linux-kernel - Re: [PATCHv4 09/14] x86/mm: Handle boot-time paging mode switching at early boot

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 17 Aug 2017 11:10:03 +0200
From:   Ingo Molnar <mingo@...nel.org>
To:     "Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>
Cc:     Linus Torvalds <torvalds@...ux-foundation.org>,
        Andrew Morton <akpm@...ux-foundation.org>, x86@...nel.org,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>,
        "H. Peter Anvin" <hpa@...or.com>, Andi Kleen <ak@...ux.intel.com>,
        Dave Hansen <dave.hansen@...el.com>,
        Andy Lutomirski <luto@...capital.net>,
        Michal Hocko <mhocko@...nel.org>, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCHv4 09/14] x86/mm: Handle boot-time paging mode switching
 at early boot


* Kirill A. Shutemov <kirill.shutemov@...ux.intel.com> wrote:

> This patch adds detection of 5-level paging at boot-time and adjusts
> virtual memory layout and folds p4d page table layer if needed.
> 
> We have to make X86_5LEVEL dependant on SPARSEMEM_VMEMMAP.
> !SPARSEMEM_VMEMMAP configuration doesn't work well with variable
> MAX_PHYSMEM_BITS.
> 
> Signed-off-by: Kirill A. Shutemov <kirill.shutemov@...ux.intel.com>
> ---
>  arch/x86/Kconfig                        |  1 +
>  arch/x86/boot/compressed/kaslr.c        | 13 +++++--
>  arch/x86/entry/entry_64.S               | 12 +++++++
>  arch/x86/include/asm/page_64_types.h    | 13 +++----
>  arch/x86/include/asm/pgtable_64_types.h | 35 +++++++++++--------
>  arch/x86/include/asm/processor.h        |  2 +-
>  arch/x86/include/asm/sparsemem.h        |  9 ++---
>  arch/x86/kernel/head64.c                | 60 ++++++++++++++++++++++++++-------
>  arch/x86/kernel/head_64.S               | 18 ++++++----
>  arch/x86/kernel/setup.c                 |  5 ++-
>  arch/x86/mm/dump_pagetables.c           |  8 +++--
>  arch/x86/mm/kaslr.c                     | 13 ++++---
>  12 files changed, 129 insertions(+), 60 deletions(-)

Please also split this patch up some more, into as many individual (but 
bisectable) changes as possible.

> diff --git a/arch/x86/include/asm/sparsemem.h b/arch/x86/include/asm/sparsemem.h
> index 1f5bee2c202f..ba67afd870b7 100644
> --- a/arch/x86/include/asm/sparsemem.h
> +++ b/arch/x86/include/asm/sparsemem.h
> @@ -26,13 +26,8 @@
>  # endif
>  #else /* CONFIG_X86_32 */
>  # define SECTION_SIZE_BITS	27 /* matt - 128 is convenient right now */
> -# ifdef CONFIG_X86_5LEVEL
> -#  define MAX_PHYSADDR_BITS	52
> -#  define MAX_PHYSMEM_BITS	52
> -# else
> -#  define MAX_PHYSADDR_BITS	44
> -#  define MAX_PHYSMEM_BITS	46
> -# endif
> +# define MAX_PHYSADDR_BITS	(p4d_folded ? 44 : 52)
> +# define MAX_PHYSMEM_BITS	(p4d_folded ? 46 : 52)
>  #endif

The kernel code size impact of these de-constification changes should be measured, 
double checked and documented as well. We are adding all kinds of overhead to 
(what I expect to be) commonly used kernels, let's do it carefully.

Thanks,

	Ingo