9 files changed, 24 insertions, 19 deletions
diff --git a/mpn/x86_64/aors_n.asm b/mpn/x86_64/aors_n.asm
index eb0b471f1..46b595429 100644
--- a/mpn/x86_64/aors_n.asm
+++ b/mpn/x86_64/aors_n.asm
@@ -1,7 +1,6 @@
-dnl  AMD64 mpn_add_n -- Add two limb vectors of the same length > 0 and store
-dnl  sum in a third limb vector.
+dnl  AMD64 mpn_add_n, mpn_sub_n
 
-dnl  Copyright 2003, 2004, 2005, 2007 Free Software Foundation, Inc.
+dnl  Copyright 2003, 2004, 2005, 2007, 2008 Free Software Foundation, Inc.
 
 dnl  This file is part of the GNU MP Library.
 
@@ -25,7 +24,8 @@ C	     cycles/limb
 C K8,K9:	 1.69
 C K10:		 2
 C P4:		10.5
-C P6-15:	 4.87
+C P6-15 (Core2): 4.87
+C P6-28 (Atom):	 4
 
 MULFUNC_PROLOGUE(mpn_add_n mpn_add_nc)
 
diff --git a/mpn/x86_64/aorsmul_1.asm b/mpn/x86_64/aorsmul_1.asm
index 7b4b49cc6..114b19107 100644
--- a/mpn/x86_64/aorsmul_1.asm
+++ b/mpn/x86_64/aorsmul_1.asm
@@ -23,7 +23,8 @@ C	     cycles/limb
 C K8,K9:	 2.5
 C K10:		 2.5
 C P4:		14.9
-C P6-15:	 5.09
+C P6-15 (Core2): 5.09
+C P6-28 (Atom):	21.3
 
 C The inner loop of this code is the result of running a code generation and
 C permutation tool suite written by David Harvey and Torbjorn Granlund.
diff --git a/mpn/x86_64/dive_1.asm b/mpn/x86_64/dive_1.asm
index 6ed48b717..d4a5ab5cc 100644
--- a/mpn/x86_64/dive_1.asm
+++ b/mpn/x86_64/dive_1.asm
@@ -24,7 +24,8 @@ C	     cycles/limb
 C K8,K9:	10
 C K10:		10
 C P4:		33
-C P6-15:	13.25
+C P6-15 (Core2):13.25
+C P6-28 (Atom):	42
 
 C A quick adoption of the 32-bit K7 code.
 
diff --git a/mpn/x86_64/divrem_1.asm b/mpn/x86_64/divrem_1.asm
index d743771c4..9700eb3af 100644
--- a/mpn/x86_64/divrem_1.asm
+++ b/mpn/x86_64/divrem_1.asm
@@ -24,7 +24,8 @@ C	    cycles/limb     cycles/limb
 C K8,K9:	17		15
 C K10:		17		15
 C P4:		83		57
-C P6-15:	32		21
+C P6-15 (Core2):32		21
+C P6-28 (Atom):	70		39
 
 
 C mp_limb_t mpn_divrem_1 (mp_ptr qp, mp_size_t fn,
diff --git a/mpn/x86_64/lshift.asm b/mpn/x86_64/lshift.asm
index cbe087a13..ece6d8bb1 100644
--- a/mpn/x86_64/lshift.asm
+++ b/mpn/x86_64/lshift.asm
@@ -24,7 +24,8 @@ C	     cycles/limb
 C K8,K9:	 2.375	(1.375 for cnt=1)
 C K10:		 2.375	(1.375 for cnt=1)
 C P4:		 8	(10.5  for cnt=1)
-C P6-15:	 2.11	(4.28 for cnt=1)
+C P6-15 (Core2): 2.11	(4.28 for cnt=1)
+C P6-28 (Atom):	 5.9
 
 
 C INPUT PARAMETERS
diff --git a/mpn/x86_64/mod_34lsub1.asm b/mpn/x86_64/mod_34lsub1.asm
index cb37c7be1..796a3c8ca 100644
--- a/mpn/x86_64/mod_34lsub1.asm
+++ b/mpn/x86_64/mod_34lsub1.asm
@@ -25,7 +25,8 @@ C	     cycles/limb
 C K8,K9:	 1.0
 C K10:		 1.12
 C P4:  		 3.25
-C P6-15:	 1.5
+C P6-15 (Core2): 1.5
+C P6-28 (Atom):	 2.5
 
 
 C INPUT PARAMETERS
diff --git a/mpn/x86_64/mode1o.asm b/mpn/x86_64/mode1o.asm
index aec9c9c18..f4fd7581b 100644
--- a/mpn/x86_64/mode1o.asm
+++ b/mpn/x86_64/mode1o.asm
@@ -25,7 +25,8 @@ C	     cycles/limb
 C K8,K9:	10
 C K10:		10
 C P4:		33
-C P6-15:	13
+C P6-15 (Core2):13
+C P6-28 (Atom):	35
 
 
 C mp_limb_t mpn_modexact_1_odd (mp_srcptr src, mp_size_t size,
diff --git a/mpn/x86_64/mul_1.asm b/mpn/x86_64/mul_1.asm
index d21ad576e..38acd1a37 100644
--- a/mpn/x86_64/mul_1.asm
+++ b/mpn/x86_64/mul_1.asm
@@ -24,6 +24,8 @@ C K8,K9:	 2.5
 C K10:		 2.5
 C P4:		 12.3
 C P6-15:	 4.0
+C P6-15 (Core2): 4.0
+C P6-28 (Atom):	19.8
 
 C The inner loop of this code is the result of running a code generation and
 C permutation tool suite written by David Harvey and Torbjorn Granlund.
diff --git a/mpn/x86_64/rshift.asm b/mpn/x86_64/rshift.asm
index 859673c89..683b90a98 100644
--- a/mpn/x86_64/rshift.asm
+++ b/mpn/x86_64/rshift.asm
@@ -24,18 +24,15 @@ C	     cycles/limb
 C K8,K9:	 2.375
 C K10:		 2.375
 C P4:		 8
-C P6-15:	 2.11
+C P6-15 (Core2): 2.11
+C P6-28 (Atom):	 5.9
 
 
 C INPUT PARAMETERS
-C rp	rdi
-C up	rsi
-C n	rdx
-C cnt	rcx
-
-define(`rp',`%rdi')
-define(`up',`%rsi')
-define(`n',`%rdx')
+define(`rp',	`%rdi')
+define(`up',	`%rsi')
+define(`n',	`%rdx')
+define(`cnt',	`%rcx')
 
 ASM_START()
 	TEXT