Hello Mario!

Replying to a message of Mario Semo to Erik Huelsmann:

 EH>> I did not install an exception handler for this secondary 
 EH>> thread, but the compiler's runtime-library does not seem to 
 EH>> install one for the main thread and does not seem to have 
 EH>> any problems in stack-expansion.... 
 MS> the main thread's stack is preallocated.
Ok, that explains some of it...

 MS> I use 1MB of not preallocated stack (allocation via guard page
 MS> handling) for all threads != main thread.  (without problems, using VACPP)
Ok, I thought it should be possible, but my compiler, as far as I now know,
does not install a handler for guard-page handling. It leaves it to the
system. (Stackhandling is not affected as I understood from you, since the
stack for thread 1 is commited.)

 EH>> I could write an exception-handler that uses the 
 EH>> guardpage-system to expand the stack, but the operating 
 EH>> system is supposed to be able to do that itself...
 MS> thats what i expect too.

 MS> maybe you are NOT touching all bytes on the stack? (if i turn up
 MS> stack-check, it could cause a problem here).
This could be a problem, but not as far as I know, since this is on the
call to a DLL, just pushing some parameters onto the stack (or within the
DLL, but I can't verify that).

 MS>  WARPED!, Mario

As I understood from Peter and you is that thread 1 stack is precommited, I
decided to follow the same policy for other threads for now.

There is another problem now: PMERR_ATOM_NOT_FOUND on WinRegisterClass.

AHAB is a previously succesfully created anchorblock.
cs is a variable of type CSTRING (a c-compatible string-type)
u is of type ULONG
callme is just there to verify the result of WinRegisterClass.
{at}ObjectWinProc is a pointer to a window-procedure for my self-defined object window.
WinName is a var of type CSTRING too, containing 'Thrd444Obj10'
The code is the following (in pascal):

 cs:=WinName;
 u:=WinGetLastError(AHAB);
 if not bool(WinRegisterClass(AHAB,cs,{at}ObjectWinProc,0,4)) then
    callme;
 u:=WinGetLastError(AHAB);

On the second call to WinGetLastError the return value is 0x81017
(errorlevel: unrecoverable, error PMERR_ATOM_NOT_FOUND). This returnvalue
actually surprises me, since I did not create, nor did I use an atomtable.

How can it return this value?

Bye, Erik!
[TeamOS/2 NL]
preferred personal reactions through e-mail
[internet: ErikH{at}hcc-gron.idn.nl]

--- FleetStreet 1.14 NR