| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <CAK7LNARNyYNMMYcEwU-pVFb-b21cKN+K_jwseLTtoOfscS2Ggw@mail.gmail.com>
Date: Thu, 21 Apr 2022 02:11:36 +0900
From: Masahiro Yamada <masahiroy@...nel.org>
To: Muchun Song <songmuchun@...edance.com>
Cc: Jonathan Corbet <corbet@....net>, mike.kravetz@...cle.com,
Andrew Morton <akpm@...ux-foundation.org>,
"Luis R. Rodriguez" <mcgrof@...nel.org>,
Kees Cook <keescook@...omium.org>,
Iurii Zaikin <yzaikin@...gle.com>,
Oscar Salvador <osalvador@...e.de>,
David Hildenbrand <david@...hat.com>,
"open list:DOCUMENTATION" <linux-doc@...r.kernel.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Linux Memory Management List <linux-mm@...ck.org>,
duanxiongchun@...edance.com, smuchun@...il.com
Subject: Re: [PATCH v8 1/4] mm: hugetlb_vmemmap: introduce CONFIG_HUGETLB_PAGE_HAS_OPTIMIZE_VMEMMAP
On Wed, Apr 13, 2022 at 11:48 PM Muchun Song <songmuchun@...edance.com> wrote:
>
> If the size of "struct page" is not the power of two but with the feature
> of minimizing overhead of struct page associated with each HugeTLB is
> enabled, then the vmemmap pages of HugeTLB will be corrupted after
> remapping (panic is about to happen in theory). But this only exists when
> !CONFIG_MEMCG && !CONFIG_SLUB on x86_64. However, it is not a conventional
> configuration nowadays. So it is not a real word issue, just the result
> of a code review. But we have to prevent anyone from configuring that
> combined configurations. In order to avoid many checks like "is_power_of_2
> (sizeof(struct page))" through mm/hugetlb_vmemmap.c. Introduce a new macro
> CONFIG_HUGETLB_PAGE_HAS_OPTIMIZE_VMEMMAP to represent the size of struct
> page is power of two and CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP is
> configured. Then make the codes of this feature depends on this new macro.
> Then we could prevent anyone do any unexpected configurations. A new
> autoconf_ext.h is introduced as well, which serves as an extension for
> autoconf.h since those special configurations (e.g.
> CONFIG_HUGETLB_PAGE_HAS_OPTIMIZE_VMEMMAP here) are rely on the autoconf.h
> (generated from Kconfig), so we cannot embed those configurations into
> Kconfig. After this change, it would be easy if someone want to do the
> similar thing (add a new CONFIG) in the future.
>
> Signed-off-by: Muchun Song <songmuchun@...edance.com>
> Suggested-by: Luis Chamberlain <mcgrof@...nel.org>
> ---
> Kbuild | 19 +++++++++++++++++++
> arch/x86/mm/init_64.c | 2 +-
> include/linux/hugetlb.h | 2 +-
> include/linux/kconfig.h | 4 ++++
> include/linux/mm.h | 2 +-
> include/linux/page-flags.h | 2 +-
> kernel/autoconf_ext.c | 26 ++++++++++++++++++++++++++
> mm/hugetlb_vmemmap.c | 8 ++------
> mm/hugetlb_vmemmap.h | 4 ++--
> mm/sparse-vmemmap.c | 4 ++--
> scripts/mod/Makefile | 2 ++
> 11 files changed, 61 insertions(+), 14 deletions(-)
> create mode 100644 kernel/autoconf_ext.c
>
> diff --git a/Kbuild b/Kbuild
> index fa441b98c9f6..83c0d5a418d1 100644
> --- a/Kbuild
> +++ b/Kbuild
> @@ -2,6 +2,12 @@
> #
> # Kbuild for top-level directory of the kernel
>
> +# autoconf_ext.h is generated last since it depends on other generated headers,
> +# however those other generated headers may include autoconf_ext.h. Use the
> +# following macro to avoid circular dependency.
> +
> +KBUILD_CFLAGS_KERNEL += -D__EXCLUDE_AUTOCONF_EXT_H
> +
> #####
> # Generate bounds.h
>
> @@ -37,6 +43,19 @@ $(offsets-file): arch/$(SRCARCH)/kernel/asm-offsets.s FORCE
> $(call filechk,offsets,__ASM_OFFSETS_H__)
>
> #####
> +# Generate autoconf_ext.h.
> +
> +autoconf_ext-file := include/generated/autoconf_ext.h
> +
> +always-y += $(autoconf_ext-file)
> +targets += kernel/autoconf_ext.s
> +
> +kernel/autoconf_ext.s: $(bounds-file) $(timeconst-file) $(offsets-file)
> +
> +$(autoconf_ext-file): kernel/autoconf_ext.s FORCE
> + $(call filechk,offsets,__LINUX_AUTOCONF_EXT_H__)
> +
> +#####
> # Check for missing system calls
>
> always-y += missing-syscalls
> diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
> index 4b9e0012bbbf..9b8dfa6e4da8 100644
> --- a/arch/x86/mm/init_64.c
> +++ b/arch/x86/mm/init_64.c
> @@ -1268,7 +1268,7 @@ static struct kcore_list kcore_vsyscall;
>
> static void __init register_page_bootmem_info(void)
> {
> -#if defined(CONFIG_NUMA) || defined(CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP)
> +#if defined(CONFIG_NUMA) || defined(CONFIG_HUGETLB_PAGE_HAS_OPTIMIZE_VMEMMAP)
> int i;
>
> for_each_online_node(i)
> diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
> index ac2ece9e9c79..d42de8abd2b6 100644
> --- a/include/linux/hugetlb.h
> +++ b/include/linux/hugetlb.h
> @@ -623,7 +623,7 @@ struct hstate {
> unsigned int nr_huge_pages_node[MAX_NUMNODES];
> unsigned int free_huge_pages_node[MAX_NUMNODES];
> unsigned int surplus_huge_pages_node[MAX_NUMNODES];
> -#ifdef CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP
> +#ifdef CONFIG_HUGETLB_PAGE_HAS_OPTIMIZE_VMEMMAP
> unsigned int optimize_vmemmap_pages;
> #endif
> #ifdef CONFIG_CGROUP_HUGETLB
> diff --git a/include/linux/kconfig.h b/include/linux/kconfig.h
> index 20d1079e92b4..00796794f177 100644
> --- a/include/linux/kconfig.h
> +++ b/include/linux/kconfig.h
> @@ -4,6 +4,10 @@
>
> #include <generated/autoconf.h>
>
> +#if defined(__KERNEL__) && !defined(__EXCLUDE_AUTOCONF_EXT_H)
> +#include <generated/autoconf_ext.h>
> +#endif
> +
Please do not do this either.
When autoconf_ext.h is updated, the kernel tree
would be rebuilt entirely.
--
Best Regards
Masahiro Yamada
Powered by blists - more mailing lists