Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932425Ab2KNCXW (ORCPT ); Tue, 13 Nov 2012 21:23:22 -0500 Received: from mail-gg0-f174.google.com ([209.85.161.174]:59535 "EHLO mail-gg0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932375Ab2KNCXU (ORCPT ); Tue, 13 Nov 2012 21:23:20 -0500 Date: Tue, 13 Nov 2012 18:23:13 -0800 (PST) From: Hugh Dickins X-X-Sender: hugh@eggly.anvils To: Johannes Weiner cc: Ingo Molnar , linux-kernel@vger.kernel.org, linux-mm@kvack.org, Paul Turner , Lee Schermerhorn , Christoph Lameter , Rik van Riel , Mel Gorman , Andrew Morton , Andrea Arcangeli , Linus Torvalds , Peter Zijlstra , Thomas Gleixner , Zhouping Liu Subject: Re: [PATCH 21/31] sched, numa, mm: Implement THP migration In-Reply-To: <20121113184835.GH10092@cmpxchg.org> Message-ID: References: <1352826834-11774-1-git-send-email-mingo@kernel.org> <1352826834-11774-22-git-send-email-mingo@kernel.org> <20121113184835.GH10092@cmpxchg.org> User-Agent: Alpine 2.00 (LNX 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 7862 Lines: 244 On Tue, 13 Nov 2012, Johannes Weiner wrote: > [Put Hugh back on CC] > > What happened to Hugh's fixes to the LRU handling? I believe it was > racy beyond affecting memcg, it's just that memcg code had a BUG_ON in > the right place to point it out. Right, yes, I made several fixes there, but never sent out the final version. And I can't see your fix to the memcg end of THP migration either - though I've not done much more than scan the Subjects so far, I've not yet got to grips with yesterday/today's latest approach. I'll send out two patches, yours and mine, to what's in the current linux-next and the current mmotm, which I have been testing with in several contexts; they probably both apply easily to Ingo's new tree. But I vehemently hope that this all very soon vanishes from linux-next, and the new stuff is worked on properly for a while, in a separate development branch of tip, hopefully converging with Mel's. Hugh > > On Tue, Nov 13, 2012 at 06:13:44PM +0100, Ingo Molnar wrote: > > From: Peter Zijlstra > > > > Add THP migration for the NUMA working set scanning fault case. > > > > It uses the page lock to serialize. No migration pte dance is > > necessary because the pte is already unmapped when we decide > > to migrate. > > > > Signed-off-by: Peter Zijlstra > > Cc: Johannes Weiner > > Cc: Mel Gorman > > Cc: Andrea Arcangeli > > Cc: Andrew Morton > > Cc: Linus Torvalds > > Link: http://lkml.kernel.org/n/tip-yv9vbiz2s455zxq1ffzx3fye@git.kernel.org > > [ Significant fixes and changelog. ] > > Signed-off-by: Ingo Molnar > > --- > > mm/huge_memory.c | 131 +++++++++++++++++++++++++++++++++++++++++++------------ > > mm/migrate.c | 2 +- > > 2 files changed, 103 insertions(+), 30 deletions(-) > > > > diff --git a/mm/huge_memory.c b/mm/huge_memory.c > > index c4c0a57..931caf4 100644 > > --- a/mm/huge_memory.c > > +++ b/mm/huge_memory.c > > @@ -742,12 +742,13 @@ void do_huge_pmd_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, > > unsigned int flags, pmd_t entry) > > { > > unsigned long haddr = address & HPAGE_PMD_MASK; > > + struct page *new_page = NULL; > > struct page *page = NULL; > > - int node; > > + int node, lru; > > > > spin_lock(&mm->page_table_lock); > > if (unlikely(!pmd_same(*pmd, entry))) > > - goto out_unlock; > > + goto unlock; > > > > if (unlikely(pmd_trans_splitting(entry))) { > > spin_unlock(&mm->page_table_lock); > > @@ -755,45 +756,117 @@ void do_huge_pmd_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, > > return; > > } > > > > -#ifdef CONFIG_NUMA > > page = pmd_page(entry); > > - VM_BUG_ON(!PageCompound(page) || !PageHead(page)); > > + if (page) { > > + VM_BUG_ON(!PageCompound(page) || !PageHead(page)); > > > > - get_page(page); > > - spin_unlock(&mm->page_table_lock); > > + get_page(page); > > + node = mpol_misplaced(page, vma, haddr); > > + if (node != -1) > > + goto migrate; > > + } > > > > - /* > > - * XXX should we serialize against split_huge_page ? > > - */ > > +fixup: > > + /* change back to regular protection */ > > + entry = pmd_modify(entry, vma->vm_page_prot); > > + set_pmd_at(mm, haddr, pmd, entry); > > + update_mmu_cache_pmd(vma, address, entry); > > > > - node = mpol_misplaced(page, vma, haddr); > > - if (node == -1) > > - goto do_fixup; > > +unlock: > > + spin_unlock(&mm->page_table_lock); > > + if (page) > > + put_page(page); > > > > - /* > > - * Due to lacking code to migrate thp pages, we'll split > > - * (which preserves the special PROT_NONE) and re-take the > > - * fault on the normal pages. > > - */ > > - split_huge_page(page); > > - put_page(page); > > return; > > > > -do_fixup: > > +migrate: > > + spin_unlock(&mm->page_table_lock); > > + > > + lock_page(page); > > spin_lock(&mm->page_table_lock); > > - if (unlikely(!pmd_same(*pmd, entry))) > > - goto out_unlock; > > -#endif > > + if (unlikely(!pmd_same(*pmd, entry))) { > > + spin_unlock(&mm->page_table_lock); > > + unlock_page(page); > > + put_page(page); > > + return; > > + } > > + spin_unlock(&mm->page_table_lock); > > > > - /* change back to regular protection */ > > - entry = pmd_modify(entry, vma->vm_page_prot); > > - if (pmdp_set_access_flags(vma, haddr, pmd, entry, 1)) > > - update_mmu_cache_pmd(vma, address, entry); > > + new_page = alloc_pages_node(node, > > + (GFP_TRANSHUGE | GFP_THISNODE) & ~__GFP_WAIT, > > + HPAGE_PMD_ORDER); > > > > -out_unlock: > > + if (!new_page) > > + goto alloc_fail; > > + > > + lru = PageLRU(page); > > + > > + if (lru && isolate_lru_page(page)) /* does an implicit get_page() */ > > + goto alloc_fail; > > + > > + if (!trylock_page(new_page)) > > + BUG(); > > + > > + /* anon mapping, we can simply copy page->mapping to the new page: */ > > + new_page->mapping = page->mapping; > > + new_page->index = page->index; > > + > > + migrate_page_copy(new_page, page); > > + > > + WARN_ON(PageLRU(new_page)); > > + > > + spin_lock(&mm->page_table_lock); > > + if (unlikely(!pmd_same(*pmd, entry))) { > > + spin_unlock(&mm->page_table_lock); > > + if (lru) > > + putback_lru_page(page); > > + > > + unlock_page(new_page); > > + ClearPageActive(new_page); /* Set by migrate_page_copy() */ > > + new_page->mapping = NULL; > > + put_page(new_page); /* Free it */ > > + > > + unlock_page(page); > > + put_page(page); /* Drop the local reference */ > > + > > + return; > > + } > > + > > + entry = mk_pmd(new_page, vma->vm_page_prot); > > + entry = maybe_pmd_mkwrite(pmd_mkdirty(entry), vma); > > + entry = pmd_mkhuge(entry); > > + > > + page_add_new_anon_rmap(new_page, vma, haddr); > > + > > + set_pmd_at(mm, haddr, pmd, entry); > > + update_mmu_cache_pmd(vma, address, entry); > > + page_remove_rmap(page); > > spin_unlock(&mm->page_table_lock); > > - if (page) > > + > > + put_page(page); /* Drop the rmap reference */ > > + > > + if (lru) > > + put_page(page); /* drop the LRU isolation reference */ > > + > > + unlock_page(new_page); > > + unlock_page(page); > > + put_page(page); /* Drop the local reference */ > > + > > + return; > > + > > +alloc_fail: > > + if (new_page) > > + put_page(new_page); > > + > > + unlock_page(page); > > + > > + spin_lock(&mm->page_table_lock); > > + if (unlikely(!pmd_same(*pmd, entry))) { > > put_page(page); > > + page = NULL; > > + goto unlock; > > + } > > + goto fixup; > > } > > > > int copy_huge_pmd(struct mm_struct *dst_mm, struct mm_struct *src_mm, > > diff --git a/mm/migrate.c b/mm/migrate.c > > index 3299949..72d1056 100644 > > --- a/mm/migrate.c > > +++ b/mm/migrate.c > > @@ -417,7 +417,7 @@ int migrate_huge_page_move_mapping(struct address_space *mapping, > > */ > > void migrate_page_copy(struct page *newpage, struct page *page) > > { > > - if (PageHuge(page)) > > + if (PageHuge(page) || PageTransHuge(page)) > > copy_huge_page(newpage, page); > > else > > copy_highpage(newpage, page); > > -- > > 1.7.11.7 > > > > -- > > To unsubscribe, send a message with 'unsubscribe linux-mm' in > > the body to majordomo@kvack.org. For more info on Linux MM, > > see: http://www.linux-mm.org/ . > > Don't email: email@kvack.org > > -- > To unsubscribe, send a message with 'unsubscribe linux-mm' in > the body to majordomo@kvack.org. For more info on Linux MM, > see: http://www.linux-mm.org/ . > Don't email: email@kvack.org > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/