> The patches in question are NOT currently under consideration for the
> distribution kernel, at least not as I understand it. Basically,
yep, there are a few known problems left, and probably a few yet
unidentified.
> I understand it, the various patches are (A) incompatible with each
> other, and (B) largely untested other than by the author.
all patches I've seen do exactly the same - define HZ to something
other than "100" - it's all you have to do in order to increase HZ.
The differences are in how far they go to hide this change from userspace;
the common things being syscalls and '/proc' process/cpu info. There
are no compatibility issues (wrt eachover). (in fact i think all of
the patches do not change _anything_ unless you request a higher HZ.)
> However, I understand that those patches have brought to light certain
> other kernel problems which have been fixed as a result...
I'm afraid fixing some of them will take some time, as the fixes sometimes
are unobvious, require changes to apis and/or are controversial ;)
[one problem, which i didn't see mentioned or handled yet, is: anything
that assumess it can turn off interrupts for less then 1/100s. If you
set HZ to 1024 the limit suddenly becomes <1ms and the chances of missing
timer ticks increases]
-
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/