Added part10-multicore - a rough multicore approach to sound

part10-multicore/Makefile

@@ -0,0 +1,22 @@
+CFILES = $(wildcard *.c)
+OFILES = $(CFILES:.c=.o)
+LLVMPATH = /opt/homebrew/opt/llvm/bin
+CLANGFLAGS = -Wall -O2 -ffreestanding -nostdinc -nostdlib -mcpu=cortex-a72+nosimd
+
+all: clean kernel8.img
+
+boot.o: boot.S
+	$(LLVMPATH)/clang --target=aarch64-elf $(CLANGFLAGS) -c boot.S -o boot.o
+
+audio.o : audio.bin
+	$(LLVMPATH)/llvm-objcopy -I binary -O elf64-littleaarch64 -B aarch64 $< $@
+
+%.o: %.c
+	$(LLVMPATH)/clang --target=aarch64-elf $(CLANGFLAGS) -c $< -o $@
+
+kernel8.img: boot.o $(OFILES) audio.o
+	$(LLVMPATH)/ld.lld -m aarch64elf -nostdlib boot.o $(OFILES) audio.o -T link.ld -o kernel8.elf
+	$(LLVMPATH)/llvm-objcopy -O binary kernel8.elf kernel8.img
+
+clean:
+	/bin/rm kernel8.elf *.o *.img > /dev/null 2> /dev/null || true

part10-multicore/Makefile.gcc

@@ -0,0 +1,22 @@
+CFILES = $(wildcard *.c)
+OFILES = $(CFILES:.c=.o)
+GCCFLAGS = -Wall -O2 -ffreestanding -nostdinc -nostdlib -nostartfiles
+GCCPATH = ../../gcc-arm-9.2-2019.12-x86_64-aarch64-none-elf/bin
+
+all: clean kernel8.img
+
+boot.o: boot.S
+	$(GCCPATH)/aarch64-none-elf-gcc $(GCCFLAGS) -c boot.S -o boot.o
+
+audio.o : audio.bin
+	$(GCCPATH)/aarch64-none-elf-objcopy -I binary -O elf64-littleaarch64 -B aarch64 $< $@
+
+%.o: %.c
+	$(GCCPATH)/aarch64-none-elf-gcc $(GCCFLAGS) -c $< -o $@
+
+kernel8.img: boot.o $(OFILES) audio.o
+	$(GCCPATH)/aarch64-none-elf-ld -nostdlib -nostartfiles boot.o $(OFILES) audio.o -T link.ld -o kernel8.elf
+	$(GCCPATH)/aarch64-none-elf-objcopy -O binary kernel8.elf kernel8.img
+
+clean:
+	/bin/rm kernel8.elf *.o *.img > /dev/null 2> /dev/null || true

part10-multicore/README.md

@@ -0,0 +1,23 @@
+Writing a "bare metal" operating system for Raspberry Pi 4 (Part 10)
+====================================================================
+
+Using multiple CPU cores
+------------------------
+Instead of a background DMA transfer, I suggested that we might use a second CPU core to play the audio whilst our main core continues on. I also said it would be hard on the Raspberry Pi 4... and it is.
+
+This code isn't great yet, but it's enough to prove the concept in principle. Importantly, you'll need to modify your _config.txt_ file on your SD card to include the following lines:
+
+```c
+kernel_old=1
+disable_commandline_tags=1
+arm_64bit=1
+```
+
+For now, the points of interest in the code are:
+
+ * The new _boot.S_ and related _boot.h_ loader
+ * The new _multicore.c_ and _multicore.h_ files
+ * A slimmed down _io.h_ and _kernel.c_ (DMA sound removed), with a new approach to `main()`
+ * A revised _link.ld_
+
+I will write more soon!

part10-multicore/boot.S

@@ -0,0 +1,87 @@
+#include "boot.h"
+
+.section ".text.boot"
+
+.globl _start
+_start:
+	ldr	x0, =SCTLR_VALUE_MMU_DISABLED
+	msr	sctlr_el1, x0
+
+        ldr     x0, =LOCAL_CONTROL
+	str     wzr, [x0]
+	mov     w1, 0x80000000
+	str     w1, [x0, #(LOCAL_PRESCALER - LOCAL_CONTROL)]
+
+        ldr     x0, =OSC_FREQ           // Sort out the timer
+	msr     cntfrq_el0, x0
+	msr     cntvoff_el2, xzr
+
+	ldr	x0, =HCR_VALUE
+	msr	hcr_el2, x0
+
+	ldr	x0, =SCR_VALUE
+	msr	scr_el3, x0
+
+	ldr	x0, =SPSR_VALUE
+	msr	spsr_el3, x0
+
+	adr	x0, in_el2
+	msr	elr_el3, x0
+
+        eret
+in_el2:
+	mrs	x6, mpidr_el1		
+	and	x6, x6,#0xFF		// Check processor id
+	cbz	x6, primary_cpu         // Hang for all non-primary CPU
+
+	adr     x5, spin_cpu0
+proc_hang:
+	wfe
+	ldr     x4, [x5, x6, lsl #3]
+	cbz     x4, proc_hang
+
+	mov     x0, #0
+        str     x0, [x5, x6, lsl #3]    // Zero the address again
+
+        ldr     x1, =__test_stack       // Get ourselves a fresh stack
+        mov     sp, x1
+boot_kernel:
+	mov     x1, #0
+	mov     x2, #0
+	mov     x3, #0
+	br      x4
+        b       proc_hang
+primary_cpu:
+	ldr	x1, =__bss_start
+	ldr	w2, =__bss_size
+memzero:
+        cbz     w2, startup
+	str     xzr, [x1], #8
+	sub     w2, w2, #1
+	cbnz    w2, memzero
+startup:
+	mov	sp, #LOW_MEMORY
+	bl	main
+	b 	proc_hang		// should never come here
+
+.ltorg
+
+.org 0xd8
+.globl spin_cpu0
+spin_cpu0:
+        .quad 0
+
+.org 0xe0
+.globl spin_cpu1
+spin_cpu1:
+        .quad 0
+
+.org 0xe8
+.globl spin_cpu2
+spin_cpu2:
+        .quad 0
+
+.org 0xf0
+.globl spin_cpu3
+spin_cpu3:
+        .quad 0

part10-multicore/boot.h

@@ -0,0 +1,58 @@
+// ***************************************
+// SCTLR_EL1, System Control Register (EL1), Page 2654 of AArch64-Reference-Manual.
+// ***************************************
+
+#define SCTLR_RESERVED                  (3 << 28) | (3 << 22) | (1 << 20) | (1 << 11)
+#define SCTLR_EE_LITTLE_ENDIAN          (0 << 25)
+#define SCTLR_EOE_LITTLE_ENDIAN         (0 << 24)
+#define SCTLR_I_CACHE_DISABLED          (0 << 12)
+#define SCTLR_D_CACHE_DISABLED          (0 << 2)
+#define SCTLR_MMU_DISABLED              (0 << 0)
+#define SCTLR_MMU_ENABLED               (1 << 0)
+
+#define SCTLR_VALUE_MMU_DISABLED	(SCTLR_RESERVED | SCTLR_EE_LITTLE_ENDIAN | SCTLR_I_CACHE_DISABLED | SCTLR_D_CACHE_DISABLED | SCTLR_MMU_DISABLED)
+
+// ***************************************
+// HCR_EL2, Hypervisor Configuration Register (EL2), Page 2487 of AArch64-Reference-Manual.
+// ***************************************
+
+#define HCR_RW	    			(1 << 31)
+#define HCR_VALUE			HCR_RW
+
+// ***************************************
+// SCR_EL3, Secure Configuration Register (EL3), Page 2648 of AArch64-Reference-Manual.
+// ***************************************
+
+#define SCR_RESERVED	    		(3 << 4)
+#define SCR_RW				(1 << 10)
+#define SCR_NS				(1 << 0)
+#define SCR_VALUE	    	    	(SCR_RESERVED | SCR_RW | SCR_NS)
+
+// ***************************************
+// SPSR_EL3, Saved Program Status Register (EL3) Page 389 of AArch64-Reference-Manual.
+// ***************************************
+
+#define SPSR_MASK_ALL 			(7 << 6)
+#define SPSR_EL1h			(5 << 0)
+#define SPSR_VALUE			(SPSR_MASK_ALL | SPSR_EL1h)
+
+// ***************************************
+// Memory management stuff
+// ***************************************
+
+#define PAGE_SHIFT	 		12
+#define TABLE_SHIFT 			9
+#define SECTION_SHIFT			(PAGE_SHIFT + TABLE_SHIFT)
+
+#define PAGE_SIZE   			(1 << PAGE_SHIFT)
+#define SECTION_SIZE			(1 << SECTION_SHIFT)	
+
+#define LOW_MEMORY              	(2 * SECTION_SIZE)         // 0x400000
+
+// ***************************************
+// Timer stuff
+// ***************************************
+
+#define LOCAL_CONTROL		        0xff800000
+#define LOCAL_PRESCALER		        0xff800008
+#define OSC_FREQ		        54000000

part10-multicore/fb.c

@@ -0,0 +1,260 @@
+#include "io.h"
+#include "mb.h"
+#include "terminal.h"
+
+unsigned int width, height, pitch, isrgb;
+unsigned char *fb;
+
+int curx = 0;
+int cury = 0;
+
+void fb_init()
+{
+    mbox[0] = 35*4; // Length of message in bytes
+    mbox[1] = MBOX_REQUEST;
+
+    mbox[2] = MBOX_TAG_SETPHYWH; // Tag identifier
+    mbox[3] = 8; // Value size in bytes
+    mbox[4] = 8; // Value size in bytes (again!)
+    mbox[5] = 1920; // Value(width)
+    mbox[6] = 1080; // Value(height)
+
+    mbox[7] = MBOX_TAG_SETVIRTWH;
+    mbox[8] = 8;
+    mbox[9] = 8;
+    mbox[10] = 1920;
+    mbox[11] = 1080;
+
+    mbox[12] = MBOX_TAG_SETVIRTOFF;
+    mbox[13] = 8;
+    mbox[14] = 8;
+    mbox[15] = 0; // Value(x)
+    mbox[16] = 0; // Value(y)
+
+    mbox[17] = MBOX_TAG_SETDEPTH;
+    mbox[18] = 4;
+    mbox[19] = 4;
+    mbox[20] = 32; // Bits per pixel
+
+    mbox[21] = MBOX_TAG_SETPXLORDR;
+    mbox[22] = 4;
+    mbox[23] = 4;
+    mbox[24] = 1; // RGB
+
+    mbox[25] = MBOX_TAG_GETFB;
+    mbox[26] = 8;
+    mbox[27] = 8;
+    mbox[28] = 4096; // FrameBufferInfo.pointer
+    mbox[29] = 0;    // FrameBufferInfo.size
+
+    mbox[30] = MBOX_TAG_GETPITCH;
+    mbox[31] = 4;
+    mbox[32] = 4;
+    mbox[33] = 0; // Bytes per line
+
+    mbox[34] = MBOX_TAG_LAST;
+
+    // Check call is successful and we have a pointer with depth 32
+    if (mbox_call(MBOX_CH_PROP) && mbox[20] == 32 && mbox[28] != 0) {
+        mbox[28] &= 0x3FFFFFFF; // Convert GPU address to ARM address
+        width = mbox[10];       // Actual physical width
+        height = mbox[11];      // Actual physical height
+        pitch = mbox[33];       // Number of bytes per line
+        isrgb = mbox[24];       // Pixel order
+        fb = (unsigned char *)((long)mbox[28]);
+    }
+}
+
+void drawPixel(int x, int y, unsigned char attr)
+{
+    int offs = (y * pitch) + (x * 4);
+    *((unsigned int*)(fb + offs)) = vgapal[attr & 0x0f];
+}
+
+void drawRect(int x1, int y1, int x2, int y2, unsigned char attr, int fill)
+{
+    int y=y1;
+
+    while (y <= y2) {
+       int x=x1;
+       while (x <= x2) {
+	  if ((x == x1 || x == x2) || (y == y1 || y == y2)) drawPixel(x, y, attr);
+	  else if (fill) drawPixel(x, y, (attr & 0xf0) >> 4);
+          x++;
+       }
+       y++;
+    }
+}
+
+void drawLine(int x1, int y1, int x2, int y2, unsigned char attr)  
+{  
+    int dx, dy, p, x, y;
+
+    dx = x2-x1;
+    dy = y2-y1;
+    x = x1;
+    y = y1;
+    p = 2*dy-dx;
+
+    while (x<x2) {
+       if (p >= 0) {
+          drawPixel(x,y,attr);
+          y++;
+          p = p+2*dy-2*dx;
+       } else {
+          drawPixel(x,y,attr);
+          p = p+2*dy;
+       }
+       x++;
+    }
+}
+
+void drawCircle(int x0, int y0, int radius, unsigned char attr, int fill)
+{
+    int x = radius;
+    int y = 0;
+    int err = 0;
+ 
+    while (x >= y) {
+	if (fill) {
+	   drawLine(x0 - y, y0 + x, x0 + y, y0 + x, (attr & 0xf0) >> 4);
+	   drawLine(x0 - x, y0 + y, x0 + x, y0 + y, (attr & 0xf0) >> 4);
+	   drawLine(x0 - x, y0 - y, x0 + x, y0 - y, (attr & 0xf0) >> 4);
+	   drawLine(x0 - y, y0 - x, x0 + y, y0 - x, (attr & 0xf0) >> 4);
+	}
+	drawPixel(x0 - y, y0 + x, attr);
+	drawPixel(x0 + y, y0 + x, attr);
+	drawPixel(x0 - x, y0 + y, attr);
+        drawPixel(x0 + x, y0 + y, attr);
+	drawPixel(x0 - x, y0 - y, attr);
+	drawPixel(x0 + x, y0 - y, attr);
+	drawPixel(x0 - y, y0 - x, attr);
+	drawPixel(x0 + y, y0 - x, attr);
+
+	if (err <= 0) {
+	    y += 1;
+	    err += 2*y + 1;
+	}
+ 
+	if (err > 0) {
+	    x -= 1;
+	    err -= 2*x + 1;
+	}
+    }
+}
+
+void drawChar(unsigned char ch, int x, int y, unsigned char attr, int zoom)
+{
+    unsigned char *glyph = (unsigned char *)&font + (ch < FONT_NUMGLYPHS ? ch : 0) * FONT_BPG;
+
+    for (int i=1;i<=(FONT_HEIGHT*zoom);i++) {
+	for (int j=0;j<(FONT_WIDTH*zoom);j++) {
+	    unsigned char mask = 1 << (j/zoom);
+	    unsigned char col = (*glyph & mask) ? attr & 0x0f : (attr & 0xf0) >> 4;
+
+	    drawPixel(x+j, y+i, col);
+	}
+	glyph += (i%zoom) ? 0 : FONT_BPL;
+    }
+}
+
+void drawString(int x, int y, char *s, unsigned char attr, int zoom)
+{
+    while (*s) {
+       if (*s == '\r') {
+          x = 0;
+       } else if(*s == '\n') {
+          x = 0; y += (FONT_HEIGHT*zoom);
+       } else {
+	  drawChar(*s, x, y, attr, zoom);
+          x += (FONT_WIDTH*zoom);
+       }
+       s++;
+    }
+}
+
+void moveRect(int oldx, int oldy, int width, int height, int shiftx, int shifty, unsigned char attr)
+{
+    unsigned int newx = oldx + shiftx, newy = oldy + shifty;
+    unsigned int xcount = 0, ycount = 0;
+    unsigned int bitmap[width][height]; // This is very unsafe if it's too big for the stack...
+    unsigned int offs;
+
+    // Save the bitmap
+    while (xcount < width) {
+       while (ycount < height) {
+          offs = ((oldy + ycount) * pitch) + ((oldx + xcount) * 4);
+
+	  bitmap[xcount][ycount] = *((unsigned int*)(fb + offs));
+	  ycount++;
+       }
+       ycount=0;
+       xcount++;
+    }
+    // Wipe it out with background colour
+    drawRect(oldx, oldy, oldx + width, oldy + width, attr, 1);
+    // Draw it again
+    for (int i=newx;i<newx + width;i++) {
+        for (int j=newy;j<newy + height;j++) {
+            offs = (j * pitch) + (i * 4);
+	    *((unsigned int*)(fb + offs)) = bitmap[i-newx][j-newy];
+        }
+    }
+}
+
+void wait_msec(unsigned int n)
+{
+    register unsigned long f, t, r;
+
+    // Get the current counter frequency
+    asm volatile ("mrs %0, cntfrq_el0" : "=r"(f));
+    // Read the current counter
+    asm volatile ("mrs %0, cntpct_el0" : "=r"(t));
+    // Calculate expire value for counter
+    t+=((f/1000)*n)/1000;
+    do{asm volatile ("mrs %0, cntpct_el0" : "=r"(r));}while(r<t);
+}
+
+int strlen(const char *str) {
+    const char *s;
+
+    for (s = str; *s; ++s);
+    return (s - str);
+}
+
+void debugstr(char *str) {
+    if (curx + (strlen(str) * 8)  >= 1920) {
+       curx = 0; cury += 8;
+    }
+    if (cury + 8 >= 1080) {
+       cury = 0;
+    }
+    drawString(curx, cury, str, 0x0f, 1);
+    curx += (strlen(str) * 8);
+}
+
+void debugcrlf(void) {
+    curx = 0; cury += 8;
+}
+
+void debugch(unsigned char b) {
+    unsigned int n;
+    int c;
+    for(c=4;c>=0;c-=4) {
+        n=(b>>c)&0xF;
+        n+=n>9?0x37:0x30;
+        debugstr((char *)&n);
+    }
+    debugstr(" ");
+}
+
+void debughex(unsigned int d) {
+    unsigned int n;
+    int c;
+    for(c=28;c>=0;c-=4) {
+        n=(d>>c)&0xF;
+        n+=n>9?0x37:0x30;
+        debugstr((char *)&n);
+    }
+    debugstr(" ");
+}

part10-multicore/fb.h

@@ -0,0 +1,13 @@
+void fb_init();
+void drawPixel(int x, int y, unsigned char attr);
+void drawChar(unsigned char ch, int x, int y, unsigned char attr, int zoom);
+void drawString(int x, int y, char *s, unsigned char attr, int zoom);
+void drawRect(int x1, int y1, int x2, int y2, unsigned char attr, int fill);
+void drawCircle(int x0, int y0, int radius, unsigned char attr, int fill);
+void drawLine(int x1, int y1, int x2, int y2, unsigned char attr);
+void moveRect(int oldx, int oldy, int width, int height, int shiftx, int shifty, unsigned char attr);
+void wait_msec(unsigned int n);
+void debugstr(char *str);
+void debugcrlf(void);
+void debugch(unsigned char b);
+void debughex(unsigned int d);

part10-multicore/io.c

@@ -0,0 +1,199 @@
+#include "io.h"
+
+// GPIO
+
+enum {
+    GPFSEL0         = PERIPHERAL_BASE + 0x200000,
+    GPSET0          = PERIPHERAL_BASE + 0x20001C,
+    GPCLR0          = PERIPHERAL_BASE + 0x200028,
+    GPPUPPDN0       = PERIPHERAL_BASE + 0x2000E4
+};
+
+enum {
+    GPIO_MAX_PIN       = 53,
+    GPIO_FUNCTION_OUT  = 1,
+    GPIO_FUNCTION_ALT5 = 2,
+    GPIO_FUNCTION_ALT3 = 7,
+    GPIO_FUNCTION_ALT0 = 4
+};
+
+enum {
+    Pull_None = 0,
+    Pull_Down = 1, // Are down and up the right way around?
+    Pull_Up = 2
+};
+
+void mmio_write(long reg, unsigned int val) { *(volatile unsigned int *)reg = val; }
+unsigned int mmio_read(long reg) { return *(volatile unsigned int *)reg; }
+
+unsigned int gpio_call(unsigned int pin_number, unsigned int value, unsigned int base, unsigned int field_size, unsigned int field_max) {
+    unsigned int field_mask = (1 << field_size) - 1;
+  
+    if (pin_number > field_max) return 0;
+    if (value > field_mask) return 0; 
+
+    unsigned int num_fields = 32 / field_size;
+    unsigned int reg = base + ((pin_number / num_fields) * 4);
+    unsigned int shift = (pin_number % num_fields) * field_size;
+
+    unsigned int curval = mmio_read(reg);
+    curval &= ~(field_mask << shift);
+    curval |= value << shift;
+    mmio_write(reg, curval);
+
+    return 1;
+}
+
+unsigned int gpio_set     (unsigned int pin_number, unsigned int value) { return gpio_call(pin_number, value, GPSET0, 1, GPIO_MAX_PIN); }
+unsigned int gpio_clear   (unsigned int pin_number, unsigned int value) { return gpio_call(pin_number, value, GPCLR0, 1, GPIO_MAX_PIN); }
+unsigned int gpio_pull    (unsigned int pin_number, unsigned int value) { return gpio_call(pin_number, value, GPPUPPDN0, 2, GPIO_MAX_PIN); }
+unsigned int gpio_function(unsigned int pin_number, unsigned int value) { return gpio_call(pin_number, value, GPFSEL0, 3, GPIO_MAX_PIN); }
+
+void gpio_useAsAlt0(unsigned int pin_number) {
+    gpio_pull(pin_number, Pull_None);
+    gpio_function(pin_number, GPIO_FUNCTION_ALT0);
+}
+
+void gpio_useAsAlt3(unsigned int pin_number) {
+    gpio_pull(pin_number, Pull_None);
+    gpio_function(pin_number, GPIO_FUNCTION_ALT3);
+}
+
+void gpio_useAsAlt5(unsigned int pin_number) {
+    gpio_pull(pin_number, Pull_None);
+    gpio_function(pin_number, GPIO_FUNCTION_ALT5);
+}
+
+void gpio_initOutputPinWithPullNone(unsigned int pin_number) {
+    gpio_pull(pin_number, Pull_None);
+    gpio_function(pin_number, GPIO_FUNCTION_OUT);
+}
+
+void gpio_setPinOutputBool(unsigned int pin_number, unsigned int onOrOff) {
+    if (onOrOff) {
+        gpio_set(pin_number, 1);
+    } else {
+        gpio_clear(pin_number, 1);
+    }
+}
+
+// UART
+
+enum {
+    AUX_BASE        = PERIPHERAL_BASE + 0x215000,
+    AUX_IRQ         = AUX_BASE,
+    AUX_ENABLES     = AUX_BASE + 4,
+    AUX_MU_IO_REG   = AUX_BASE + 64,
+    AUX_MU_IER_REG  = AUX_BASE + 68,
+    AUX_MU_IIR_REG  = AUX_BASE + 72,
+    AUX_MU_LCR_REG  = AUX_BASE + 76,
+    AUX_MU_MCR_REG  = AUX_BASE + 80,
+    AUX_MU_LSR_REG  = AUX_BASE + 84,
+    AUX_MU_MSR_REG  = AUX_BASE + 88,
+    AUX_MU_SCRATCH  = AUX_BASE + 92,
+    AUX_MU_CNTL_REG = AUX_BASE + 96,
+    AUX_MU_STAT_REG = AUX_BASE + 100,
+    AUX_MU_BAUD_REG = AUX_BASE + 104,
+    AUX_UART_CLOCK  = 500000000,
+    UART_MAX_QUEUE  = 16 * 1024
+};
+
+#define AUX_MU_BAUD(baud) ((AUX_UART_CLOCK/(baud*8))-1)
+
+unsigned char uart_output_queue[UART_MAX_QUEUE];
+unsigned int uart_output_queue_write = 0;
+unsigned int uart_output_queue_read = 0;
+
+void uart_init() {
+    mmio_write(AUX_ENABLES, 1); //enable UART1
+    mmio_write(AUX_MU_IER_REG, 0);
+    mmio_write(AUX_MU_CNTL_REG, 0);
+    mmio_write(AUX_MU_LCR_REG, 3); //8 bits
+    mmio_write(AUX_MU_MCR_REG, 0);
+    mmio_write(AUX_MU_IER_REG, 0);
+    mmio_write(AUX_MU_IIR_REG, 0xC6); //disable interrupts
+    mmio_write(AUX_MU_BAUD_REG, AUX_MU_BAUD(115200));
+    gpio_useAsAlt5(14);
+    gpio_useAsAlt5(15);
+    mmio_write(AUX_MU_CNTL_REG, 3); //enable RX/TX
+}
+
+unsigned int uart_isOutputQueueEmpty() {
+    return uart_output_queue_read == uart_output_queue_write;
+}
+
+unsigned int uart_isReadByteReady()  { return mmio_read(AUX_MU_LSR_REG) & 0x01; }
+unsigned int uart_isWriteByteReady() { return mmio_read(AUX_MU_LSR_REG) & 0x20; }
+
+unsigned char uart_readByte() {
+    while (!uart_isReadByteReady());
+    return (unsigned char)mmio_read(AUX_MU_IO_REG);
+}
+
+void uart_writeByteBlockingActual(unsigned char ch) {
+    while (!uart_isWriteByteReady()); 
+    mmio_write(AUX_MU_IO_REG, (unsigned int)ch);
+}
+
+void uart_loadOutputFifo() {
+    while (!uart_isOutputQueueEmpty() && uart_isWriteByteReady()) {
+        uart_writeByteBlockingActual(uart_output_queue[uart_output_queue_read]);
+        uart_output_queue_read = (uart_output_queue_read + 1) & (UART_MAX_QUEUE - 1); // Don't overrun
+    }
+}
+
+void uart_writeByteBlocking(unsigned char ch) {
+    unsigned int next = (uart_output_queue_write + 1) & (UART_MAX_QUEUE - 1); // Don't overrun
+
+    while (next == uart_output_queue_read) uart_loadOutputFifo();
+
+    uart_output_queue[uart_output_queue_write] = ch;
+    uart_output_queue_write = next;
+}
+
+void uart_writeText(char *buffer) {
+    while (*buffer) {
+       if (*buffer == '\n') uart_writeByteBlockingActual('\r');
+       uart_writeByteBlockingActual(*buffer++);
+    }
+}
+
+void uart_drainOutputQueue() {
+    while (!uart_isOutputQueueEmpty()) uart_loadOutputFifo();
+}
+
+void uart_update() {
+    uart_loadOutputFifo();
+
+    if (uart_isReadByteReady()) {
+       unsigned char ch = uart_readByte();
+       if (ch == '\r') uart_writeText("\n"); else uart_writeByteBlocking(ch);
+    }
+}
+
+void uart_hex(unsigned int d) {
+    unsigned int n;
+    int c;
+    for(c=28;c>=0;c-=4) {
+        // get highest tetrad
+        n=(d>>c)&0xF;
+        // 0-9 => '0'-'9', 10-15 => 'A'-'F'
+        n+=n>9?0x37:0x30;
+
+        uart_writeByteBlockingActual(n);
+    }
+}
+
+void uart_byte(unsigned char b) {
+    unsigned int n;
+    int c;
+    for(c=4;c>=0;c-=4) {
+        // get highest tetrad
+        n=(b>>c)&0xF;
+        // 0-9 => '0'-'9', 10-15 => 'A'-'F'
+        n+=n>9?0x37:0x30;
+
+        uart_writeByteBlockingActual(n);
+    }
+    uart_writeByteBlockingActual(' ');
+}

part10-multicore/io.h

@@ -0,0 +1,15 @@
+#define PERIPHERAL_BASE 0xFE000000
+
+void uart_init();
+void uart_writeText(char *buffer);
+void uart_loadOutputFifo();
+unsigned char uart_readByte();
+unsigned int uart_isReadByteReady();
+void uart_writeByteBlockingActual(unsigned char ch);
+void uart_update();
+void mmio_write(long reg, unsigned int val);
+unsigned int mmio_read(long reg);
+void gpio_useAsAlt0(unsigned int pin_number);
+void gpio_useAsAlt3(unsigned int pin_number);
+void uart_hex(unsigned int d);
+void uart_byte(unsigned char b);

part10-multicore/kernel.c

@@ -0,0 +1,120 @@
+#include "fb.h"
+#include "io.h"
+#include "multicore.h"
+
+#define PWM_BASE        (PERIPHERAL_BASE + 0x20C000 + 0x800) /* PWM1 register base address on RPi4 */
+#define CLOCK_BASE      (PERIPHERAL_BASE + 0x101000)
+
+#define BCM2711_PWMCLK_CNTL 40
+#define BCM2711_PWMCLK_DIV  41
+#define PM_PASSWORD 0x5A000000 
+
+#define BCM2711_PWM_CONTROL 0
+#define BCM2711_PWM_STATUS  1
+#define BCM2711_PWM_DMAC    2
+#define BCM2711_PWM0_RANGE  4
+#define BCM2711_PWM0_DATA   5
+#define BCM2711_PWM_FIFO    6
+#define BCM2711_PWM1_RANGE  8
+#define BCM2711_PWM1_DATA   9
+
+#define BCM2711_PWM1_USEFIFO  0x2000  /* Data from FIFO */
+#define BCM2711_PWM1_ENABLE   0x0100  /* Channel enable */
+#define BCM2711_PWM0_USEFIFO  0x0020  /* Data from FIFO */
+#define BCM2711_PWM0_ENABLE   0x0001  /* Channel enable */
+#define BCM2711_PWM_ENAB      0x80000000  /* PWM DMA Configuration: DMA Enable (bit 31 set) */
+
+#define BCM2711_GAPO2 0x20
+#define BCM2711_GAPO1 0x10
+#define BCM2711_RERR1 0x8
+#define BCM2711_WERR1 0x4
+#define BCM2711_FULL1 0x1
+#define ERRORMASK (BCM2711_GAPO2 | BCM2711_GAPO1 | BCM2711_RERR1 | BCM2711_WERR1)
+
+volatile unsigned* clk = (void*)CLOCK_BASE;
+volatile unsigned* pwm = (void*)PWM_BASE;
+
+static void audio_init(void)
+{   
+    gpio_useAsAlt0(40); // Ensure PWM1 is mapped to GPIO 40/41
+    gpio_useAsAlt0(41);
+    wait_msec(2);
+
+    // Setup clock
+    
+    *(clk + BCM2711_PWMCLK_CNTL) = PM_PASSWORD | (1 << 5); // Stop clock
+    wait_msec(2);
+
+    int idiv = 2;
+    *(clk + BCM2711_PWMCLK_DIV)  = PM_PASSWORD | (idiv<<12);
+    *(clk + BCM2711_PWMCLK_CNTL) = PM_PASSWORD | 16 | 1;  // Osc + Enable
+    wait_msec(2);
+
+    // Setup PWM
+    
+    *(pwm + BCM2711_PWM_CONTROL) = 0;
+    wait_msec(2);
+    
+    *(pwm+BCM2711_PWM0_RANGE) = 0x264; // 44.1khz, Stereo, 8-bit (54Mhz / 44100 / 2)
+    *(pwm+BCM2711_PWM1_RANGE) = 0x264;
+    
+    *(pwm+BCM2711_PWM_CONTROL) =
+          BCM2711_PWM1_USEFIFO |
+          BCM2711_PWM1_ENABLE | 
+          BCM2711_PWM0_USEFIFO | 
+          BCM2711_PWM0_ENABLE | 1<<6;
+}
+
+void playaudio_cpu()
+{
+    int i=0;
+    long status;
+
+    extern unsigned char _binary_audio_bin_start[];
+    extern unsigned char _binary_audio_bin_size[];
+    
+    unsigned int size = (long)&_binary_audio_bin_size;
+    unsigned char *data = &(_binary_audio_bin_start[0]);
+
+    // Write data out to FIFO
+
+    while (i < size) {
+        status = *(pwm + BCM2711_PWM_STATUS);
+        if (!(status & BCM2711_FULL1)) {
+            *(pwm+BCM2711_PWM_FIFO) = *(data + i);
+            i++;
+            *(pwm+BCM2711_PWM_FIFO) = *(data + i);
+            i++;
+        }
+        if ((status & ERRORMASK)) {
+            *(pwm+BCM2711_PWM_STATUS) = ERRORMASK;
+        }
+    }
+}
+
+void core1_main(void)
+{
+    debugstr("Playing on CPU Core #1... ");
+    playaudio_cpu();
+    debugstr(" done"); debugcrlf();
+}
+
+void core0_main(void)
+{
+    while (1) {
+       wait_msec(0x100000);
+       debugstr("x");
+    }
+}
+
+void main(void)
+{
+    fb_init();
+
+    debugstr("Initialising audio unit... ");
+    audio_init();
+    debugstr("done"); debugcrlf();
+
+    start_core1(core1_main);
+    core0_main();
+}

part10-multicore/link.ld

@@ -0,0 +1,24 @@
+SECTIONS
+{
+    .text : { KEEP(*(.text.boot)) *(.text .text.* .gnu.linkonce.t*) }
+    .rodata : { *(.rodata .rodata.* .gnu.linkonce.r*) }
+    PROVIDE(_data = .);
+    .data : { *(.data .data.* .gnu.linkonce.d*) }
+    .bss (NOLOAD) : {
+        . = ALIGN(16);
+        __bss_start = .;
+        *(.bss .bss.*)
+        *(COMMON)
+        __bss_end = .;
+        __bss_size = (__bss_end - __bss_start)>>3;
+    }
+    .testStack :
+    {
+        . = ALIGN(16);
+        . = . + 512;
+        __test_stack  = .;
+        *(.testStack)
+    }
+    _end = .;
+   /DISCARD/ : { *(.comment) *(.gnu*) *(.note*) *(.eh_frame*) }
+}

part10-multicore/mb.c

@@ -0,0 +1,39 @@
+#include "io.h"
+
+// The buffer must be 16-byte aligned as only the upper 28 bits of the address can be passed via the mailbox
+volatile unsigned int __attribute__((aligned(16))) mbox[36];
+
+enum {
+    VIDEOCORE_MBOX = (PERIPHERAL_BASE + 0x0000B880),
+    MBOX_READ      = (VIDEOCORE_MBOX + 0x0),
+    MBOX_POLL      = (VIDEOCORE_MBOX + 0x10),
+    MBOX_SENDER    = (VIDEOCORE_MBOX + 0x14),
+    MBOX_STATUS    = (VIDEOCORE_MBOX + 0x18),
+    MBOX_CONFIG    = (VIDEOCORE_MBOX + 0x1C),
+    MBOX_WRITE     = (VIDEOCORE_MBOX + 0x20),
+    MBOX_RESPONSE  = 0x80000000,
+    MBOX_FULL      = 0x80000000,
+    MBOX_EMPTY     = 0x40000000
+};
+
+unsigned int mbox_call(unsigned char ch)
+{
+    // 28-bit address (MSB) and 4-bit value (LSB)
+    unsigned int r = ((unsigned int)((long) &mbox) &~ 0xF) | (ch & 0xF);
+
+    // Wait until we can write
+    while (mmio_read(MBOX_STATUS) & MBOX_FULL);
+    
+    // Write the address of our buffer to the mailbox with the channel appended
+    mmio_write(MBOX_WRITE, r);
+
+    while (1) {
+        // Is there a reply?
+        while (mmio_read(MBOX_STATUS) & MBOX_EMPTY);
+
+        // Is it a reply to our message?
+        if (r == mmio_read(MBOX_READ)) return mbox[1]==MBOX_RESPONSE; // Is it successful?
+           
+    }
+    return 0;
+}

part10-multicore/mb.h

@@ -0,0 +1,34 @@
+extern volatile unsigned int mbox[36];
+
+enum {
+    MBOX_REQUEST  = 0
+};
+
+enum {
+    MBOX_CH_POWER = 0,
+    MBOX_CH_FB    = 1,
+    MBOX_CH_VUART = 2,
+    MBOX_CH_VCHIQ = 3,
+    MBOX_CH_LEDS  = 4,
+    MBOX_CH_BTNS  = 5,
+    MBOX_CH_TOUCH = 6,
+    MBOX_CH_COUNT = 7,
+    MBOX_CH_PROP  = 8 // Request from ARM for response by VideoCore
+};
+
+enum {
+    MBOX_TAG_SETPOWER   = 0x28001,
+    MBOX_TAG_SETCLKRATE = 0x38002,
+
+    MBOX_TAG_SETPHYWH   = 0x48003,
+    MBOX_TAG_SETVIRTWH  = 0x48004,
+    MBOX_TAG_SETVIRTOFF = 0x48009,
+    MBOX_TAG_SETDEPTH   = 0x48005,
+    MBOX_TAG_SETPXLORDR = 0x48006,
+    MBOX_TAG_GETFB      = 0x40001,
+    MBOX_TAG_GETPITCH   = 0x40008,
+
+    MBOX_TAG_LAST       = 0
+};
+
+unsigned int mbox_call(unsigned char ch);

part10-multicore/multicore.c

@@ -0,0 +1,29 @@
+#include "multicore.h"
+
+void store32(unsigned long address, unsigned long value)
+{
+    *(unsigned long *) address = value;
+}
+
+unsigned long load32(unsigned long address)
+{
+    return *(unsigned long *) address;
+}
+
+void start_core1(void (*func)(void))
+{
+    store32((unsigned long)&spin_cpu1, (unsigned long)func);
+    asm volatile ("sev");
+}
+
+void start_core2(void (*func)(void))
+{
+    store32((unsigned long)&spin_cpu2, (unsigned long)func);
+    asm volatile ("sev");
+}
+
+void start_core3(void (*func)(void))
+{
+    store32((unsigned long)&spin_cpu3, (unsigned long)func);
+    asm volatile ("sev");
+}

part10-multicore/multicore.h

@@ -0,0 +1,10 @@
+extern unsigned int spin_cpu0;
+extern unsigned int spin_cpu1;
+extern unsigned int spin_cpu2;
+extern unsigned int spin_cpu3;
+
+unsigned long load32(unsigned long address);
+void start_core1(void (*func)(void));
+void start_core2(void (*func)(void));
+void start_core3(void (*func)(void));
+unsigned int get_core_number(void);

part10-multicore/terminal.h

@@ -0,0 +1,253 @@
+unsigned int vgapal[] = {
+    0x000000,
+    0x0000AA,
+    0x00AA00,
+    0x00AAAA,
+    0xAA0000,
+    0xAA00AA,
+    0xAA5500,
+    0xAAAAAA,
+    0x555555,
+    0x5555FF,
+    0x55FF55,
+    0x55FFFF,
+    0xFF5555,
+    0xFF55FF,
+    0xFFFF55,
+    0xFFFFFF
+};
+
+enum {
+    FONT_WIDTH     = 8,
+    FONT_HEIGHT    = 8,
+    FONT_BPG       = 8,  // Bytes per glyph
+    FONT_BPL       = 1,  // Bytes per line
+    FONT_NUMGLYPHS = 224
+};
+
+unsigned char font[FONT_NUMGLYPHS][FONT_BPG] = {
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0000 (nul)
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0001
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0002
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0003
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0004
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0005
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0006
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0007
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0008
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0009
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+000A
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+000B
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+000C
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+000D
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+000E
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+000F
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0010
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0011
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0012
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0013
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0014
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0015
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0016
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0017
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0018
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0019
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+001A
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+001B
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+001C
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+001D
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+001E
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+001F
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0020 (space)
+    { 0x18, 0x3C, 0x3C, 0x18, 0x18, 0x00, 0x18, 0x00},   // U+0021 (!)
+    { 0x36, 0x36, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0022 (")
+    { 0x36, 0x36, 0x7F, 0x36, 0x7F, 0x36, 0x36, 0x00},   // U+0023 (#)
+    { 0x0C, 0x3E, 0x03, 0x1E, 0x30, 0x1F, 0x0C, 0x00},   // U+0024 ($)
+    { 0x00, 0x63, 0x33, 0x18, 0x0C, 0x66, 0x63, 0x00},   // U+0025 (%)
+    { 0x1C, 0x36, 0x1C, 0x6E, 0x3B, 0x33, 0x6E, 0x00},   // U+0026 (&)
+    { 0x06, 0x06, 0x03, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0027 (')
+    { 0x18, 0x0C, 0x06, 0x06, 0x06, 0x0C, 0x18, 0x00},   // U+0028 (()
+    { 0x06, 0x0C, 0x18, 0x18, 0x18, 0x0C, 0x06, 0x00},   // U+0029 ())
+    { 0x00, 0x66, 0x3C, 0xFF, 0x3C, 0x66, 0x00, 0x00},   // U+002A (*)
+    { 0x00, 0x0C, 0x0C, 0x3F, 0x0C, 0x0C, 0x00, 0x00},   // U+002B (+)
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x0C, 0x0C, 0x06},   // U+002C (,)
+    { 0x00, 0x00, 0x00, 0x3F, 0x00, 0x00, 0x00, 0x00},   // U+002D (-)
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x0C, 0x0C, 0x00},   // U+002E (.)
+    { 0x60, 0x30, 0x18, 0x0C, 0x06, 0x03, 0x01, 0x00},   // U+002F (/)
+    { 0x3E, 0x63, 0x73, 0x7B, 0x6F, 0x67, 0x3E, 0x00},   // U+0030 (0)
+    { 0x0C, 0x0E, 0x0C, 0x0C, 0x0C, 0x0C, 0x3F, 0x00},   // U+0031 (1)
+    { 0x1E, 0x33, 0x30, 0x1C, 0x06, 0x33, 0x3F, 0x00},   // U+0032 (2)
+    { 0x1E, 0x33, 0x30, 0x1C, 0x30, 0x33, 0x1E, 0x00},   // U+0033 (3)
+    { 0x38, 0x3C, 0x36, 0x33, 0x7F, 0x30, 0x78, 0x00},   // U+0034 (4)
+    { 0x3F, 0x03, 0x1F, 0x30, 0x30, 0x33, 0x1E, 0x00},   // U+0035 (5)
+    { 0x1C, 0x06, 0x03, 0x1F, 0x33, 0x33, 0x1E, 0x00},   // U+0036 (6)
+    { 0x3F, 0x33, 0x30, 0x18, 0x0C, 0x0C, 0x0C, 0x00},   // U+0037 (7)
+    { 0x1E, 0x33, 0x33, 0x1E, 0x33, 0x33, 0x1E, 0x00},   // U+0038 (8)
+    { 0x1E, 0x33, 0x33, 0x3E, 0x30, 0x18, 0x0E, 0x00},   // U+0039 (9)
+    { 0x00, 0x0C, 0x0C, 0x00, 0x00, 0x0C, 0x0C, 0x00},   // U+003A (:)
+    { 0x00, 0x0C, 0x0C, 0x00, 0x00, 0x0C, 0x0C, 0x06},   // U+003B (;)
+    { 0x18, 0x0C, 0x06, 0x03, 0x06, 0x0C, 0x18, 0x00},   // U+003C (<)
+    { 0x00, 0x00, 0x3F, 0x00, 0x00, 0x3F, 0x00, 0x00},   // U+003D (=)
+    { 0x06, 0x0C, 0x18, 0x30, 0x18, 0x0C, 0x06, 0x00},   // U+003E (>)
+    { 0x1E, 0x33, 0x30, 0x18, 0x0C, 0x00, 0x0C, 0x00},   // U+003F (?)
+    { 0x3E, 0x63, 0x7B, 0x7B, 0x7B, 0x03, 0x1E, 0x00},   // U+0040 (@)
+    { 0x0C, 0x1E, 0x33, 0x33, 0x3F, 0x33, 0x33, 0x00},   // U+0041 (A)
+    { 0x3F, 0x66, 0x66, 0x3E, 0x66, 0x66, 0x3F, 0x00},   // U+0042 (B)
+    { 0x3C, 0x66, 0x03, 0x03, 0x03, 0x66, 0x3C, 0x00},   // U+0043 (C)
+    { 0x1F, 0x36, 0x66, 0x66, 0x66, 0x36, 0x1F, 0x00},   // U+0044 (D)
+    { 0x7F, 0x46, 0x16, 0x1E, 0x16, 0x46, 0x7F, 0x00},   // U+0045 (E)
+    { 0x7F, 0x46, 0x16, 0x1E, 0x16, 0x06, 0x0F, 0x00},   // U+0046 (F)
+    { 0x3C, 0x66, 0x03, 0x03, 0x73, 0x66, 0x7C, 0x00},   // U+0047 (G)
+    { 0x33, 0x33, 0x33, 0x3F, 0x33, 0x33, 0x33, 0x00},   // U+0048 (H)
+    { 0x1E, 0x0C, 0x0C, 0x0C, 0x0C, 0x0C, 0x1E, 0x00},   // U+0049 (I)
+    { 0x78, 0x30, 0x30, 0x30, 0x33, 0x33, 0x1E, 0x00},   // U+004A (J)
+    { 0x67, 0x66, 0x36, 0x1E, 0x36, 0x66, 0x67, 0x00},   // U+004B (K)
+    { 0x0F, 0x06, 0x06, 0x06, 0x46, 0x66, 0x7F, 0x00},   // U+004C (L)
+    { 0x63, 0x77, 0x7F, 0x7F, 0x6B, 0x63, 0x63, 0x00},   // U+004D (M)
+    { 0x63, 0x67, 0x6F, 0x7B, 0x73, 0x63, 0x63, 0x00},   // U+004E (N)
+    { 0x1C, 0x36, 0x63, 0x63, 0x63, 0x36, 0x1C, 0x00},   // U+004F (O)
+    { 0x3F, 0x66, 0x66, 0x3E, 0x06, 0x06, 0x0F, 0x00},   // U+0050 (P)
+    { 0x1E, 0x33, 0x33, 0x33, 0x3B, 0x1E, 0x38, 0x00},   // U+0051 (Q)
+    { 0x3F, 0x66, 0x66, 0x3E, 0x36, 0x66, 0x67, 0x00},   // U+0052 (R)
+    { 0x1E, 0x33, 0x07, 0x0E, 0x38, 0x33, 0x1E, 0x00},   // U+0053 (S)
+    { 0x3F, 0x2D, 0x0C, 0x0C, 0x0C, 0x0C, 0x1E, 0x00},   // U+0054 (T)
+    { 0x33, 0x33, 0x33, 0x33, 0x33, 0x33, 0x3F, 0x00},   // U+0055 (U)
+    { 0x33, 0x33, 0x33, 0x33, 0x33, 0x1E, 0x0C, 0x00},   // U+0056 (V)
+    { 0x63, 0x63, 0x63, 0x6B, 0x7F, 0x77, 0x63, 0x00},   // U+0057 (W)
+    { 0x63, 0x63, 0x36, 0x1C, 0x1C, 0x36, 0x63, 0x00},   // U+0058 (X)
+    { 0x33, 0x33, 0x33, 0x1E, 0x0C, 0x0C, 0x1E, 0x00},   // U+0059 (Y)
+    { 0x7F, 0x63, 0x31, 0x18, 0x4C, 0x66, 0x7F, 0x00},   // U+005A (Z)
+    { 0x1E, 0x06, 0x06, 0x06, 0x06, 0x06, 0x1E, 0x00},   // U+005B ([)
+    { 0x03, 0x06, 0x0C, 0x18, 0x30, 0x60, 0x40, 0x00},   // U+005C (\)
+    { 0x1E, 0x18, 0x18, 0x18, 0x18, 0x18, 0x1E, 0x00},   // U+005D (])
+    { 0x08, 0x1C, 0x36, 0x63, 0x00, 0x00, 0x00, 0x00},   // U+005E (^)
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xFF},   // U+005F (_)
+    { 0x0C, 0x0C, 0x18, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+0060 (`)
+    { 0x00, 0x00, 0x1E, 0x30, 0x3E, 0x33, 0x6E, 0x00},   // U+0061 (a)
+    { 0x07, 0x06, 0x06, 0x3E, 0x66, 0x66, 0x3B, 0x00},   // U+0062 (b)
+    { 0x00, 0x00, 0x1E, 0x33, 0x03, 0x33, 0x1E, 0x00},   // U+0063 (c)
+    { 0x38, 0x30, 0x30, 0x3e, 0x33, 0x33, 0x6E, 0x00},   // U+0064 (d)
+    { 0x00, 0x00, 0x1E, 0x33, 0x3f, 0x03, 0x1E, 0x00},   // U+0065 (e)
+    { 0x1C, 0x36, 0x06, 0x0f, 0x06, 0x06, 0x0F, 0x00},   // U+0066 (f)
+    { 0x00, 0x00, 0x6E, 0x33, 0x33, 0x3E, 0x30, 0x1F},   // U+0067 (g)
+    { 0x07, 0x06, 0x36, 0x6E, 0x66, 0x66, 0x67, 0x00},   // U+0068 (h)
+    { 0x0C, 0x00, 0x0E, 0x0C, 0x0C, 0x0C, 0x1E, 0x00},   // U+0069 (i)
+    { 0x30, 0x00, 0x30, 0x30, 0x30, 0x33, 0x33, 0x1E},   // U+006A (j)
+    { 0x07, 0x06, 0x66, 0x36, 0x1E, 0x36, 0x67, 0x00},   // U+006B (k)
+    { 0x0E, 0x0C, 0x0C, 0x0C, 0x0C, 0x0C, 0x1E, 0x00},   // U+006C (l)
+    { 0x00, 0x00, 0x33, 0x7F, 0x7F, 0x6B, 0x63, 0x00},   // U+006D (m)
+    { 0x00, 0x00, 0x1F, 0x33, 0x33, 0x33, 0x33, 0x00},   // U+006E (n)
+    { 0x00, 0x00, 0x1E, 0x33, 0x33, 0x33, 0x1E, 0x00},   // U+006F (o)
+    { 0x00, 0x00, 0x3B, 0x66, 0x66, 0x3E, 0x06, 0x0F},   // U+0070 (p)
+    { 0x00, 0x00, 0x6E, 0x33, 0x33, 0x3E, 0x30, 0x78},   // U+0071 (q)
+    { 0x00, 0x00, 0x3B, 0x6E, 0x66, 0x06, 0x0F, 0x00},   // U+0072 (r)
+    { 0x00, 0x00, 0x3E, 0x03, 0x1E, 0x30, 0x1F, 0x00},   // U+0073 (s)
+    { 0x08, 0x0C, 0x3E, 0x0C, 0x0C, 0x2C, 0x18, 0x00},   // U+0074 (t)
+    { 0x00, 0x00, 0x33, 0x33, 0x33, 0x33, 0x6E, 0x00},   // U+0075 (u)
+    { 0x00, 0x00, 0x33, 0x33, 0x33, 0x1E, 0x0C, 0x00},   // U+0076 (v)
+    { 0x00, 0x00, 0x63, 0x6B, 0x7F, 0x7F, 0x36, 0x00},   // U+0077 (w)
+    { 0x00, 0x00, 0x63, 0x36, 0x1C, 0x36, 0x63, 0x00},   // U+0078 (x)
+    { 0x00, 0x00, 0x33, 0x33, 0x33, 0x3E, 0x30, 0x1F},   // U+0079 (y)
+    { 0x00, 0x00, 0x3F, 0x19, 0x0C, 0x26, 0x3F, 0x00},   // U+007A (z)
+    { 0x38, 0x0C, 0x0C, 0x07, 0x0C, 0x0C, 0x38, 0x00},   // U+007B ({)
+    { 0x18, 0x18, 0x18, 0x00, 0x18, 0x18, 0x18, 0x00},   // U+007C (|)
+    { 0x07, 0x0C, 0x0C, 0x38, 0x0C, 0x0C, 0x07, 0x00},   // U+007D (})
+    { 0x6E, 0x3B, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+007E (~)
+    { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},   // U+007F
+    { 0x1E, 0x33, 0x03, 0x33, 0x1E, 0x18, 0x30, 0x1E},   // U+00C7 (C cedille)
+    { 0x00, 0x33, 0x00, 0x33, 0x33, 0x33, 0x7E, 0x00},   // U+00FC (u umlaut)
+    { 0x38, 0x00, 0x1E, 0x33, 0x3F, 0x03, 0x1E, 0x00},   // U+00E9 (e aigu)
+    { 0x7E, 0xC3, 0x3C, 0x60, 0x7C, 0x66, 0xFC, 0x00},   // U+00E2 (a circumflex)
+    { 0x33, 0x00, 0x1E, 0x30, 0x3E, 0x33, 0x7E, 0x00},   // U+00E4 (a umlaut)
+    { 0x07, 0x00, 0x1E, 0x30, 0x3E, 0x33, 0x7E, 0x00},   // U+00E0 (a grave)
+    { 0x0C, 0x0C, 0x1E, 0x30, 0x3E, 0x33, 0x7E, 0x00},   // U+00E5 (a ring)
+    { 0x00, 0x00, 0x1E, 0x03, 0x03, 0x1E, 0x30, 0x1C},   // U+00E7 (c cedille)
+    { 0x7E, 0xC3, 0x3C, 0x66, 0x7E, 0x06, 0x3C, 0x00},   // U+00EA (e circumflex)
+    { 0x33, 0x00, 0x1E, 0x33, 0x3F, 0x03, 0x1E, 0x00},   // U+00EB (e umlaut)
+    { 0x07, 0x00, 0x1E, 0x33, 0x3F, 0x03, 0x1E, 0x00},   // U+00E8 (e grave)
+    { 0x33, 0x00, 0x0E, 0x0C, 0x0C, 0x0C, 0x1E, 0x00},   // U+00EF (i umlaut)
+    { 0x3E, 0x63, 0x1C, 0x18, 0x18, 0x18, 0x3C, 0x00},   // U+00EE (i circumflex)
+    { 0x07, 0x00, 0x0E, 0x0C, 0x0C, 0x0C, 0x1E, 0x00},   // U+00EC (i grave)
+    { 0x63, 0x1C, 0x36, 0x63, 0x7F, 0x63, 0x63, 0x00},   // U+00C4 (A umlaut)
+    { 0x0C, 0x0C, 0x00, 0x1E, 0x33, 0x3F, 0x33, 0x00},   // U+00C5 (A ring)
+    { 0x07, 0x00, 0x3F, 0x06, 0x1E, 0x06, 0x3F, 0x00},   // U+00C8 (E grave)
+    { 0x00, 0x00, 0xFE, 0x30, 0xFE, 0x33, 0xFE, 0x00},   // U+00E6 (ae)
+    { 0x7C, 0x36, 0x33, 0x7F, 0x33, 0x33, 0x73, 0x00},   // U+00C6 (AE)
+    { 0x1E, 0x33, 0x00, 0x1E, 0x33, 0x33, 0x1E, 0x00},   // U+00F4 (o circumflex)
+    { 0x00, 0x33, 0x00, 0x1E, 0x33, 0x33, 0x1E, 0x00},   // U+00F6 (o umlaut)
+    { 0x00, 0x07, 0x00, 0x1E, 0x33, 0x33, 0x1E, 0x00},   // U+00F2 (o grave)
+    { 0x1E, 0x33, 0x00, 0x33, 0x33, 0x33, 0x7E, 0x00},   // U+00FB (u circumflex)
+    { 0x00, 0x07, 0x00, 0x33, 0x33, 0x33, 0x7E, 0x00},   // U+00F9 (u grave)
+    { 0x00, 0x33, 0x00, 0x33, 0x33, 0x3E, 0x30, 0x1F},   // U+00FF (y umlaut)
+    { 0xC3, 0x18, 0x3C, 0x66, 0x66, 0x3C, 0x18, 0x00},   // U+00D6 (O umlaut)
+    { 0x33, 0x00, 0x33, 0x33, 0x33, 0x33, 0x1E, 0x00},   // U+00DC (U umlaut)
+    { 0x18, 0x18, 0x7E, 0x03, 0x03, 0x7E, 0x18, 0x18},   // U+00A2 (dollarcents)
+    { 0x1C, 0x36, 0x26, 0x0F, 0x06, 0x67, 0x3F, 0x00},   // U+00A3 (pound sterling)
+    { 0x33, 0x33, 0x1E, 0x3F, 0x0C, 0x3F, 0x0C, 0x0C},   // U+00A5 (yen)
+    { 0x7C, 0xC6, 0x1C, 0x36, 0x36, 0x1C, 0x33, 0x1E},   // U+00A7 (paragraph)
+    { 0x70, 0xD8, 0x18, 0x3C, 0x18, 0x18, 0x1B, 0x0E},   // U+0192 (dutch florijn)
+    { 0x38, 0x00, 0x1E, 0x30, 0x3E, 0x33, 0x7E, 0x00},   // U+00E1 (a aigu)
+    { 0x1C, 0x00, 0x0E, 0x0C, 0x0C, 0x0C, 0x1E, 0x00},   // U+00ED (i augu)
+    { 0x00, 0x38, 0x00, 0x1E, 0x33, 0x33, 0x1E, 0x00},   // U+00F3 (o aigu)
+    { 0x00, 0x38, 0x00, 0x33, 0x33, 0x33, 0x7E, 0x00},   // U+00FA (u aigu)
+    { 0x00, 0x1F, 0x00, 0x1F, 0x33, 0x33, 0x33, 0x00},   // U+00F1 (n ~)
+    { 0x3F, 0x00, 0x33, 0x37, 0x3F, 0x3B, 0x33, 0x00},   // U+00D1 (N ~)
+    { 0x3C, 0x36, 0x36, 0x7C, 0x00, 0x00, 0x00, 0x00},   // U+00AA (superscript a)
+    { 0x1C, 0x36, 0x36, 0x1C, 0x00, 0x00, 0x00, 0x00},   // U+00BA (superscript 0)
+    { 0x0C, 0x00, 0x0C, 0x06, 0x03, 0x33, 0x1E, 0x00},   // U+00BF (inverted ?)
+    { 0x00, 0x00, 0x00, 0x3F, 0x03, 0x03, 0x00, 0x00},   // U+2310 (gun pointing right)
+    { 0x00, 0x00, 0x00, 0x3F, 0x30, 0x30, 0x00, 0x00},   // U+00AC (gun pointing left)
+    { 0xC3, 0x63, 0x33, 0x7B, 0xCC, 0x66, 0x33, 0xF0},   // U+00BD (1/2)
+    { 0xC3, 0x63, 0x33, 0xBD, 0xEC, 0xF6, 0xF3, 0x03},   // U+00BC (1/4)
+    { 0x18, 0x18, 0x00, 0x18, 0x18, 0x18, 0x18, 0x00},   // U+00A1 (inverted !)
+    { 0x00, 0xCC, 0x66, 0x33, 0x66, 0xCC, 0x00, 0x00},   // U+00AB (<<)
+    { 0x00, 0x33, 0x66, 0xCC, 0x66, 0x33, 0x00, 0x00},   // U+00BB (>>)
+    { 0x55, 0x00, 0xAA, 0x00, 0x55, 0x00, 0xAA, 0x00},   // U+2591 (25% solid)
+    { 0x55, 0xAA, 0x55, 0xAA, 0x55, 0xAA, 0x55, 0xAA},   // U+2592 (50% solid)
+    { 0xFF, 0xAA, 0xFF, 0x55, 0xFF, 0xAA, 0xFF, 0x55},   // U+2593 (75% solid)
+    { 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08},   // U+2502 (thin vertical)
+    { 0x08, 0x08, 0x08, 0x08, 0x0f, 0x08, 0x08, 0x08},   // U+2524 (down L, left L, up L)
+    { 0x08, 0x08, 0x08, 0x0F, 0x08, 0x0F, 0x08, 0x08},   // U+2561 (up L, down L, left D)
+    { 0x14, 0x14, 0x14, 0x14, 0x17, 0x14, 0x14, 0x14},   // U+2562 (up D, down D, left L)
+    { 0x00, 0x00, 0x00, 0x00, 0x1F, 0x14, 0x14, 0x14},   // U+2556 (down D, left L)
+    { 0x00, 0x00, 0x00, 0x0F, 0x08, 0x0F, 0x08, 0x08},   // U+2555 (down L, left D)
+    { 0x14, 0x14, 0x14, 0x17, 0x10, 0x17, 0x14, 0x14},   // U+2563 (up D, down D, left D)
+    { 0x14, 0x14, 0x14, 0x14, 0x14, 0x14, 0x14, 0x14},   // U+2551 (double vertical)
+    { 0x00, 0x00, 0x00, 0x1F, 0x10, 0x17, 0x14, 0x14},   // U+2557 (down D, left D)
+    { 0x14, 0x14, 0x14, 0x17, 0x10, 0x1F, 0x00, 0x00},   // U+255D (up D, left D)
+    { 0x14, 0x14, 0x14, 0x14, 0x1F, 0x00, 0x00, 0x00},   // U+255C (up D, left L)
+    { 0x08, 0x08, 0x08, 0x0F, 0x08, 0x0F, 0x00, 0x00},   // U+255B (up L, left D)
+    { 0x00, 0x00, 0x00, 0x00, 0x0f, 0x08, 0x08, 0x08},   // U+2510 (down L, left L)
+    { 0x08, 0x08, 0x08, 0x08, 0xf8, 0x00, 0x00, 0x00},   // U+2514 (up L, right L)
+    { 0x08, 0x08, 0x08, 0x08, 0xff, 0x00, 0x00, 0x00},   // U+2534 (up L, right L, left L)
+    { 0x00, 0x00, 0x00, 0x00, 0xff, 0x08, 0x08, 0x08},   // U+252C (down L, right L, left L)
+    { 0x08, 0x08, 0x08, 0x08, 0xf8, 0x08, 0x08, 0x08},   // U+251C (down L, right L, up L)
+    { 0x00, 0x00, 0x00, 0x00, 0xFF, 0x00, 0x00, 0x00},   // U+2500 (thin horizontal)
+    { 0x08, 0x08, 0x08, 0x08, 0xff, 0x08, 0x08, 0x08},   // U+253C (up L, right L, left L, down L)
+    { 0x08, 0x08, 0x08, 0xF8, 0x08, 0xF8, 0x08, 0x08},   // U+255E (up L, down L, right D)
+    { 0x14, 0x14, 0x14, 0x14, 0xF4, 0x14, 0x14, 0x14},   // U+255F (up D, down D, right L)
+    { 0x14, 0x14, 0x14, 0xF4, 0x04, 0xFC, 0x00, 0x00},   // U+255A (up D, right D)
+    { 0x00, 0x00, 0x00, 0xFC, 0x04, 0xF4, 0x14, 0x14},   // U+2554 (down D, right D)
+    { 0x14, 0x14, 0x14, 0xF7, 0x00, 0xFF, 0x00, 0x00},   // U+2569 (left D, right D, up D)
+    { 0x00, 0x00, 0x00, 0xFF, 0x00, 0xF7, 0x14, 0x14},   // U+2566 (left D, right D, down D)
+    { 0x14, 0x14, 0x14, 0xF4, 0x04, 0xF4, 0x14, 0x14},   // U+2560 (up D, down D, right D)
+    { 0x00, 0x00, 0x00, 0xFF, 0x00, 0xFF, 0x00, 0x00},   // U+2550 (double horizontal)
+    { 0x14, 0x14, 0x14, 0xF7, 0x00, 0xF7, 0x14, 0x14},   // U+256C (left D, right D, down D, up D)
+    { 0x08, 0x08, 0x08, 0xFF, 0x00, 0xFF, 0x00, 0x00},   // U+2567 (left D, right D, up L)
+    { 0x14, 0x14, 0x14, 0x14, 0xFF, 0x00, 0x00, 0x00},   // U+2568 (left L, right L, up D)
+    { 0x00, 0x00, 0x00, 0xFF, 0x00, 0xFF, 0x08, 0x08},   // U+2564 (left D, right D, down L)
+    { 0x00, 0x00, 0x00, 0x00, 0xFF, 0x14, 0x14, 0x14},   // U+2565 (left L, right L, down D)
+    { 0x14, 0x14, 0x14, 0x14, 0xFC, 0x00, 0x00, 0x00},   // U+2559 (up D, right L)
+    { 0x08, 0x08, 0x08, 0xF8, 0x08, 0xF8, 0x00, 0x00},   // U+2558 (up L, right D)
+    { 0x00, 0x00, 0x00, 0xF8, 0x08, 0xF8, 0x08, 0x08},   // U+2552 (down L, right D)
+    { 0x00, 0x00, 0x00, 0x00, 0xFC, 0x14, 0x14, 0x14},   // U+2553 (down D, right L)
+    { 0x14, 0x14, 0x14, 0x14, 0xFF, 0x14, 0x14, 0x14},   // U+256B (left L, right L, down D, up D)
+    { 0x08, 0x08, 0x08, 0xFF, 0x08, 0xFF, 0x08, 0x08},   // U+256A (left D, right D, down L, up L)
+    { 0x08, 0x08, 0x08, 0x08, 0x0f, 0x00, 0x00, 0x00},   // U+2518 (up L, left L)
+    { 0x00, 0x00, 0x00, 0x00, 0xf8, 0x08, 0x08, 0x08},   // U+250C (down L, right L)
+    { 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF},   // U+2588 (solid)
+    { 0x00, 0x00, 0x00, 0x00, 0xFF, 0xFF, 0xFF, 0xFF},   // U+2584 (bottom half)
+    { 0x0F, 0x0F, 0x0F, 0x0F, 0x0F, 0x0F, 0x0F, 0x0F},   // U+258C (left half)
+    { 0xF0, 0xF0, 0xF0, 0xF0, 0xF0, 0xF0, 0xF0, 0xF0},   // U+2590 (right half)
+    { 0xFF, 0xFF, 0xFF, 0xFF, 0x00, 0x00, 0x00, 0x00},   // U+2580 (top half)
+};

part9-sound/README.md

@@ -74,4 +74,6 @@ As we enable the DMA engine, playback begins. Notably, however, we're returned t
 
 Conclusion
 ----------
-We're now ready to integrate sound into our Breakout game using background DMA transfers! Coming soon...
+We're now ready to integrate sound into our Breakout game. But rather than do that, let's go down a rabbit hole and see if we can't get CPU playback running on a separate core! I said it was hard, but I love a challenge.
+
+[Go to part10-multicore >](../part10-multicore)