On Wed, 7 Feb 2007, Christoph Lameter wrote: > Clearing reference bits? Ummm... That is a pretty inaccurate measure since > reclaim can remove pages and revert the reference bits. It can never work > reliably. >
It's not intended to work precisely, it's intended to give a good estimate of task memory footprint. There are several different scenarios that can interfere with getting a precise measurement via this method; then again /proc/pid/smaps is so expensive already because of the iteration through VMA's that it's not something you'd do regularly. Any other suggestions on how to obtain this data is certainly welcome. > Would it be possible to sync up with the people doing the page table > interface? > (+pauld) Is there an update on Paul Davies' implementation that would allow us to iterate through a set of pte's in a vm_area_struct range? > Could we somehow consolidate smaps and numa_maps? > Sure, but that's beyond the scope of this patchset. My intention in extracting away a new pte_walker was to prevent having two identical implementations and it could easily be extracted even further to lib for the ioremap case that Paul mentioned. David - 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/