Mike Kravetz <mike.kravetz@xxxxxxxxxx> writes: > On 05/17/2018 11:24 PM, Michal Hocko wrote: >> On Fri 18-05-18 11:03:16, Huang, Ying wrote: >> [...] >>> The patch is a generic optimization which should benefit quite some >>> workloads, not for a specific use case. To demonstrate the performance >>> benefit of the patch, we tested it with vm-scalability run on >>> transparent huge page. >> >> It is also adds quite some non-intuitive code. So is this worth? Does >> any _real_ workload benefits from the change? > > One way to 'add less code' would be to create a helper routine that > indicates the order in which sub-pages are to be copied. IIUC, you > added the same algorithm for sub-page ordering to copy_huge_page() > that was previously added to clear_huge_page(). Correct? Yes. > If so, then perhaps a common helper could be used by both the clear > and copy huge page routines. It would also make maintenance easier. That's a good idea. But this may need to turn copy_user_highpage()/clear_user_highpage() calling in copy_user_huge_page()/clear_huge_page() from direct call to indirect call. I don't know whether this will incur some overhead. Will try to measure this. Best Regards, Huang, Ying
