Differences

This shows you the differences between two versions of the page.

--- sd:star_forth_test_suite [2026/03/13 06:48] – appledog
+++ sd:star_forth_test_suite [2026/03/14 13:20] (current) – appledog
@@ Line 16: / Line 16: @@
 | 1.0\\ 260313 | PASS | '': FIZZBUZZ 21 1 DO I 15 MOD 0 = IF ." FizzBuzz" ELSE I 3 MOD 0 = IF ." Fizz" ELSE I 5 MOD 0 = IF ." Buzz" ELSE I . THEN THEN THEN CR LOOP ;'' | not shown (but it works) |
 | 1.0\\ 260313 | PASS | '': SIMPLE 500000 0 DO LOOP ;''\\ ''SIMPLE'' | ~6 seconds (170,000 w/s) at 4MHz |
-| 1.0\\ 260313 | FAIL | '': BENCH 500000 0 DO I DUP * DROP LOOP ;''\\ ''BENCH'' | CRASH |
+| 1.0\\ 260313 | PASS | '': BENCH 500000 0 DO I DUP * DROP LOOP ;''\\ ''BENCH'' | 22 seconds (140,000 w/s) at 4Mhz |
+| 1.0\\ 260314 | PASS | '': #WORDS 0 $030003 P@ BEGIN DUP WHILE SWAP 1 + SWAP P@ REPEAT DROP . ." words" CR ;'' | Shows 73 words (see WORDS, which shows 72). |
+Note that AS-IS, SD/FORTH 1.0 is not FORTH-79 compliant. That's fine, the following self-hosted words act as a sort of test suite, and also bring us up squarely into FORTH-79.
+<codify forth>
+    : 1+ 1 + ;
+    : 1- 1 - ;
+    : 2+ 2 + ;
+    : 2- 2 - ;
+    : 0= 0 = ;
+    : 0< 0 < ;
+    : 0> 0 > ;
+    : NEGATE 0 SWAP - ;
+    : ABS DUP 0< IF NEGATE THEN ;
+    : NIP SWAP DROP ;
+    : TUCK SWAP OVER ;
+    : ?DUP DUP IF DUP THEN ;
+    : ROT >R SWAP R> SWAP ;
+    : 2DUP OVER OVER ;
+    : MAX 2DUP < IF SWAP THEN DROP ;
+    : MIN 2DUP > IF SWAP THEN DROP ;
+    : SPACE 32 EMIT ;
+    : SPACES 0 DO SPACE LOOP ;
+    : /MOD 2DUP MOD >R / R> SWAP ;
+    : +! DUP @ ROT + SWAP ! ;
+    : TYPE 0 DO DUP C@ EMIT 1 + LOOP DROP ;
+    : TYPE ( addr len -- ) 0 DO DUP C@ EMIT 1 + LOOP DROP ;
+    : MOVE CMOVE ;
+</codify>
+== FORTH-79 TESTS
+You could also try this; if you see all PASS and then ABC and then A on the next line underneath the space beside the C, it works!
+<codify Forth>
+: TEST= = IF ." PASS " ELSE ." FAIL " THEN ;
+( Arithmetic )
+1+ 6 TEST=
+1- 4 TEST=
+2+ 7 TEST=
+2- 3 TEST=
+NEGATE -5 TEST=
+-7 NEGATE 7 TEST=
+-5 ABS 5 TEST=
+ABS 5 TEST=
+( Comparisons )
+0= 1 TEST=
+0= 0 TEST=
+-1 0< 1 TEST=
+0< 0 TEST=
+0> 1 TEST=
+-1 0> 0 TEST=
+( Stack ops )
+2 NIP 2 TEST=
+2 TUCK 2 TEST= DROP DROP
+2 TUCK DROP 1 TEST= DROP
+2 TUCK DROP DROP 2 TEST=
+?DUP + 10 TEST=
+?DUP 0 TEST=
+2 3 ROT 1 TEST= DROP DROP
+4 2DUP + >R + R> + 14 TEST=
+( Math )
+3 MAX 7 TEST=
+7 MAX 7 TEST=
+3 MIN 3 TEST=
+7 MIN 3 TEST=
+3 /MOD SWAP 1 TEST=
+3 /MOD 2 TEST=
+DROP
+( Variable and +! )
+VARIABLE TV
+TV !
+TV +!
+TV @ 13 TEST=
+( SPACE and SPACES )
+EMIT 66 EMIT 67 EMIT CR
+SPACES 65 EMIT CR
+( Bit Shifting Ops )
+0 LSHIFT 1 TEST=
+1 LSHIFT 2 TEST=
+4 LSHIFT 16 TEST=
+8 LSHIFT 256 TEST=
+$FF 4 LSHIFT $FF0 TEST=
+1 RSHIFT 8 TEST=
+8 RSHIFT 1 TEST=
+$FF0 4 RSHIFT $FF TEST=
+1 RSHIFT 0 TEST=
+</codify>
+== Benchmarking
+The SIMPLE and BENCH tests indicate SD/FORTH is:
+* 75x faster than Stellar BASIC (1,800 lines/sec) × ~5 ops/line ≈ 9,000 ops/sec)
+* 5x to 7x faster than a real C64 running DurexForth (~25,000 words/sec)
+* Comparable to a 4 MHz Z80 running Forth850 or an 8MHz Z80B running CamelForth.
+However, this is an upper end. The following MUL and DIV heavy benchmarks operate a bit slower. Here's a summary of the followinf programs:
+| Benchmark | Words/Sec. | Notes |
+| Bare Loop | 176,000 | Only dispatch overhead |
+| Stackbench | 160,000 | Only stack operations |
+| DUP * DROP | 136,000 | Multiply, Stack Ops |
+| ADD only | 120,000 | Same as Mixed but no DIV or MUL |
+| Mixed no DIV | 100,000 | Same as Mixed but no DIV |
+| FizzBench | 97,000 | 3x MOD, branches |
+| Mixed +-*/ | 90,000 | Division, Heavy Stack |
+The estimate is that 70% overhead is dispatch. So with a touch of opt I think the system wants to run at 180,000 words/sec, but gets tinted down by various factors. Stack operations are cheap, then come other types of operation, then MUL, DIV and MOD are considered 'heavy'. This is not unusual, and is in fact a great sign; this is a very fast FORTH.
+=== STACKBENCH
+<codify Forth>
+    : STACKBENCH ( -- )
+2 3
+0 DO
+        DUP DROP SWAP OVER DROP
+      LOOP
+      DROP DROP DROP ;
+    STACKBENCH
+</codify>
+=== FIZZBENCH
+<codify Forth>
+    : FIZZBUZZ 21 1 DO I 15 MOD 0 = IF ." FizzBuzz" ELSE I 3 MOD
+= IF ." Fizz" ELSE I 5 MOD
+= IF ." Buzz" ELSE I . THEN THEN THEN CR LOOP ;
+    : FIZZBENCH 0 10000 0
+        DO I 15 MOD 0 = IF 1 + ELSE
+            I 3 MOD 0 = IF 1 + ELSE
+                I 5 MOD 0 = IF 1 + THEN
+                THEN
+            THEN
+        LOOP
+    ;
+    : FB10 10 0 DO FIZZBENCH LOOP . ." hits" CR ;
+    FB10
+</codify>
+This executes in 18.5 seconds, indicating 97,000 words/sec.
+=== BENCH-ARITH
+<codify FORTH>
+    : ARITH ( n -- n )
+      DUP 42 +
+      DUP 7 -
+      DUP 3 *
+      DUP 2 /
+      DROP DROP DROP DROP ;
+    : BENCH-ARITH ( -- )
+      ." Arithmetic benchmark" CR
+0 DO
+        I ARITH DROP
+      LOOP
+      ." Done" CR ;
+    BENCH-ARITH
+</codify>
+Same without DIV:
+<codify Forth>
+    : ARITH3 ( n -- n )
+      DUP 42 +
+      DUP 7 -
+      DUP 3 *
+      DROP DROP DROP ;
+    : BENCH3 ( -- )
+      ." Add/Sub/Mul benchmark" CR
+0 DO
+        I ARITH3 DROP
+      LOOP
+      ." Done" CR ;
+    BENCH3
+</codify>