On Mon, Jan 28, 2008 at 06:10:39PM +0200, Izik Eidus wrote: > i dont understand how is that better than notification on tlb flush?
I certainly agree. The quoted call wasn't actually the one that could be moved in a single place in the .h file though. But the 4/4 patch could be reduced to a few liner patch and it would not require any change to ksm and mm/*.c internals that way etc... If Christoph prefers to expand the two bits I added in the .h file, all over the .c files that's ok with us. There is no good excuse for that IMHO because a __ptep_clear_flush could be added to optimize away the notifiers if it's followed by the invalidate_range (not the common case), and if we want to just mark a spte readonly instead of invalidating it we could implement new ptep_ function that would invoke a new mmu notifier method. So there's full room for optimizations without requiring the expansion. In the end it's only a coding style issue but one that will become visible in the VM (i.e. ksm has to be modified to add a mmu_notifier call after ptep_clear_flush, it won't work automatically without changes anymore). So I'd like to know what can we do to help to merge the 4 patches from Christoph in mainline, I'd appreciate comments on them so we can help to address any outstanding issue! It's very important to have this code in 2.6.25 final. KVM requires mmu notifiers for reliable swapping, madvise/ballooning, ksm etc... so it's high priority to get something merged to provide this functionality regardless of its coding style ;). Thanks! Andrea -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
