Re: Linus rollup

On Thu, Jan 30, 2003 at 09:37:06AM -0800, Stephen Hemminger wrote:

quoted

you certainly mean wmb() not rmb(), right? If yes, then yes.

I actually didn't notice the write_begin/end, not sure who could need
them, I would suggest removing them, rather than to revert the mb()
there too.

The write_begin/end was suggested by Andrew as a simplification for use
when using this to update values already write-locked by other means.

One possible usage was to fix the race issues with non-atomic update
of 64 bit i_size.

It looks overdesign to me, you don't need the spinlock for that, the
i_sem is explicit too. The generalized abstraction is worthwhile when
you have to use it 99% of the time, the missing 1% doesn't need to be
abstracted, forward porting my implementation is the best for such
specific case IMHO.

Andrea
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/