I hadn't coutned them, but I'm not surprised: when Alan started feeding me
the cleanup patches it fairly quickly became obvious that something was
wrong.
I don't think all the PROC_FS config issues have to be gone, but yes, they
should probably shrink by an order of magnitude. I don't think that should
be an impossible goal.
> (the worst #ifdef offender btw. is MODULES, but those are mostly clean and
> isolated. Could be solved with a section trick though ...)
I don't worry about MODULE ifconfig's, because those are "local"
decisions, really - an issue of whether a single driver is compiled as a
module or as a integral part of the kernel. As such it doesn't break any
concept of modularity - the difference could be seen as the difference
between a global config that impacts things all over (CONFIG_PROC_FS) and
tons of small local configs that just share a common concept (MODULE).
One thing I _have_ considered using a "section" trick for is the "init"
routines. Right now one of the uglier files is drivers/net/Space.c, which
is just endless "#ifconfig"s - and it should be possible to reasonably
easily create a section for "device driver probe" addresses, and thus
spread out these things into the drivers that actually implement them.
(The same is true of filesystems and memory devices, for example)
However, that's definitely a 2.3 issue. In contrast, the CONFIG_PROC_FS
thing could probably be considered just a cleanup patch (I'd think it
should be pretty trivial to remove quite a noticeable portion of them with
an "obviously correct" kind of patch).
Linus
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/