Steve Capper <steve.capper@xxxxxxxxxx> writes: > On Tue, Oct 14, 2014 at 04:27:53PM +0530, Aneesh Kumar K.V wrote: >> get_user_pages_fast attempts to pin user pages by walking the page >> tables directly and avoids taking locks. Thus the walker needs to be >> protected from page table pages being freed from under it, and needs >> to block any THP splits. >> >> One way to achieve this is to have the walker disable interrupts, and >> rely on IPIs from the TLB flushing code blocking before the page table >> pages are freed. >> >> On some platforms we have hardware broadcast of TLB invalidations, thus >> the TLB flushing code doesn't necessarily need to broadcast IPIs; and >> spuriously broadcasting IPIs can hurt system performance if done too >> often. >> >> This problem has been solved on PowerPC and Sparc by batching up page >> table pages belonging to more than one mm_user, then scheduling an >> rcu_sched callback to free the pages. This RCU page table free logic >> has been promoted to core code and is activated when one enables >> HAVE_RCU_TABLE_FREE. Unfortunately, these architectures implement >> their own get_user_pages_fast routines. >> >> The RCU page table free logic coupled with a an IPI broadcast on THP >> split (which is a rare event), allows one to protect a page table >> walker by merely disabling the interrupts during the walk. >> >> This patch provides a general RCU implementation of get_user_pages_fast >> that can be used by architectures that perform hardware broadcast of >> TLB invalidations. >> >> It is based heavily on the PowerPC implementation. >> >> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@xxxxxxxxxxxxxxxxxx> >> --- >> >> NOTE: I kept the description of patch as it is and also retained the documentation. >> I also dropped the tested-by and other SOB, because i was not sure whether they want >> to be blamed for the bugs here. Please feel free to update. > > Hi Aneesh, > Thank you for coding this up. > > I've compiled and briefly tested this on arm (with and without LPAE), > and arm64. I ran a custom futex on THP tail test, and this passed. > I'll test this a little more aggressively with ltp. > > I think Linus has already pulled in the RCU gup I posted, could you > please instead write a patch against? Will do that. > 2667f50 mm: introduce a general RCU get_user_pages_fast() In that patch don't you need to check the _PAGE_PRESENT details in case of gup_huge_pmd ? -aneesh -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>