Re: [PATCH 14/19] mm: Add user_landing in mm_struct

[Dmitry Safonov]

On 11/8/20 7:04 PM, Andy Lutomirski wrote:
> On Sat, Nov 7, 2020 at 9:18 PM Dmitry Safonov <dima@xxxxxxxxxx> wrote:
>>
>> Instead of having every architecture to define vdso_base/vdso_addr etc,
>> provide a generic mechanism to track landing in userspace.
>> It'll minimize per-architecture difference, the number of callbacks to
>> provide.
>>
>> Originally, it started from thread [1] where the need for .close()
>> callback on vm_special_mapping was pointed, this generic code besides
>> removing duplicated .mremap() callbacks provides a cheaper way to
>> support munmap() on vdso mappings without introducing .close() callbacks
>> for every architecture (with would bring even more code duplication).
> 
> I find the naming odd.  It's called "user_landing", which is
> presumably a hard-to-understand shorthand for "user mode landing pad
> for return from a signal handler if SA_RESTORER is not set".  But,
> looking at the actual code, it's not this at all -- it's just the vDSO
> base address.

Agree. Originally, I tried to track the actual landing address on the
vdso, but .mremap() seemed simpler when tracking the vma base.

> So how about just calling it vdso_base?  I'm very much in favor of
> consolidating and cleaning up, and improving the vdso remap/unmap
> code, but I'm not convinced that we should call it anything other than
> the vdso base.

Sure.

Thanks,
         Dmitry