Peter Samuelson writes:
> Not that any of us who don't do embedded projects ought to care very
> much, but I was curious.  I grepped test9pre6 for globals initialized
> to 0 or NULL and came up with 2495 lines, first iteration.

I did kick it off, but I've not had the time recently to go through and
find all the others; yes, there are a lot of them around in the kernel
tree.  The main problem anyone will face with this is that a lot of people
remember when there wasn't a zero initialised bss in the kernel, and
the tendency to initialise to zero has become a habbit, or is left in
some ancient documents that new people read.  But whatever the reason,
patches still get generated which add code with explicit zero initialisers.

> A lot of recent Linus patches seem to have this conversion in them, so
> it seems someone *does* care about image size.  (Presumably people
> running the kernel in-place from flash, right?)  Would anyone be
> interested in patches to uninitialize these variables?

Probably the best thing to do would be to send them to linux-kernel,
other relevent mailing list or maintainer (see MAINTAINERS/CREDITS file)
in small per-file chunks and copy them to Linus.

> (A related question: __initdata *does* have to be initialized, right?)

Yes.  Anything that does into __initdata by definition has to be initialised,
so there will be some zero initialisers left over.
   _____
  |_____| ------------------------------------------------- ---+---+-
  |   |         Russell King        [EMAIL PROTECTED]      --- ---
  | | | | http://www.arm.linux.org.uk/personal/aboutme.html   /  /  |
  | +-+-+                                                     --- -+-
  /   |               THE developer of ARM Linux              |+| /|\
 /  | | |                                                     ---  |
    +-+-+ -------------------------------------------------  /\\\  |
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
Please read the FAQ at http://www.tux.org/lkml/

Reply via email to