From: Mark Rutland <mark.rutland@arm.com> To: Yu Zhao <yuzhao@google.com>, Anshuman Khandual <anshuman.khandual@arm.com> Cc: Catalin Marinas <catalin.marinas@arm.com>, Will Deacon <will.deacon@arm.com>, "Aneesh Kumar K . V" <aneesh.kumar@linux.vnet.ibm.com>, Andrew Morton <akpm@linux-foundation.org>, Nick Piggin <npiggin@gmail.com>, Peter Zijlstra <peterz@infradead.org>, Joel Fernandes <joel@joelfernandes.org>, "Kirill A . Shutemov" <kirill@shutemov.name>, Ard Biesheuvel <ard.biesheuvel@linaro.org>, Chintan Pandya <cpandya@codeaurora.org>, Jun Yao <yaojun8558363@gmail.com>, Laura Abbott <labbott@redhat.com>, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, linux-arch@vger.kernel.org, linux-mm@kvack.org Subject: Re: [PATCH v3 3/3] arm64: mm: enable per pmd page table lock Date: Mon, 11 Mar 2019 12:12:28 +0000 [thread overview] Message-ID: <20190311121147.GA23361@lakrids.cambridge.arm.com> (raw) In-Reply-To: <20190310011906.254635-3-yuzhao@google.com> Hi, On Sat, Mar 09, 2019 at 06:19:06PM -0700, Yu Zhao wrote: > Switch from per mm_struct to per pmd page table lock by enabling > ARCH_ENABLE_SPLIT_PMD_PTLOCK. This provides better granularity for > large system. > > I'm not sure if there is contention on mm->page_table_lock. Given > the option comes at no cost (apart from initializing more spin > locks), why not enable it now. > > We only do so when pmd is not folded, so we don't mistakenly call > pgtable_pmd_page_ctor() on pud or p4d in pgd_pgtable_alloc(). (We > check shift against PMD_SHIFT, which is same as PUD_SHIFT when pmd > is folded). Just to check, I take it pgtable_pmd_page_ctor() is now a NOP when the PMD is folded, and this last paragraph is stale? > Signed-off-by: Yu Zhao <yuzhao@google.com> > --- > arch/arm64/Kconfig | 3 +++ > arch/arm64/include/asm/pgalloc.h | 12 +++++++++++- > arch/arm64/include/asm/tlb.h | 5 ++++- > 3 files changed, 18 insertions(+), 2 deletions(-) > > diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig > index cfbf307d6dc4..a3b1b789f766 100644 > --- a/arch/arm64/Kconfig > +++ b/arch/arm64/Kconfig > @@ -872,6 +872,9 @@ config ARCH_WANT_HUGE_PMD_SHARE > config ARCH_HAS_CACHE_LINE_SIZE > def_bool y > > +config ARCH_ENABLE_SPLIT_PMD_PTLOCK > + def_bool y if PGTABLE_LEVELS > 2 > + > config SECCOMP > bool "Enable seccomp to safely compute untrusted bytecode" > ---help--- > diff --git a/arch/arm64/include/asm/pgalloc.h b/arch/arm64/include/asm/pgalloc.h > index 52fa47c73bf0..dabba4b2c61f 100644 > --- a/arch/arm64/include/asm/pgalloc.h > +++ b/arch/arm64/include/asm/pgalloc.h > @@ -33,12 +33,22 @@ > > static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr) > { > - return (pmd_t *)__get_free_page(PGALLOC_GFP); > + struct page *page; > + > + page = alloc_page(PGALLOC_GFP); > + if (!page) > + return NULL; > + if (!pgtable_pmd_page_ctor(page)) { > + __free_page(page); > + return NULL; > + } > + return page_address(page); > } > > static inline void pmd_free(struct mm_struct *mm, pmd_t *pmdp) > { > BUG_ON((unsigned long)pmdp & (PAGE_SIZE-1)); > + pgtable_pmd_page_dtor(virt_to_page(pmdp)); > free_page((unsigned long)pmdp); > } It looks like arm64's existing stage-2 code is inconsistent across alloc/free, and IIUC this change might turn that into a real problem. Currently we allocate all levels of stage-2 table with __get_free_page(), but free them with p?d_free(). We always miss the ctor and always use the dtor. Other than that, this patch looks fine to me, but I'd feel more comfortable if we could first fix the stage-2 code to free those stage-2 tables without invoking the dtor. Anshuman, IIRC you had a patch to fix the stage-2 code to not invoke the dtors. If so, could you please post that so that we could take it as a preparatory patch for this series? Thanks, Mark. > diff --git a/arch/arm64/include/asm/tlb.h b/arch/arm64/include/asm/tlb.h > index 106fdc951b6e..4e3becfed387 100644 > --- a/arch/arm64/include/asm/tlb.h > +++ b/arch/arm64/include/asm/tlb.h > @@ -62,7 +62,10 @@ static inline void __pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte, > static inline void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmdp, > unsigned long addr) > { > - tlb_remove_table(tlb, virt_to_page(pmdp)); > + struct page *page = virt_to_page(pmdp); > + > + pgtable_pmd_page_dtor(page); > + tlb_remove_table(tlb, page); > } > #endif > > -- > 2.21.0.360.g471c308f928-goog >
WARNING: multiple messages have this Message-ID (diff)
From: Mark Rutland <mark.rutland@arm.com> To: Yu Zhao <yuzhao@google.com>, Anshuman Khandual <anshuman.khandual@arm.com> Cc: linux-arch@vger.kernel.org, Ard Biesheuvel <ard.biesheuvel@linaro.org>, Peter Zijlstra <peterz@infradead.org>, Catalin Marinas <catalin.marinas@arm.com>, Will Deacon <will.deacon@arm.com>, linux-kernel@vger.kernel.org, Nick Piggin <npiggin@gmail.com>, Jun Yao <yaojun8558363@gmail.com>, linux-mm@kvack.org, "Aneesh Kumar K . V" <aneesh.kumar@linux.vnet.ibm.com>, Chintan Pandya <cpandya@codeaurora.org>, Joel Fernandes <joel@joelfernandes.org>, "Kirill A . Shutemov" <kirill@shutemov.name>, Andrew Morton <akpm@linux-foundation.org>, Laura Abbott <labbott@redhat.com>, linux-arm-kernel@lists.infradead.org Subject: Re: [PATCH v3 3/3] arm64: mm: enable per pmd page table lock Date: Mon, 11 Mar 2019 12:12:28 +0000 [thread overview] Message-ID: <20190311121147.GA23361@lakrids.cambridge.arm.com> (raw) In-Reply-To: <20190310011906.254635-3-yuzhao@google.com> Hi, On Sat, Mar 09, 2019 at 06:19:06PM -0700, Yu Zhao wrote: > Switch from per mm_struct to per pmd page table lock by enabling > ARCH_ENABLE_SPLIT_PMD_PTLOCK. This provides better granularity for > large system. > > I'm not sure if there is contention on mm->page_table_lock. Given > the option comes at no cost (apart from initializing more spin > locks), why not enable it now. > > We only do so when pmd is not folded, so we don't mistakenly call > pgtable_pmd_page_ctor() on pud or p4d in pgd_pgtable_alloc(). (We > check shift against PMD_SHIFT, which is same as PUD_SHIFT when pmd > is folded). Just to check, I take it pgtable_pmd_page_ctor() is now a NOP when the PMD is folded, and this last paragraph is stale? > Signed-off-by: Yu Zhao <yuzhao@google.com> > --- > arch/arm64/Kconfig | 3 +++ > arch/arm64/include/asm/pgalloc.h | 12 +++++++++++- > arch/arm64/include/asm/tlb.h | 5 ++++- > 3 files changed, 18 insertions(+), 2 deletions(-) > > diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig > index cfbf307d6dc4..a3b1b789f766 100644 > --- a/arch/arm64/Kconfig > +++ b/arch/arm64/Kconfig > @@ -872,6 +872,9 @@ config ARCH_WANT_HUGE_PMD_SHARE > config ARCH_HAS_CACHE_LINE_SIZE > def_bool y > > +config ARCH_ENABLE_SPLIT_PMD_PTLOCK > + def_bool y if PGTABLE_LEVELS > 2 > + > config SECCOMP > bool "Enable seccomp to safely compute untrusted bytecode" > ---help--- > diff --git a/arch/arm64/include/asm/pgalloc.h b/arch/arm64/include/asm/pgalloc.h > index 52fa47c73bf0..dabba4b2c61f 100644 > --- a/arch/arm64/include/asm/pgalloc.h > +++ b/arch/arm64/include/asm/pgalloc.h > @@ -33,12 +33,22 @@ > > static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr) > { > - return (pmd_t *)__get_free_page(PGALLOC_GFP); > + struct page *page; > + > + page = alloc_page(PGALLOC_GFP); > + if (!page) > + return NULL; > + if (!pgtable_pmd_page_ctor(page)) { > + __free_page(page); > + return NULL; > + } > + return page_address(page); > } > > static inline void pmd_free(struct mm_struct *mm, pmd_t *pmdp) > { > BUG_ON((unsigned long)pmdp & (PAGE_SIZE-1)); > + pgtable_pmd_page_dtor(virt_to_page(pmdp)); > free_page((unsigned long)pmdp); > } It looks like arm64's existing stage-2 code is inconsistent across alloc/free, and IIUC this change might turn that into a real problem. Currently we allocate all levels of stage-2 table with __get_free_page(), but free them with p?d_free(). We always miss the ctor and always use the dtor. Other than that, this patch looks fine to me, but I'd feel more comfortable if we could first fix the stage-2 code to free those stage-2 tables without invoking the dtor. Anshuman, IIRC you had a patch to fix the stage-2 code to not invoke the dtors. If so, could you please post that so that we could take it as a preparatory patch for this series? Thanks, Mark. > diff --git a/arch/arm64/include/asm/tlb.h b/arch/arm64/include/asm/tlb.h > index 106fdc951b6e..4e3becfed387 100644 > --- a/arch/arm64/include/asm/tlb.h > +++ b/arch/arm64/include/asm/tlb.h > @@ -62,7 +62,10 @@ static inline void __pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte, > static inline void __pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmdp, > unsigned long addr) > { > - tlb_remove_table(tlb, virt_to_page(pmdp)); > + struct page *page = virt_to_page(pmdp); > + > + pgtable_pmd_page_dtor(page); > + tlb_remove_table(tlb, page); > } > #endif > > -- > 2.21.0.360.g471c308f928-goog > _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
next prev parent reply other threads:[~2019-03-11 12:12 UTC|newest] Thread overview: 120+ messages / expand[flat|nested] mbox.gz Atom feed top 2019-02-14 21:16 [PATCH] arm64: mm: enable per pmd page table lock Yu Zhao 2019-02-14 21:16 ` Yu Zhao 2019-02-18 15:12 ` Will Deacon 2019-02-18 15:12 ` Will Deacon 2019-02-18 19:49 ` Yu Zhao 2019-02-18 19:49 ` Yu Zhao 2019-02-18 20:48 ` Yu Zhao 2019-02-18 20:48 ` Yu Zhao 2019-02-19 4:09 ` Anshuman Khandual 2019-02-19 4:09 ` Anshuman Khandual 2019-02-18 23:13 ` [PATCH v2 1/3] arm64: mm: use appropriate ctors for page tables Yu Zhao 2019-02-18 23:13 ` Yu Zhao 2019-02-18 23:13 ` [PATCH v2 2/3] arm64: mm: don't call page table ctors for init_mm Yu Zhao 2019-02-18 23:13 ` Yu Zhao 2019-02-26 15:13 ` Mark Rutland 2019-02-26 15:13 ` Mark Rutland 2019-03-09 3:52 ` Yu Zhao 2019-03-09 3:52 ` Yu Zhao 2019-02-18 23:13 ` [PATCH v2 3/3] arm64: mm: enable per pmd page table lock Yu Zhao 2019-02-18 23:13 ` Yu Zhao 2019-02-19 4:21 ` [PATCH v2 1/3] arm64: mm: use appropriate ctors for page tables Anshuman Khandual 2019-02-19 4:21 ` Anshuman Khandual 2019-02-19 5:32 ` Yu Zhao 2019-02-19 5:32 ` Yu Zhao 2019-02-19 6:17 ` Anshuman Khandual 2019-02-19 6:17 ` Anshuman Khandual 2019-02-19 22:28 ` Yu Zhao 2019-02-19 22:28 ` Yu Zhao 2019-02-20 10:27 ` Anshuman Khandual 2019-02-20 10:27 ` Anshuman Khandual 2019-02-20 12:24 ` Matthew Wilcox 2019-02-20 12:24 ` Matthew Wilcox 2019-02-20 20:22 ` Yu Zhao 2019-02-20 20:22 ` Yu Zhao 2019-02-20 20:59 ` Matthew Wilcox 2019-02-20 20:59 ` Matthew Wilcox 2019-02-20 20:59 ` Matthew Wilcox 2019-02-20 1:34 ` Matthew Wilcox 2019-02-20 1:34 ` Matthew Wilcox 2019-02-20 1:34 ` Matthew Wilcox 2019-02-20 3:20 ` Anshuman Khandual 2019-02-20 3:20 ` Anshuman Khandual 2019-02-20 21:03 ` Matthew Wilcox 2019-02-20 21:03 ` Matthew Wilcox 2019-02-20 21:03 ` Matthew Wilcox 2019-02-26 15:12 ` Mark Rutland 2019-02-26 15:12 ` Mark Rutland 2019-03-09 4:01 ` Yu Zhao 2019-03-09 4:01 ` Yu Zhao 2019-03-09 4:01 ` Yu Zhao 2019-03-10 1:19 ` [PATCH v3 " Yu Zhao 2019-03-10 1:19 ` Yu Zhao 2019-03-10 1:19 ` Yu Zhao 2019-03-10 1:19 ` [PATCH v3 2/3] arm64: mm: don't call page table ctors for init_mm Yu Zhao 2019-03-10 1:19 ` Yu Zhao 2019-03-10 1:19 ` Yu Zhao 2019-03-10 1:19 ` [PATCH v3 3/3] arm64: mm: enable per pmd page table lock Yu Zhao 2019-03-10 1:19 ` Yu Zhao 2019-03-10 1:19 ` Yu Zhao 2019-03-11 8:28 ` Anshuman Khandual 2019-03-11 8:28 ` Anshuman Khandual 2019-03-11 23:10 ` Yu Zhao 2019-03-11 23:10 ` Yu Zhao 2019-03-11 12:12 ` Mark Rutland [this message] 2019-03-11 12:12 ` Mark Rutland 2019-03-11 12:57 ` Anshuman Khandual 2019-03-11 12:57 ` Anshuman Khandual 2019-03-11 23:11 ` Yu Zhao 2019-03-11 23:11 ` Yu Zhao 2019-03-11 23:11 ` Yu Zhao 2019-03-11 7:45 ` [PATCH v3 1/3] arm64: mm: use appropriate ctors for page tables Anshuman Khandual 2019-03-11 7:45 ` Anshuman Khandual 2019-03-11 23:23 ` Yu Zhao 2019-03-11 23:23 ` Yu Zhao 2019-03-12 0:57 ` [PATCH v4 1/4] " Yu Zhao 2019-03-12 0:57 ` Yu Zhao 2019-03-12 0:57 ` Yu Zhao 2019-03-12 0:57 ` [PATCH v4 2/4] arm64: mm: don't call page table ctors for init_mm Yu Zhao 2019-03-12 0:57 ` Yu Zhao 2019-03-12 0:57 ` Yu Zhao 2019-03-12 0:57 ` [PATCH v4 3/4] arm64: mm: call ctor for stage2 pmd page Yu Zhao 2019-03-12 0:57 ` Yu Zhao 2019-03-12 0:57 ` Yu Zhao 2019-03-12 2:19 ` [PATCH] KVM: ARM: Remove pgtable page standard functions from stage-2 page tables Anshuman Khandual 2019-03-12 2:19 ` Anshuman Khandual 2019-03-12 2:40 ` Yu Zhao 2019-03-12 2:40 ` Yu Zhao 2019-03-12 10:37 ` Suzuki K Poulose 2019-03-12 10:37 ` Suzuki K Poulose 2019-03-12 11:31 ` Anshuman Khandual 2019-03-12 11:31 ` Anshuman Khandual 2019-03-12 11:43 ` Suzuki K Poulose 2019-03-12 11:43 ` Suzuki K Poulose 2019-03-12 13:25 ` [PATCH V2] " Anshuman Khandual 2019-03-12 13:25 ` Anshuman Khandual 2019-04-01 16:16 ` Will Deacon 2019-04-01 16:16 ` Will Deacon 2019-04-01 16:16 ` Will Deacon 2019-04-01 18:34 ` Yu Zhao 2019-04-01 18:34 ` Yu Zhao 2019-04-01 18:34 ` Yu Zhao 2019-04-02 9:03 ` Will Deacon 2019-04-02 9:03 ` Will Deacon 2019-04-02 9:03 ` Will Deacon 2019-04-08 14:22 ` Will Deacon 2019-04-08 14:22 ` Will Deacon 2019-04-08 14:22 ` Will Deacon 2019-04-08 14:22 ` Will Deacon 2019-04-08 17:18 ` Yu Zhao 2019-04-08 17:18 ` Yu Zhao 2019-04-08 17:18 ` Yu Zhao 2019-04-08 9:09 ` Marc Zyngier 2019-04-08 9:09 ` Marc Zyngier 2019-04-08 9:09 ` Marc Zyngier 2019-04-08 9:09 ` Marc Zyngier 2019-03-12 0:57 ` [PATCH v4 4/4] arm64: mm: enable per pmd page table lock Yu Zhao 2019-03-12 0:57 ` Yu Zhao 2019-03-12 0:57 ` Yu Zhao 2019-02-19 3:08 ` [PATCH] " Anshuman Khandual 2019-02-19 3:08 ` 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=20190311121147.GA23361@lakrids.cambridge.arm.com \ --to=mark.rutland@arm.com \ --cc=akpm@linux-foundation.org \ --cc=aneesh.kumar@linux.vnet.ibm.com \ --cc=anshuman.khandual@arm.com \ --cc=ard.biesheuvel@linaro.org \ --cc=catalin.marinas@arm.com \ --cc=cpandya@codeaurora.org \ --cc=joel@joelfernandes.org \ --cc=kirill@shutemov.name \ --cc=labbott@redhat.com \ --cc=linux-arch@vger.kernel.org \ --cc=linux-arm-kernel@lists.infradead.org \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=npiggin@gmail.com \ --cc=peterz@infradead.org \ --cc=will.deacon@arm.com \ --cc=yaojun8558363@gmail.com \ --cc=yuzhao@google.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.