Re: [PATCH v2] Introduce a version6 of autofs interface, to fix design error.

[David Miller <davem@davemloft.net>]

From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Fri, 27 Apr 2012 11:55:12 -0700

> On Fri, Apr 27, 2012 at 11:42 AM, Linus Torvalds
> <torvalds@linux-foundation.org> wrote:
>>
>> There's no question that systemd is broken.
> 
> Actually, I'll take that back.
> 
> Yes, systemd has breakage. But it's actually automount that is the
> truly broken piece of sh*t.
> 
> I think that 'automount' is even more broken. The fact that the
> automount maintainers knew about this, and added TOTALLY BROKEN code
> to their automount source tree, over five years ago, because the
> authors clearly did not understand what the f*ck they were doing,
> that's the real problem.

I respectfully disagree.

It's ugly as shit, but it is the only one place where one can be
absolutely sure that we are dealing with a pipe passing those v5
things around.

All these hacks we have been talking about, assuming the mount means
the pipe is for passing structure so-and-so around, and now trying
to find some other check such as one on current->comm...

That's better?

Only the application really knows.  And I bet the person who wrote
that automountd code you find so distasteful analyzed this and came
to realize how difficult the kernel side would be to get right.

So they decided that the datastructure on 32-bit changes based upon
whether we run on a 64-bit kernel or not, that's the interface the
kernel has always exported, and therefore that's the interface it
choose to use.