diff --git a/config.s b/config.s
index 511d33f..051360c 100644
--- a/config.s
+++ b/config.s
@@ -1,5 +1,5 @@
-NUMSEG equ 1 ; TODO build system should calculate this value
+NUMSEG equ 4; TODO build system should calculate this value
 
 ; might be good to move one of these into the high memory area
-K_ADDR    equ 0x500  ; kernel load address
-STACK_SEG equ 0x1050
+KERNEL_SEG equ 0x50
+STACK_SEG  equ 0x1050
diff --git a/src/boot.s b/src/boot.s
index 974e70e..ace3604 100644
--- a/src/boot.s
+++ b/src/boot.s
@@ -5,7 +5,7 @@
 [org 0x7C00]
 [bits 16]
 
-mov ax, K_ADDR >> 4
+mov ax, KERNEL_SEG
 push ax ; setup for ds
 mov es, ax
 mov gs, ax
@@ -13,7 +13,7 @@ mov fs, ax
 
 xor cx, cx
 mov ds, cx
-mov word [0x21*4], 2 ; see kernel_entry
+mov word [0x21*4], 2 ; see kjmp in kernel.s
 mov word [0x21*4+2], ax
 
 mov ax, 0x200 | NUMSEG
@@ -34,7 +34,7 @@ push ds
 push ax
 retf
 
-; TODO we have lots of space for init code here
+; TODO filesystem and reserve space for mbr partition table
 
 times 510 - ($-$$) db 0
 dw 0xAA55
diff --git a/src/int21.s b/src/int21.s
index 4538e74..c26d717 100644
--- a/src/int21.s
+++ b/src/int21.s
@@ -45,7 +45,13 @@ tmp: dw 0
 retcode: dw 0
 verify: db 0
 
-rdin_echo:; ah=01h
+; AH = 01h
+; read a character from stdin and print it to stdout
+; inputs:
+; 	none
+; outputs:
+; 	AL: character read from stdin
+rdin_echo:
 	push ax
 	mov ah, 8
 	int 0x21
@@ -55,10 +61,10 @@ rdin_echo:; ah=01h
 ; AH = 4dh
 ; get return code
 ; inputs:
-; none
+; 	none
 ; outputs:
-; AH: termination type (0 = normal, 1 = control-C abort, 2 = critical error abort, 3 = terminate and stay resident)
-; AL: return code
+; 	AH: termination type (0 = normal, 1 = control-C abort, 2 = critical error abort, 3 = terminate and stay resident)
+; 	AL: return code
 getret:
 	xor ax, ax
 	xchg [retcode], ax
@@ -67,10 +73,10 @@ getret:
 ; AH = 30h
 ; get the DOS version number
 ; inputs:
-; none
+; 	none
 ; outputs:
-; AL: major version
-; AH: minor version
+; 	AL: major version
+; 	AH: minor version
 getver:
 	mov ax, 8 ; if it is not zero indexed this indicates windows ME
 	xor bx, bx ; update: what does that comment mean
@@ -81,9 +87,9 @@ getver:
 ; AH = 54h
 ; get disk verify flag
 ; inputs:
-; none
+; 	none
 ; outputs:
-; AL: 0 if off, 1 if on
+; 	AL: 0 if off, 1 if on
 getverify:
 	mov al, [verify]
 	ret
@@ -91,9 +97,9 @@ getverify:
 ; AH = 2eh
 ; set disk verify flag
 ; inputs:
-; AL: 0 if off, 1 if on
+; 	AL: 0 if off, 1 if on
 ; outputs:
-; none
+; 	none
 setverify:
 	mov [verify], al
 	ret
@@ -101,9 +107,9 @@ setverify:
 ; AH = 35h
 ; get interrupt vector
 ; inputs:
-; AL: interrupt number
+; 	AL: interrupt number
 ; outputs:
-; ES:BX: current interrupt handler
+; 	ES:BX: current interrupt handler
 getint:
 	push ds
 	push di
@@ -123,29 +129,29 @@ getint:
 ; AH = 25h
 ; set interrupt vector
 ; inputs:
-; AL: interrupt number
-; DS:DX: new interrupt handler
+; 	AL: interrupt number
+; 	DS:DX: new interrupt handler
 ; outputs:
-; none
+; 	none
 setint:
 	pusha
 	xor ah, ah
 	shl ax, 2
 	mov di, ax
 	xor ax, ax
-	push ds
-	mov ds, ax
-	mov word [di], dx
-	mov word [di+2], ds
-	pop ds
+	push es
+	mov es, ax
+	mov word es:[di], dx
+	mov word es:[di+2], ds
+	pop es
 	popa
 	ret
 
 ; read from CMOS register
 ; inputs:
-; AL: register
+; 	AL: register
 ; outputs:
-; AL: value
+; 	AL: value
 ; TODO: handle bcd here to abstract it from the kernel
 rdcmos:
 	cli
@@ -165,10 +171,10 @@ rdcmos:
 
 ; write to CMOS register
 ; inputs:
-; BL: value
-; BH: register
+; 	BL: value
+; 	BH: register
 ; outputs:
-; none
+; 	none
 wrcmos:
 	cli
 	push ax
@@ -184,11 +190,11 @@ wrcmos:
 ; AH = 2ch
 ; read system time from the CMOS
 ; inputs:
-; none
+; 	none
 ; outputs:
-; CH: hours
-; CL: minutes
-; DH: seconds
+; 	CH: hours
+; 	CL: minutes
+; 	DH: seconds
 gettime:
 	push ax
 	xor dl, dl
@@ -215,11 +221,11 @@ gettime:
 ; AH = 2dh
 ; set system time in the CMOS
 ; inputs:
-; CH: hours
-; CL: minutes
-; DH: seconds
+; 	CH: hours
+; 	CL: minutes
+; 	DH: seconds
 ; outputs:
-; none
+; 	none
 settime:
 	push bx
 	mov bl, ch
diff --git a/src/kernel.s b/src/kernel.s
index ebf0418..f70633c 100644
--- a/src/kernel.s
+++ b/src/kernel.s
@@ -1,6 +1,10 @@
 %include "config.s"
 
-kernel_entry: ; MUST be a short jump due to loader config
+kjmp: ; MUST be a short jump due to loader config
+	jmp kernel_entry
+%include "int21.s"
+
+kernel_entry:
 	mov si, hello_string
 	mov bl, 0x70
 	call print_string
@@ -8,5 +12,4 @@ kernel_entry: ; MUST be a short jump due to loader config
 
 hello_string: db "hello world!", 13, 10, '$'
 
-%include "int21.s"
 %include "vga.s"
diff --git a/src/vga.s b/src/vga.s
index fbf1559..32aa442 100644
--- a/src/vga.s
+++ b/src/vga.s
@@ -1,15 +1,11 @@
 ; print a string to the screen
 ; inputs:
-; DS:SI: pointer to '$'-terminated string
-; BL: attribute
+; 	DS:SI: pointer to '$'-terminated string
+; 	BL: attribute
 ; outputs:
-; none
+; 	none
 print_string:
-	push ax
-	push bx
-	push cx
-	push dx
-	push bp
+	pusha
 	push es
 
 	; get cursor position in DX
@@ -20,27 +16,22 @@ print_string:
 	; calculate string length by iterating over it until we reach '$'
 	; this sucks but we have to do it because the bios expects to be passed the string length directly
 	xor cx, cx
-	push si
+	mov bp, si
+	jmp .start_loop
 .size_loop:
 	inc cx
 	inc si
+.start_loop:
 	cmp byte [si], '$'
 	jnz .size_loop
-	pop si
-	mov bp, si
 
 	; print string and update cursor position
 	mov ax, ds
 	mov es, ax
-	mov ah, 0x13
-	mov al, 0x01
+	mov ax, 0x1301
 	mov bh, 0x00
 	int 0x10
 
 	pop es
-	pop bp
-	pop dx
-	pop cx
-	pop bx
-	pop ax
+	popa
 	ret