Hi Przemek,
Przemyslaw Czerpak ha scritto:
On Sun, 12 Oct 2008, Francesco Saverio Giudice wrote:
Hi Francesco,
I'm using last SVN
2008-10-11 21:49 UTC+0200 Viktor Szakats (harbour.01 syenar hu)
thank you for your efforts.
Tell me if I can help in any way.
Let's try to reduce the problem.
In tests/mt/ there is set of MT test programs.
Can you replicate the problem with each of them?
If not then which ones fails?
How exactly does the screen output look?
Which GT you are using?
Can someone replicate it in non VISTA Windows or it's only local
to this OS?
For tests you can easy disable GT locking.
It's enough to comment in rtl/hbgtcore.c line 119:
pGT->pMutex = hb_threadMutexCreate( TRUE );
Mutex will not be allocated what efectively disable any GT locking.
Does it help to resolve the problem?
I looked at the code and it seems to be OK.
In Linux it works without any problems so it's hard to me guess
were to look for possible problems when I do not know where exactly
the problem is. It could be sth stupid like problems with synchronization
mechanism final cleanup code strictly oriented to given OS.
BTW Does the problem happen in each application linked with
hbvmmt even if it does not create new threads?
I have rebuilt with last SVN :
2008-10-11 16:59 UTC+0200 Viktor Szakats (harbour.01 syenar hu)
and tested all files in tests\mt
and all show correctly on the screen, but when program ends console is
locked.
Just to be clear, using as sample mttest01 this what I see:
J:\cvs\harbourSVN_trunk\tests\mt>mttest01
Harbour 1.1.0dev (Rev. 9615)
join: .T.
result: Hello World!!! Hello World!!! Hello World!!!
static var type: B
12345.678
12346.678
with cursor blinking at right of last 8 (exactly behavior with all others).
I have to kill mttest01.exe from task manager and console returns to prompt.
I have also commented out line 119 in rtl\hbgtcore.c as you suggested,
but the result is the same.
Best regards
Francesco
_______________________________________________
Harbour mailing list
Harbour@harbour-project.org
http://lists.harbour-project.org/mailman/listinfo/harbour
