diff --git a/Makefile b/Makefile
index 9026dc4..c5460d7 100644
--- a/Makefile
+++ b/Makefile
@@ -57,8 +57,8 @@ info:
 	@echo "    $(MACS)"
 	@echo "  PRNG functions:"
 	@echo "    $(PRNGS)"
-	@echo "  LC functions:"
-	@echo "    $(ALGORITHMS_LC)"
+#	@echo "  LC functions:"
+#	@echo "    $(ALGORITHMS_LC)"
 	
 #	echo $(ALGORITHMS_TEST_BIN_MAIN)
 #	echo $(ALGORITHMS)
@@ -102,9 +102,13 @@ tests: $(ALGORITHMS_TEST_BIN) \
        $(ALGORITHMS_TEST_BIN_MAIN_HEX)
 
 .PHONY:  stats
-stats: $(patsubst %, %_size.txt, $(ALGORITHMS_LC))
-	$(RUBY) sumsize.rb $^ > sizestats.txt
-
+stats: $(SIZESTAT_FILE)
+#$(patsubst %, %_size.txt, $(ALGORITHMS_LC))
+	
+	
+$(SIZESTAT_FILE): $(patsubst %, %_size.txt, $(ALGORITHMS_LC))
+	$(RUBY) sumsize.rb $^ > $(SIZESTAT_FILE)
+	@cat $(SIZESTAT_FILE)	
 #-------------------------------------------------------------------------------	
 	
 define OBJ_TEMPLATE
@@ -156,7 +160,7 @@ all: $(foreach algo, $(ALGORITHMS), $(algo)_OBJ)
 .PHONY: clean
 clean:
 	rm -rf *.o *.elf *.eps *.png *.pdf *.bak *_size.txt
-	rm -rf *.lst *.map $(EXTRA_CLEAN_FILES)
+	rm -rf *.lst *.map $(EXTRA_CLEAN_FILES) $(SIZESTAT_FILE)
 
 flash:
 	$(ERASECMD)
diff --git a/avr-makefile.inc b/avr-makefile.inc
index c3db1ce..544f60b 100644
--- a/avr-makefile.inc
+++ b/avr-makefile.inc
@@ -15,6 +15,8 @@ override CFLAGS	= -pedantic -std=c99 -Wall -Wstrict-prototypes  $(OPTIMIZE) -mmc
 override LDFLAGS       = -Wl,-Map,
 override ASFLAGS	   = -mmcu=$(MCU_TARGET)
 
+SIZESTAT_FILE = sizestats.txt
+
 OBJCOPY	= avr-objcopy
 OBJDUMP	= avr-objdump
 SIZE = avr-size
diff --git a/grain.c b/grain.c
index b4bfb2d..4cedb31 100644
--- a/grain.c
+++ b/grain.c
@@ -19,10 +19,23 @@
 #define S(i) ((ctx->lfsr[9-((i)/8)])>>(7-((i)%8)))
 /* b0, b1, b2, ..., b78, b79 */
 #define B(i) ((ctx->nfsr[9-((i)/8)])>>(7-((i)%8)))
+#define _B(i) (((ctx->nfsr[9-((i)/8)])>>(7-((i)%8)))&1)
 
 
 uint8_t h_lut[4] PROGMEM = {0x4C, 0xB6, 0xD3, 0x26};
 
+#ifdef GRAIN_BADOPTIMISATION
+uint8_t g_lut[128] PROGMEM = {
+        0xF0, 0xA5, 0x0F, 0x5A, 0x0F, 0x5A, 0xF0, 0xA5, 0x0F, 0x5A, 0xF0, 0xA5, 0xF0, 0x5A, 0x0F, 0x0F, 
+        0xC3, 0x96, 0x3C, 0x69, 0x3C, 0x69, 0xC3, 0x96, 0x9C, 0xC9, 0x63, 0x36, 0x63, 0xC9, 0x9C, 0x9C, 
+        0x0F, 0x5A, 0x0F, 0x5A, 0xF0, 0xA5, 0xF0, 0x5A, 0xF0, 0xA5, 0xF0, 0xA5, 0x0F, 0xA5, 0x0F, 0xF0, 
+        0x3C, 0x69, 0x3C, 0x69, 0xC3, 0x96, 0xC3, 0x69, 0x63, 0x36, 0x63, 0x36, 0x9C, 0x36, 0x9C, 0x63, 
+        0x0F, 0xD2, 0xF0, 0x2D, 0xF0, 0x2D, 0x0F, 0xD2, 0xF0, 0x2D, 0x0F, 0xD2, 0x0F, 0x2D, 0xF0, 0x78, 
+        0x3C, 0xE1, 0xC3, 0x1E, 0xC3, 0x1E, 0x3C, 0xE1, 0x63, 0xBE, 0x9C, 0x41, 0x9C, 0xBE, 0x63, 0xEB, 
+        0x00, 0xDD, 0x00, 0xDD, 0xFF, 0x22, 0xFF, 0xDD, 0xFF, 0x22, 0xFF, 0x22, 0x00, 0x22, 0xF0, 0x87, 
+        0xF3, 0x2E, 0xF3, 0x2E, 0x0C, 0xD1, 0x0C, 0x2E, 0xAC, 0x71, 0xAC, 0x71, 0x53, 0x71, 0xA3, 0xD4  };
+#endif
+
 uint8_t grain_enc(grain_ctx_t* ctx){
 	uint8_t s80, s0, c1, c2;
 	uint8_t i;
@@ -37,22 +50,55 @@ uint8_t grain_enc(grain_ctx_t* ctx){
 		c1 = c2;
 	}
 	/* clock the NFSR */
-	uint8_t b80, a,b,d,e;
+	uint8_t b80;
+/*	778 Byte in this variant / 617 clks enc_time */
+#ifndef GRAIN_BADOPTIMISATION
+    uint8_t a,b,d,e;
 	b80 = B(62) ^ B(60) ^ B(52) ^ B(45) ^ 
 	      B(37) ^ B(33) ^ B(28) ^ B(21) ^ 
 	      B(14) ^ B( 9) ^ B( 0) ^ s0;
 	b80 ^= (a = B(63) & B(60));
 	b80 ^= (b = B(37) & B(33));
-	b80 ^= B(15) & B( 9); /* c */
+	b80 ^= B(15) & B( 9); // c 
 	b80 ^= (d = B(60) & B(52) & B(45));
 	b80 ^= (e = B(33) & B(28) & B(21));
-	b80 ^= B(63) & B(45) & B(28) & B(9); /* f */
+	b80 ^= B(63) & B(45) & B(28) & B(9); // f 
 	/* -- */
-	b80 ^= b & B(60) & B(52); /* g */
-	b80 ^= a & B(21) & B(15); /* h */
-	b80 ^= d & B(63) & B(37); /* i */
-	b80 ^= e & B(15) & B( 9); /* j */
-	b80 ^= e & B(52) & B(45) & B(37); /* k */
+	b80 ^= b & B(60) & B(52); // g 
+	b80 ^= a & B(21) & B(15); // h 
+	b80 ^= d & B(63) & B(37); // i 
+	b80 ^= e & B(15) & B( 9); // j 
+	b80 ^= e & B(52) & B(45) & B(37); // k
+#else
+	/* let's reorder the bits */
+	uint16_t x; 
+
+/*
+	x  = _B(21); x<<=1;
+	x |= _B(33); x<<=1;
+	x |= _B(9) ; x<<=1;
+	x |= _B(45); x<<=1;
+	x |= _B(52); x<<=1;
+	x |= _B(37); x<<=1;
+	x |= _B(60); x<<=1;
+	x |= _B(28); x<<=1;
+	x |= _B(15); x<<=1;
+	x |= _B(63);
+*/
+	x  = ((ctx->nfsr[8])&0x41)<<1; // B15 & B09
+	x |= ((ctx->nfsr[2])&0x09);    // B63 & B60 
+//	x |= ((ctx->nfsr[4])&0x04)<<4; // B45
+	x |= (((ctx->nfsr[5])&0x44) | 
+	      ((ctx->nfsr[3])&0x08) | 
+	      (((((ctx->nfsr[7])&0x04)<<3) |((ctx->nfsr[4])&0x04))<<2) )<<2; // B37 & B33
+//	x |= ((ctx->nfsr[3])&0x08)<<2; // B52
+	x |= ((ctx->nfsr[6])&0x08)>>1; // B28
+//	x |= ((ctx->nfsr[7])&0x04)<<7; // B21 
+
+
+	b80 = pgm_read_byte(g_lut+(x/8))>>(x%8);
+	b80 ^= s0 ^ B(62) ^ B(14) ^ B(0);
+#endif
 	c1 = b80 & 1;
 	for(i=0; i<10; ++i){
 		c2 = (ctx->nfsr[i])>>7;
diff --git a/grain_nfsr_lutgen.c b/grain_nfsr_lutgen.c
new file mode 100644
index 0000000..9b9277d
--- /dev/null
+++ b/grain_nfsr_lutgen.c
@@ -0,0 +1,91 @@
+/**
+ * 
+ * author: Daniel Otte
+ * email:  daniel.otte@rub.de
+ * license: GPLv3
+ *
+ * this program generate a lookuptable for the nfsr-feedback-function in grain 
+ */
+
+#include <stdint.h>
+#include <stdio.h>
+
+#define X(i) ((x)>>((i)))
+#define B63 X(0)
+#define B60 X(3)
+#define B52 X(5)
+#define B45 X(6)
+#define B37 X(4)
+#define B33 X(8)
+#define B28 X(2)
+#define B21 X(9)
+#define B15 X(1)
+#define B09 X(7)
+
+uint8_t g(uint16_t x){
+	uint8_t a,b,d,e;
+	uint8_t ret;
+
+	ret = B60 ^ B52 ^ B45 ^ B37 ^ B33 ^ B28 ^ B21 ^ B09;
+	ret ^= (a = B63 & B60);
+	ret ^= (b = B37 & B33);
+	ret ^= B15 & B09;
+	ret ^= (d = B60 & B52 & B45);
+	ret ^= (e = B33 & B28 & B21);
+	ret ^= B63 & B45 & B28 & B09;
+	ret ^= b & B60 & B52;
+	ret ^= a & B21 & B15;
+	ret ^= d & B63 & B37;
+	ret ^= e & B15 & B09;
+	ret ^= e & B52 & B45 & B37;
+	
+	return ret&1;
+}
+
+int main(void){
+	uint16_t i; 
+	uint8_t t, lut[128]={0}; /* 2**10 / 8 == 2**(10-3) == 2**7 == 128 */
+	puts(
+	"/* \n"
+	" * author: Daniel Otte \n"
+	" * email:  daniel.otte@rub.de  \n"
+	" * license: GPLv3 \n"
+	" *  \n"
+	" * this program generate a lookuptable for the h-function in grain  \n"
+	" *  \n"
+	" */ \n");
+	puts("/* \n"
+	     " * b63 b15 b28 b60 b37 b52 b45 b09 b33 b21 - g");
+	
+	for(i=0; i<0x0400; ++i){
+		t = g(i);
+		printf(" *  %c   %c   %c   %c   %c   %c   %c   %c   %c   %c  - %c\n",
+			(i&0x01)?'1':'0',
+			(i&0x02)?'1':'0',
+			(i&0x04)?'1':'0',
+			(i&0x08)?'1':'0',
+			(i&0x10)?'1':'0',
+			(i&0x20)?'1':'0',
+			(i&0x40)?'1':'0',
+			(i&0x80)?'1':'0',
+			(i&0x0100)?'1':'0',
+			(i&0x0200)?'1':'0',
+			t?'1':'0' );
+		lut[i/8] |= t<<(i%8);
+//		if(i%4==3){	
+//			puts(" * --");
+//		}
+	}
+	puts(" */\n");
+	
+	printf(" uint8_t g_lut[128]= {");
+	for(i=0; i<128; ++i){
+		if(i%16==0){
+			printf("\n\t");
+		}
+		printf("0x%2.2X%c ", lut[i], (i!=127)?',':' ');
+	}
+	printf("};\n\n");
+	return 0; 
+}
+
diff --git a/prng.mk b/prng.mk
index c31f69d..9114ce8 100644
--- a/prng.mk
+++ b/prng.mk
@@ -4,7 +4,7 @@ ALGO_NAME := PRNG
 # comment out the following line for removement of PRNG from the build process
 PRNGS += $(ALGO_NAME)
 
-$(ALGO_NAME)_OBJ      := prng.o
+$(ALGO_NAME)_OBJ      := prng.o sha256-asm.o
 $(ALGO_NAME)_TEST_BIN := main.o debug.o uart.o serial-tools.o sha256-asm.o \
                          xtea-asm.o arcfour-asm.o prng.o cast5.o