Re: [patch] mm, oom: add comment for why oom_adj exists

[Michal Hocko <mhocko@kernel.org>]

On Wed 04-11-15 12:32:14, David Rientjes wrote:
> /proc/pid/oom_adj exists solely to avoid breaking existing userspace 
> binaries that write to the tunable.
> 
> Add a comment in the only possible location within the kernel tree to 
> describe the situation and motivation for keeping it around.

I am not sure this is really needed but it certainly is not harmful.
If this is a way to suppress any attempts for changes like
http://lkml.kernel.org/r/1f80189385e540c2a5b2747a7a265d8c%40SHMBX01.spreadtrum.com
then it does not explain why those are not desirable.

> Signed-off-by: David Rientjes <rientjes@google.com>
> ---
>  fs/proc/base.c | 10 ++++++++++
>  1 file changed, 10 insertions(+)
> 
> diff --git a/fs/proc/base.c b/fs/proc/base.c
> --- a/fs/proc/base.c
> +++ b/fs/proc/base.c
> @@ -1032,6 +1032,16 @@ static ssize_t oom_adj_read(struct file *file, char __user *buf, size_t count,
>  	return simple_read_from_buffer(buf, count, ppos, buffer, len);
>  }
>  
> +/*
> + * /proc/pid/oom_adj exists solely for backwards compatibility with previous
> + * kernels.  The effective policy is defined by oom_score_adj, which has a
> + * different scale: oom_adj grew exponentially and oom_score_adj grows linearly.
> + * Values written to oom_adj are simply mapped linearly to oom_score_adj.
> + * Processes that become oom disabled via oom_adj will still be oom disabled
> + * with this implementation.
> + *
> + * oom_adj cannot be removed since existing userspace binaries use it.

This is a bit strong wording. I think the knob can be removed in the future.

* oom_adj is kept for compatibility reasons. There are still few
* projects which use oom_adj only. We have tried to convert all of them
* which could be found but it will take some time until all those changes
* bubble up to all users. We might try to remove the knob in few years
* if the situtation changes.

> + */
>  static ssize_t oom_adj_write(struct file *file, const char __user *buf,
>  			     size_t count, loff_t *ppos)
>  {

-- 
Michal Hocko
SUSE Labs