[avr-gcc-list] Now to do Tail Recursions?

[Bob Paddock]

Some gcc targets have an option of "-mtail-call" to optimize
calls followed immediately by a return into a jump to the subroutine.

Example:

rcall hex8
ret

becomes

rjmp hex8

I tried -mtail-call with avr-gcc and I get "invalid option" from cc1.

Did I not use it correctly, or more likely it is is not implemented?

Anyone have a AWK or perl script that does it the hard way at hand already?

In a related size question why are there two 'ret' instructions, one right  
after
the other?  'ret ret' wastes two bytes.

  22:hexprint.c    **** uint8_t convert_to_hex( uint8_t a ) /* Return hex  
character of passed nibble */
  23:hexprint.c    **** {
  78                    .LM1:
  79                    /* prologue: frame size=0 */
  80                    /* prologue end (size=0) */
  24:hexprint.c    ****     if( a < 10 )
  82                    .LM2:
  83 0000 8A30                  cpi r24,lo8(10)
  84 0002 18F4                  brsh .L2
  25:hexprint.c    **** 	return( (uint8_t) (a + (uint8_t) '0') );		/* 0 ->  
9 */
  86                    .LM3:
  87 0004 805D                  subi r24,lo8(-(48))
  88 0006 9927                  clr r25
  89 0008 0895                  ret
  90                    .L2:
  26:hexprint.c    ****     else
  27:hexprint.c    **** 	return( (uint8_t) ((a + 7) + (uint8_t) '0') );	/*  
A -> F */
  92                    .LM4:
  93 000a 895C                  subi r24,lo8(-(55))
  94 000c 9927                  clr r25
  28:hexprint.c    **** }
  96                    .LM5:
  97 000e 0895                  ret
  98                    /* epilogue: frame size=0 */
  99 0010 0895                  ret
 100                    /* epilogue end (size=1) */
 101                    /* function convert_to_hex size 9 (8) */