf(dev, GFP_KERNEL, "%s_ccerrint",
>  					  dev_name(dev));
>  		if (!irq_name) {
>  			ret = -ENOMEM;
>  			goto err_disable_pm;
>  		}
>  
>  		ret = devm_request_irq(dev, irq, dma_ccerr_handler, 0, irq_name,
>  				       ecc);
> -		if (ret) {
> -			dev_err(dev, "CCERRINT (%d) failed --> %d\n", irq, ret);
> +		if (ret)
>  			goto err_disable_pm;
> -		}
>  		ecc->ccerrint = irq;
>  	}

[Severity: High]
This isn't a bug introduced by this patch, but the error interrupt setup
appears to suffer from the same problem.

Would an unhandled probe deferral from platform_get_irq_byname() result
in silently swallowed errors here as well?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260709135846.97972-1-panchuang@vivo.com?part=24