Re: [Regression]: mm: nommu: Memory leak introduced with commit "mm/nommu: use alloc_pages_exact() rather than its own implementation"

From: Maxime Coquelin
Date: Mon Feb 16 2015 - 06:45:30 EST


Hello Joonsoo,

2015-02-16 5:43 GMT+01:00 Joonsoo Kim <js1304@xxxxxxxxx>:
>
> Hello,
>
> Sorry for my mistake.
> Problem happens because when we allocate memory through
> __get_free_pages(), refcount of each pages is not 1 except
> head page. Below modification will fix your problem. Could you
> test it, please?

I just tested it, and confirm it fixes the regression.

You can add my:
Tested-by: Maxime Coquelin <mcoquelin.stm32@xxxxxxxxx>

Thanks for the quick fix!
Maxime

>
> Thanks.
>
> ------------>8-------------
> diff --git a/mm/nommu.c b/mm/nommu.c
> index 28bd8c4..ff6c1e2 100644
> --- a/mm/nommu.c
> +++ b/mm/nommu.c
> @@ -1189,11 +1189,9 @@ static int do_mmap_private(struct vm_area_struct *vma,
> if (sysctl_nr_trim_pages && total - point >= sysctl_nr_trim_pages) {
> total = point;
> kdebug("try to alloc exact %lu pages", total);
> - base = alloc_pages_exact(len, GFP_KERNEL);
> - } else {
> - base = (void *)__get_free_pages(GFP_KERNEL, order);
> }
>
> + base = alloc_pages_exact(total << PAGE_SHIFT, GFP_KERNEL);
> if (!base)
> goto enomem;
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/