RE: ROUND_UP|DOWN macro cleanup in arch/ia64,x86_64

> So as ALIGN macro does the same work as ROUNDUP,

Although it is mathematically the same operation, the
semantic associations of the name are important too.
If I have an I/O device that works in blocks of a given
size, I don't think that I'm "aligning" a request
to make
it match the capabilities of the device, I think that I'm
"rounding up" to a multiple of the size.

Maybe this is because I started out in mathematics before
discovering that computers were so much fun, or maybe it
is a British-English bias ... I can't tell, but it makes
sense to me to use ROUNDUP in some places, and ALIGN in
others.

I haven't scanned through all usages to see whether
existing
usage matches my bias ... but changing them all to use just
one name doesn't feel like the right thing to do.  It makes
some code less clear (to me).

>      is at a common place
>      & is accessible to everyone

If ROUNDUP isn't available everywhere, then it should be.
To
avoid code duplication perhaps we should add:

    #define ROUNDUP(size, len) ALIGN((size), (len))

and delete the previous ROUNDUP definition?

-Tony
-
To unsubscribe from this list: send the line
"unsubscribe linux-ia64" in
the body of a message to majordomovger.kernel.org
More majordomo info at  http://vge
r.kernel.org/majordomo-info.html

* From: "Luck, Tony"
* Date: Fri, 13 Apr 2007 09:45:49 -0700

>> So as ALIGN macro does the same work as ROUNDUP,
>
> Although it is mathematically the same operation, the
> semantic associations of the name are important too.
[]
> Maybe this is because I started out in mathematics
before
> discovering that computers were so much fun, or maybe
it
> is a British-English bias ... I can't tell, but it
makes
> sense to me to use ROUNDUP in some places, and ALIGN
in
> others.

I agree. I came from neither one (still coming); first look
and vague
semantics of ALIGN macro, English confusion, a bit of
laziness, led me to
stupid messages about aligning, Linus being so kind to
reply, a couple of
months ago. Anyway i did that after some patching and
re-patching that
macro by gurus .

> If ROUNDUP isn't available everywhere, then it should
be. To
> avoid code duplication perhaps we should add:
>
>     #define ROUNDUP(size, len) ALIGN((size), (len))
>
> and delete the previous ROUNDUP definition?

And i would vote for "align" via
"round".
____
-
To unsubscribe from this list: send the line
"unsubscribe linux-ia64" in
the body of a message to majordomovger.kernel.org
More majordomo info at  http://vge
r.kernel.org/majordomo-info.html