Re: [PATCH 5/7] mm/damon/sysfs-schemes: support probe_hits_wsum damos core filter

[SJ Park <sj@kernel.org>]

On Thu, 10 Sep 2026 14:30:13 +0000 sashiko-bot@kernel.org wrote:

> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Medium] Time-of-Check to Time-of-Use race when validating probe_hits_wsum range allows invalid filter configurations to bypass validation.
> --
> 
> commit 27bc99697c36a87c8e4b72fe6006955a564a4bb2
> Author: SJ Park <sj@kernel.org>
> 
> mm/damon/sysfs-schemes: support probe_hits_wsum damos core filter
> 
> Extend DAMON sysfs interface to support probe_hits_wsum input. Also
> update sysfs input based scheme build logic to setup the min/max probe
> hits weighted sum range as user provided.
> 
> > diff --git a/mm/damon/sysfs-schemes.c b/mm/damon/sysfs-schemes.c
> > index 4d9147d2a269e..d9dd4ee0442c7 100644
> > --- a/mm/damon/sysfs-schemes.c
> > +++ b/mm/damon/sysfs-schemes.c
> 
> [ ... ]
> 
> > @@ -2844,6 +2847,14 @@ static int damon_sysfs_add_scheme_filters(struct damos *scheme,
> >  			}
> >  			filter->sz_range.min = sysfs_filter->range_min;
> >  			filter->sz_range.max = sysfs_filter->range_max;
> > +		} else if (filter->type == DAMOS_FILTER_TYPE_PROBE_HITS_WSUM) {
> > +			if (sysfs_filter->range_min >
> > +					sysfs_filter->range_max) {
> > +				damos_destroy_filter(filter);
> > +				return -EINVAL;
> > +			}
> > +			filter->range_min = sysfs_filter->range_min;
> > +			filter->range_max = sysfs_filter->range_max;
> 
> [Severity: Medium]
> Are these repeated reads of sysfs_filter->range_min and sysfs_filter->range_max
> susceptible to a time-of-check to time-of-use race?
> 
> If a privileged user concurrently writes to a filter's min or max sysfs
> attribute between the validation check and the assignments:
> 
>     if (sysfs_filter->range_min > sysfs_filter->range_max) {
>         // Validation passes
>     }
> 
>     // Concurrent sysfs write updates range_min here
> 
>     filter->range_min = sysfs_filter->range_min; // Now invalid
> 
> Could this allow an invalid configuration to bypass the check and be added to
> the DAMON scheme? Would it be safer to read the values into local variables
> first, validate those local variables, and then assign them to the filter?

Good finding, I agree this is better to be fixed.  And maybe we could move the
validation to the core layer parameters validation logic.

I will fix this in the next revision.

> 
> >  		}
> >  
> >  		damos_add_filter(scheme, filter);
> 
> -- 
> Sashiko AI review · https://sashiko.dev/#/patchset/20260910141849.170896-1-sj@kernel.org?part=5


Thanks,
SJ