Still applies. I'm doing my tests on 2.2.1 as well.
> 2. I'd really like to try the page coloring stuff, but I don't want
> to spend a lot of time testing/tweaking other stuff (IE the devfs)
> stuff.
The page colouring stuff doesn't depend on devfs. Extract the diffs
for the following files:
linux/Documentation/page-colouring.txt
linux/fs/proc/mem.c
linux/include/linux/proc_fs.h
linux/include/linux/sched.h
linux/mm/memory.c
> 3. the ioctl() approach to turn this on (per process) seems like it
> might well be a reasonable approach to avoid the overhead for
> non-cpu bound processes that aren't going to give a squat about
> this. I'd assume that I could write a simple piece of code that
> does the ioctl() call, then does an exec() to the real program (ie
> 'color xyz' would enable coloring and then execute xyz)???
I think that should work. It's easy enough to test.
> 4. What's the chance of your extracting just the coloring code and
> doing a separate patch that is easier to play with?
I'd rather not. It's a pain for me to maintain the different patches
I'm already maintaining. And it's easy enough to take an editor to the
msr patch and just use the page colouring bits.
I now also consider the devfs stuff to be high quality and low
instrusion, so it shouldn't cause problems for those who don't want to
test devfs itself. To be really sure, just pass "devfs=nomount" on
your boot line and it won't even get mounted.
Regards,
Richard....
-
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/