Re: [PATCH v4 12/12] drm/xe/xe_ras: Report correctable errors using SIGID

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] xe_ras: WARN on valid XE_RAS_COMP_RESERVED telemetry
--

commit ce7ef0dce322046ec43ef7cd45243457eb7f0974
Author: Badal Nilawar <badal.nilawar@intel.com>

drm/xe/xe_ras: Report correctable errors using SIGID

This commit updates the reporting of correctable RAS errors to use the
xe_log_from helper instead of xe_warn. It introduces mapping functions to
convert RAS severity and components to the appropriate logging types and IDs.

> diff --git a/drivers/gpu/drm/xe/xe_ras.c b/drivers/gpu/drm/xe/xe_ras.c
> --- a/drivers/gpu/drm/xe/xe_ras.c
> +++ b/drivers/gpu/drm/xe/xe_ras.c
[ ... ]
> @@ -459,8 +467,11 @@ void xe_ras_counter_threshold_crossed(struct xe_device *xe,
>  		if (!ras_counter_is_valid(xe, &errors[id]))
>  			continue;
>  
> -		xe_warn(xe, "[RAS]: %s %s detected\n",
> -			comp_to_str(component), sev_to_str(severity));
> +		xe_log_from(xe, ras_sev_to_cper_sev(severity),
> +			    ras_comp_to_hw_sigid(component),
> +			    ras_comp_to_log_comp(component),
> +			    &errors[id], sizeof(errors[id]),
> +			    "error detected\n");

[Severity: Medium]
Could this trigger a WARN for the XE_RAS_COMP_RESERVED component ID?

Looking at ras_counter_is_valid(), it validates the component using:
in_range(component, XE_RAS_COMP_NOT_SUPPORTED + 1, XE_RAS_COMP_MAX - 1)

Since in_range(val, start, len) evaluates val >= start && val < start + len,
passing in component 3 (XE_RAS_COMP_RESERVED) returns true, allowing it to
pass the validity check.

When this reserved component is passed to xe_log_from(), it triggers the
logging subsystem. Because the xe_log macro framework lacks a definition
for HARDWARE, 3, log_component_prefix() falls back to
log_unknown_component_prefix(), which triggers a kernel WARN().

This warning will crash systems configured with panic_on_warn when they
encounter expected or reserved hardware telemetry data. Should we explicitly
filter out reserved component IDs before passing them to the logging framework?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261009104602.1075405-14-badal.nilawar@intel.com?part=12