On Fri, Apr 17, 2020 at 02:05:50PM -0500, Eric W. Biederman wrote:

Alexey Gladkov [off-list ref] writes:

quoted

The hidepid parameter values are becoming more and more and it becomes
difficult to remember what each new magic number means.

So I relooked at the code.  And I think I was misreading things.
However I think it is a legitimate concern.

Can you please mention in your description of this change that
switching from fsparam_u32 to fs_param_string is safe even when
using the new mount api because fsparam_u32 and fs_param_string
both are sent from userspace with "fsconfig(fd, FSCONFIG_SET_STRING, ...)".

Sure.

Or words to that effect.  Ideally you will even manually test that case
to confirm.

I will add a selftest for this.

-- 
Rgrds, legion