Re: [PATCH v5 19/33] drm/xe: Report all probe errors using SIGID

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] Unconditionally logging all probe failures as fatal errors causes normal probe deferrals (-EPROBE_DEFER) to be incorrectly logged as fatal hardware errors.
--

--- Patch [19]: [PATCH v5 19/33] drm/xe: Report all probe errors using SIGID ---
commit c6ce1e9226a28ea2318a390f9b11cf4044410b13
Author: Michal Wajdeczko <[email protected]>

drm/xe: Report all probe errors using SIGID

This commit catches and reports all probe errors by splitting xe_pci_probe()
and using xe_log_err_fatal() on the errors returned from __xe_pci_probe().

> diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
> index 4068b6bf9eaa7..1e04e8ef2611f 100644
> --- a/drivers/gpu/drm/xe/xe_pci.c
> +++ b/drivers/gpu/drm/xe/xe_pci.c
[ ... ]
> @@ -1180,6 +1175,26 @@ static int xe_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
>  	if (xe_display_driver_probe_defer(pdev))
>  		return -EPROBE_DEFER;
>  
> +	err = __xe_pci_probe(pdev, desc);
> +	if (err) {
> +		xe_log_err_fatal(pdev, PROBE, err, "driver loading failed for device '%04x'\n",
> +				 pdev->device);

[Severity: Medium]
Could __xe_pci_probe() still return -EPROBE_DEFER from other initialization
steps?

While the explicit xe_display_driver_probe_defer() check was kept outside to
avoid logging -EPROBE_DEFER as a fatal error, __xe_pci_probe() calls into
several other subsystems (such as xe_display_probe() or xe_pm_probe()) which
might also legitimately return -EPROBE_DEFER.

If they do, wouldn't this unconditionally log normal probe deferrals as
fatal errors?

> +		return err;
> +	}
> +
> +	return 0;
> +}

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=19