From: Michal Hocko <mhocko@kernel.org> To: Anshuman Khandual <anshuman.khandual@arm.com> Cc: linux-mm@kvack.org, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, suzuki.poulose@arm.com, punit.agrawal@arm.com, will.deacon@arm.com, Steven.Price@arm.com, catalin.marinas@arm.com, mike.kravetz@oracle.com, n-horiguchi@ah.jp.nec.com Subject: Re: [PATCH 1/4] mm/hugetlb: Enable PUD level huge page migration Date: Tue, 2 Oct 2018 14:39:09 +0200 [thread overview] Message-ID: <20181002123909.GS18290@dhcp22.suse.cz> (raw) In-Reply-To: <1538482531-26883-2-git-send-email-anshuman.khandual@arm.com> On Tue 02-10-18 17:45:28, Anshuman Khandual wrote: > Architectures like arm64 have PUD level HugeTLB pages for certain configs > (1GB huge page is PUD based on ARM64_4K_PAGES base page size) that can be > enabled for migration. It can be achieved through checking for PUD_SHIFT > order based HugeTLB pages during migration. Well a long term problem with hugepage_migration_supported is that it is used in two different context 1) to bail out from the migration early because the arch doesn't support migration at all and 2) to use movable zone for hugetlb pages allocation. I am especially concerned about the later because the mere support for migration is not really good enough. Are you really able to find a different giga page during the runtime to move an existing giga page out of the movable zone? So I guess we want to split this into two functions arch_hugepage_migration_supported and hugepage_movable. The later would be a reasonably migrateable subset of the former. Without that this patch migth introduce subtle regressions when somebody relies on movable zone to be really movable. > Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com> > --- > include/linux/hugetlb.h | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h > index 6b68e34..9c1b77f 100644 > --- a/include/linux/hugetlb.h > +++ b/include/linux/hugetlb.h > @@ -483,7 +483,8 @@ static inline bool hugepage_migration_supported(struct hstate *h) > { > #ifdef CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION > if ((huge_page_shift(h) == PMD_SHIFT) || > - (huge_page_shift(h) == PGDIR_SHIFT)) > + (huge_page_shift(h) == PUD_SHIFT) || > + (huge_page_shift(h) == PGDIR_SHIFT)) > return true; > else > return false; > -- > 2.7.4 -- Michal Hocko SUSE Labs
WARNING: multiple messages have this Message-ID (diff)
From: mhocko@kernel.org (Michal Hocko) To: linux-arm-kernel@lists.infradead.org Subject: [PATCH 1/4] mm/hugetlb: Enable PUD level huge page migration Date: Tue, 2 Oct 2018 14:39:09 +0200 [thread overview] Message-ID: <20181002123909.GS18290@dhcp22.suse.cz> (raw) In-Reply-To: <1538482531-26883-2-git-send-email-anshuman.khandual@arm.com> On Tue 02-10-18 17:45:28, Anshuman Khandual wrote: > Architectures like arm64 have PUD level HugeTLB pages for certain configs > (1GB huge page is PUD based on ARM64_4K_PAGES base page size) that can be > enabled for migration. It can be achieved through checking for PUD_SHIFT > order based HugeTLB pages during migration. Well a long term problem with hugepage_migration_supported is that it is used in two different context 1) to bail out from the migration early because the arch doesn't support migration at all and 2) to use movable zone for hugetlb pages allocation. I am especially concerned about the later because the mere support for migration is not really good enough. Are you really able to find a different giga page during the runtime to move an existing giga page out of the movable zone? So I guess we want to split this into two functions arch_hugepage_migration_supported and hugepage_movable. The later would be a reasonably migrateable subset of the former. Without that this patch migth introduce subtle regressions when somebody relies on movable zone to be really movable. > Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com> > --- > include/linux/hugetlb.h | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h > index 6b68e34..9c1b77f 100644 > --- a/include/linux/hugetlb.h > +++ b/include/linux/hugetlb.h > @@ -483,7 +483,8 @@ static inline bool hugepage_migration_supported(struct hstate *h) > { > #ifdef CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION > if ((huge_page_shift(h) == PMD_SHIFT) || > - (huge_page_shift(h) == PGDIR_SHIFT)) > + (huge_page_shift(h) == PUD_SHIFT) || > + (huge_page_shift(h) == PGDIR_SHIFT)) > return true; > else > return false; > -- > 2.7.4 -- Michal Hocko SUSE Labs
next prev parent reply other threads:[~2018-10-02 12:39 UTC|newest] Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top 2018-10-02 12:15 [PATCH 0/4] arm64/mm: Enable HugeTLB migration Anshuman Khandual 2018-10-02 12:15 ` Anshuman Khandual 2018-10-02 12:15 ` [PATCH 1/4] mm/hugetlb: Enable PUD level huge page migration Anshuman Khandual 2018-10-02 12:15 ` Anshuman Khandual 2018-10-02 12:38 ` Suzuki K Poulose 2018-10-02 12:38 ` Suzuki K Poulose 2018-10-02 12:56 ` Anshuman Khandual 2018-10-02 12:56 ` Anshuman Khandual 2018-10-02 12:56 ` Anshuman Khandual 2018-10-03 10:22 ` Suzuki K Poulose 2018-10-03 10:22 ` Suzuki K Poulose 2018-10-03 10:22 ` Suzuki K Poulose 2018-10-03 11:10 ` Anshuman Khandual 2018-10-03 11:10 ` Anshuman Khandual 2018-10-03 11:10 ` Anshuman Khandual 2018-10-03 11:17 ` Suzuki K Poulose 2018-10-03 11:17 ` Suzuki K Poulose 2018-10-03 11:17 ` Suzuki K Poulose 2018-10-03 11:27 ` Michal Hocko 2018-10-03 11:27 ` Michal Hocko 2018-10-02 12:39 ` Michal Hocko [this message] 2018-10-02 12:39 ` Michal Hocko 2018-10-03 2:16 ` Anshuman Khandual 2018-10-03 2:16 ` Anshuman Khandual 2018-10-03 6:58 ` Michal Hocko 2018-10-03 6:58 ` Michal Hocko 2018-10-03 9:58 ` Anshuman Khandual 2018-10-03 9:58 ` Anshuman Khandual 2018-10-03 10:59 ` Michal Hocko 2018-10-03 10:59 ` Michal Hocko 2018-10-03 11:37 ` Anshuman Khandual 2018-10-03 11:37 ` Anshuman Khandual 2018-10-03 11:48 ` Michal Hocko 2018-10-03 11:48 ` Michal Hocko 2018-10-03 13:06 ` Anshuman Khandual 2018-10-03 13:06 ` Anshuman Khandual 2018-10-03 13:36 ` Michal Hocko 2018-10-03 13:36 ` Michal Hocko 2018-10-05 7:34 ` Anshuman Khandual 2018-10-05 7:34 ` Anshuman Khandual 2018-10-09 14:14 ` Michal Hocko 2018-10-09 14:14 ` Michal Hocko 2018-10-10 3:09 ` Anshuman Khandual 2018-10-10 3:09 ` Anshuman Khandual 2018-10-10 9:39 ` Michal Hocko 2018-10-10 9:39 ` Michal Hocko 2018-10-11 3:16 ` Anshuman Khandual 2018-10-11 3:16 ` Anshuman Khandual 2018-10-02 12:15 ` [PATCH 2/4] mm/hugetlb: Enable arch specific huge page size support for migration Anshuman Khandual 2018-10-02 12:15 ` Anshuman Khandual 2018-10-02 12:15 ` [PATCH 3/4] arm64/mm: Enable HugeTLB migration Anshuman Khandual 2018-10-02 12:15 ` Anshuman Khandual 2018-10-02 12:15 ` [PATCH 4/4] arm64/mm: Enable HugeTLB migration for contiguous bit HugeTLB pages Anshuman Khandual 2018-10-02 12:15 ` Anshuman Khandual
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=20181002123909.GS18290@dhcp22.suse.cz \ --to=mhocko@kernel.org \ --cc=Steven.Price@arm.com \ --cc=anshuman.khandual@arm.com \ --cc=catalin.marinas@arm.com \ --cc=linux-arm-kernel@lists.infradead.org \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=mike.kravetz@oracle.com \ --cc=n-horiguchi@ah.jp.nec.com \ --cc=punit.agrawal@arm.com \ --cc=suzuki.poulose@arm.com \ --cc=will.deacon@arm.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.