Re: MMU notifiers review and some proposals

[Nick Piggin <npiggin@suse.de>]

On Sat, Jul 26, 2008 at 01:38:13PM +0200, Andrea Arcangeli wrote:
> On Sat, Jul 26, 2008 at 05:08:10AM +0200, Nick Piggin wrote:
> > Anyway, I just voice my opinion and let Andrew and Linus decide. To be
> > clear: I have not found any actual bugs in Andrea's -mm patchset, only
> > some dislikes of the approach.
> 
> Yes, like I said I think this is a matter of taste of what you like of
> the tradeoff. There are disadvantages and advantages in both and if we
> wait forever to please everyone taste, it'll never go in.

And for this item, I think there has been a bit too much emphasis
on pleasing the taste of the drivers and not enough on the core
VM. My concern about adding a new TLB flushing design to core VM
was never taken seriously, for example. Nor was my request for
performance numbers. And I did ask early in the year.

I believe when you don't have any real numbers for justification,
the only sane thing to do is go with the most minimal and simplest
implementation first, and add complexity if/when it can be justified.

--
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/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>