[PATCH] lib/string: fix memchr_inv() for large ranges

[PATCH] lib/string: fix memchr_inv() for large ranges

[Bradley Morgan]

memchr_inv() takes a size_t length but counts 8 byte words in an
unsigned int. At 32GiB that count wraps, so the scan can quietly miss
most of the range.

Use size_t for the word count.

Fixes: 798248206b59 ("lib/string.c: introduce memchr_inv()")
Signed-off-by: Bradley Morgan <include@grrlz.net>
---
 lib/string.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/lib/string.c b/lib/string.c
index 1f9297e9776a..8ea7d4b9f0c0 100644
--- a/lib/string.c
+++ b/lib/string.c
@@ -837,7 +837,8 @@ void *memchr_inv(const void *start, int c, size_t bytes)
 {
 	u8 value = c;
 	u64 value64;
-	unsigned int words, prefix;
+	size_t words;
+	unsigned int prefix;
 
 	if (bytes <= 16)
 		return check_bytes8(start, value, bytes);
-- 
2.53.0

Re: [PATCH] lib/string: fix memchr_inv() for large ranges

[Andy Shevchenko]

On Sun, Jun 21, 2026 at 12:11:33PM +0000, Bradley Morgan wrote:
> memchr_inv() takes a size_t length but counts 8 byte words in an
> unsigned int. At 32GiB that count wraps, so the scan can quietly miss
> most of the range.

Wow! Is it real life use case?

> Use size_t for the word count.

...

> @@ -837,7 +837,8 @@ void *memchr_inv(const void *start, int c, size_t bytes)
>  {
>  	u8 value = c;
>  	u64 value64;
> -	unsigned int words, prefix;
> +	size_t words;
> +	unsigned int prefix;

Perhaps move towards reversed xmas tree while at it?

	unsigned int prefix;
	size_t words;
	u8 value = c;
	u64 value64;

(same lines touched, no extra).

>  	if (bytes <= 16)
>  		return check_bytes8(start, value, bytes);

-- 
With Best Regards,
Andy Shevchenko