On Sun 04-11-18 07:44:56, Yangtao Li wrote:
> WARN_ON() already contains an unlikely(), so it's not necessary to use
> unlikely.
We should just get rid of this ugliness altogether. It no longer serves
its purpose. This is a historical artifact from 2005 where do_brk
was called outside of the core mm. We do have a proper abstraction in
vm_brk_flags and that one does the locking properly.
>
> Signed-off-by: Yangtao Li <tiny.windzz@xxxxxxxxx>
> ---
> mm/mmap.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/mm/mmap.c b/mm/mmap.c
> index 6c04292e16a7..2077008ade0c 100644
> --- a/mm/mmap.c
> +++ b/mm/mmap.c
> @@ -2965,10 +2965,8 @@ SYSCALL_DEFINE5(remap_file_pages, unsigned long, start, unsigned long, size,
> static inline void verify_mm_writelocked(struct mm_struct *mm)
> {
> #ifdef CONFIG_DEBUG_VM
> - if (unlikely(down_read_trylock(&mm->mmap_sem))) {
> - WARN_ON(1);
> + if (WARN_ON(down_read_trylock(&mm->mmap_sem)))
> up_read(&mm->mmap_sem);
> - }
> #endif
> }
>
> --
> 2.17.0
>
--
Michal Hocko
SUSE Labs
