fix expericodegen crash related to shortcircuiting

codeGenIntermediate/src/prog8/codegen/intermediate/ExpressionGen.kt

@@ -17,8 +17,6 @@ internal class ExpressionCodeResult(val chunks: IRCodeChunks, val dt: IRDataType
         if(resultReg!=-1) require(resultFpReg==-1)
         if(resultFpReg!=-1) require(resultReg==-1)
     }
-
-    fun lastInstruction() = chunks.last().instructions.last()
 }
 
 internal class ExpressionGen(private val codeGen: IRCodeGen) {

codeGenIntermediate/src/prog8/codegen/intermediate/IRCodeGen.kt

@@ -1026,7 +1026,8 @@ class IRCodeGen(
         val condition = ifElse.condition as PtBinaryExpression
         val leftTr = expressionEval.translateExpression(condition.left)
         addToResult(result, leftTr, leftTr.resultReg, -1)
-        val requireCompareZero = leftTr.lastInstruction().opcode !in OpcodesThatSetStatusbits
+        val lastInstruction = leftTr.chunks.last().instructions.lastOrNull()
+        val requireCompareZero = lastInstruction?.opcode !in OpcodesThatSetStatusbits
         when(condition.operator) {
             "==" -> {
                 if(requireCompareZero)

compiler/test/vm/TestCompilerVirtual.kt

@@ -473,4 +473,17 @@ instructions {
         compileText(VMTarget(), false, src, writeAssembly = true) shouldNotBe null
     }
 
+    test("IR codegen for while loop with shortcircuit") {
+        val src="""
+main {
+    sub start() {
+        cx16.r0L=1
+        while cx16.r0L < 10 and cx16.r0L>0 {
+            cx16.r0L++
+        }
+    }
+}"""
+        compileText(VMTarget(), true, src, writeAssembly = true) shouldNotBe null
+    }
+
 })

docs/source/todo.rst

@@ -9,6 +9,8 @@ Future Things and Ideas
 ^^^^^^^^^^^^^^^^^^^^^^^
 Compiler:
 
+- get rid of the noshortcircuit fallback option and code.
+- What happens when we make all subs return a boolean not as ubyte in A, but in the cpu's Carry flag?
 - Multidimensional arrays and chained indexing, purely as syntactic sugar over regular arrays.
 - make a form of "manual generics" possible like: varsub routine(T arg)->T  where T is expanded to a specific type
   (this is already done hardcoded for several of the builtin functions)
@@ -24,6 +26,7 @@ Compiler:
     - (need separate step in codegen and IR to write the "golden" variables)
 
 - do we need (array)variable alignment tag instead of block alignment tag? You want to align the data, not the code in the block?
+- ir: proper code gen for the CALLI instruction and that it (optionally) returns a word value that needs to be assigned to a reg
 - ir: getting it in shape for code generation
 - ir: related to the one above: block alignment doesn't translate well to variables in the block (the actual stuff that needs to be aligned in memory)  but: need variable alignment tag instead of block alignment tag, really
 - ir: idea: (but LLVM IR simply keeps the variables, so not a good idea then?...): replace all scalar variables by an allocated register. Keep a table of the variable to register mapping (including the datatype)
@@ -44,7 +47,7 @@ Compiler:
 
 Libraries:
 
-- once a VAL_1 implementation is merged into the X16 kernal properly, remove all the workarounds in cx16 floats.parse_f()
+- once a VAL_1 implementation is merged into the X16 kernal properly, remove all the workarounds in cx16 floats.parse_f()  .   Prototype parse routine in examples/cx16/floatparse.p8
 - fix the problems in atari target, and flesh out its libraries.
 - c128 target: make syslib more complete (missing kernal routines)?
 - pet32 target: make syslib more complete (missing kernal routines)?
@@ -52,8 +55,6 @@ Libraries:
 
 Optimizations:
 
-- give a warning for variables that could be a const - or even make them a const (if not @shared)?
-- treat every scalar variable decl with initialization value, as const by default, unless the variable gets assigned to somewhere (or has its address taken, or is @shared)
 - VariableAllocator: can we think of a smarter strategy for allocating variables into zeropage, rather than first-come-first-served?
   for instance, vars used inside loops first, then loopvars, then uwords used as pointers, then the rest
 - various optimizers skip stuff if compTarget.name==VMTarget.NAME.  Once 6502-codegen is done from IR code,
@@ -76,9 +77,6 @@ What if we were to re-introduce Structs in prog8? Some thoughts:
 Other language/syntax features to think about
 ---------------------------------------------
 
-- module directive to set the text encoding for that whole file (iso, petscii, etc.)
-- chained assignments   `x=y=z=99`
-- declare multiple variables    `ubyte x,y,z`    (if init value present, all get that init value)
-- chained comparisons   `10<x<20` ,   `x==y==z`   (desugars to  `10<x and x<20`,   `x==y and y==z`)
-- postincrdecr as expression, preincrdecr expression   (`y = x++`,   `y = ++x`)    .... is this even possible, expression with side effects like this?
-- negative array index to refer to an element from the end of the array.  Python `[-1]` or Raku syntax `[\*-1]`  , `[\*/2]` .... \*=size of the array
+- add (rom/ram)bank support to romsub.   A call will then automatically switch banks, use callfar and something else when in banked ram.
+  challenges: how to not make this too X16 specific? How does the compiler know what bank to switch (ram/rom)?
+  How to make it performant when we want to (i.e. NOT have it use callfar/auto bank switching) ?

examples/test.p8

@@ -1,118 +1,13 @@
 %import textio
-%zeropage dontuse
+%zeropage basicsafe
+
+; Note: this program can be compiled for multiple target systems.
 
 main {
     sub start() {
-        ubyte @shared x = 10
-        if value(12) < x < value(100)
-            txt.print("gottem")
-
-        sub value(ubyte v) -> ubyte {
-            cx16.r0++
-            txt.print("value(): ")
-            txt.print_ub(v)
-            txt.nl()
-            return v
+        cx16.r0L=1
+        while cx16.r0L < 10 and cx16.r0L>0 {
+            cx16.r0L++
         }
-
-        bigtest()
-    }
-
-    ubyte @shared a1 = 10
-    ubyte @shared a2 = 20
-    ubyte @shared x1 = 30
-    ubyte @shared x2 = 40
-    ubyte @shared zero = 0
-
-    sub bigtest () {
-        txt.print("1a:\n")
-        if calc_a1()<calc_x1() and calc_a2()<=calc_x2()
-            txt.print("* 1a and ok\n")
-
-        txt.print("\n1b:\n")
-        if calc_a1()<calc_x1() and calc_a2()>calc_x2()
-            txt.print("* 1b and fail\n")
-
-        txt.print("\n1c:\n")
-        if calc_a1()>calc_x1() and calc_a2()<=calc_x2()
-            txt.print("* 1c and fail\n")
-
-        txt.print("\n2a:\n")
-        if calc_a1()<calc_x1() or calc_a2()<=calc_x2()
-            txt.print("* 2a or ok\n")
-
-        txt.print("\n2b:\n")
-        if calc_a1()<calc_x1() or calc_a2()>calc_x2()
-            txt.print("* 2b or ok\n")
-
-        txt.print("\n3a:\n")
-        if calc_a1()>calc_x1() or calc_a2()<=calc_x2()
-            txt.print("* 3a or ok\n")
-
-        txt.print("\n3b:\n")
-        if calc_a1()>calc_x1() or calc_a2()>calc_x2()
-            txt.print("* 3b or fail\n")
-
-        txt.print("\n4a:\n")
-        bool result = calc_a1()<calc_x1() or calc_a2()>calc_x2()
-        txt.print_ub(result)
-        txt.nl()
-        txt.print("\n4b:\n")
-        result = calc_a1()>=calc_x1() and calc_a2()>calc_x2()
-        txt.print_ub(result)
-        txt.nl()
-        @($4000) &= 22
-
-        txt.print("\n5a:\n")
-        result = bool_true() or bool_false()
-        txt.print("\n5b:\n")
-        result = bool_true() and bool_false()
-        txt.print("\n5c:\n")
-        result = bool_false() and bool_true()
-        txt.print("\n5d:\n")
-        result = bool_false() xor bool_true()
-
-        txt.print("augmented and shortcut:\n")
-        bool @shared b1 = false
-        cx16.r0++
-        b1 = b1 and bool_true()
-        txt.print("augmented and no shortcut:\n")
-        b1 = true
-        cx16.r0++
-        b1 = b1 and bool_true()
-
-        txt.print("augmented or shortcut:\n")
-        b1 = true
-        cx16.r0++
-        b1 = b1 or bool_true()
-        txt.print("augmented or no shortcut:\n")
-        b1 = false
-        cx16.r0++
-        b1 = b1 or bool_true()
-    }
-
-    sub bool_true() -> bool {
-        txt.print("bool_true\n")
-        return true
-    }
-    sub bool_false() -> bool {
-        txt.print("bool_false\n")
-        return false
-    }
-    sub calc_a1() -> ubyte {
-        txt.print("calc_a1\n")
-        return a1+zero
-    }
-    sub calc_a2() -> ubyte {
-        txt.print("calc_a2\n")
-        return a2+zero
-    }
-    sub calc_x1() -> ubyte {
-        txt.print("calc_x1\n")
-        return x1+zero
-    }
-    sub calc_x2() -> ubyte {
-        txt.print("calc_x2\n")
-        return x2+zero
     }
 }