Re: [PATCH] tools/testing/shared: catch a failed posix_memalign()

[Vishal Moola (Fractile)]

On Mon, Oct 19, 2026 at 11:31:28PM -0400, Liam R. Howlett (Google) wrote:
> posix_memalign() returns zero or a positive error number and never sets
> errno, so the checks for a negative return in kmem_cache_alloc_lru() and
> kmem_cache_alloc_bulk() is not possible.  A failed aligned allocation
> carries on with the pointer unset, into the constructor or the memset().
> 
> Test for any non-zero return.
> 
> Signed-off-by: Liam R. Howlett (Google) <liam@xxxxxxxxxxxxx>

Reviewed-by: Vishal Moola (Fractile) <vishal.moola@xxxxxxxxx>

It's not really a big deal, but seems this was introduced in commit
8801c35c3672c ("tools: fix -Wunused-result in linux.c"), so +CC the
author Shuah Khan.

>  tools/testing/shared/linux.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/tools/testing/shared/linux.c b/tools/testing/shared/linux.c
> index e0a0693df08f5..a37d2f86d3254 100644
> --- a/tools/testing/shared/linux.c
> +++ b/tools/testing/shared/linux.c
> @@ -83,7 +83,7 @@ void *kmem_cache_alloc_lru(struct kmem_cache *cachep, struct list_lru *lru,
>  	} else {
>  		pthread_mutex_unlock(&cachep->lock);
>  		if (cachep->align) {
> -			if (posix_memalign(&p, cachep->align, cachep->size) < 0)
> +			if (posix_memalign(&p, cachep->align, cachep->size))
>  				return NULL;
>  		} else {
>  			p = malloc(cachep->size);
> @@ -191,7 +191,7 @@ bool kmem_cache_alloc_bulk(struct kmem_cache *cachep, gfp_t gfp, size_t size,
>  
>  			if (cachep->align) {
>  				if (posix_memalign(&p[i], cachep->align,
> -					       cachep->size) < 0)
> +					       cachep->size))
>  					break;
>  			} else {
>  				p[i] = malloc(cachep->size);
> -- 
> 2.47.3
>