Wine release 981108
Alexandre Julliard
developers only release. There are many bugs and many unimplemented API
features. Most applications still do not work correctly.
Patches should be submitted to "jull...@lrc.epfl.ch". Please don't
forget to include a ChangeLog entry.
WHAT'S NEW with Wine-981108: (see ChangeLog for details)
- More common controls.
- Better separation of X11-dependent code.
- More DOS support.
- Winsock IPX support.
- Lots of bug fixes.
See the README file in the distribution for installation instructions.
Because of lags created by using mirror, this message may reach you before
the release is available at the ftp sites. The sources will be available
from the following locations:
ftp://sunsite.unc.edu/pub/Linux/ALPHA/wine/development/Wine-981108.tar.gz
ftp://tsx-11.mit.edu/pub/linux/ALPHA/Wine/development/Wine-981108.tar.gz
ftp://ftp.infomagic.com/pub/mirrors/linux/sunsite/ALPHA/wine/development/Wine-981108.tar.gz
ftp://ftp.progsoc.uts.edu.au/pub/Wine/development/Wine-981108.tar.gz
It should also be available from any site that mirrors tsx-11 or sunsite.
You can also get the current source directly from the CVS tree. Check
http://www.winehq.com/cvs.html for details.
If you submitted a patch, please check to make sure it has been
included in the new release.
If you want to get the new releases faster, you can subscribe to the
wine-patches mailing list by sending a mail containing 'subscribe
wine-patches your_address' to majo...@tiger.informatik.hu-berlin.de.
Wine is available thanks to the work of many people. See the file
AUTHORS in the distribution for the complete list.
--
Alexandre Julliard
jull...@lrc.epfl.ch
robert w hall
<jull...@lrcsun15.epfl.ch> writes
cvs version - Turnpike OK, but MathCad 7 still dies thus:-
(full post last week)
X Error of failed request: BadWindow (invalid Window parameter)
Major opcode of failed request: 10 (X_UnmapWindow)
Resource id in failed request: 0x28005c3
Serial number of failed request: 8869
Current serial number in output stream: 8870
I gather 1 or 2 other progs also go out this way
--
robert w hall
Tom Bech
Borland Delphi 3 C/S now not only loads, but also
displays something in the component tabs. So, now it
is actually possible to create forms and add various
components. I created a small "hello world"
program with a few edit boxes and buttons,
and to my suprise it compiled with no problem. Trying to
run it, however, crashed wine. Here's the bt:
[snip]
err:resource:FindResourceEx32W 0x41b90000(teedb30.dpl) "#08db"("#0006") not found!
err:resource:FindResourceEx32W 0x41f10000(tee30.dpl) "#08db"("#0006") not found!
err:resource:FindResourceEx32W 0x50330000(dclqrt30.dpl) "#08db"("#0006") not found!
fixme:ddeml:DdeNameService32 (0,49794,0,2): stub
fixme:ddeml:DdeNameService32 (0,49794,0,1): stub
fixme:ddeml:DdeNameService32 (0,49794,0,2): stub
fixme:ddeml:DdeNameService32 (0,49794,0,1): stub
fixme:module:CreateProcess32A (C:\PROG~BWX\BORLAND\DELP~45T\BIN\Project1.exe,...): bInheritHandles ignored
fixme:module:CreateProcess32A (C:\PROG~BWX\BORLAND\DELP~45T\BIN\Project1.exe,...): DEBUG_PROCESS ignored
fixme:module:CreateProcess32A (C:\PROG~BWX\BORLAND\DELP~45T\BIN\Project1.exe,...): DEBUG_ONLY_THIS_PROCESS ignored
fixme:module:CreateProcess32A (C:\PROG~BWX\BORLAND\DELP~45T\BIN\Project1.exe,...): CREATE_NEW_CONSOLE ignored
fixme:module:CreateProcess32A (C:\PROG~BWX\BORLAND\DELP~45T\BIN\Project1.exe,...): NORMAL_PRIORITY_CLASS ignored
fixme:module:CreateProcess32A (C:\PROG~BWX\BORLAND\DELP~45T\BIN\Project1.exe,...): STARTF_FORCEOFFFEEDBACK ignored
Unexpected Windows program segfault - opcode = 8
Page fault in 32-bit code (0x080c18fc).
Fault address is 0x85a294b0
Loading symbols: wine /usr/X11R6/lib/libSM.so.6 /usr/X11R6/lib/libICE.so.6
/usr/X11R6/lib/libXpm.so.4 /usr/X11R6/lib/libXext.so.6
/usr/X11R6/lib/libX11.so.6 /lib/libdl.so.2 /lib/libm.so.6 /lib/libc.so.6
/lib/ld-linux.so.2 /lib/libnss_files.so.1
KEYBOARD COMMCTRL USER GDI KERNEL WINEPS WPROCS DISPLAY SYSTEM
dfwdbk32.dll rlink32.dll ibevnt30.dpl dss30.dpl dcldss30.dpl teeui30.dpl
tee30.dpl teedb30.dpl dcltee30.dpl qrpt30.dpl dclqrt30.dpl dclocx30.dpl
vclsmp30.dpl dclsmp30.dpl dclmid30.dpl WSOCK32 inet30.dpl inetdb30.dpl
dclisp30.dpl dclnet30.dpl dcl31w30.dpl DbX30.dpl vcldbx30.dpl vcldb30.dpl
dcldb30.dpl dclstd30.dpl exptdemo.dll stdvcs32.dll dfwkbd.dll DFWEDIT.DLL
OLEPRO32.DLL dcc.DLL SHELL32 WINMM vclx30.dpl VERSION delphimm.dll
WINSPOOL MSVCRT20.dll oledlg.dll COMDLG32 COMCTL32.dll MPR ole32.dll
OLEAUT32.dll ADVAPI32 vcl30.dpl DELPHI32.EXE USER32 GDI32 KERNEL32
DeferredDebug for: OLEAUT32.dll ole32.dllUnable to open .DBG file ole32.dbg
COMCTL32.dll OLEPRO32.DLL
In 32 bit mode.
Register dump:
CS:0023 SS:002b DS:002b ES:002b FS:07f7 GS:0000
EIP:080c18fc ESP:435dfc9c EBP:435dfcb8 EFLAGS:00010246( R- 00 I Z- -P1 )
EAX:42f294b0 EBX:42b00000 ECX:42b00000 EDX:00000000
ESI:42f2d000 EDI:435f0930
Stack dump:
0x435dfc9c (dfwkbd.dll..reloc+0x50fc9c): 4029a0ec 435f0928 40299d10 42d90000 42b00000 080c17be 435f0928 435dfd68
0x435dfcbc (dfwkbd.dll..reloc+0x50fcbc): 080c17df 4029a0ec 435dfd9c 430e37a8 00000000 0000080f 435dfce0 4029a628
0x435dfcdc (dfwkbd.dll..reloc+0x50fcdc): 40288764 00003488 04782569 505c3a43 7e474f52 5c585742 4c524f42 5c444e41
0x435dfcfc (dfwkbd.dll..reloc+0x50fcfc):
Backtrace:
=>0 0x080c18fc (PE_InitTls+0x6c [pe_image.c:967])
1 0x080c17df (PE_CreateProcess+0x13b(name=0x435dfd9c, cmd_line=0x430e37a8, env=0x0, startup=0x430e319c, info=0x430e31e0) [pe_image.c:891])
2 0x080bf2c0 (CreateProcess32A+0x780(lpApplicationName=0x430e3760, lpCommandLine=0x430e37a8, lpProcessAttributes=0x0, lpThreadAttributes=0x0, bInheritHandles=0x1, dwCreationFlags=0x33, lpEnvironment=0x0, lpCurrentDirectory=0x0, lpStartupInfo=0x430e319c, lpProcessInfo=0x430e31e0) [module.c:680])
3 0x42dc4d69 (dfwdbk32.dll._DBKINITPROC+0x3cad)
4 0x42dc7c7b (dfwdbk32.dll._DBKINITPROC+0x6bbf)
5 0x42dda7c1 (dfwdbk32.dll.@__unlockDebuggerData$qv+0x45d5)
6 0x0816d163 (THREAD_Start+0x4f(thdb=0x402881b8) [thread.c:369])
7 0x0816c717 (SYSDEPS_StartThread+0x1f(thdb=0x402881b8) [sysdeps.c:86])
8 0x4016ae6e (thread_start+0x4 [sysctl.c])
0x080c18fc (PE_InitTls+0x6c [pe_image.c:967]): movl %edx,0x0(%ebx,%eax,1)
Wine-dbg>
I can provide more specific -debugmsg +whatever if anybody's interested.
Tom
--
Ove Kaaven
I get a similar crash in Eudora 3.01 16-bit whenever a dialog box is
closed (like that "You have new mail" message box). I think it was
X_PutImage before, not sure, at least now it's X_UnmapWindow for me
too.
Another issue: I also hate this resource leak that makes my 16-bit
apps run out of resources and crash after a while (if I download about
50-100 emails in Eudora, the graphics suddenly vanishes and then it
hangs if there's no progress bar... if the progress bar is shown, this
happens after downloading about 10-20 emails, I think).
gerard patel
>Hi,
>
>Borland Delphi 3 C/S now not only loads, but also
>displays something in the component tabs. So, now it
>is actually possible to create forms and add various
>components. I created a small "hello world"
>program with a few edit boxes and buttons,
>and to my suprise it compiled with no problem. Trying to
>run it, however, crashed wine. Here's the bt:
>
>[snip]
I think that you tried to run it in the IDE, did you ?
I fear that's impossible for now. Win32 debugging Api is badly
documented and the Wine emulation is not complete.
Furthermore some exception handling is disabled for now
so the wine debugger can work. Try to run outside of the
Ide. I have already loaded without great problems small
apps written in C++Builder, so it should work with
Delphi 3 programs.
Gerard
Marcus Meissner
...
>Backtrace:
>=>0 0x080c18fc (PE_InitTls+0x6c [pe_image.c:967])
>
>Wine-dbg>
>
>I can provide more specific -debugmsg +whatever if anybody's interested.
No need.
The problem is that PE_InitTls relocates the pointers in the directory
even though they already have been relocated (by do_relocations).
I am not sure why the last one changed the function, here is a
fix.
Ciao, Marcus
Index: loader/pe_image.c
===================================================================
RCS file: /home/wine/wine/loader/pe_image.c,v
retrieving revision 1.8
diff -u -r1.8 pe_image.c
--- pe_image.c 1998/11/07 12:35:50 1.8
+++ pe_image.c 1998/11/08 15:48:48
@@ -939,6 +939,13 @@
}
}
+/************************************************************************
+ * PE_InitTls (internal)
+ *
+ * If included, initialises the thread local storages of modules.
+ * Pointers in those structs are not RVAs but real pointers which have been
+ * relocated by do_relocations() already.
+ */
void PE_InitTls(THDB *thdb)
{
WINE_MODREF *wm;
@@ -964,27 +971,21 @@
if (!(pem->flags & PE_MODREF_TLS_ALLOCED)) {
pem->tlsindex = THREAD_TlsAlloc(thdb);
- *(LPDWORD)AdjustPtr(pdir->AddressOfIndex,delta)
- =pem->tlsindex;
+ *pdir->AddressOfIndex=pem->tlsindex;
}
pem->flags |= PE_MODREF_TLS_ALLOCED;
datasize= pdir->EndAddressOfRawData-pdir->StartAddressOfRawData;
size = datasize + pdir->SizeOfZeroFill;
mem=VirtualAlloc(0,size,MEM_RESERVE|MEM_COMMIT,PAGE_READWRITE);
- memcpy(mem,
- AdjustPtr(pdir->StartAddressOfRawData,delta),
- datasize);
-
- /* don't use TlsSetValue, we are in the wrong thread */
+ memcpy(mem,(LPVOID)pdir->StartAddressOfRawData,datasize);
if (pdir->AddressOfCallBacks) {
PIMAGE_TLS_CALLBACK *cbs =
- (PIMAGE_TLS_CALLBACK *)
- AdjustPtr(pdir->AddressOfCallBacks, delta);
+ (PIMAGE_TLS_CALLBACK *)pdir->AddressOfCallBacks;
- if (*cbs) {
+ if (*cbs)
FIXME(win32, "TLS Callbacks aren't going to be called\n");
- }
}
+ /* Don't use TlsSetValue, we are in the wrong thread */
thdb->tls_array[pem->tlsindex] = mem;
}
}
Patrik Stridvall
<9Bc6BBA4...@n-cantrell.demon.co.uk>...
>In article <1su30a3...@lrcsun15.epfl.ch>, Alexandre Julliard
><jull...@lrcsun15.epfl.ch> writes
>cvs version - Turnpike OK, but MathCad 7 still dies thus:-
>(full post last week)
>X Error of failed request: BadWindow (invalid Window parameter)
> Major opcode of failed request: 10 (X_UnmapWindow)
> Resource id in failed request: 0x28005c3
> Serial number of failed request: 8869
> Current serial number in output stream: 8870
>
>I gather 1 or 2 other progs also go out this way
Solved in my unpublished patch moving X11 dependent thing to x11drv.
I didn't want to publish it before Wine release 981108
since it doesn't add something and reorganizes some things.
Unfortunaly I have added some new code (non working),
so I can't send you a diff yet.
But I will explain what I think solved the problem.
In file windows/event.c in function EVENT_DestroyWindow
add pWnd->window = None at the end of the function;
If it works please don't post a patch of the solution
in wine-patches since it will confuse Alexandre
when he hopefully adds my patch on saturday.
EVENT_DestroyWindow will be in a different file (windows/x11drv/event.c).
Please post it in the newsgroup instead.
Uwe Bonnes
: <jull...@lrcsun15.epfl.ch> writes
: cvs version - Turnpike OK, but MathCad 7 still dies thus:-
: (full post last week)
: X Error of failed request: BadWindow (invalid Window parameter)
: Major opcode of failed request: 10 (X_UnmapWindow)
: Resource id in failed request: 0x28005c3
: Serial number of failed request: 8869
: Current serial number in output stream: 8870
: I gather 1 or 2 other progs also go out this way
Hallo,
this is with
./wine -debugmsg +relay,+snoop,+x11 -sync /dosd/cae/pasic71/spde/spde.exe
Call USER32.305: GetWindowLongA(000016d4,fffffffc) ret=00672f65 fs=022f
Ret USER32.305: GetWindowLongA() retval=006726fa ret=00672f65 fs=022f
Call KERNEL32.693: TlsGetValue(00000002) ret=00690efa fs=022f
Ret KERNEL32.693: TlsGetValue() retval=40a22740 ret=00690efa fs=022f
Call USER32.18: CallWindowProcA(081aeed8,000016d4,00000082,00000000,00000000)
ret=00673062 fs=022f
trace:relay:WINPROC_CallWndProc32 (wndproc=0x81aeed8,hwnd=000016d4,
msg=WM_NCDESTROY,wp=00000000,lp=00000000)
trace:relay:WINPROC_CallWndProc32 (wndproc=0x6756d7,hwnd=000016d4,
msg=WM_NCDESTROY,wp=00000000,lp=00000000)
Ret USER32.18: CallWindowProcA() retval=00000000 ret=00673062 fs=022f
Call USER32.305: GetWindowLongA(000016d4,fffffffc) ret=00672f78 fs=022f
Ret USER32.305: GetWindowLongA() retval=006726fa ret=00672f78 fs=022f
Call USER32.517: SetWindowLongA(000016d4,fffffffc,081aeed8) ret=00672f97
fs=022f
Ret USER32.517: SetWindowLongA() retval=006726fa ret=00672f97 fs=022f
Call KERNEL32.693: TlsGetValue(00000002) ret=00690efa fs=022f
Ret KERNEL32.693: TlsGetValue() retval=40a22740 ret=00690efa fs=022f
Call KERNEL32.195: EnterCriticalSection(0075f1a8) ret=00690fbc fs=022f
Ret KERNEL32.195: EnterCriticalSection() retval=9e8dd147 ret=00690fbc fs=022f
Call KERNEL32.195: EnterCriticalSection(0075f1a8) ret=00690bb5 fs=022f
Ret KERNEL32.195: EnterCriticalSection() retval=9e8dd147 ret=00690bb5 fs=022f
Call KERNEL32.494: LeaveCriticalSection(0075f1a8) ret=00690be7 fs=022f
Ret KERNEL32.494: LeaveCriticalSection() retval=00000000 ret=00690be7 fs=022f
Call KERNEL32.494: LeaveCriticalSection(0075f1a8) ret=00691004 fs=022f
Ret KERNEL32.494: LeaveCriticalSection() retval=ffffffff ret=00691004 fs=022f
Call KERNEL32.693: TlsGetValue(00000002) ret=00690efa fs=022f
Ret KERNEL32.693: TlsGetValue() retval=40a22740 ret=00690efa fs=022f
trace:x11:TSXDeleteContext Call XDeleteContext
trace:x11:TSXDeleteContext Ret XDeleteContext
trace:x11:TSXDestroyWindow Call XDestroyWindow
trace:x11:TSXDestroyWindow Ret XDestroyWindow
trace:x11:TSXCheckWindowEvent Call XCheckWindowEvent
trace:x11:TSXCheckWindowEvent Ret XCheckWindowEvent
Ret USER32.18: CallWindowProcA() retval=00000000 ret=00673062 fs=022f
trace:x11:TSXUnmapWindow Call XUnmapWindow
X Error of failed request: BadWindow (invalid Window parameter)
Major opcode of failed request: 10 (X_UnmapWindow)
Resource id in failed request: 0x34008b3
Serial number of failed request: 21496
Current serial number in output stream: 21497
--
Uwe Bonnes b...@elektron.ikp.physik.tu-darmstadt.de
Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------
David Cuthbert
> I think that you tried to run it in the IDE, did you ?
> I fear that's impossible for now. Win32 debugging Api is badly
> documented
Hmm? I've been successfully using the Win32 debugging API for well over a
year now with only the Microsoft documentation. What aspect of it is
"badly documented"?
The only thing I haven't been able to do is single-step through individual
CPU instructions, and that's more because I haven't bothered to dig
through the i386 documentation enough to set things up appropriately.
Is there some aspect that I can provide insight on?
Huw D M Davies
Ove Kaaven <ov...@arcticnet.no> wrote:
>
>Another issue: I also hate this resource leak that makes my 16-bit
>apps run out of resources and crash after a while (if I download about
>50-100 emails in Eudora, the graphics suddenly vanishes and then it
>hangs if there's no progress bar... if the progress bar is shown, this
>happens after downloading about 10-20 emails, I think).
Run with -debugmsg +local and look at the heap dump when the heap becomes full.
You can either work out which type of object is causing the problem by looking
at the size of the blocks or add a bit of code that prints the magic value of
each object - assuming it's the gdi heap that's exhausted. Then you just have
to find where the leak is.
Huw.
--
Dr. Huw D M Davies | Clarendon Laboratory
h.da...@physics.ox.ac.uk | Parks Road
Tel: +44 1865 272390 | Oxford OX1 3PU
Fax: +44 1865 272400 | UK
robert w hall
<p...@leissner.se> writes
>><jull...@lrcsun15.epfl.ch> writes
>>cvs version - Turnpike OK, but MathCad 7 still dies thus:-
>>(full post last week)
>>X Error of failed request: BadWindow (invalid Window parameter)
>>
>
>
>In file windows/event.c in function EVENT_DestroyWindow
>add pWnd->window = None at the end of the function;
>
>
Patrik -thanks, that appears to cure the immediate problem. Bob
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
--
robert w hall