> This seems to assume that ffs is 1-relative in the bit numbers. My testing
> shows it is 0-relative, and the manual bears me out.
Linux follows the libc convention of returning a 1-relative bit
offset, or 0 for 'no bits set'.
Maybe your ffs() routine is different?
From: ron minnich <rminnich@la...> - 2004-06-29 17:47:56
On Tue, 29 Jun 2004, Keir Fraser wrote:
> Linux follows the libc convention of returning a 1-relative bit
> offset, or 0 for 'no bits set'.
this is weird. I'm beginning to worry about vmware. I had a printk on this
ffs() stuff in vmware and for ffs(1) I got a zero; on my laptop it just
Also, I fixed another problem in hypervisor_callback by removing
PUSHW ES (push %es for linux folks).
Yep, if the callback pushed ES it would trap, otherwise not.
arg. I really need to get off vmware soon :-)