[SOLVED] Stack Overflow without decernable Reason

[MMarie]

Hi all,
I'm currently working on implementing a "Compiler" for a custom language and have been running into a Stack Overflow without a decernable reason. I'm not doing recursions and have checked that, and even upping the Stack-Size still doesn't help.

(This is my first post so please just let me know if I have done anything wrong with Formatting or anything)

Here is my code

Code: Pascal  [Select][+][-]

1. {$mode objfpc}
2. unit uASM;
3.  
4. interface
5.  
6.   uses StrUtils, SysUtils, Types, uPRECodes, uUtil;
7.  
8.   type
9.     TInformation = record
10.       path, outpath: String;
11.       _file: TextFile;
12.       cline: Integer;
13.       sline: String;
14.     end;
15.     TPREAssembler = class
16.     private
17.       FInterfaceData: TInterfaceDataArr;
18.  
19.       { FeEc 16-11-2022: idc that this violates my own Styleguide, its gonna get changed later on }
20.       info: TInformation;
21.     public
22.       function Assemble: TByteArray;
23.       procedure Run(const AFile: String);
24.     end;
25.  
26. implementation
27.   function TPREAssembler.Assemble: TByteArray;
28.   var
29.     instruction: String;
30.     tmp, out: TByteArray;
31.     split, parameters: Types.TStringDynArray;
32.   begin
33.     out := [];
34.  
35.     while not eof(info._file) do
36.     begin
37.       info.cline := info.cline + 1;
38.       Readln(info._file, info.sline);
39.       info.sline := Trim(info.sline);
40.  
41.       // Parse Line
42.       if (info.sline[1] = ';') then continue;
43.  
44.       split := SplitString(info.sline, ' ');
45.       instruction := LowerCase(split[0]);
46.      
47.       if (Length(split) > 1) then
48.         parameters := Copy(split, 1, Length(split)-1)
49.       else
50.         parameters := [''];
51.      
52.       // Assemble Instruction
53.       try
54.         case instruction of
55.           'halt': begin
56.             out := ConcatByteArrs(out, [OP_HALT]);
57.           end;
58.           'sys': begin
59.             out := ConcatByteArrs(out, [OP_SYS]);
60.           end;
61.           'ld': begin
62.             {
63.               Possible Parameters
64.               Param1
65.                 Hardcoded Byte / Byte Const
66.               Param2
67.                 Hardcoded Any Type / Any Type Const
68.                 Variable
69.             }
70.             tmp := ParseASMParam(parameters[0], [prmByte], FInterfaceData); // Causes Stack Overflow, why???
71.             tmp := ConcatByteArrs(tmp, ParseASMParam(parameters[1], [prmAll], FInterfaceData));
72.             out := ConcatByteArrs(out, ConcatByteArrs([OP_LD], tmp));
73.           end;
74.         else
75.         begin
76.           writeln('presdk-asm: ', info.path, ' (line ', IntToStr(info.cline), '): error: Invalid Instruction "', instruction, '"');
77.         end;
78.       end
79.       except
80.         on e: EInvalidParameterException do
81.         begin
82.           writeln('presdk-asm: ', info.path, ' (line ', IntToStr(info.cline), '): error: ',
83.           e.Message);
84.         end;
85.       end;
86.     end;
87.  
88.     result := out;
89.   end;
90.  
91.   procedure TPREAssembler.Run(const AFile: String);
92.   var
93.     bytecode: TByteArray;
94.     i: Integer;
95.   begin
96.     if not FileExists(AFile) then
97.     begin
98.       writeln('PREASM: error: File "', AFile, '" does not exist!');
99.       exit;
100.     end;
101.  
102.     info.path := AFile;
103.     info.outpath := ExtractFilePath(AFile)+'/'+ExtractFileName(AFile)+'.prex';
104.  
105.     Assign(info._file, info.path);
106.     ReSet(info._file);
107.     info.sline := '';
108.     info.cline := 0;
109.  
110.     // Interface Section
111.     while not eof(info._file) and (info.sline <> '.start') do
112.     begin
113.       info.cline := info.cline + 1;
114.       Readln(info._file, info.sline);
115.     end;
116.  
117.     // Source Code
118.     bytecode := Assemble;
119.     for i := 0 to Length(bytecode)-1 do
120.     begin
121.       write(Format('%.2x ', [bytecode[i]]));
122.       if (i mod 8 = 0) and (i > 0) then writeln;
123.     end;
124.     writeln;
125.   end;
126. end.
127.  

Code: Pascal  [Select][+][-]

1. {$mode objfpc}
2. unit uUtil;
3.  
4. interface
5.   uses StrUtils, SysUtils, uASCII, uPRECodes;
6.  
7.   type
8.     EInvalidParameterException = class(Exception);
9.     EInvalidNamespaceException = class(Exception);
10.     TParamTypes = set of (prmAll, prmBool, prmByte, prmShort, prmInt, prmLong, prmFloat, prmStr, prmVar);
11.    
12.     { Holds information on a Source-File, used for multisource }
13.     TInterfaceData = record
14.       name: String;
15.       path: String;
16.       procNames: TStringDynArray;
17.       procIdents: TIntegerArray;
18.       constNames: TStringDynArray;
19.       constVals: TByteArrayArray;
20.     end;
21.     TInterfaceDataArr = array of TInterfaceData;
22.  
23.   { Converison Funcs }
24.   function IntToBytes(const AInt: Integer): TByteArray;
25.  
26.   { Parsing Funcs }
27.   function ParseASMParam(const AParam: String; const acceptableTypes: TParamTypes; var AInterfaceData: TInterfaceDataArr): TByteArray;
28.  
29.   { Other Funcs }
30.   function ConcatByteArrs(AArr1: TByteArray; const AArr2: TByteArray): TByteArray;
31.   function GetInterfaceData(const AArray: TInterfaceDataArr; const AName: String): TInterfaceData;
32. implementation
33.   function IntToBytes(const AInt: Integer): TByteArray;
34.   begin
35.     exit([
36.           AInt and $FF00000000000000, AInt and $00FF000000000000, AInt and $0000FF0000000000,
37.           AInt and $000000FF00000000, AInt and $00000000FF000000, AInt and $0000000000FF0000,
38.           AInt and $000000000000FF00, AInt and $00000000000000FF
39.         ]);
40.   end;
41.  
42.   function ParseASMParam(const AParam: String; const acceptableTypes: TParamTypes; var AInterfaceData: TInterfaceDataArr): TByteArray;
43.   var
44.     prefix, value: String;
45.     tmpInt: Integer;
46.     cutcomma: Boolean;
47.   begin
48.     prefix := AParam[1];
49.     value := Copy(AParam, 2, Length(AParam));
50.     if EndsStr(',', value) then
51.     begin
52.       value := Copy(value, 1, Length(value)-1);
53.       cutcomma := True;
54.     end;
55.  
56.     if (prefix = '#') then
57.     begin
58.       if (value = 'true') or (value = 'false') then
59.       begin
60.         if prmBool in acceptableTypes then
61.           if value = 'true' then exit([1]) else exit([0])
62.         else
63.           raise EInvalidParameterException.Create('Invalid Parameter Type!');
64.       end;
65.  
66.       if ((Byte(value[1]) >= CHR_0) and (Byte(value[1]) <= CHR_9)) or (value[1] = '$') then // Number
67.         exit(IntToBytes(StrToInt(value)))
68.       else // Other Constant
69.       begin
70.         GetInterfaceData(AInterfaceData, 'test');
71.         exit([$D0, $00]);
72.       end;
73.  
74.       raise EInvalidParameterException.Create('Unrecognised Parameter Type! Parameter: '+value);
75.     end;
76.   end;
77.  
78.   function ConcatByteArrs(AArr1: TByteArray; const AArr2: TByteArray): TByteArray;
79.   var
80.     i: Integer;
81.   begin
82.     SetLength(AArr1, Length(AArr1) + Length(AArr2));
83.     for i := 0 to Length(AArr2)-1 do
84.       AArr1[i+Length(AArr2)] := AArr2[i];
85.  
86.     result := AArr1;
87.   end;
88.  
89.   function GetInterfaceData(const AArray: TInterfaceDataArr; const AName: String): TInterfaceData;
90.   var
91.     intdat: TInterfaceData;
92.   begin
93.     for intdat in AArray do
94.     begin
95.       if (intdat.name = AName) then exit(intdat);
96.     end;
97.  
98.     raise EInvalidNamespaceException.Create('Can''t find Interface-Data for Namespace: '+AName);
99.   end;
100. end.
101.  

[Thaddy]

Use range checking and overflow checking to debug, you can also use stack thrashing. Also use line info. See compiler options.

[Martin_fr]

What is your OS?

Do you compile with optimizations? -O2 or above? Does the error happen if you switch optimization off?

What does the debugger show?
=> For this make sure (if you don't already) you compile with debug info. And with range and stack checking (as Thaddy said).
=> Then open the stack window (View => debug windows) and get the content.

If you compile from commandline (i.e. if you do not use the Lazarus IDE) then run in whatever debugger you use (gdb/lldb) and get the stack from there.

[MarkMLl]

Welcome to the forum. As others have hinted, you'd make things a bit easier if you showed your command line and precise results, since otherwise we don't know what range checking etc. is being applied.

One thing that does stand out is that a hex number like $ff00000000000000 will be interpreted as a signed literal (i.e. Int64) even though that's not your intention, you need to use qword($ff00000000000000) to get the correct bit pattern.

MarkMLl

[jamie]

Please look at the definitions of TbyteArray, TIntegerArray etc.

Those types in my opinion were created to be used in a Casting or pointer situation.
 
 Maybe what you want to use is all dynamic instead?

[MMarie]

What is your OS?

Do you compile with optimizations? -O2 or above? Does the error happen if you switch optimization off?

What does the debugger show?
=> For this make sure (if you don't already) you compile with debug info. And with range and stack checking (as Thaddy said).
=> Then open the stack window (View => debug windows) and get the content.

If you compile from commandline (i.e. if you do not use the Lazarus IDE) then run in whatever debugger you use (gdb/lldb) and get the stack from there.

Im on x86_64 Windows 11, the error also occurs when not compiling with optimizations. I have specified {$R+} and {$Q+} in the programs entry point.

Here is my compile command for debugging:
fpc src/PRESDK.pas -g -gt -Ct -Fu"../inc/" -FE"out/"

And here is the Stack info I can get from gdb:

Code: Text  [Select][+][-]

1. (gdb) x/x $esp
2. 0x432a70:       0x00000000
3. (gdb) info frame
4. Stack level 0, frame at 0x142fa78:
5.  eip = 0x415830 in PARSEASMPARAM (src/uUtil.pas:47); saved eip 0x414ede
6.  called by frame at 0x142fde8
7.  source language pascal.
8.  Arglist at 0x142fa70, args: APARAM=..., ACCEPTABLETYPES=..., AINTERFACEDATA=0xc6e58955, result=0x0
9.  Locals at 0x142fa70, Previous frame's sp is 0x142fa78
10.  Saved registers:
11.   ebp at 0x142fa70, edi at 0x142fa6c, eip at 0x142fa74

Edit: Meant to say that I specified {$R+} and {$Q+} within the main program unit, not inside the actual entry-point.

[MMarie]

Please look at the definitions of TbyteArray, TIntegerArray etc.

Those types in my opinion were created to be used in a Casting or pointer situation.
 
 Maybe what you want to use is all dynamic instead?

These type definitions come from another Unit within the project, which are dynamic.

[Zvoni]

Don't you have to set the Length of the Dynamic Array-Result before exiting?
Or is "Exit" doing it in the Background?

because that's the only thing i can see, that OP never sets the Length of the dynamic array, the resulting Byte-Array as well as that "var" Argument FInterface or whatever it's called

[MMarie]

Don't you have to set the Length of the Dynamic Array-Result before exiting?
Or is "Exit" doing it in the Background?

because that's the only thing i can see, that OP never sets the Length of the dynamic array, the resulting Byte-Array as well as that "var" Argument FInterface or whatever it's called

Atleast not in my experience, calling exit like e.g. exit([$00, $fe]); hasn't caused problems for me. I assume its set in the background

[jamie]

Please look at the definitions of TbyteArray, TIntegerArray etc.

Those types in my opinion were created to be used in a Casting or pointer situation.
 
 Maybe what you want to use is all dynamic instead?

These type definitions come from another Unit within the project, which are dynamic.

Please rest your mouse over those types in your code, all of your code to ensure that is true. These types also exist in the SysUtil unit too and order of units are important.

[MMarie]

Please look at the definitions of TbyteArray, TIntegerArray etc.

Those types in my opinion were created to be used in a Casting or pointer situation.
 
 Maybe what you want to use is all dynamic instead?

These type definitions come from another Unit within the project, which are dynamic.

Please rest your mouse over those types in your code, all of your code to ensure that is true. These types also exist in the SysUtil unit too and order of units are important.

Since I'm not using an IDE I've just played "uPRECodes", which is the unit that  contains my definitions, nothing changes

[jamie]

I also noticed you are not using {$H+} strings, which means most likely all of your strings are shortstring which are 255 in length, without index sizing.

[MMarie]

I just noticed that, very talently, misunderstood how exactly compiler flags like {$R+} work... Now with having properly added them, I get an ERangeError. Can't get any useful debug info anymore though, because the program immediatly exits on this error.

[Zvoni]

No IDE?
Then writeln each value at each step. So you can see what‘s the real cause of the exception

[MMarie]

No IDE?
Then writeln each value at each step. So you can see what‘s the real cause of the exception

The values are all OK, the Error definetly comes from the method call unless i'm missing something