[PATCH] lib: interval_tree_test: validate benchmark parameters

[PATCH] lib: interval_tree_test: validate benchmark parameters

[Samuel Moelius]

The interval tree runtime test accepts module parameters that are later
used as divisors while generating randomized intervals and while
reporting average timings. For example, max_endpoint=1 makes the
generated interval end value zero and the next modulo operation divides
by that zero value.

Reject non-positive counts and require max_endpoint to provide at least
one non-zero generated endpoint before the test allocates state or
starts the benchmark.

Assisted-by: Codex:gpt-5.5-cyber-preview
Signed-off-by: Samuel Moelius <sam.moelius@trailofbits.com>
---
 lib/interval_tree_test.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/lib/interval_tree_test.c b/lib/interval_tree_test.c
index 16200feacbf3..58f6f3a307f4 100644
--- a/lib/interval_tree_test.c
+++ b/lib/interval_tree_test.c
@@ -311,6 +311,27 @@ static inline int span_iteration_check(void) {return 0; }
 
 static int interval_tree_test_init(void)
 {
+	if (nnodes <= 0) {
+		pr_warn("nnodes must be positive\n");
+		return -EINVAL;
+	}
+	if (nsearches <= 0) {
+		pr_warn("nsearches must be positive\n");
+		return -EINVAL;
+	}
+	if (perf_loops <= 0) {
+		pr_warn("perf_loops must be positive\n");
+		return -EINVAL;
+	}
+	if (search_loops <= 0) {
+		pr_warn("search_loops must be positive\n");
+		return -EINVAL;
+	}
+	if (max_endpoint < 2) {
+		pr_warn("max_endpoint must be at least 2\n");
+		return -EINVAL;
+	}
+
 	nodes = kmalloc_objs(struct interval_tree_node, nnodes);
 	if (!nodes)
 		return -ENOMEM;
-- 
2.43.0

Re: [PATCH] lib: interval_tree_test: validate benchmark parameters

[Andrew Morton]

On Tue,  9 Jun 2026 00:54:47 +0000 Samuel Moelius <sam.moelius@trailofbits.com> wrote:

> The interval tree runtime test accepts module parameters that are later
> used as divisors while generating randomized intervals and while
> reporting average timings. For example, max_endpoint=1 makes the
> generated interval end value zero and the next modulo operation divides
> by that zero value.
> 
> Reject non-positive counts and require max_endpoint to provide at least
> one non-zero generated endpoint before the test allocates state or
> starts the benchmark.

That makes sense, thanks.

> --- a/lib/interval_tree_test.c
> +++ b/lib/interval_tree_test.c
> @@ -311,6 +311,27 @@ static inline int span_iteration_check(void) {return 0; }
>  
>  static int interval_tree_test_init(void)
>  {
> +	if (nnodes <= 0) {
> +		pr_warn("nnodes must be positive\n");
> +		return -EINVAL;
> +	}
> +	if (nsearches <= 0) {
> +		pr_warn("nsearches must be positive\n");
> +		return -EINVAL;
> +	}
> +	if (perf_loops <= 0) {
> +		pr_warn("perf_loops must be positive\n");
> +		return -EINVAL;
> +	}
> +	if (search_loops <= 0) {
> +		pr_warn("search_loops must be positive\n");
> +		return -EINVAL;
> +	}
> +	if (max_endpoint < 2) {
> +		pr_warn("max_endpoint must be at least 2\n");
> +		return -EINVAL;
> +	}
> +
>  	nodes = kmalloc_objs(struct interval_tree_node, nnodes);
>  	if (!nodes)
>  		return -ENOMEM;

Perhaps this can be done within the __param macro:

#define __param(type, name, init, msg)					\
	static type name = init;					\
	module_param(name, type, 0444);					\
	if (name <= 0) {						\
		pr_warn(__stringify(name) " must be positive\n");	\
		name = init;						\
	}								\
	MODULE_PARM_DESC(name, msg);

I'll add your patch as-is.  If you prefer the above approach, please send
along a replacement patch.

Re: [PATCH] lib: interval_tree_test: validate benchmark parameters

[Samuel Moelius]

On Tue, Jun 9, 2026 at 9:13 PM Andrew Morton <akpm@linux-foundation.org> wrote:
>
> On Tue,  9 Jun 2026 00:54:47 +0000 Samuel Moelius <sam.moelius@trailofbits.com> wrote:
>
> > The interval tree runtime test accepts module parameters that are later
> > used as divisors while generating randomized intervals and while
> > reporting average timings. For example, max_endpoint=1 makes the
> > generated interval end value zero and the next modulo operation divides
> > by that zero value.
> >
> > Reject non-positive counts and require max_endpoint to provide at least
> > one non-zero generated endpoint before the test allocates state or
> > starts the benchmark.
>
> That makes sense, thanks.
>
> > --- a/lib/interval_tree_test.c
> > +++ b/lib/interval_tree_test.c
> > @@ -311,6 +311,27 @@ static inline int span_iteration_check(void) {return 0; }
> >
> >  static int interval_tree_test_init(void)
> >  {
> > +     if (nnodes <= 0) {
> > +             pr_warn("nnodes must be positive\n");
> > +             return -EINVAL;
> > +     }
> > +     if (nsearches <= 0) {
> > +             pr_warn("nsearches must be positive\n");
> > +             return -EINVAL;
> > +     }
> > +     if (perf_loops <= 0) {
> > +             pr_warn("perf_loops must be positive\n");
> > +             return -EINVAL;
> > +     }
> > +     if (search_loops <= 0) {
> > +             pr_warn("search_loops must be positive\n");
> > +             return -EINVAL;
> > +     }
> > +     if (max_endpoint < 2) {
> > +             pr_warn("max_endpoint must be at least 2\n");
> > +             return -EINVAL;
> > +     }
> > +
> >       nodes = kmalloc_objs(struct interval_tree_node, nnodes);
> >       if (!nodes)
> >               return -ENOMEM;
>
> Perhaps this can be done within the __param macro:
>
> #define __param(type, name, init, msg)                                  \
>         static type name = init;                                        \
>         module_param(name, type, 0444);                                 \
>         if (name <= 0) {                                                \
>                 pr_warn(__stringify(name) " must be positive\n");       \
>                 name = init;                                            \
>         }                                                               \
>         MODULE_PARM_DESC(name, msg);
>
> I'll add your patch as-is.  If you prefer the above approach, please send
> along a replacement patch.

I looked at this, but it would seem to require a more complicated
__param macro. For example, max_endpoint must be 2 or more. Also, the
macro would need to emit a checking function.

If it's okay, I would like to please go with this patch.