diff --git a/sha1.asm b/sha1.asm
index 88769ba..8f2dcf6 100644
--- a/sha1.asm
+++ b/sha1.asm
@@ -2,14 +2,14 @@
 	mcopy	sha1.macros
 
 * Direct page locations	
-;chunk	gequ	0
-a_	gequ	8	; elements of state
-b	gequ	12
-c	gequ	16
-d	gequ	20
-e	gequ	24
-idx	gequ	28
-unused	gequ	30
+length	gequ	0
+extra	gequ	8
+idx	gequ	10
+a_	gequ	12	; elements of state
+b	gequ	16
+c	gequ	20
+d	gequ	24
+e	gequ	28
 f_plus_k gequ	32
 temp	gequ	36
 h0	gequ	40
@@ -49,6 +49,12 @@ SHA1_INIT start
 	sta	h4
 	lda	#$C3D2
 	sta	h4+2
+	
+	stz	length
+	stz	length+2
+	stz	length+4
+	stz	length+6
+	stz	extra
 	rtl
 	end
 
diff --git a/sha1.cc b/sha1.cc
new file mode 100644
index 0000000..7bdb77a
--- /dev/null
+++ b/sha1.cc
@@ -0,0 +1,139 @@
+#pragma noroot
+#pragma lint -1
+#pragma optimize -1
+
+#include "sha1.h"
+#include <string.h>
+
+#define length_offset 0
+#define extra_offset 8
+#define hash_offset 40
+#define data_offset 60
+
+extern void SHA1_PROCESSCHUNK(void);
+
+void sha1_update(struct sha1_context *context,
+                 const unsigned char *data,
+                 unsigned long length)
+{
+    unsigned int extra = context->extra;
+
+    context->length += length;
+
+    if (extra > 0) {
+        if (length >= 64 - extra) {
+            memcpy(&context->chunk[extra], data, 64 - extra);
+            sha1_processchunk(context);
+            length -= 64 - extra;
+            data += 64 - extra;
+        } else {
+            memcpy(&context->chunk[extra], data, length);
+            context->extra += length;
+            return;
+        }
+    }
+    
+    while (length >= 64) {
+        memcpy(&context->chunk, data, 64);
+        sha1_processchunk(context);
+        length -= 64;
+        data += 64;
+    }
+
+    memcpy(&context->chunk, data, length);
+    context->extra = length;
+}
+
+
+void sha1_finalize(struct sha1_context *context)
+{
+    unsigned int extra = context->extra;
+
+    context->chunk[extra++] = 0x80;
+    
+    memset(&context->chunk[extra], 0, 64 - extra);
+    
+    if (extra > 64 - 8) {
+        sha1_processchunk(context);
+        memset(&context->chunk, 0, 64);
+    }
+    
+    /* Append total length in bits */
+    asm {
+        phd
+        lda context
+        tcd
+        
+        asl length_offset
+        rol length_offset+2
+        rol length_offset+4
+        rol length_offset+6
+        asl length_offset
+        rol length_offset+2
+        rol length_offset+4
+        rol length_offset+6
+        asl length_offset
+        rol length_offset+2
+        rol length_offset+4
+        rol length_offset+6
+        
+        lda length_offset+6
+        xba
+        sta data_offset+56
+        lda length_offset+4
+        xba
+        sta data_offset+58
+        lda length_offset+2
+        xba
+        sta data_offset+60
+        lda length_offset
+        xba
+        sta data_offset+62
+        
+        jsl SHA1_PROCESSCHUNK
+
+        lda hash_offset
+        xba
+        tay
+        lda hash_offset+2
+        xba
+        sta hash_offset
+        sty hash_offset+2
+        
+        lda hash_offset+4
+        xba
+        tay
+        lda hash_offset+4+2
+        xba
+        sta hash_offset+4
+        sty hash_offset+4+2
+
+        lda hash_offset+8
+        xba
+        tay
+        lda hash_offset+8+2
+        xba
+        sta hash_offset+8
+        sty hash_offset+8+2
+
+        lda hash_offset+12
+        xba
+        tay
+        lda hash_offset+12+2
+        xba
+        sta hash_offset+12
+        sty hash_offset+12+2
+
+        lda hash_offset+16
+        xba
+        tay
+        lda hash_offset+16+2
+        xba
+        sta hash_offset+16
+        sty hash_offset+16+2
+
+        pld
+    }
+}
+
+#append "sha1.asm"
diff --git a/sha1.h b/sha1.h
index dcf0481..39ebb12 100644
--- a/sha1.h
+++ b/sha1.h
@@ -1,11 +1,16 @@
 struct sha1_context {
-	unsigned char reserved1[8];
-	unsigned long a,b,c,d,e;
-	unsigned char reserved2[12];
-	unsigned long h0,h1,h2,h3,h4;
+	unsigned long length;
+	unsigned long length2;
+	unsigned short extra;
+	unsigned char reserved1[30];
+	unsigned char hash[20];
 	unsigned char chunk[64];
-	unsigned char reserved3[300];
+	unsigned char reserved2[16];
 };
 
-int sha1_init(struct sha1_context *context);
-int sha1_processchunk(struct sha1_context *context);
+void sha1_init(struct sha1_context *context);
+void sha1_processchunk(struct sha1_context *context);
+
+void sha1_update(struct sha1_context *context, const unsigned char *data, unsigned long length);
+
+void sha1_finalize(struct sha1_context *context);
diff --git a/sha1test.c b/sha1test.c
index 20dc787..0298ae6 100644
--- a/sha1test.c
+++ b/sha1test.c
@@ -3,14 +3,15 @@
 #include <MiscTool.h>
 #include <Memory.h>
 #include <orca.h>
+#include <string.h>
 
-int main(void) {
+int main(int argc, char **argv) {
     unsigned int i;
     unsigned long tick_count;
     long double bytes_per_sec;
     
     struct sha1_context *context, **context_hndl;
-    struct sha1_context context_init = {{0}, 0,0,0,0,0, {0}, 0,0,0,0,0,
+    struct sha1_context context_init = {0,0,0, {0}, {0},
         {0x61,0x62,0x63,0x80,
          0x00,0x00,0x00,0x00,
          0x00,0x00,0x00,0x00,
@@ -40,8 +41,12 @@ int main(void) {
     sha1_init(context);
     sha1_processchunk(context);
     
-    printf("abcde = %08lx %08lx %08lx %08lx %08lx\n", context->a, context->b, context->c, context->d, context->e);
-    printf("h[..] = %08lx %08lx %08lx %08lx %08lx\n", context->h0, context->h1, context->h2, context->h3, context->h4);
+    printf("h[..] = %02x%02x%02x%02x %02x%02x%02x%02x %02x%02x%02x%02x %02x%02x%02x%02x %02x%02x%02x%02x\n", 
+           context->hash[3], context->hash[2], context->hash[1], context->hash[0], 
+	   context->hash[7], context->hash[6], context->hash[5], context->hash[4], 
+	   context->hash[11], context->hash[10], context->hash[9], context->hash[8], 
+	   context->hash[15], context->hash[14], context->hash[13], context->hash[12], 
+	   context->hash[19], context->hash[18], context->hash[17], context->hash[16]);
 
     tick_count = GetTick();
     for (i = 0; i < 1000; i++) {
@@ -51,4 +56,25 @@ int main(void) {
     
     bytes_per_sec = (long double)1000 * 64 * 60 / tick_count;
     printf("Time for 1000 iters = %lu ticks (%lf bytes/sec)\n", tick_count, bytes_per_sec);
+
+    tick_count = GetTick();
+    sha1_init(context);
+    sha1_update(context, (void*)0x030000, 64000);
+    sha1_finalize(context);
+    tick_count = GetTick() - tick_count;
+    bytes_per_sec = (long double)1000 * 64 * 60 / tick_count;
+    printf("Append time = %lu ticks (%lf bytes/sec)\n", tick_count, bytes_per_sec);
+
+    if (argc > 1) {
+        sha1_init(context);
+        sha1_update(context, argv[1], strlen(argv[1]));
+        sha1_finalize(context);
+
+        printf("h[..] = %02x%02x%02x%02x %02x%02x%02x%02x %02x%02x%02x%02x %02x%02x%02x%02x %02x%02x%02x%02x\n", 
+               context->hash[0], context->hash[1], context->hash[2], context->hash[3], 
+    	       context->hash[4], context->hash[5], context->hash[6], context->hash[7], 
+    	       context->hash[8], context->hash[9], context->hash[10], context->hash[11], 
+    	       context->hash[12], context->hash[13], context->hash[14], context->hash[15], 
+    	       context->hash[16], context->hash[17], context->hash[18], context->hash[19]);
+    }
 }