OpenEuphoria

RE: EuGrid - was Runs w/I FOR JONAS

• Posted by Jonas Temple <jktemple at yhti.net> Dec 03, 2002
• 892 views

Derek/Phil,

No innuendo on my part either.  My statement of "tk_mem does a peek or 
poke against that memory safe.e shows an error, which is not true" has 
to deal with the fact that safe.e will show a read/write error if the 
memory was not allocated directly with allocate().  In tk_mem since 
memory is allocated through Windows functions and then peeked/poked with 
Euphoria routines safe.e thinks it's an error, which in actuality it is 
not. 

My latest angle on finding the bug has been to open a text file and 
write something to it at certain points in the program.  I was thinking 
that since puts(1,"test") sends the output to the screen that if I could 
avoid screen i/o it would still bomb.  It proved me wrong!  Adding code 
to open/write/close a text file got the program working.  Maybe I should 
just leave it in there :)

Jonas

Phil Russell wrote:
> Hi Derek,
> 
> I intended no innuendo, honest! I was working from Jonas' statement that
> 
> <snip>
> And yes, I did try using safe.e in place of machine.e. Of course the 
> problem you run into when using win32lib is that tk_mem uses the Windows 
> 
> 
> memory management routines so when a routine in tk_mem does a peek or 
> poke against that memory safe.e shows an error, which is not true. My 
> solution was to create a version of tk_mem that uses Euphoria's allocate 
> 
> 
> and free routines instead (I also had to use register_block in other 
> places in my code as well). Of course the bound program ran fine 
> then...
> </snip>
> 
> Which, although not in anyway conclusive, suggests to me that this is, 
> at heart, a memory allocation problem of some sort.  What I am 
> suggesting is to see whether or not this problem is arising as some sort 
> 
> of interaction *between* tk_mem and EuGrid in bound programs. 
> 
> Memory allocation is a fairly minor part of EuGrid and it seems to be to 
> 
> be worth trying a modification to a) see if it fixes the problem and b) 
> stop Jonas from having to use a modified version of a standard Eu lib.
> 
> Regards,
> 
> Phil