[PATCH v2 4/6] x86: Add clear_page_nocache

[Jan Beulich]

>>> On 09.08.12 at 17:03, "Kirill A. Shutemov" <kirill.shutemov at linux.intel.com> wrote:
> From: Andi Kleen <ak at linux.intel.com>
> 
> Add a cache avoiding version of clear_page. Straight forward integer variant
> of the existing 64bit clear_page, for both 32bit and 64bit.

While on 64-bit this is fine, I fail to see how you avoid using the
SSE2 instruction on non-SSE2 systems.

> Also add the necessary glue for highmem including a layer that non cache
> coherent architectures that use the virtual address for flushing can
> hook in. This is not needed on x86 of course.
> 
> If an architecture wants to provide cache avoiding version of clear_page
> it should to define ARCH_HAS_USER_NOCACHE to 1 and implement
> clear_page_nocache() and clear_user_highpage_nocache().
> 
> Signed-off-by: Andi Kleen <ak at linux.intel.com>
> Signed-off-by: Kirill A. Shutemov <kirill.shutemov at linux.intel.com>
> ---
>  arch/x86/include/asm/page.h          |    2 ++
>  arch/x86/include/asm/string_32.h     |    5 +++++
>  arch/x86/include/asm/string_64.h     |    5 +++++
>  arch/x86/lib/Makefile                |    1 +
>  arch/x86/lib/clear_page_nocache_32.S |   30 ++++++++++++++++++++++++++++++
>  arch/x86/lib/clear_page_nocache_64.S |   29 +++++++++++++++++++++++++++++

Couldn't this more reasonably go into clear_page_{32,64}.S?

>  arch/x86/mm/fault.c                  |    7 +++++++
>  7 files changed, 79 insertions(+), 0 deletions(-)
>  create mode 100644 arch/x86/lib/clear_page_nocache_32.S
>  create mode 100644 arch/x86/lib/clear_page_nocache_64.S
>...
>--- /dev/null
>+++ b/arch/x86/lib/clear_page_nocache_32.S
>@@ -0,0 +1,30 @@
>+#include <linux/linkage.h>
>+#include <asm/dwarf2.h>
>+
>+/*
>+ * Zero a page avoiding the caches
>+ * rdi	page

Wrong comment.

>+ */
>+ENTRY(clear_page_nocache)
>+	CFI_STARTPROC
>+	mov    %eax,%edi

You need to pick a different register here (e.g. %edx), since
%edi has to be preserved by all functions called from C.

>+	xorl   %eax,%eax
>+	movl   $4096/64,%ecx
>+	.p2align 4
>+.Lloop:
>+	decl	%ecx
>+#define PUT(x) movnti %eax,x*8(%edi) ; movnti %eax,x*8+4(%edi)

Is doing twice as much unrolling as on 64-bit really worth it?

Jan