Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758138AbdLVJs6 (ORCPT ); Fri, 22 Dec 2017 04:48:58 -0500 Received: from mx2.suse.de ([195.135.220.15]:33630 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757979AbdLVJsw (ORCPT ); Fri, 22 Dec 2017 04:48:52 -0500 Date: Fri, 22 Dec 2017 10:48:49 +0100 From: Michal Hocko To: Mike Kravetz , Naoya Horiguchi Cc: "linux-mm@kvack.org" , Andrew Morton , LKML Subject: Re: [RFC PATCH 0/5] mm, hugetlb: allocation API and migration improvements Message-ID: <20171222094849.GO4831@dhcp22.suse.cz> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <7cf6978c-5bf2-cbe4-6f7f-ba09998f482d@ah.jp.nec.com> <659e21c7-ebed-8b64-053a-f01a31ef6e25@oracle.com> User-Agent: Mutt/1.9.1 (2017-09-22) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 5474 Lines: 152 On Thu 21-12-17 15:35:28, Mike Kravetz wrote: [...] > You can add, > > Reviewed-by: Mike Kravetz > > I had some concerns about transferring huge page state during migration > not specific to this patch, so I did a bunch of testing. On Fri 22-12-17 08:58:48, Naoya Horiguchi wrote: [...] > Yes, I tested again with additional changes below, and hugetlb migration > works fine from mbind(2). Thank you very much for your work. > > Reviewed-by: Naoya Horiguchi > > for the series. Thanks a lot to both of you! I have added the changelog to the last patch. I am currently busy as hell so I will unlikely send the whole thing before new year but please double check the changelog if you find some more time. --- >From 65e7dda5dfe90fa656285729a270855e93637caa Mon Sep 17 00:00:00 2001 From: Michal Hocko Date: Fri, 22 Dec 2017 10:31:18 +0100 Subject: [PATCH] hugetlb, mempolicy: fix the mbind hugetlb migration do_mbind migration code relies on alloc_huge_page_noerr for hugetlb pages. alloc_huge_page_noerr uses alloc_huge_page which is a highlevel allocation function which has to take care of reserves, overcommit or hugetlb cgroup accounting. None of that is really required for the page migration because the new page is only temporal and either will replace the original page or it will be dropped. This is essentially as for other migration call paths and there shouldn't be any reason to handle mbind in a special way. The current implementation is even suboptimal because the migration might fail just because the hugetlb cgroup limit is reached, or the overcommit is saturated. Fix this by making mbind like other hugetlb migration paths. Add a new migration helper alloc_huge_page_vma as a wrapper around alloc_huge_page_nodemask with additional mempolicy handling. alloc_huge_page_noerr has no more users and it can go. Reviewed-by: Mike Kravetz Reviewed-by: Naoya Horiguchi Signed-off-by: Michal Hocko --- include/linux/hugetlb.h | 5 ++--- mm/hugetlb.c | 33 +++++++++++++++++++-------------- mm/mempolicy.c | 3 +-- 3 files changed, 22 insertions(+), 19 deletions(-) diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index e035002d3fb6..08a4af411e25 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -345,10 +345,9 @@ struct huge_bootmem_page { struct page *alloc_huge_page(struct vm_area_struct *vma, unsigned long addr, int avoid_reserve); struct page *alloc_huge_page_node(struct hstate *h, int nid); -struct page *alloc_huge_page_noerr(struct vm_area_struct *vma, - unsigned long addr, int avoid_reserve); struct page *alloc_huge_page_nodemask(struct hstate *h, int preferred_nid, nodemask_t *nmask); +struct page *alloc_huge_page_vma(struct vm_area_struct *vma, unsigned long address); int huge_add_to_page_cache(struct page *page, struct address_space *mapping, pgoff_t idx); @@ -526,7 +525,7 @@ struct hstate {}; #define alloc_huge_page(v, a, r) NULL #define alloc_huge_page_node(h, nid) NULL #define alloc_huge_page_nodemask(h, preferred_nid, nmask) NULL -#define alloc_huge_page_noerr(v, a, r) NULL +#define alloc_huge_page_vma(vma, address) NULL #define alloc_bootmem_huge_page(h) NULL #define hstate_file(f) NULL #define hstate_sizelog(s) NULL diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 4426c5b23a20..e00deabe6d17 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -1672,6 +1672,25 @@ struct page *alloc_huge_page_nodemask(struct hstate *h, int preferred_nid, return alloc_migrate_huge_page(h, gfp_mask, preferred_nid, nmask); } +/* mempolicy aware migration callback */ +struct page *alloc_huge_page_vma(struct vm_area_struct *vma, unsigned long address) +{ + struct mempolicy *mpol; + nodemask_t *nodemask; + struct page *page; + struct hstate *h; + gfp_t gfp_mask; + int node; + + h = hstate_vma(vma); + gfp_mask = htlb_alloc_mask(h); + node = huge_node(vma, address, gfp_mask, &mpol, &nodemask); + page = alloc_huge_page_nodemask(h, node, nodemask); + mpol_cond_put(mpol); + + return page; +} + /* * Increase the hugetlb pool such that it can accommodate a reservation * of size 'delta'. @@ -2077,20 +2096,6 @@ struct page *alloc_huge_page(struct vm_area_struct *vma, return ERR_PTR(-ENOSPC); } -/* - * alloc_huge_page()'s wrapper which simply returns the page if allocation - * succeeds, otherwise NULL. This function is called from new_vma_page(), - * where no ERR_VALUE is expected to be returned. - */ -struct page *alloc_huge_page_noerr(struct vm_area_struct *vma, - unsigned long addr, int avoid_reserve) -{ - struct page *page = alloc_huge_page(vma, addr, avoid_reserve); - if (IS_ERR(page)) - page = NULL; - return page; -} - int alloc_bootmem_huge_page(struct hstate *h) __attribute__ ((weak, alias("__alloc_bootmem_huge_page"))); int __alloc_bootmem_huge_page(struct hstate *h) diff --git a/mm/mempolicy.c b/mm/mempolicy.c index f604b22ebb65..96823fa07f38 100644 --- a/mm/mempolicy.c +++ b/mm/mempolicy.c @@ -1121,8 +1121,7 @@ static struct page *new_page(struct page *page, unsigned long start, int **x) } if (PageHuge(page)) { - BUG_ON(!vma); - return alloc_huge_page_noerr(vma, address, 1); + return alloc_huge_page_vma(vma, address); } else if (thp_migration_supported() && PageTransHuge(page)) { struct page *thp; -- 2.15.1 -- Michal Hocko SUSE Labs