On Sun, Jul 27, 2008 at 02:45:20AM -0700, Andrew Morton wrote: > On Thu, 24 Jul 2008 16:39:49 +0200 Nick Piggin <npiggin@xxxxxxx> wrote: > > > I think everybody is hoping to have a workable mmu notifier scheme > > merged in 2.6.27 (myself included). However I do have some concerns > > about the implementation proposed (in -mm). > > > > I apologise for this late review, before anybody gets too upset, > > most of my concerns have been raised before, but I'd like to state > > my case again and involving everyone. > > Nick, having read through this discussion and the code (yet again) I > think I'll go ahead and send it all in to Linus. On the basis that You call. As I said, I was OK for you to use your judgement. > - the code is fairly short and simple > > - has no known bugs > > - seems to be needed by some folks ;) > > - you already have a protopatch which partially addresses your > concerns and afaik there's nothing blocking future improvements to > this implementation? Seems like I was able to get my head around KVM and GRU more easily than I had feared. I don't expect help, but I will say that good simplifications will always get merged unless there is a reasonable performance case not to, so I will work on patches. > And a late-breaking review comment: given that about 0.000000000000001% > of people will actually use mm_take_all_locks(), could we make its > compilation conditional on something? Such as CONFIG_MMU_NOTIFIER? Well I prefer that because I don't want any new callers to pop up. -- To unsubscribe from this list: send the line "unsubscribe linux-arch" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html