some minor improvments and bug fixes

Makefile_m644_conf.inc

@@ -14,7 +14,7 @@ override ASFLAGS	   = -mmcu=$(MCU_TARGET) -Wa,--gdwarf-2
 
 #PROGRAMMER     = avr911
 #PROG_PORT      = /dev/ttyUSB1
-PROGRAMMER     = usbasp
+PROGRAMMER     = avrispmkII
 PROG_PORT      = usb
 DEFS           = -D$(call uc, $(MCU_TARGET)) -DF_CPU=$(F_CPU)
 FLASHCMD       = avrdude -p $(MCU_TARGET) -P $(PROG_PORT) -c $(PROGRAMMER) -U flash:w:# no space at the end

bcal/bcal-performance.c

@@ -32,15 +32,17 @@
 #include "performance_test.h"
 #include "stack_measuring.h"
 #include "cli.h"
+#include "uart_i.h"
 #include <stdint.h>
 #include <stdlib.h>
 #include <string.h>
+#include <stdio.h>
 #include <avr/pgmspace.h>
 
 #define PATTERN_A 0xAA
 #define PATTERN_B 0x55
 
-
+/*
 static
 void printvalue(unsigned long v){
 	char str[20];
@@ -51,6 +53,7 @@ void printvalue(unsigned long v){
 	}
 	cli_putstr(str);
 }
+*/
 
 void bcal_performance(const bcdesc_t* bcd){
 	bcdesc_t bc;
@@ -62,30 +65,22 @@ void bcal_performance(const bcdesc_t* bcd){
 	uint64_t t;
 	uint8_t i;
 
-	if(bc.type!=BCDESC_TYPE_BLOCKCIPHER)
+	if(bc.type != BCDESC_TYPE_BLOCKCIPHER)
 		return;
 	calibrateTimer();
 	print_overhead();
-	cli_putstr_P(PSTR("\r\n\r\n === "));
+	printf_P(PSTR("\n\n === %S"), bc.name);
-	cli_putstr_P(bc.name);
+	printf_P(PSTR(" performance === \n"
-	cli_putstr_P(PSTR(" performance === "
+	              "    type:             blockcipher\n"
-	                  "\r\n    type:             blockcipher"
+	              "    keysize (bits):     %5"PRIu16"\n"), keysize);
-	                  "\r\n    keysize (bits):     "));
+	printf_P(PSTR("    ctxsize (bytes):    %5"PRIu16"\n"), bc.ctxsize_B);
-	printvalue(keysize);
+	printf_P(PSTR("    blocksize (bits):   %5"PRIu16"\n"), bc.blocksize_b);
-
+	uart0_flush();
-	cli_putstr_P(PSTR("\r\n    ctxsize (bytes):    "));
-	printvalue(bc.ctxsize_B);
-
-	cli_putstr_P(PSTR("\r\n    blocksize (bits):   "));
-	printvalue(bc.blocksize_b);
-
-
-
 	t=0;
 	if(bc.init.init1){
-		if((bc.flags&BC_INIT_TYPE)==BC_INIT_TYPE_1){
+		if((bc.flags & BC_INIT_TYPE) == BC_INIT_TYPE_1){
 			for(i=0; i<32; ++i){
-				startTimer(0);
+				startTimer(1);
 				START_TIMER;
 				(bc.init.init1)(key, &ctx);
 				STOP_TIMER;
@@ -96,7 +91,7 @@ void bcal_performance(const bcdesc_t* bcd){
 			}
 		} else {
 			for(i=0; i<32; ++i){
-				startTimer(0);
+				startTimer(1);
 				START_TIMER;
 				(bc.init.init2)(key, keysize, &ctx);
 				STOP_TIMER;
@@ -107,9 +102,10 @@ void bcal_performance(const bcdesc_t* bcd){
 			}
 		}
 		t>>=5;
-		cli_putstr_P(PSTR("\r\n    init (cycles):      "));
+		printf_P(PSTR("    init (cycles):      %5"PRIu16"\n"), t);
-		printvalue(t);
 	}
+
+    uart0_flush();
 	t=0;
 	for(i=0; i<32; ++i){
 		startTimer(0);
@@ -119,9 +115,9 @@ void bcal_performance(const bcdesc_t* bcd){
 		t += stopTimer();
 	}
 	t>>=5;
-	cli_putstr_P(PSTR("\r\n    encrypt (cycles):   "));
+	printf_P(PSTR("    encrypt (cycles):   %5"PRIu16"\n"), t);
-	printvalue(t);
 
+    uart0_flush();
 	t=0;
 	for(i=0; i<32; ++i){
 		startTimer(0);
@@ -131,10 +127,11 @@ void bcal_performance(const bcdesc_t* bcd){
 		t += stopTimer();
 	}
 	t>>=5;
-	cli_putstr_P(PSTR("\r\n    decrypt (cycles):   "));
+	printf_P(PSTR("    decrypt (cycles):   %5"PRIu16"\n"), t);
-	printvalue(t);
+    uart0_flush();
 
 	if(bc.free){
+	    uart0_flush();
 		bc.free(&ctx);
 	}
 }
@@ -147,17 +144,17 @@ void bcal_stacksize(const bcdesc_t* bcd){
 	uint8_t data[(bc.blocksize_b+7)/8];
 	uint16_t keysize = get_keysize(bc.valid_keysize_desc);
 	uint8_t key[(keysize+7)/8];
-	uint16_t t1, t2;
+	uint16_t t1 = 0, t2 = 0;
 
-	if(bc.type!=BCDESC_TYPE_BLOCKCIPHER)
+	if(bc.type != BCDESC_TYPE_BLOCKCIPHER)
 		return;
-	cli_putstr_P(PSTR("\r\n\r\n === "));
+	printf_P(PSTR("\n === %S stack-usage ===\n"),bc.name);
-	cli_putstr_P(bc.name);
+
-	cli_putstr_P(PSTR(" stack-usage === "));
+	uart0_flush();
 
 	if(bc.init.init1){
-		if((bc.flags&BC_INIT_TYPE)==BC_INIT_TYPE_1){
+		if((bc.flags & BC_INIT_TYPE) == BC_INIT_TYPE_1){
-			cli();
+	    	cli();
 			stack_measure_init(&smctx, PATTERN_A);
 			bc.init.init1(&ctx, key);
 			t1 = stack_measure_final(&smctx);
@@ -166,18 +163,18 @@ void bcal_stacksize(const bcdesc_t* bcd){
 			t2 = stack_measure_final(&smctx);
 			sei();
 		} else {
-			cli();
+	    	cli();
 			stack_measure_init(&smctx, PATTERN_A);
 			bc.init.init2(&ctx, keysize, key);
 			t1 = stack_measure_final(&smctx);
 			stack_measure_init(&smctx, PATTERN_B);
 			bc.init.init2(&ctx, keysize, key);
-			t2 = stack_measure_final(&smctx);
+            t2 = stack_measure_final(&smctx);
-			sei();
+            sei();
-		}
+        }
+
 		t1 = (t1>t2)?t1:t2;
-		cli_putstr_P(PSTR("\r\n    init (bytes):       "));
+		printf_P(PSTR("    init (bytes):       %5"PRIu16"\n"), t1);
-		printvalue((unsigned long)t1);
 	}
 	cli();
 	stack_measure_init(&smctx, PATTERN_A);
@@ -189,8 +186,7 @@ void bcal_stacksize(const bcdesc_t* bcd){
 	sei();
 
 	t1 = (t1>t2)?t1:t2;
-	cli_putstr_P(PSTR("\r\n    encBlock (bytes):   "));
+	printf_P(PSTR("    encBlock (bytes):   %5"PRIu16"\n"), t1);
-	printvalue((unsigned long)t1);
 
 	cli();
 	stack_measure_init(&smctx, PATTERN_A);
@@ -202,8 +198,7 @@ void bcal_stacksize(const bcdesc_t* bcd){
 	sei();
 
 	t1 = (t1>t2)?t1:t2;
-	cli_putstr_P(PSTR("\r\n    decBlock (bytes):   "));
+	printf_P(PSTR("    decBlock (bytes):   %5"PRIu16"\n"), t1);
-	printvalue((unsigned long)t1);
 
 	if(bc.free){
 		bc.free(&ctx);
@@ -215,7 +210,7 @@ void bcal_performance_multiple(const bcdesc_t* const* bcd_list){
 	for(;;){
 		bcd = (void*)pgm_read_word(bcd_list);
 		if(!bcd){
-			cli_putstr_P(PSTR("\r\n\r\n End of performance figures\r\n"));
+			puts_P(PSTR("\n End of performance figures\n"));
 			return;
 		}
 		bcal_performance(bcd);

bigint/bigint.c

@@ -315,7 +315,7 @@ int8_t bigint_cmp_u(const bigint_t* a, const bigint_t* b){
 	if(a->length_W < b->length_W){
 		return -1;
 	}
-	if(a->length_W==0){
+	if(a->length_W == 0){
 		return 0;
 	}
 	uint16_t i;

mkfiles/001_cli_std.mk

@@ -1,3 +1,8 @@
 CLI_STD =  cli-stub.o cli-basics.o cli-core.o cli-hexdump.o debug.o hexdigit_tab.o \
            dbz_strings.o string-extras-asm.o uart_i-asm.o circularbytebuffer-asm.o \
            main-test-common.o
+
+#CLI_STD = cli.o debug.o hexdigit_tab.o \
+#          dbz_strings.o string-extras-asm.o uart_i-asm.o circularbytebuffer-asm.o \
+#          main-test-common.o
+

stack_measuring.S

@@ -25,12 +25,23 @@
 stack_measure_init:
 	movw r30, r24
 	lds r20, __brkval
-	lds r21, __brkval+1
+	lds r21, __brkval + 1
 	in r0, _SFR_IO_ADDR(SREG)
 	cli
 	in r26, _SFR_IO_ADDR(SPL)
 	out _SFR_IO_ADDR(SREG), r0
 	in r27, _SFR_IO_ADDR(SPH)
+	cp r20, r26
+	cpc r21, r27
+	brlo 10f
+	ldi r20, lo8(__bss_end)
+	ldi r21, hi8(__bss_end)
+	cp r20, r26
+	cpc r21, r27
+	brlo 10f
+	ldi r21, 1
+	clr r20
+10:
 	st Z+, r20
 	st Z+, r21
 	st Z+, r26
@@ -41,23 +52,31 @@ stack_measure_init:
 	sbc r25, r21
 	sbiw r24, 1
 	st X, r22
-1:  st -X, r22
+50: st -X, r22
 	sbiw r24, 1
-	brne 1b
+	brne 50b
 	ret
 
 .global stack_measure_final
 stack_measure_final:
 	movw r30, r24
 	lds r20, __brkval
-	lds r21, __brkval+1
+	lds r21, __brkval + 1
 	ld r26, Z+
 	ld r27, Z+
+
 	cp r20, r26
 	cpc r21, r27
-	brlo 1f
+	brlo 10f
-	movw r26, r20
+	ldi r20, lo8(__bss_end)
-1:
+	ldi r21, hi8(__bss_end)
+	cp r20, r26
+	cpc r21, r27
+	brlo 10f
+	ldi r21, 1
+	clr r20
+10:
+
 	adiw r26, 1
 	ld r24, Z+
 	ld r25, Z+

test_src/cli-core.S

@@ -392,6 +392,7 @@ cli_auto_help:
 99:
 	pop_range 28, 29
 	pop_range 14, 17
+	ori r24, 1
 	ret
 ahelphead_str:
 .asciz "\r\n[auto help] available commands:\r\n <command> - <params> - <address>\r\n"

test_src/cli-stub.c

@@ -98,7 +98,7 @@ int8_t cmd_interface(PGM_VOID_P cmd_desc){
 		c = cli_rx();
 		switch (c){
 		case CLI_ENTER:
-			if((exit_code=search_and_call(cli_buffer, maxcmdlength, cmd_desc))<=0){
+			if((exit_code = search_and_call(cli_buffer, maxcmdlength, cmd_desc)) <=0 ){
 				free(cli_buffer);
 				return exit_code;
 			}

test_src/cli.h

@@ -57,15 +57,15 @@ extern uint8_t cli_echo;
 void cli_putc(char c);
 uint16_t cli_getc(void);
 uint16_t cli_getc_cecho(void);
-uint8_t cli_getsn(char* s, uint16_t n);
+uint8_t cli_getsn(char* s, size_t n);
 uint8_t cli_getsn_cecho(char* s, uint16_t n);
 void cli_putstr(const char* s);
 void cli_putstr_P(PGM_P s);
 void cli_hexdump_byte(uint8_t byte);
-void cli_hexdump(const void* data, uint16_t length);
+void cli_hexdump(const void* data, size_t length);
-void cli_hexdump_rev(const void* data, uint16_t length);
+void cli_hexdump_rev(const void* data, size_t length);
-void cli_hexdump2(const void* data, uint16_t length);
+void cli_hexdump2(const void* data, size_t length);
-void cli_hexdump_block(const void* data, uint16_t length, uint8_t indent, uint8_t width);
+void cli_hexdump_block(const void* data, size_t length, uint8_t indent, uint8_t width);
 
 void echo_ctrl(char* s);
 int8_t cmd_interface(PGM_VOID_P cmd_desc);

test_src/main-test-common.c

@@ -52,5 +52,5 @@ void welcome_msg(const char* algoname){
 	cli_putstr(__TIME__);
 	cli_putstr_P(PSTR(")\r\nloaded and running\r\n"));
 */
-    printf_P(PSTR("\n\nAVR-Crypto-Lib VS(%s; %s %s)\nloaded and running\n"), algoname, __DATE__, __TIME__);
+    printf_P(PSTR("\n\nAVR-Crypto-Lib VS (%s; %s %s)\nloaded and running\n"), algoname, __DATE__, __TIME__);
 }

test_src/uart_i-asm.S

@@ -192,6 +192,11 @@ uart0_init:
 	std Z+UART0_TXON_OFFSET, r24
 	std Z+UART0_RXON_OFFSET, r24
 #endif	
+#if UART0_HOOK
+	std Z+UART0_HOOK_OFFSET,   r1
+	std Z+UART0_HOOK_OFFSET+1, r1
+	std Z+UART0_HOOKR_OFFSET,  r1
+#endif
 	ldi r24, UBRRH_VALUE
 	STORE_IO UBRR0H, r24
 	ldi r24, UBRRL_VALUE
@@ -496,5 +501,16 @@ uart0_sethook:
 	st X+, r25
 	ret
 #endif
+
+.global uart0_flush
+uart0_flush:
+10:
+ 	ldi r24, lo8(uart0_ctx+UART0_CBB_TX_OFFSET)
+ 	ldi r25, hi8(uart0_ctx+UART0_CBB_TX_OFFSET)
+	rcall circularbytebuffer_cnt
+	tst r24
+	brne 10b
+	ret
+
 	
 #endif /* UART0_I */

test_src/uart_i.h

@@ -115,7 +115,7 @@ typedef struct{
 #endif
 #if UART1_SWFLOWCTRL
 	volatile uint8_t txon;  /**< flag indicating if we are allowed to send data */
-	volatile uint8_t rxon;  /**< flag indicating if we have send an \a XOFF */
+	volatile uint8_t rxon;  /**< flag indicating if we have send a XOFF */
 #endif
 } uart1_ctx_t;
 
@@ -156,6 +156,15 @@ uint16_t uart0_getc(void);
  */
 uint8_t uart0_dataavail(void);
 
+/**
+ * \brief flushes the internal transmit buffer.
+ *
+ * This function function waits until all data was send (the transmit-buffer
+ * does not contain any data).
+ */
+void uart0_flush(void);
+
+
 #if	UART0_HOOK
 /**
  * \brief sets the hook for uart0.