November 2004 - Wine-Devel - list.winehq.org

Is it time to delete wineinstall?
by Dan Kegel Nov. 18, 2004

Nov. 18, 2004

Is wineinstall needed anymore? README seems to indicate you can install Wine just fine without it. - Dan -- Trying to get a job as a c++ developer? See http://kegel.com/academy/getting-hired.html

11 21

Re: dialogex resource loading on sparc
by Dmitry Timoshkov Nov. 18, 2004

Nov. 18, 2004

"Eric Frias" <efrias(a)syncad.com> wrote: > --- windows/dialog.c (.../vendor/wine/current) (revision 29312) > +++ windows/dialog.c (.../branches/wine/2004.11.16-cvs-update) (revision 29312) > @@ -378,8 +378,10 @@ > { > const WORD *p = (const WORD *)template; > > - result->style = GET_DWORD(p); p += 2; > - if (result->style == 0xffff0001) /* DIALOGEX resource */ > + WORD signature = GET_WORD(p); p++; > + WORD dlgver = GET_WORD(p); p++; This won't compile with older C compilers, you are mixing data declarations with code. -- Dmitry.

2 1

replace a wrong fixme in dlls/winmm/playsound.c
by Rémi Assailly Nov. 18, 2004

Nov. 18, 2004

I tested a software ( a monopoly game ) and the error was that a sound was still playing so I think that is not a wine bug. Before submitting this patch to wine-patches I would like to know what do you think about it, what should be WARN, FIXME, if choosing ERR to all is correct or if some cases are useless.

1 0

Re: Is it time to delete wineinstall?
by Ivan Leo Puoti Nov. 17, 2004

Nov. 17, 2004

> Besides, on the wine-users alias, new users who > install RPMs have no idea there is a config file > because the RPMs don't create it for them. The config file is meant to go ASAP, so that is a good thing. Most users use RPMs and are quite happy with them, and it causes less support hassle. > It may be a redundant feature, but "configure/ make/ > make install" is definately not a replacement for > "ease of use." ;) Then as other have said wineinstall should just contain ./configure && make && su -c 'make install' Ivan. ____________________________________________________________ Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazione. Abbonati subito su http://www.libero.it

2 1

profiling GTA4. question about polling
by Brian Gunlogson Nov. 17, 2004

Nov. 17, 2004

GTA4 (Grand Theft Auto: Vice City) runs on Wine with no additional tinkering! Albeit horribly slow. A profile of GTA shows that only one D3D function (drawStridedSlow) needs major optimization. According to the oprofile results, polling seems to be the major bottleneck. (the top of a flat profile generated with oprofile) samples % image name app name symbol name 2177620 14.0147 d3d8.dll.so wine-preloader drawStridedSlow 1681661 10.8228 vmlinux vmlinux do_pollfd 1025198 6.5980 vmlinux vmlinux pipe_poll 515461 3.3174 vmlinux vmlinux add_wait_queue 423342 2.7245 vmlinux vmlinux poll_freewait 423185 2.7235 vmlinux vmlinux __pollwait 354490 2.2814 vmlinux vmlinux fput 285473 1.8372 vmlinux vmlinux sys_poll I'd like to optimize this. Any hints? Brian G __________________________________ Do you Yahoo!? The all-new My Yahoo! - Get yours free! http://my.yahoo.com

1 0

wineprefixcreate hang
by Olive Esseret Nov. 17, 2004

Nov. 17, 2004

I do not know if this question is appropriate here; so please excuse me if it is not (I have not had any answer elsewhere) On Mdk10.1 the following command (which is called by wineprefixcreate and so also by wineinstall) never terminate. wine rundll32 setupapi.dll,InstallHinfSection DefaultInstall 128 wine.inf Although this happens most of the time, it does not appears every time and did not appeard on Mdk10.0. If it is not possible to make it works, anyone has an idea on how to do the things by hands? Olive

2 3

Supported Apps
by Tom Nov. 17, 2004

Nov. 17, 2004

Hello All, Would anyone reject to the idea of changing "Tucows top" to "License" and listing the licence on the supported page? I was looking over this page and the Tucows rankings have changed, and some of the apps that we have listed are no longer in the top 100 ranking at Tucows. Tom

1 0

incomplete relay when accessed from winedbg
by Bill Medland Nov. 17, 2004

Nov. 17, 2004

Guys Any chance of help on this? I am trying to figure out why some numbers are misformatted through Crystal Reports ToText function. (e.g. 6..0 instead of 599.95) I tried to turn +relay on from within winedbg so that I could see just what functions got called. All that relay gave me was "window proc", "hook proc" and wineps calls; no ntdll or anything else. Any idea why it isn't giving me everything? (wine 20040914) -- Bill Medland mailto:billmedland(a)mercuryspeed.com http://webhome.idirect.com/~kbmed

3 2

access PE headers in winelib shared library
by Jia L Wu Nov. 17, 2004

Nov. 17, 2004

Hi, Can somebody provide some example or information on how to access PE headers(e.g. PIMAGE_DOS_HEADER, PIMAGE_NT_HEADER) wrapped in winelib shared library(e.g. winelib dll)? Thanks a lot! Wu

3 2

appdb linux links
by James Hawkins Nov. 17, 2004

Nov. 17, 2004

One of my conceptions of wine is as a stepping stone for users switching over to linux from windows that still only feel comfortable with windows apps or those that don't know what linux apps are available. What do you think about adding a 'Similar Linux Application' link or section to each windows app where applicable that would point them in the direction of the linux equivalent of the windows app and thus help with the transition process? They might not necessarily use the app, but at least they'll be more aware. What are some ideas about this? -- James Hawkins

4 7

wineprefixcreate (and so wineinstall) hang
by Olive Esseret Nov. 17, 2004

Nov. 17, 2004

I do not know if this question is appropriate here; so please excuse me if it is not (I have not had any answer elsewhere) On Mdk10.1 and with wine-20041019 the following command (which is called by wineprefixcreate and so also by wineinstall) never terminate. wine rundll32 setupapi.dll,InstallHinfSection DefaultInstall 128 wine.inf Although this happens most of the time, it does not appears every time and did not appeard on Mdk10.0 (although there does not seems to be essential differences between the two distributions) If it is not possible to make it works, anyone has an idea on how to do the things by hands? If you need more informations about my system configuration, please let me know (I have no idea what is appropriate...) Olive P.S. Please excuse if this message apperas twice; I had posted it two hours ago and it has still not appeared. __________________________________ Do you Yahoo!? The all-new My Yahoo! - Get yours free! http://my.yahoo.com

1 0

Winetest questions
by Dan Kegel Nov. 17, 2004

Nov. 17, 2004

1. Why are all the archived posts to wine-test-results empty? See http://www.winehq.org/hypermail/wine-tests-results/2004/11 Every message body is empty! What's up with that? 2. The URL http://test.winehq.org/data/ appears to contain digests of all winetest results received by the server, but this information is not documented anywhere I could find at www.winehq.org. Seems like a top level page is in order. 3. winetest.exe currently assumes you're running under Windows. See main.c line 577: (putenv ("WINETEST_PLATFORM=windows") || Wouldn't it be useful to allow Linux results to be computed and reported via winetest, too? Presumably it could detect that it's running on Linux somehow. (perhaps by running uname -a...) - Dan -- Trying to get a job as a c++ developer? See http://kegel.com/academy/getting-hired.html

3 5

Install OCX over wine
by Francisco Nauber Bernardo Gois Nov. 17, 2004

Nov. 17, 2004

I need install ocx file using wine.How i can do thist? Francisco Nauber Bernardo Gois Unidade de Relacionamento com Clientes Administração Financeira (SUNAF) Serviço Federal de Processamento de Dados (SERPRO) Fone: (85) 88443794

2 1

Could someone look at this simple code and explain why it doesnt work?
by luis lenders Nov. 16, 2004

Nov. 16, 2004

Hi, i found this very simple app which just pops up a window: #include <windows.h> LRESULT CALLBACK WndProc(HWND hwnd, UINT msg, WPARAM wparam, LPARAM lparam) { switch (msg) { case WM_DESTROY: PostQuitMessage(0); return 0; default: return DefWindowProc(hwnd, msg, wparam, lparam); } } int WINAPI WinMain(HINSTANCE instance, HINSTANCE prev, LPSTR cmdLine, int cmdShow) { WNDCLASSEX class; MSG msg; HWND window; /* this stuff is just boilerplate */ memset(&class, 0, sizeof(WNDCLASSEX)); class.cbSize = sizeof(WNDCLASSEX); class.hInstance = instance; class.lpszClassName = "MainWnd"; class.lpfnWndProc = WndProc; class.hbrBackground = (HBRUSH) COLOR_WINDOW + 1; if (!RegisterClassEx(&class)) { MessageBox(NULL, "Could not register window class", "", MB_OK | MB_ICONERROR); return 1; } window = CreateWindowEx(0, "MainWnd", "Main Window", WS_OVERLAPPEDWINDOW & ~WS_MAXIMIZEBOX, CW_USEDEFAULT, CW_USEDEFAULT, 200, 150, NULL, NULL, instance, NULL); ShowWindow(window, cmdShow); /* Every thread that creates a window must pump the message loop */ while (GetMessage(&msg, NULL, 0, 0)) { TranslateMessage(&msg); DispatchMessage(&msg); } return 0; } Now i tried to implement this in a dll (avicap32.dll) in capCreatecapturewindow like this: /*********************************************************************** * capCreateCaptureWindowW (AVICAP32.@) */ HWND VFWAPI capCreateCaptureWindowW(LPCWSTR lpszWindowName, DWORD dwStyle, INT x, INT y, INT nWidth, INT nHeight, HWND hWnd, INT nID) { HWND ret; HINSTANCE dummy; LRESULT CALLBACK WndProc(HWND hwnd, UINT msg, WPARAM wparam, LPARAM lparam) { switch (msg) { case WM_DESTROY: PostQuitMessage(0); return 0; default: return DefWindowProcW(hwnd, msg, wparam, lparam); } } int WINAPI Try(HINSTANCE instance, HINSTANCE prev, LPSTR cmdLine, int cmdShow) { WNDCLASSEXW class; MSG msg; HWND window; /* this stuff is just boilerplate */ memset(&class, 0, sizeof(WNDCLASSEXW)); class.cbSize = sizeof(WNDCLASSEXW); class.hInstance = instance; dummy=instance; class.lpszClassName = "MainWnd"; class.lpfnWndProc = WndProc; class.hbrBackground = (HBRUSH) COLOR_WINDOW + 1; if (!RegisterClassExW(&class)) { MessageBoxW(NULL, "Could not register window class", "", MB_OK | MB_ICONERROR); return 1; } window = CreateWindowExW(0, "MainWnd", "Main Window", WS_OVERLAPPEDWINDOW & ~WS_MAXIMIZEBOX, CW_USEDEFAULT, CW_USEDEFAULT, 200, 150, NULL, NULL, instance, NULL); ret=window; ShowWindow(window, cmdShow); /* Every thread that creates a window must pump the message loop */ while (GetMessageW(&msg, NULL, 0, 0)) { TranslateMessage(&msg); DispatchMessageW(&msg);} return 0; } Try(dummy,NULL,NULL,0); return ret; } Now when capcreatecapturewindow is called no window shows up at all. What did i do wrong? Greetings Robbert : ___________________________________________________________ Moving house? Beach bar in Thailand? New Wardrobe? Win £10k with Yahoo! Mail to make your dream a reality. Get Yahoo! Mail www.yahoo.co.uk/10k

2 1

GradientFill revisited
by Arjen Nienhuis Nov. 16, 2004

Nov. 16, 2004

I created a patch! Hooray! http://www.nienhuisbeheer.nl/~arjen/gradientfill.patch Does anyone have a test for this function. What is the best way to create a program to test these things? I've never written a windows program from scratch. Does anyone have a simple app, where I can insert some calls to GradientFill?

1 0

Re: Kernel 2.6.9 - Issue Found
by Dan Kegel Nov. 16, 2004

Nov. 16, 2004

I couldn't find the patch at either of the two links Jesse posted, but I did see something similar here: http://marc.theaimsgroup.com/?l=bk-commits-head&m=110048429304053&w=2 -- Trying to get a job as a c++ developer? See http://kegel.com/academy/getting-hired.html

3 3

Re: Kernel 2.6.9 - Issue Found
by Paul Rupe Nov. 16, 2004

Nov. 16, 2004

> It doesn't seem to fix the copy protection problem. So I'm wondering a few > things. Does wine need to be patched? This new patch seems to be related to > a chunk of code added in patch #2 mentioned in the original posting. I've > tried 2.6.10-rc2 with patch #1 reversed and patch 1+3 reversed. No difference > this time. No luck with Remedy either. I tried reversing all three patches and also tried 2.6.10-rc2. Same crash as before. -- Paul Rupe <prupe(a)myrealbox.com>

2 1

Developer Cheatsheet formatting...
by Dan Kegel Nov. 15, 2004

Nov. 15, 2004

http://www.winehq.com/site/developer-cheatsheet is 1.5 times as wide as my screen in Firefox and Mozilla. Seems like it needs a bit of adjustment... -- Trying to get a job as a c++ developer? See http://kegel.com/academy/getting-hired.html

4 5

WINELIB Use References
by Dagan Shai Nov. 15, 2004

Nov. 15, 2004

Hi, I'm new to WINE / WINELIB. We're considering using WINELIB in order to port a commercial telephony application (service node). I would appreciate some information from someone with experience porting commercial server-side applications with WINELIB regarding the problems that arise, resources used and recommendations. Anyone that can provide such information please contact me privately. Thanks, Shai Dagan

1 0

rsaenh.dll
by Michael Jung Nov. 15, 2004

Nov. 15, 2004

Hi, the attached patch incorporates code from the public domain library LibTomCrypt (http://libtomcrypt.org) into rsaenh.dll. This allows us to get rid of the compile and run-time dependencies on OpenSSL, and - as far as I understand it - might solve some license problems for ReactOS (http://www.reactos.com:8080/pipermail/ros-general/2004-February/000706.html) It would be nice to get some comments on what remains to be done before I submit the patch to wine-patches. I renamed the implossl.[ch] (OpenSSL glue logic) files to the more neutral implglue.[ch]. How to I get diff to produce a patch that reflects this change? Greetings, Michael

1 0

How to port c++ program to linux using winelib
by Jia L Wu Nov. 15, 2004

Nov. 15, 2004

Hi, I have a win32 program written in C++. I am trying to port it to linux using winelib. The software contanis one exe file and several dll files. I would like to keep this structure when moving them to linux. That is, I still would like to generate several corresponding shared libraries (instead of statically linking all files together) in llinux and they will be loaded at the runtime. However using winelib tools(winemaker, configure, make) for shared library generation, I get unresolved symbols error. It seems that I have to write a spec file for each dll to import(and export) functions provide by other dlls. However, since they all written in c++ and what need to be imported can either be class or class method, and parameters can be class either. SO I don't know how to handle this in spec file. Beside each dll file exports many functions or classes. It would be tedious to write a lengthy spec file manually. My question is that if it is possible that linking is not invoked during shared library generation under wine (but during the runtime)? How should I do it? Thanks. Wu

4 8

"Software Engineering with Wine" tutorial
by Dan Kegel Nov. 14, 2004

Nov. 14, 2004

I'm putting together a mini-lab course on "Software Engineering with Wine". So far, it only shows how to build wine and run the conformance test, and troubleshoot common problems (in particular, the ones that caused hundreds of test failures when I ran them a couple weeks ago). It's at http://kegel.com/wine/sweng/ Comments welcome. (In particular, are there any other test suite gotchas I should tell the students about?) - Dan -- Trying to get a job as a c++ developer? See http://kegel.com/academy/getting-hired.html

1 0

Re: Kernel 2.6.9
by Jerry Jenkins Nov. 14, 2004

Nov. 14, 2004

Paul Rupe wrote: > On Friday November 12 2004 11:48 am, Michael Stefaniuc wrote: > > >>Btw. does wine crash only with Remedy or does winemine crash too? > > > Most apps (winemine included) are ok. When I get home I'll try to find a > downloadable app that reproduces the problem consistently. > > Usually wine just hangs until I kill -9 everything, but this time I managed > to get a stack trace. It looks like a recursive loop: > =>1 0x7759b155 cxx_frame_handler+0x401 in msvcrt (0x77927f90) > 2 0x7759b5c3 in msvcrt (+0xb5c3) (0x77927fb4) > 3 0x77ee7be1 __wine_call_from_32_regs+0xb5 in ntdll (0x7792831c) > 4 0x7759ac79 __CxxFrameHandler+0x5 in msvcrt (0x77928358) > 5 0x77ec63b8 EXC_RtlRaiseException+0x18c in ntdll (0x779283ec) > 6 0x77edf32b raise_segv_exception+0x2f in ntdll (0x77928408) > 7 0x77ee7c50 __wine_call_from_32_restore_regs+0x0 in ntdll (0x779287d4) > ... > 90 0x77edf32b raise_segv_exception+0x2f in ntdll (0x7792f7d8) > Frames 2-7 repeat all the way down to 90. This is after a 'make clean' and > a full rebuild of Wine, btw. > > Does increasing stack size help? For example: Index: dlls/kernel/process.c =================================================================== RCS file: /var/cvs/wine/dlls/kernel/process.c,v retrieving revision 1.80 diff -u -r1.80 process.c --- dlls/kernel/process.c 30 Oct 2004 02:11:33 -0000 1.80 +++ dlls/kernel/process.c 15 Nov 2004 11:36:47 -0000 @@ -1183,7 +1183,7 @@ set_library_wargv( __wine_main_argv ); if (!build_command_line( __wine_main_wargv )) goto error; - stack_size = RtlImageNtHeader(peb->ImageBaseAddress)->OptionalHeader.SizeOfStackReserve; + stack_size = 10 * 1024 * 1024; /* allocate main thread stack */ if (!THREAD_InitStack( NtCurrentTeb(), stack_size )) goto error; __________________________________ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.com

1 0

Re: Last-chance deactivation of MDI Child before destruction
by Dmitry Timoshkov Nov. 14, 2004

Nov. 14, 2004

"Zach Gorman" <zach(a)archetypeauction.com> wrote: > Changelog: > If, during destruction of an MDI child frame, the application > re-activates the window that is about to be destroyed, de-activate > it one last time. > > Background: > This resolves an obscure problem caused by a difference in Windows > and Wine behavior. I have an MDI MFC app in which one of the MDI > child frames forces itself to stay at the back of the Z-order by > responding to MDI Activate messages, and re-activating the old > window (yes I know this is nasty). When Wine tries to switch the > active MDI child during the active child's destruction, this causes > the about-to-be-destroyed window to become active again. Windows > will issue one last WM_MDIACTIVATE to deactivate the window that > will be destroyed in this case, but Wine won't. This patch corrects > that difference in behavior. What happens if that additional WM_MDIACTIVATE won't deactivate an MDI child either, or an app doesn't allow deactivation for some reason? Anyway, please add another MDI message test case showing what Windows does in that case because your change is not obvious at all. -- Dmitry.

2 1

Kernel 2.6.9 - Issue Found
by Jesse Allen Nov. 14, 2004

Nov. 14, 2004

Hi, For the past two weeks I've been tracking an issue that I found with kernels greater than 2.6.8 / 2.6.8.1 and wine. Starting with kernel 2.6.9-rc1, Warcraft III copy protection would not work -- ie please insert disc. I found changes to the cdrom driver in rc1. Reversing them did not work. For a while I was lost to what could have changed. Then I had an idea: search for changes that could affect the way processes run. That led me to reverse changes related to ptrace, and I found a series of patches that break Warcraft III copy protection code. Patch 1: ptrace single-stepping fix Included in 2.6.9-rc1 By Davide Libenzi? http://linux.bkbits.net:8080/linux-2.6/cset(a)1.1803.1803.144.55 This change includes a note that it could break debuggers. Reversing this change would allow War3 to work in rc1, but kernels greater than rc1 would still not work. Patch 2: make single-step into signal delivery stop in handler Included in 2.6.9-rc2 By Roland McGrath http://linux.bkbits.net:8080/linux-2.6/cset(a)1.1832.59.196 The next patch may depend on patch #2 because it is a bug-fix based on 2. Patch 3: Fix PTRACE_CONT after single-step into signal delivery Also 2.6.9-rc2 By Roland McGrath http://linux.bkbits.net:8080/linux-2.6/cset(a)1.1847 Reversing patches 1, 2, and 3 in kernels 2.6.9-rc2 and greater (through 2.6.10-rc1) will allow War3 to work. I tried 1 & 3 without 2, but it failed leading me to believe that 1 & 2 are required. I didn't try 1 & 2 alone because I think reversing 2 would require 3 reversed. All the patch changes seem to be on ptrace single-stepping. However, there is at least one ptrace single-stepping patch that seems to make no difference at all: http://linux.bkbits.net:8080/linux-2.6/cset(a)1.1803.144.70 There are other single-stepping patches for different non-i386 architectures but I have no idea on these because I don't run the others. They are meaningless for wine anyways. Now that I've identified a problem, I'm curious where else it's cropping up. Paul Rupe, could you try reversing these patches to see if this fixes your crashing problem? And how about those people having trouble with debugging? Then I guess the next step is to find out if this is a wine bug or not. If it's not, then it may be a kernel bug and we should report it to the kernel devs. Jesse

4 5

"Crystallography programs under Wine"
by Dan Kegel Nov. 14, 2004

Nov. 14, 2004

I just ran into http://www.ccp14.ac.uk/solution/wine/ which lists about ten free Windows apps used for crystalography, and how to run them under Wine. I guess these deserve an appdb entry; they don't seem to have one yet. That page describes workarounds needed to run with Wine as of about three years ago. Things may have progressed since then. For instance, the "you must rename files to have lower case" workaround no longer seems neccessary for the INDX program. I wrote the maintainer about that, haven't heard back from him yet. - Dan -- Trying to get a job as a c++ developer? See http://kegel.com/academy/getting-hired.html

1 0

Re: taskmodal messagebox implementation, take 3
by Ivan Leo Puoti Nov. 14, 2004

Nov. 14, 2004

> No problem. I'll give a lengthy reply in the hope it can help other folks > dealing with this sort of thing too. Well, your code works, and I've made it into a patch. I've also removed what I beleive is a useless if check from MSGBOX_OnInit(). Due to my lame webmail client you code arrived with no indentation, so in the future please attach any code you send to me. If you think the attached patch is OK I'll forward it to wine-patches. Ivan. ____________________________________________________________ Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazione. Abbonati subito su http://www.libero.it

2 1

Re: [D3D] Some surfaces related fixes
by Christian Costa Nov. 14, 2004

Nov. 14, 2004

Mike Hearn wrote: >On Sun, 14 Nov 2004 13:40:15 +0000, Christian Costa wrote: > > >>+ if ((pDDSD->dwFlags & DDSD_LPSURFACE) && (pDDSD->lpSurface == NULL)) { >>+ /* Frank Herbert's Dune specifies a null pointer for the surface, ignore the LPSURFACE field */ >>+ FIXME("Null surface pointer specified, ignore it!\n"); >>+ pDDSD->dwFlags &= ~DDSD_LPSURFACE; >>+ } >>+ >> >> > >Apps passing in bogus data should normally be a trace or warn at most, >unles there is actually something to fix here it doesn't make sense to use >a fixme as it'll inevitably get users asking about it .... > >thanks -mike > > > > > Your're right, Mike. :-) I will resend a patch for that. Thanks. Christian

1 0

Re: taskmodal messagebox implementation, take 3
by Juan Lang Nov. 14, 2004

Nov. 14, 2004

Hi Ivan. --- Ivan Leo Puoti <puoti(a)inwind.it> wrote: > Please consider I'm no programming wiz and I'm learning as I go along. No problem. I'll give a lengthy reply in the hope it can help other folks dealing with this sort of thing too. > Now, I don't see how else I could get MSGBOX_EnumProc() to remember the > values > of those that now are static variables, apart from using the heap for > everything, that seems a bit inefficent as we know how much data we want > to > store, or global variables, do you think I should use global ones > instead? > If not suggestions are welcome. Globals are also not thread-safe. Generally for thread safety, I'd suggest the following: declare a structure that has all the data you want to modify in MSGBOX_EnumProc, something like: struct MsgBoxEnumProcParams { int pass; int counter1; int counter2; int heapcheck; HWND *handles; }; In MessageBoxIndirectW, create one of these and pass its address as the LPARAM value to EnumThreadWindows(). However, this ignores style issues. I have problems with names like "counter1" and "counter2"; they don't tell me what they mean. Also, I don't like the magic values 0, 1, 2, 3, 4 for the LPARAM value; at the least these should be symbolic constants. Pass 3 doesn't modify the parameters at all, so it really should get its own callback function. It looks like heapcheck's purpose is to see whether handles has in fact been allocated. You could instead make sure to set handles to NULL when it isn't allocated. Also, there's the possibility of a memory leak in pass 4. I'd suggest the caller should be responsible for memory allocation. With these suggestions, I've come up with the following changes. I haven't checked whether this code works (or even compiles), it's just my attempt at rewriting what you're trying to do: struct ThreadWindows { UINT numHandles; UINT handlesLen; HWND *handles; }; BOOL CALLBACK MSGBOX_CountDisabledThreadWindows(HWND hwnd, LPARAM lParam) { if (!lParam) return FALSE; if (!IsWindowEnabled(hwnd)) *(UINT *)lParam++; return TRUE; } BOOL CALLBACK MSGBOX_EnumDisabledThreadWindows(HWND hwnd, LPARAM lParam) { if (!lParam) return FALSE; if (!IsWindowEnabled(hwnd)) { struct ThreadWindows *threadWindows = (struct ThreadWindows *)lParam; if (threadWindows->numHandles < threadWindows->handlesLen) threadWindows->handles[threadWindows->numHandles++] = hwnd; } return TRUE; } INT WINAPI MessageBoxIndirectW( LPMSGBOXPARAMSW msgbox ) { LPVOID tmplate; HRSRC hRes; int ret; UINT i; struct ThreadWindows threadWindows = { 0, 0, 0 }; static const WCHAR msg_box_res_nameW[] = { 'M','S','G','B','O','X',0 }; if (!(hRes = FindResourceExW(user32_module, (LPWSTR)RT_DIALOG, msg_box_res_nameW, msgbox->dwLanguageId))) return 0; if (!(tmplate = (LPVOID)LoadResource(user32_module, hRes))) return 0; /*handle task modal message boxes*/ if ((msgbox->dwStyle & MB_TASKMODAL) && (msgbox->hwndOwner==NULL)) { EnumThreadWindows(GetCurrentThreadId(), MSGBOX_CountDisabledThreadWindows, (LPARAM)&threadWindows.numHandles); if (threadWindows.numHandles) { threadWindows.handles = (HWND *)HeapAlloc(GetProcessHeap(), 0, threadWindows.numHandles * sizeof(HWND)); if (threadWindows.handles) threadWindows.handlesLen = threadWindows.numHandles; threadWindows.numHandles = 0; } EnumThreadWindows(GetCurrentThreadId(), MSGBOX_EnumDisabledThreadWindows, (LPARAM)&threadWindows); for (i = 0; i < threadWindows.numHandles; i++) DIALOG_DisableOwner(threadWindows.handles[i]); } ret=DialogBoxIndirectParamW(msgbox->hInstance, tmplate, msgbox->hwndOwner, MSGBOX_DlgProc, (LPARAM)msgbox); if ((msgbox->dwStyle & MB_TASKMODAL) && (msgbox->hwndOwner==NULL)) { for (i = 0; i < threadWindows.numHandles; i++) DIALOG_EnableOwner(threadWindows.handles[i]); if (threadWindows.handles) HeapFree(GetProcessHeap(), 0, threadWindows.handles); } return ret; } The changes I've made: I've renamed counter1 and counter2 to numHandles and handlesLen. handlesLen is the allocated number of handles, while numHandles is the number of handles actually stored. (Your code already dealt with the possibility of their being different: well done!) I've also used handles as its own guard; if it's NULL, it's not allocated. Otherwise it is. Like I suggested, I pass a pointer as an LPARAM to the two EnumThreadWindows callback functions. It seems two calls to EnumThreadWindows were unnecessary, since we already knew the window handles, so I just iterate over the values in handles rather than calling EnumThreadWindows again. If I'm incorrect in my understanding, of course change it. And, the caller is always responsible for memory allocation and freeing. I think this is safer than having the callback function do it. > >Any reason you aren't using HeapAlloc instead? > As this heap is only used by MSGBOX_EnumProc(), is there any reason why > I should > use HeapAlloc? It's sometimes easier to debug memory problems with the Windows heaps than with the libc ones. --Juan __________________________________ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.com

1 0

winzip 9 doesn't run under wine :-(
by Dan Kegel Nov. 14, 2004

Nov. 14, 2004

While putting together a tutorial on how to contribute test cases to Wine (http://kegel.com/wine/sweng/) I noticed that the current version 9 of WinZip doesn't install under vanilla Wine; it aborts with the message WinZip was unable to locate a recent version of the Windows HTML help viewer, which is required to display WinZip help. ... Seems like a bit of a showstopper. What's the plan for addressing this? Is it a 0.9 feature? - Dan -- Trying to get a job as a c++ developer? See http://kegel.com/academy/getting-hired.html

3 8

ListView DrawText Flags
by Zach Gorman Nov. 14, 2004

Nov. 14, 2004

Dimitrie - You have a patch back in 2002 that cleans up the listview DrawText() code in comctl32. When calling DrawText from a single-line edit control, you don't pass DT_NOPREFIX. Is this intentional, or just an oversight? My testing on Windows shows that & characters do not get converted to _ even in the single-line edit control case (indicating that DT_NOPREFIX should be passed). Thanks, Zach

2 1

Missing file names in WINEDBG
by lats Nov. 14, 2004

Nov. 14, 2004

In the last week WINEDBG has stopped outputting the file/path and linenumber of entries in the backtrace. Does anyone know what happened? Jeff

3 3

Re: taskmodal messagebox implementation, take 3
by Juan Lang Nov. 13, 2004

Nov. 13, 2004

Ivan, a couple comments on this. You wrote: +BOOL CALLBACK MSGBOX_EnumProc(HWND hwnd, LPARAM lParam) +{ + static int counter1, counter2, heapcheck=0; + static HWND *handles; static variables aren't thread safe. Two threads attempting to do task modal dialogs (yes, this is probably strange) could result in weird failures. + handles=(HWND*)malloc(counter1*sizeof(HWND)); Any reason you aren't using HeapAlloc instead? --Juan __________________________________ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.com

1 0

dlls/commdlg/fontdlg.c regression
by Gerald Pfeifer Nov. 13, 2004

Nov. 13, 2004

The following change to dlls/commdlg/fontdlg.c revision 1.70 date: 2004/08/14 00:42:35; author: julliard; state: Exp; lines: +2 -5 Jacek Caban <jack(a)itma.pwr.wroc.pl> Fix a bug in passing parameters to CFn_WMInitDialog and CFn_WMCommand in FormatCharDlgProcW. causes the following warning with GCC 3.3.2 on FreeBSD 4.10: fontdlg.c:1161: warning: passing arg 4 of `CFn_WMCommand' from incompatible pointer type Would you mind fixing this? Thanks, Gerald -- Gerald Pfeifer (Jerry) gerald(a)pfeifer.com http://www.pfeifer.com/gerald/

2 3

Using a DIB to reduce number of calls to X
by Arjen Nienhuis Nov. 13, 2004

Nov. 13, 2004

GradientFill can be implemented by drawing to a DIB, and then blting that DIB to the screen. This will reduce the number of calls to X from O(n) to O(1). 1. Is that a good idea? Any caveats? 2. Can that method be used elsewhere? I'll try to implement it right away. P.S. We do implement the alpha part of GradientFill the same as Windows. The comments might sugest otherwise.

2 1

For the appdb hackers
by Ivan Leo Puoti Nov. 13, 2004

Nov. 13, 2004

It would be great if I could delete/approve apps and go to the next one without having to go back to adminAppQueue.php Thanks for the work you're doing, it was badly needed. Ivan. ____________________________________________________________ Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazione. Abbonati subito su http://www.libero.it

3 3

Kernel 2.6.9
by Paul Rupe Nov. 13, 2004

Nov. 13, 2004

Has anyone tried Wine with Linux kernel 2.6.9? I'm getting crashes with this version that don't happen when I reboot back to 2.6.8.1. I haven't had much luck debugging it so far, but each time, ps shows a <defunct> wine-preloader process and wineserver still running. -- Paul Rupe <prupe(a)myrealbox.com>

7 10

Re: ws2tcpip.h: Define socklen_t
by felix.nawothnig＠t-online.de Nov. 13, 2004

Nov. 13, 2004

On 11/13/2004 02:22:18 AM, I wrote: > We don't have socklen_t defined anywhere and that's the place where > it's defined in the official headers so... WASTE compiles fine now... > (I'm sure there is a reason why it isn't defined but I can't see it - > so tell me why this will be rejected... err... if it will be... :-) Ooops... WINE doesn't build anymore, I guess that's a good reason. :) Sorry... Anyway, what's the correct way to get socklen_t then? Would it make sense to do... #ifndef __socklen_t_defined typedef int socklen_t; # define __socklen_t_defined #endif ...or prefix it with WS() maybe? Or should I just include unistd.h in winelib apps? (Not sure if that's possible even since I'm using MSVCRT...) -flx

1 0

enhmfdrv: ExtSelectClipRgn
by Jon Griffiths Nov. 13, 2004

Nov. 13, 2004

Hi, My app gives me a FIXME from dlls/gdi/enhmetafile.c at line 1144 complaining that ExtSelectClipRgn is not implemented (and then proceeds to draw rather corrupted looking pictures). Looking at the code there is an implementation, but its been #ifdef'd out. removing the #if 0 doesn't produce any noticable change in my app. Does any one have any idea why this was removed, or what is needed to make it work? Cheers, Jon ===== "Don't wait for the seas to part, or messiahs to come; Don't you sit around and waste this chance..." - Live jon_p_griffiths(a)yahoo.com __________________________________ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.com

1 0

Re: Checking up header files ( take 2 : mmsystem.h )
by Eric Pouech Nov. 13, 2004

Nov. 13, 2004

Rémi Assailly a écrit : > Changelog: > * set MAXERRORLENGTH to 256 > * set MMSYSERR_LASTERROR to (MMSYSERR_BASE + 20) > * set MIDIERR_LASTERROR to (MIDIERR_BASE + 7) > * add some structure tags I think it's worthless bumping the LASTERROR values if you don't add the missing error values in the .h file A+

1 0

Re: Toolbar: Cleanup Patch
by Dimitrie O. Paun Nov. 12, 2004

Nov. 12, 2004

On Mon, Nov 08, 2004 at 11:27:44AM -0600, Robert Shearman wrote: > Changelog: > - bNtfUnicode is a dup' of bUnicode so remove it and fix the few places > where it was used. bUnicode is a bad name for this flag, bNtfUnicode would be better, as bUnicode is sometimes use to denote how the text is stored in the control, which is different from how notifications should be sent. -- Dimi.

2 2

Re: Kernel 2.6.9
by Ivan Leo Puoti Nov. 12, 2004

Nov. 12, 2004

>What could possibly be different between our systems that would cause >this problem? Anything but wine and the kernel, could you try this binary package and see if it works? Like this you can find out if it's a runtime or compile time problem. http://prdownloads.sourceforge.net/wine/wine-20041019-mdk.i586.rpm?download Ivan. ____________________________________________________________ Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazione. Abbonati subito su http://www.libero.it

3 3

Re: just a little thing in dlls/winmm/lolvldrv.c
by Bill Medland Nov. 12, 2004

Nov. 12, 2004

On November 12, 2004 12:17 pm, Rémi Assailly wrote: > Changelog: > * just replace all (sizeof(MM_MLDrvs) / sizeof(MM_MLDrvs[0]) by > MAX_MM_MLDRVS to be more readable. ( I think it is ) I think it isn't. (sizeof(a)/sizeof(a[0]) or (sizeof(a)/sizeof(*a)) is a standard statement that means "the number of elements in a". Some people make a macro such as COUNTOF(a) to make it more readable but the expanded version is quite standard. The altered version is less readable because one then has to go back to the definition of MAX_MM_MLDRVS and its uses to see that indeed it is (or at least is supposed to be) the number of elements in the array that the statements are protecting. -- Bill Medland mailto:billmedland(a)mercuryspeed.com http://webhome.idirect.com/~kbmed

1 0

Re: Kernel 2.6.9
by Ivan Leo Puoti Nov. 12, 2004

Nov. 12, 2004

> I forgot to mention that I'm using latest wine CVS. So am I and wine works fine, on mdk 10.1 kernel 2.6.8.1-12mdk. Ivan. ____________________________________________________________ Libero ADSL: navighi gratis a 1.2 Mega, senza canone e costi di attivazione. Abbonati subito su http://www.libero.it

2 1

Re: Dbghelp: enhance PDB support
by Dimitrie O. Paun Nov. 12, 2004

Nov. 12, 2004

On Thu, Nov 11, 2004 at 09:25:46PM +0100, Eric Pouech wrote: > This (huge) patch does: > - apart from restructuring the code (COFF handling goes to new coff.c > file, Codeview and PDB defines and types to new mscvpdb.h file) ^^^^^^^^^ Shouldn't this be named msvcpdb.h? -- Dimi.

3 2

Tables For Standard Caption Buttons' Metrics?
by William Poetra Yoga H Nov. 12, 2004

Nov. 12, 2004

In windows/nonclient.c, the captions on caption buttons are calculated (according to the button size, which is determined by some system metrics). This is due to the fact that we don't have a replacement for the Marlett font used in Windows. But simple formulae can't reproduce the captions correctly. I've taken some screenshots of the caption buttons in Windows XP (with the classic theme, so I think the results can be reproduced in earlier versions). I think we can reproduce Windows' behavior with a table. For every value of SM_CYCAPTION, we have one table entry. This way, we can save space (the table is a lot smaller than a table of bitmaps) and still reproduce correct behavior. Anyone has comments on this? __________________________________ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.com

2 3

Enabling gcov patch not accepted
by Aaron Arvey Nov. 11, 2004

Nov. 11, 2004

Hi, I recently sent in a patch concerning enabling gcov in wine (http://www.winehq.org/hypermail/wine-patches/2004/11/0136.html) and it doesn't seem to have been incorporated in the main cvs branch :(. I have never contributed to wine before so I'm not sure how to get feedback on my patch in order to improve it so that it is accepted. If either Alexandre or anyone else believes they know why my patch wasn't accepted, please let me know. I personally have found gcov to be very helpful and I think other developers may find that to be true too. If you would like to know more about gcov, please read the documentation included in my patch (url above). Thank you, Aaron

3 2

shreg tests on xp vs 95/98
by James Hawkins Nov. 11, 2004

Nov. 11, 2004

I've been working on figuring out the behaviour of SHQueryValueEx using XP for testing. I know that on XP SHQueryValueEx just calls SHRegGetValue, so any results from testing SHQueryValueEx match those documented about SHRegGetValue and even those not documented. What I'm wondering is if SHQueryValueEx behaves the same on other platforms ie 95/98/NT etc. If you'd like to test it, run the attached test on your platform and report the results back. Thankyou for taking the time to help figure this one out. BTW you can ignore this error for now (I'll look into it after this): shreg.c:270: Test failed: didn't open copy -- James Hawkins

1 0

Cxbx
by Dimitrie O. Paun Nov. 11, 2004

Nov. 11, 2004

Hi Folks, Willie Sippel suggested I port the Bxbx Xbox emulator to Winelib. And I did, patch attached. It seems like we need a multimon.h, any takers? More on the emulator: Home: http://www.caustik.com/cxbx/ http://sourceforge.net/projects/cxbx/ -- Dimi.

1 0

ODBC - MSSQL
by Pablo Szuster Nov. 11, 2004

Nov. 11, 2004

Hi, i´m trying to access a MSSQL server (in a windows machine in the network) from a Visual Basic Application running under wine, using an ODBC driver. Since there aren´t any MSSQL drivers for linux i decided to use de native ones, so i installed mdac_typ.exe . Then i configured the DSN using odbcad32.exe but when i tested the connection (configured using tcp/ip) an error message appeared ("[Microsoft][ODBC SQL Server Driver][TCP/IP Sockets] General network error. Check your network documentation"). Do i have to install some network drivers for wine?? Has anyone been able to connect to a MSSQL server remotely using odbc under wine?? __________________________________________________ Todavía no tenés tu Ciudad Internet Mail? Obtenelo ahora! - http://webmail.ciudad.com.ar Descargá Gratis el nuevo Internet Explorer 6.0, el mejor software para actualizar tu PC. http://www.ciudad.com.ar/ar/servicios/ie/

2 1