User: Password:
Subscribe / Log in / New account

Re: [PATCH 5/6] workqueue: introduce NR_WORKER_POOLS and for_each_worker_pool()

From:  Linus Torvalds <>
To:  Tejun Heo <>
Subject:  Re: [PATCH 5/6] workqueue: introduce NR_WORKER_POOLS and for_each_worker_pool()
Date:  Fri, 13 Jul 2012 22:00:10 -0700
Message-ID:  <>
Archive-link:  Article

On Fri, Jul 13, 2012 at 9:44 PM, Tejun Heo <> wrote:
> nr_running is atomic_t (*nr_running)[2].  Ignoring the pointer to
> array part, it's just returning the address of N'th element of the
> array.  ARRAY + N == &ARRAY[N].

None of this matters one whit.

You did "&(x)[0]".

That's insane. It's crazy. It doesn't even matter what "x" is in
between, it's crazy regardless.

It's just a really confused way of saying "x" (*). Except it makes the
code look like an insane monkey on crack got a-hold of your keyboard
when you weren't looking.

And to make it worse, "x" itself was the result of doing "*&y". Which
was probably written by the insane monkey's older brother, Max, who
has been chewing Quaaludes for a few years, and as a result _his_
brain really isn't doing too well either. Even for a monkey. And now
you're letting *him* at your keyboard too?

So you had two separately (but similarly) insane ways of complicating
the code so that it was really obfuscated. When it really just
computed "y" to begin with, it just added all those "x=*&y" and
"&(x)[0]" games around it to make it look complicated.


(*) Technically, "&(x)[0]" is actually a really confused way of saying
"(x+0)" while making sure that "x" was a valid pointer. It basically
guarantees that if "x" started out as an array, it has now been
demoted to a pointer - but since arrays will be demoted to pointers by
pretty much any subsequent operation except for "sizeof()" and a
couple of other special cases anyway, you can pretty much just say
that "&(x)[0]" is "(x+0)" is "x".

And "*&y" really is exactly the same as "y", except for again some
syntactic checking (ie it is basically an odd way to verify that "y"
is an lvalue, since you cannot do an address-of of a non-lvalue).

(Log in to post comments)

Re: [PATCH 5/6] workqueue: introduce NR_WORKER_POOLS and for_each_worker_pool()

Posted Jul 21, 2012 10:03 UTC (Sat) by mgedmin (subscriber, #34497) [Link]

I was unable to find the parent post to this via GMane's web UI. All I get is Nobody here but us chickens". Looks like gmane can't handle cross-posting when parts of the thread aren't cross-posted to the original newsgroups.

Copyright © 2012, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds