From: Babu Moger <babu.moger@oracle.com>
Date: Wed, 18 Jan 2017 01:21:44 +0000 (-0800)
Subject: arch/sparc: Add a dedicated clear_page and clear_user_page for M7
X-Git-Tag: v4.1.12-92~1^2~15
X-Git-Url: https://www.infradead.org/git/?a=commitdiff_plain;h=b96a3766a473d0ed72ac544830600ccb827961f6;p=users%2Fjedix%2Flinux-maple.git

arch/sparc: Add a dedicated clear_page and clear_user_page for M7

Adding a dedicated clear_page and clear_user_page for M7.
Avoids multiple checks which are really not required.
This eliminates about 30 instructions for each call.
Seen about 3 to 4 percent latency reduction in some cases.

Orabug: 25456049

Signed-off-by: Babu Moger <babu.moger@oracle.com>
Reviewed-by: Rob Gardner <rob.gardner@oracle.com>
Signed-off-by: Allen Pais <allen.pais@oracle.com>
---

diff --git a/arch/sparc/lib/M7clear_page.S b/arch/sparc/lib/M7clear_page.S
new file mode 100644
index 0000000000000..1215ed7d661ae
--- /dev/null
+++ b/arch/sparc/lib/M7clear_page.S
@@ -0,0 +1,36 @@
+/*
+ * M7clear_page.S: SPARC M7 optimized clear_page and clear_user_page.
+ *
+ * Copyright (c) 2016, Oracle and/or its affiliates.  All rights reserved.
+ */
+
+#include <asm/asi.h>
+#include <asm/page.h>
+
+#define ASI_STBI_P      ASI_BLK_INIT_QUAD_LDD_P
+
+	.text
+	.register       %g3, #scratch
+	.align		32
+	.globl		M7clear_page
+	.globl		M7clear_user_page
+M7clear_page:	/* %o0=dest */
+M7clear_user_page:	/* %o0=dest */
+	set	PAGE_SIZE, %o2
+
+.wrzero_loop:
+
+	mov     64, %g3
+	stxa    %g0, [%o0]ASI_STBI_P
+	subcc   %o2, 256, %o2
+	stxa    %g0, [%o0+%g3]ASI_STBI_P
+	add     %o0, 256, %o0
+	sub     %g3, 192, %g3
+	stxa    %g0, [%o0+%g3]ASI_STBI_P
+	add	%g3, 64, %g3
+	bg,pt	%ncc, .wrzero_loop
+	stxa    %g0, [%o0+%g3]ASI_STBI_P
+	retl
+	membar  #StoreStore	! required by use of Block Store Init
+	.size		M7clear_page,.-M7clear_page
+	.size		M7clear_user_page,.-M7clear_user_page
diff --git a/arch/sparc/lib/M7memset.S b/arch/sparc/lib/M7memset.S
index e880bfc2227bb..1a431cda40597 100644
--- a/arch/sparc/lib/M7memset.S
+++ b/arch/sparc/lib/M7memset.S
@@ -98,20 +98,6 @@
 	.globl		M7memop_enable
 M7memop_enable:	.word 15
 
-/*
- * Define clear_page(dest) as memset(dest, 0, PAGE_SIZE)
- * (can create a more optimized version later.)
- */
-	.globl		M7clear_page
-	.globl		M7clear_user_page
-M7clear_page:		/* clear_page(dest) */
-M7clear_user_page:
-	set PAGE_SIZE, %o1
-	/* fall through into bzero code */
-
-	.size		M7clear_page,.-M7clear_page
-	.size		M7clear_user_page,.-M7clear_user_page
-
 /*
  * Define bzero(dest, n) as memset(dest, 0, n)
  * (can create a more optimized version later.)
diff --git a/arch/sparc/lib/Makefile b/arch/sparc/lib/Makefile
index c4db9e1ebe04e..62501c25eeb8d 100644
--- a/arch/sparc/lib/Makefile
+++ b/arch/sparc/lib/Makefile
@@ -35,7 +35,7 @@ lib-$(CONFIG_SPARC64) +=  NG2patch.o
 lib-$(CONFIG_SPARC64) += NG4memcpy.o NG4copy_from_user.o NG4copy_to_user.o
 lib-$(CONFIG_SPARC64) +=  NG4patch.o NG4copy_page.o NG4clear_page.o NG4memset.o
 
-lib-$(CONFIG_SPARC64) += M7memcpy.o M7memset.o M7copy_to_user.o M7copy_from_user.o M7patch.o
+lib-$(CONFIG_SPARC64) += M7memcpy.o M7memset.o M7copy_to_user.o M7copy_from_user.o M7patch.o M7clear_page.o
 
 lib-$(CONFIG_SPARC64) += GENmemcpy.o GENcopy_from_user.o GENcopy_to_user.o
 lib-$(CONFIG_SPARC64) += GENpatch.o GENpage.o GENbzero.o