From f3a6fc344ca057ab5fdda41b9c2cb1206b37ac68 Mon Sep 17 00:00:00 2001
From: Alexander von Below <Alex@vonBelow.Com>
Date: Sun, 2 Aug 2020 21:41:58 +0200
Subject: [PATCH] [Chapter 13] Fixed the MUL.4H and MLA.4H syntax

---
 Chapter 13/debug.s          | 31 -------------------------------
 Chapter 13/matrixmultneon.s | 18 +++++++++---------
 README.md                   | 20 +++++++++++++++++---
 3 files changed, 26 insertions(+), 43 deletions(-)
 delete mode 100644 Chapter 13/debug.s

diff --git a/Chapter 13/debug.s b/Chapter 13/debug.s
deleted file mode 100644
index cecbd52..0000000
--- a/Chapter 13/debug.s	
+++ /dev/null
@@ -1,31 +0,0 @@
-@ Various macros to help with debugging
-
-@ These macros preseve all registers.
-@ Beware they will change cpsr.
-
-.macro  printReg    reg
-	push	    {r0-r4, lr} @ save regs
-	mov	    r2, R\reg	@ for the %d
-	mov	    r3, R\reg	@ for the %x
-	mov	    r1, #\reg	
-	add	    r1, #'0'	@ for %c
-        ldr  	    r0, =ptfStr @ printf format str
-        bl	    _printf	@ call printf
-	pop	    {r0-r4, lr} @ restore regs
-.endm
-
-.macro	printStr    str
-	push	    {r0-r4, lr}	@ save regs
-	ldr	    r0, =1f	@ load print str
-	bl	    _printf	@ call printf
-	pop	    {r0-r4, lr}	@ restore regs
-	b	    2f		@ branch around str
-1:	.asciz	    "\str\n"
-	.align	    4
-2:
-.endm
-
-.data
-ptfStr: .asciz	"R%c = %16d, 0x%08x\n"
-.align 4
-.text
diff --git a/Chapter 13/matrixmultneon.s b/Chapter 13/matrixmultneon.s
index 4600679..fca6bb0 100644
--- a/Chapter 13/matrixmultneon.s	
+++ b/Chapter 13/matrixmultneon.s	
@@ -33,15 +33,15 @@ main:
 	LDP	D3, D4, [X0], #16
 	LDR	D5, [X0]
 
-	MUL	V6().4H, V0.4H, V3().4H[0]
-	MLA	V6().4H, V1.4H, V3().4H[1]
-	MLA	V6().4H, V2.4H, V3().4H[2]
-	MUL	V7().4H, V0.4H, V4().4H
-	MLA	V7().4H, V1.4H, V4().4H
-	MLA	V7().4H, V2.4H, V4().4H
-	MUL	V8().4H, V0.4H, V5().4H
-	MLA	V8().4H, V1.4H, V5().4H
-	MLA	V8().4H, V2.4H, V5().4H
+.macro mulcol ccol bcol
+	MUL.4H	\ccol\(), V0, \bcol\()[0]
+	MLA.4H	\ccol\(), V1, \bcol\()[1]
+	MLA.4H	\ccol\(), V2, \bcol\()[2]
+.endm
+
+	mulcol	V6, V3	// process first column
+	mulcol	V7, V4	// process second column
+	mulcol	V8, V5	// process third column
 
 	ADRP	X1, C@PAGE	// Address of C
 	ADD	X1, X1, C@PAGEOFF
diff --git a/README.md b/README.md
index 058e351..d500c3d 100644
--- a/README.md
+++ b/README.md
@@ -7,11 +7,13 @@ In this repository, I will code along with the book [Programming with 64-Bit ARM
 
 ## Latest News
 
+Pop the Champagne! 🍾 All the code is running!
+
 Once you found the bug, you feel stupid for not noticing it before. Only after some debugging I realized that Darwin has a different value for `AT_FDCWD`. This means: Chapter 7 is ready!
 
-Chapter 13 stil has an open issue, though.
+And after some disassembly and reading [Documentation](https://community.arm.com/developer/tools-software/oss-platforms/b/android-blog/posts/arm-neon-programming-quick-reference), Chapter 13 is ready as well.
 
-I got a mention in Stephen Smith's [blog](https://smist08.wordpress.com/2020/07/31/is-apple-silicon-really-arm/)!
+Last but not least, I got a mention in Stephen Smith's [blog](https://smist08.wordpress.com/2020/07/31/is-apple-silicon-really-arm/)!
 
 ### Prerequisites
 
@@ -283,7 +285,19 @@ Like in Chapter 11, all the chages have been introduced already. Nothing new her
 
 ## Chapter 13
 
-This chapter is still in the works; it compiles, but the output is wrong. There is one [issue](https://github.com/below/HelloSilicon/issues/15) I'd like to fix
+Once again, the Clang assembler seems to want a slightly different syntax: Where gcc accepts
+
+```
+MUL V6.4H, V0.4H, V3.4H[0]
+```
+
+the Clang assembler expects
+
+```
+MUL.4H V6, V0, V3[0]
+```
+
+All other changes to the code should be trivial at this point.
 
 ## Chapter 14