Thread (1 message) 1 message, 1 author, 2014-12-05

Re: [PATCH 1/9] kernel: Provide READ_ONCE and ASSIGN_ONCE

From: George Spelvin <hidden>
Date: 2014-12-05 21:38:05
Also in: lkml

Possibly related (same subject, not in this thread)

Of prefer it to match the put_user model, which is (val, ptr). But as long
as there is your safety (and the whole point of the macro is that it
figures out the type from the pointer), I guess it doesn't matter too much
in practice.

I think my original suggestion also wrote it in lower case, since it
actually works like a function (well, template, whatever). Again kind of
like the user copy "functions".

But I don't care *that* strongly.
That's an excellent point.  But it means that Christian, you're free to
do whatever seems good to you.

I'm not rabid about it, either; I just figure that now is a good time
to think carefully about it.

Personally, after running through some different names that work
with (src,dst) order I'd try the pairs:

	READ_ONCE(var) / WRITE_ONCE(value,var)
or
	LOAD_ONCE(var) / STORE_ONCE(value,var)

Of the two, I think I prefer the latter, because that's what the
operations are called at the hardware level, so it's most evocative of
the actual semantics.  "Assignment" is a higher-level concept.

But as I said, whatever you think best.  I'd just like to advocate for
some actual thinking, because it's going to be all over the kernel and
a pain to change.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help