1. Ver. 4.0 Debugging

Matt and Derek: 
I've tried running these Watcom debuggers but I'am having all kinds 
problems not knowing the commands and the source code. 
 
I think that you can debug the problem using XP or WIN2000 because 
if you create a debug version with MANAGED_MEM=1 option. 
 
Then run the the example: 
 
printf(1,"%d \n",{3217040639}) 
 
if getc(0) end if 
 
The result will print: 
 
-2147483648 
 
Maybe you could try debugging it on XP or Win2000. 
 
I'll keep trying to learn how to use this debugger. 
 
Bernie 
 
new topic     » topic index » view message » categorize

2. Re: Ver. 4.0 Debugging

bernie said...

I've tried running these Watcom debuggers but I'am having all kinds problems not knowing the commands and the source code.

I think that you can debug the problem using XP or WIN2000 because if you create a debug version with MANAGED_MEM=1 option.

Maybe you could try debugging it on XP or Win2000.

I'll keep trying to learn how to use this debugger.

I have created a MANAGED_MEM binary, but can't duplicate your bug. Perhaps if you posted a binary, we could try to duplicate it.

Matt

new topic     » goto parent     » topic index » view message » categorize

3. Re: Ver. 4.0 Debugging

bernie said...
Matt and Derek: 
I've tried running these Watcom debuggers but I'am having all kinds 
problems not knowing the commands and the source code. 
 
I think that you can debug the problem using XP or WIN2000 because 
if you create a debug version with MANAGED_MEM=1 option. 

As I have already told you, I did create a managed_mem edition of the Euphoria and it works just fine. There is NOTHING to debug because it works.

You seem to be the only one with this problem. I cannot recreate it. In fact, no-one seems to be able to recreate it.

If I was you, I'd be putting 'printf' statements in the C source to get some more info about the problem.

new topic     » goto parent     » topic index » view message » categorize

4. Re: Ver. 4.0 Debugging

mattlewis said...
bernie said...

I've tried running these Watcom debuggers but I'am having all kinds problems not knowing the commands and the source code.

I think that you can debug the problem using XP or WIN2000 because if you create a debug version with MANAGED_MEM=1 option.

Maybe you could try debugging it on XP or Win2000.

I'll keep trying to learn how to use this debugger.

I have created a MANAGED_MEM binary, but can't duplicate your bug. Perhaps if you posted a binary, we could try to duplicate it.

Matt

Matt:
Be sure you are using exw.exe that has managed memory turned on
and not a exw.exe that is set in your path.
The bug shows up on WIN2000 when I run it with the WIN98 binary
created in the source directory.
How do I post a binary ?

How do I single step through my code with this Watcom dbugger
I can set a break to FormatItem but I want to single step through
the function.

new topic     » goto parent     » topic index » view message » categorize

5. Re: Ver. 4.0 Debugging

bernie said...

Be sure you are using exw.exe that has managed memory turned on
and not a exw.exe that is set in your path.
The bug shows up on WIN2000 when I run it with the WIN98 binary
created in the source directory.
How do I post a binary ?

How do I single step through my code with this Watcom dbugger
I can set a break to FormatItem but I want to single step through
the function.

I was definitely using the correct binary. I was running right from the source directory where I built it. You could use a web page that you control, or upload it to RapidEuphoria.com and mark it for temporary storage only (I've never done this, but I think you just need to put a comment for Rob telling him not to post it).

I've only used the Watcom debugger a little. Have you looked at the documentation that comes with it? I think you just have to navigate to the place where you want to set a break point, and double click it or something.

Matt

new topic     » goto parent     » topic index » view message » categorize

6. Re: Ver. 4.0 Debugging

Bernie,

Thank you for all of your effort in telling us and running all of the tests. But the problem seems to be with your build environment rather than the source code if you can reproduce it with your binary on more than one machine but we cannot. If you get the binaries from euphoria.blogdns.com or jeremy.cowgar.com/eubins and try with those and you still have a problem post it here.

If I were you I would get the latest binary I could and try again.

Shawn

new topic     » goto parent     » topic index » view message » categorize

7. Re: Ver. 4.0 Debugging

SDPringle said...

Bernie,

Thank you for all of your effort in telling us and running all of the tests. But the problem seems to be with your build environment rather than the source code if you can reproduce it with your binary on more than one machine but we cannot. If you get the binaries from euphoria.blogdns.com or jeremy.cowgar.com/eubins and try with those and you still have a problem post it here.

If I were you I would get the latest binary I could and try again.

Shawn

Shawn:
Are there any binaries on your site built for WIN98 ?
If there are not any could you please build some for me.
I will try running them and let you know what happens.
Thank you very much.

new topic     » goto parent     » topic index » view message » categorize

8. Re: Ver. 4.0 Debugging SOLVED


  • DO NOT COMPILE WIN98 binaries on a WIN2000 machine using
    OPEN WATCOM version " open-watcom-c-win32-1.6.exe " because
    they will cause intermittent memory exception errors and
    printf and sprinf will return wrong values. This may also
    happen when compiling WIN98 binaries on vista, I have no way of testing.

  • COMPILE WIN98 binaries on the following version which I have tested.
    OPEN WATCOM version " open-watcom-c-win32-1.7a.exe " on
    a WIN2000 machine on a WIN98 machine.

  • CONFIGURE.BAT is broken when used on WIN98. But it works on WIN2000.
 
If you use the following command on WIN98 : 
 
configure --eubin d:\euphoria\bin --managed-mem 
 
This what the config.wat will contain. 
 
-- == start of config.wat == -- 
 
# Configuration for Watcom  
EU3=1  
PWD=  
EU3=1  
PWD=  
EU3=1  
PWD=  
 
-- == end of config.wat == -- 
 
 
new topic     » goto parent     » topic index » view message » categorize

9. Re: Ver. 4.0 Debugging SOLVED

bernie said...
  • DO NOT COMPILE ... using "open-watcom-c-win32-1.6.exe"

It was assumed that the latest Open Watcom would be being used. Not sure why you were using an old version of it.

bernie said...
  • CONFIGURE.BAT is broken when used on WIN98. But it works on WIN2000.

Explain to me again why using Win98 is a good idea.

In any case, why don't you write a Win98 version of the configure.bat script and pass that to us.

new topic     » goto parent     » topic index » view message » categorize

10. Re: Ver. 4.0 Debugging SOLVED

bernie said...
  • DO NOT COMPILE ... using "open-watcom-c-win32-1.6.exe"
DerekParnell said...

It was assumed that the latest Open Watcom would be being used. Not sure why you were using an old version of it.

Why would anyone expect a compiler to have bugs in it ? Euphoria version 4.0 doesn't have any.

bernie said...
  • CONFIGURE.BAT is broken when used on WIN98. But it works on WIN2000.
DerekParnell said...

Explain to me again why using Win98 is a good idea. In any case, why don't you write a Win98 version of the configure.bat script and pass that to us.

I am working with Win98 for the users that would like to use Euphoria on Win98 and are not wealthy enough to upgrade to the latest MS OS.

Maybe we should do away with Window2000 and XP
Then we would only have to worry about Vista.

-- This is what I used as a configure.bat for WIN98 
-- These are the only parameters that I know what they are used for 
-- since I don't have any other information. 
--  
@echo off 
echo # Configuration for Watcom > config.wat 
echo EUBIN=D:\euphoria\bin >> config.wat 
echo MANAGED_MEM=1 >> config.wat 
if not exist transobj.wat copy transobj.dst transobj.wat 
if not exist intobj.wat copy intobj.dst intobj.wat 
if not exist backobj.wat copy backobj.dst backobj.wat 
if not exist dosobj.wat copy dosobj.dst dosobj.wat 
if not exist dosbkobj.wat copy dosbkobj.dst dosbkobj.wat 
if not exist dostrobj.wat copy dostrobj.dst dostrobj.wat 
 
new topic     » goto parent     » topic index » view message » categorize

11. Re: Ver. 4.0 Debugging

The new binaries are for Windows are meant to work with versions Windows 95 and later. The latest archive is eubins-win-1213.zip the unit tests html is 1213-win.html. If I build something that is not windows98 compatible I will name it eubins-winnt-1213.zip.

Shawn

new topic     » goto parent     » topic index » view message » categorize

12. Re: Ver. 4.0 Debugging SOLVED

bernie said...

Why would anyone expect a compiler to have bugs in it ? Euphoria version 4.0 doesn't have any.

Buggy compilers are the norm, not the exception to the rule.. I don't think I can name a single bug free compiler.

I am working with Win98 for the users that would like to use Euphoria on Win98 and are not wealthy enough to upgrade to the latest MS OS.

[/quote]

This is a fine goal. Bernie, I applaud you for doing your best on this.

A lot of users (such as myself) went to Linux and WINE, which sadly is better supported than W98 nowadays.

bernie said...

Maybe we should do away with Window2000 and XP
Then we would only have to worry about Vista.

I second this motion.

bernie said...
-- This is what I used as a configure.bat for WIN98 
-- These are the only parameters that I know what they are used for 
-- since I don't have any other information. 
--  
@echo off 
echo # Configuration for Watcom > config.wat 
echo EUBIN=D:\euphoria\bin >> config.wat 
echo MANAGED_MEM=1 >> config.wat 
if not exist transobj.wat copy transobj.dst transobj.wat 
if not exist intobj.wat copy intobj.dst intobj.wat 
if not exist backobj.wat copy backobj.dst backobj.wat 
if not exist dosobj.wat copy dosobj.dst dosobj.wat 
if not exist dosbkobj.wat copy dosbkobj.dst dosbkobj.wat 
if not exist dostrobj.wat copy dostrobj.dst dostrobj.wat 
 

I'm not familiar with the EUBIN parameter. I do know about DEBUG and INT_CODES, as well as the (obsolete) EU3/PWD parameters. (You should never need INT_CODES, and DEBUG only makes sense if you are debugging C code of the interpreter or runtime library.)

new topic     » goto parent     » topic index » view message » categorize

Search



Quick Links

User menu

Not signed in.

Misc Menu