error A2071: initializer magnitude too large for specified size

[jj2007]

The line no identified (171) points to the - readrec  endp  statement

Now it gets mysterious - there is nothing wrong with that line.
On my PC, the attached source works fine with Masm 6.14, 6.15, 9.0, 10.0 and JWasm.

I added an error check after fopen.

[MichaelW]

I also cannot find anything wrong anywhere in the source after I commented out the include gwMacros.inc statement. What assembler command line are you using?

[GoneFishing]

Last night when I was looking through pgw001's posts I noticed that he uses MS VS as IDE.
My last suggestion is to try assembling the code  in pure batch environment i.e. in QEDITOR
Best

[jj2007]

Using \Masm32\qeditor.exe could help indeed. I used RichMasm without any problems.

There are several logical errors to fix. Inter alia,

1.
LOCAL cloc  :DWORD                          ; current location variable
is never initialised, and therefore will set the file pointer initially to an arbitrary value ->CRASH.
Test it by commenting out the mov cloc, fseek(hFile,cloc,0)

2. hMem is a local, too, and therefore this will fail in an unpredictable way:
    LOCAL hMem  :DWORD                          ; allocated memory handle
...
   cmp hMem, 0                           ; Check if file open
   jne labGO                           ; Jump if so

You should read The LOCAL variables trap carefully.

[TWell]

broken rdLine: too
- no null checking
- no output buffer size checking

[pgw001]

Folks, thanks to all for your perseverance

The observation that I use the Visual Studio 2013 IDE and the recommendation I try qeditor.exe proved to be the guiding lights

The code, including the gwmacros.inc file, assembled and linked without errors or warnings using qeditor - I am not sure
what this says about the VS environment albeit I have now copied the code that made up the 'readrec proc' into a different
test harness which subsequently went on to assemble without error under VS.

Again thanks for your contributions and support - should I advise Microsoft?

[jj2007]

should I advise Microsoft?

Honestly, Santa Claus doesn't exist, and Microsoft doesn't take advice from its clients ;-)

[Gunther]

Again thanks for your contributions and support - should I advise Microsoft?

you can try it, but the success will be questionable. So, let it be.

Gunther

[raymond]

labCR:                                 ; CR found
   mov[edi],al                           ; store CR char
   inc edi                              ; incr out pointer
   mov al,[esi]
   inc esi
   cmp al,10                           ; check if LF char
   je labLF

labLF:                                 ; LF found
   mov[edi],al                              ; store LF char
   inc edi                                 ; incr out pointer

Although the above piece of code will not throw an error, it could mean trouble if the CR character is not followed by an LF character. Your intention should be to skip storing the character if it is NOT an LF but your
code simply continues to process that code you would want to skip! :icon_eek:

[MichaelW]

Honestly, Santa Claus doesn't exist, and Microsoft doesn't take advice from its clients ;-)

They do for the big customers, like the one that requested a Word macro system that can be programmed to silently format your hard drive

[Gunther]

Michael,

They do for the big customers, like the one that requested a Word macro system that can be programmed to silently format your hard drive

but that's the world of the big money, not the normal users world.

Gunther