added Min64x4 prime numbers example

CHANGELOG.md

@@ -13,6 +13,8 @@ Changes that are planned but not implemented yet:
   * Disallowed operands
   * missing `:` after labels
   * unknown labels
+* Add named label scopes. This would allow a label to be defined in a specific scope that can be shared across files.
+* Create a "align if needed" preprocessor directive paid that generates an `.align` directive is the bytecode in between the paid isn't naturally on the same page and can fit on the same page if aligned. An error would be benerated if the block of code can't fit on the same page regardless of alignment.
 
 ## [Unreleased]
 *  Added support for The Minimal 64x4 Home Computer with an example and updated assembler functionality to support it.

examples/slu4-minimal-64x4/slu4-minimal-64x4.yaml

@@ -2700,6 +2700,23 @@ macros:
         - "phs"
         - "ldi BYTE3(@ARG(0))"
         - "phs"
+  phs4a:
+    # push 4 bytes of an absolute address to the stack
+    # stack is arranged as big-endian, address is little-endian
+    - operands:
+        count: 1
+        operand_sets:
+          list:
+            - absolute_address
+      instructions:
+        - "ldb @ARG(0)+0"
+        - "phs"
+        - "ldb @ARG(0)+1"
+        - "phs"
+        - "ldb @ARG(0)+2"
+        - "phs"
+        - "ldb @ARG(0)+3"
+        - "phs"
   phsptr:
     # push an immediate absolute address to the stack per MinOS calling convention
     - operands:
@@ -2724,6 +2741,39 @@ macros:
         - "phs"
         - "lds @ARG(0)+0+1"
         - "phs"
+  phs4s:
+    # push 4 bytes at (current) stack offset to the stack
+    - operands:
+        count: 1
+        operand_sets:
+          list:
+            - offset
+      instructions:
+        - "lds @ARG(0)+3+0"
+        - "phs"
+        - "lds @ARG(0)+2+1"
+        - "phs"
+        - "lds @ARG(0)+1+2"
+        - "phs"
+        - "lds @ARG(0)+0+3"
+        - "phs"
+  phsq:
+    # push a zero page long to the stack
+    # stack is arranged as big-endian, long in zero-page is little-endian
+    - operands:
+        count: 1
+        operand_sets:
+          list:
+            - zero_page
+      instructions:
+        - "ldz @ARG(0)+0"
+        - "phs"
+        - "ldz @ARG(0)+1"
+        - "phs"
+        - "ldz @ARG(0)+2"
+        - "phs"
+        - "ldz @ARG(0)+3"
+        - "phs"
   pls2:
     # pull 2 bytes from the stack
     - operands:
@@ -2803,6 +2853,42 @@ macros:
         - "sts @ARG(1)+2"
         - "ldz @ARG(0)+0"
         - "sts @ARG(1)+3"
+  mls4:
+    # move a long (4 bytes) from abs address to stack starting at passed offset
+    # stack is arranged as big-endian, long at abs address is little-endian
+    - operands:
+        count: 2
+        operand_sets:
+          list:
+            - absolute_address
+            - offset
+      instructions:
+        - "ldb @ARG(0)+3"
+        - "sts @ARG(1)+0"
+        - "ldb @ARG(0)+2"
+        - "sts @ARG(1)+1"
+        - "ldb @ARG(0)+1"
+        - "sts @ARG(1)+2"
+        - "ldb @ARG(0)+0"
+        - "sts @ARG(1)+3"
+  ms4l:
+    # move a long (4 bytes) from stack starting at passed offset to abs address
+    # stack is arranged as big-endian, long at abs address is little-endian
+    - operands:
+        count: 2
+        operand_sets:
+          list:
+            - offset
+            - absolute_address
+      instructions:
+        - "lds @ARG(0)+3"
+        - "stb @ARG(1)+0"
+        - "lds @ARG(0)+2"
+        - "stb @ARG(1)+1"
+        - "lds @ARG(0)+1"
+        - "stb @ARG(1)+2"
+        - "lds @ARG(0)+0"
+        - "stb @ARG(1)+3"
   aqq:
     # add two zero-page longs. result in second zero page operand
     - operands:
@@ -2844,3 +2930,16 @@ macros:
       instructions:
         - "mvv @ARG(0)+0,@ARG(1)+0"
         - "mvv @ARG(0)+2,@ARG(1)+2"
+  mll:
+    # move long from abs address to abs address
+    - operands:
+        count: 2
+        operand_sets:
+          list:
+            - absolute_address
+            - absolute_address
+      instructions:
+        - "mbb @ARG(0)+0,@ARG(1)+0"
+        - "mbb @ARG(0)+1,@ARG(1)+1"
+        - "mbb @ARG(0)+2,@ARG(1)+2"
+        - "mbb @ARG(0)+3,@ARG(1)+3"

examples/slu4-minimal-64x4/software/mathlib32.min64x4

@@ -1,3 +1,6 @@
+; Math Library 32-bit
+;  This library provides 32-bit math functions for unsigned and signed integers.
+;
 #require "slu4-min64x4-asm >= 1.1.0"
 ; Zero Page Usage
 .memzone ZERO_PAGE_APPS
@@ -118,9 +121,9 @@ _multiply:
     ; take twos complement of 8-byte results
     noq _working_mem8+0
     noq _working_mem8+4
-    inw _working_mem8+0 fcc .copy_results       ; if only INQ set the flags :-(
-    inw _working_mem8+2 fcc .copy_results
-    inw _working_mem8+4 fcc .copy_results
+    inw _working_mem8+0 bcc .copy_results       ; if only INQ set the flags :-(
+    inw _working_mem8+2 bcc .copy_results
+    inw _working_mem8+4 bcc .copy_results
     inw _working_mem8+6
 .copy_results:
     ; the entire working memory is the 64-bit results

examples/slu4-minimal-64x4/software/primes.min64x4

@@ -0,0 +1,129 @@
+; Prime Number Calculator
+;  This program calculates prime numbers using the algorithm documented here:
+;       https://en.wikipedia.org/wiki/Primality_test
+;
+#require "slu4-min64x4-asm >= 1.1.0"
+
+.memzone ZERO_PAGE_APPS
+#mute
+_n_value:            .zero 4
+
+
+#unmute
+
+.org 0 "USER_APPS"
+start:
+    ; init N-value to 1
+    clq _n_value
+    miz 1,_n_value+0
+.n_loop:
+    phsi 0
+    phsq _n_value
+    jps is_prime32
+    pls4
+    pls cpi 1 beq .print_is_prime
+    jpa .increment_n
+.print_is_prime:
+    phsq _n_value jps print_uint32 pls4
+    phsptr is_prime_str jps _PrintPtr pls2
+    ; jps _WaitInput
+.increment_n:
+    inq _n_value
+    jpa .n_loop
+
+is_prime_str: .cstr " is prime!\n"
+is_not_prime_str: .cstr " is not prime\n"
+
+; is_prime32
+;   determines wither the passed uint32 is a prime
+;
+;   Arguments
+;       sp+3 : the value to determine if prime (4 bytes)
+;       sp+7 : a place holder for return boolean
+;
+;   returns
+;       sp+7 : 0 or 1 depending on whether N  is prime
+is_prime32:
+    ; first check for 2 or 3
+    lds 3 cpi 0 bne .modulo_two        ; check top byte of N for 0
+    lds 4 cpi 0 bne .modulo_two        ; check top byte of N for 0
+    lds 5 cpi 0 bne .modulo_two        ; check top byte of N for 0
+    lds 6 cpi 3 beq .is_prime          ; check N==3
+    cpi 2 beq .is_prime                ; check N==2
+    cpi 1 beq .is_not_prime            ; check N==1
+    cpi 0 beq .is_not_prime            ; check N==0
+.modulo_two:
+    lds 6 lr1 bcc .is_not_prime        ; see if N's least signficant bit is even or odd
+.modulo_three:
+    phs4i 3                           ; place divisor on stack
+    phs4s (3+4)                       ; place dividend on stack (from stack)
+    jps divide32
+    pls4
+    ; check if remainder is 0
+    phs4i 0 jps compare_uint32ss pls4
+    pls4
+    beq .is_not_prime
+.loop_init:
+    ; set i-value (long) to 5
+    cll .current_i_val          ; clear all of I
+    mib 5, .current_i_val+0     ; copy 5 in to LSB
+.loop:
+    phs4a .current_i_val
+    phs4a .current_i_val
+    jps multiply_uint32
+    ; high 4 bytes of result should be 0 since we are only doing 32 bit N
+    lds 1 cpi 0 bne .iteration_loop_done
+    lds 2 cpi 0 bne .iteration_loop_done
+    lds 3 cpi 0 bne .iteration_loop_done
+    lds 4 cpi 0 bne .iteration_loop_done
+    pls4
+    ms4l 1,.isquared
+    pls4
+    phs4a .isquared phs4s 3+4 jps compare_uint32ss pls4 pls4
+    ; if I*I > N, we are done
+    bgt .loop_done_is_prime
+.n_gte_i_squared:
+    ; now check various modulos.
+    ; check N % I == 0
+    phs4a .current_i_val   ; I
+    phs4s (3+4)            ; N
+    jps divide32
+    pls4                   ; quotient
+    phs4i 0
+    jps compare_uint32ss
+    pls4
+    pls4
+    beq .loop_done_is_not_prime
+    ; check N % (I+2) == 0
+    mll .current_i_val,.temp_val
+    ail 2, .temp_val
+    phs4a .temp_val
+    phs4s (3+4)
+    jps divide32
+    pls4                   ; quotient
+    phs4i 0
+    jps compare_uint32ss
+    pls4                   ; zero
+    pls4                   ; remainder
+    beq .loop_done_is_not_prime
+    ; add 6 to I and loop
+    ail 6, .current_i_val
+    jpa .loop
+.iteration_loop_done:
+    ; get rid of I*I stack
+    pls4
+    pls4
+.loop_done_is_prime:
+.is_prime:
+    ldi 1 sts 7
+    rts
+.loop_done_is_not_prime:
+.is_not_prime:
+    ldi 0 sts 7
+    rts
+.current_i_val: .4byte 0
+.temp_val: .4byte 0
+.isquared: .4byte 0
+
+#include "mathlib32.min64x4"
+#include "stringlib.min64x4"

examples/slu4-minimal-64x4/software/stringlib.min64x4

@@ -1,3 +1,6 @@
+; String Library
+;   various routines for manipulating strings
+;
 #require "slu4-min64x4-asm >= 1.1.0"
 ; Zero Page Usage
 .memzone ZERO_PAGE_APPS
@@ -23,6 +26,7 @@ _working_mem5:          .zero 5
 ;
 ; TODO - check to see if there is enough rroom in teh stack to do this operation.
 ;
+.align
 cstr_reverse:
     lds 3+0 stz _tmp_Ptr1+1
     lds 3+1 stz _tmp_Ptr1+0
@@ -118,7 +122,6 @@ uint32_to_decimal_cstr:
     inv _tmp_Ptr1
     jpa .outer_loop
 .last_digit:
-    jps _Print "last digit!\n"
     ; get character of remainder
     ldz _working_mem5+0 adi $30
     ; store character in buffer
@@ -158,6 +161,26 @@ uint32_to_decimal_cstr:
     jpa _Prompt
 
 
+; print_uint32
+;   prints the passed unsigned 4 byte integer in decimal format
+;
+;   Arguments:
+;       sp + 3 : 32-bit value to print (4 bytes) big endian
+;
+BUFFER_SIZE = 32
+print_uint32:
+    phsi BUFFER_SIZE
+    phsptr .buffer
+    phs4s (3+3)
+    jps uint32_to_decimal_cstr
+    pls4
+    pls2
+    pls
+    phs2i .buffer jps _PrintPtr pls2
+    rts
+.buffer: .zero BUFFER_SIZE
+
+
 _print_working_memory:
     jps _Print "working mem = "
     ldz _working_mem5+4 jas _PrintHex