September 2007 - Wine-Devel

[AppDB] Bug: Let objectManager delete bug links

by Alexander Nicolaysen Sørnes

Bug: Let objectManager delete bug links Regards, Alexander N. Sørnes

Sept. 17, 2007

make crosstest is missing -lcabinet

by Pedro Araújo

Hi all, I've been trying to cross-compile the Wine tests, but I keep getting this error: $ ./configure && make crosstest [SNIP] i586-mingw32msvc-gcc advpack.cross.o files.cross.o install.cross.o testlist.cross.o -o advpack_crosstest.exe -lcabinet -luser32 -ladvapi32 -lkernel32 /usr/lib/gcc/i586-mingw32msvc/3.4.5/../../../../i586-mingw32msvc/bin/ld: cannot find -lcabinet collect2: ld returned 1 exit status make[2]: *** [advpack_crosstest.exe] Error 1 make[2]: Leaving directory `/.../dlls/advpack/tests' make[1]: *** [advpack/tests/__crosstest__] Error 2 make[1]: Leaving directory `/.../dlls' make: *** [dlls/__crosstest__] Error 2 Am I missing anything? Thanks in advance. -- - Pedro.

Sept. 17, 2007

Re: user32: better fix how to handle GetWindowLongPtr[AW](..., GWLP_WNDPROC) for builtin winprocs [resend]

by Alexandre Julliard

"Mikolaj Zalewski" <mikolajz(a)google.com> writes: > Is there something wrong with this patch? > From 50eacd2465df53e5cfc15458ff5ab80f0cc07e7e Mon Sep 17 00:00:00 2001 > From: Mikolaj Zalewski <mikolaj(a)zalewski.pl> > Date: Mon, 10 Sep 2007 14:03:49 -0700 > Subject: [PATCH] user32: better fix how to handle GetWindowLongPtr[AW](..., GWLP_WNDPROC) for builtin winprocs There are several other places where winprocs are returned, like SetWindowLong or GetClassLong. You probably have to handle them too, or at least write test cases for them. -- Alexandre Julliard julliard(a)winehq.org

Sept. 17, 2007

Re: An enumeration of network resources

by Juan Lang

Hi Konstantin, > The information about loaded dll is absent in the wine registry. Whether it > means, what there are no preparations of DLL's for work with a network > (similar ntlanman.dll in Windows)? That's correct, this hasn't been implemented in Wine yet. Regards, --Juan

Sept. 17, 2007

Re: msi(1/3): Don't crash if rec is NULL

by Juan Lang

> The public APIs check for bad records and return ERROR_INVALID_HANDLE, > so they will never send in a NULL rec. Something internally is > sending in a NULL rec, and that needs to be fixed. On the contrary, the record is not NULL, it has 0 parameters. I've resent in the fix with a test showing it's correct. --Juan

Sept. 17, 2007

Re: old dos/windows program crashing

by Kuba Ober

On Monday 17 September 2007, you wrote: > On Monday 17 September 2007 15:10:39 Kuba Ober wrote: > > > I obviously tried WINE then, but it didn't work, > > > however, dosbox works with it (though slow, obviously) > > > > Wine will not be any faster. Dosbox doesn't emulate the CPU (at least > > when run on x86 machines), it's your old good VM environment, just like > > if you ran DOS under win9x. > Isn't the whole point of DOSBox to emulate an old computer with all the > necessar stuff like SoundBlaster 16? Last I checked it had a setting to > set the number of emulated cycles. Hmm, I guess I was very much wrong then. I thought it was a VM thing with virtualization of hardware. I guess the fact that I ran it long ago on 486 hardware and it was still fast enough back then made me assume it had to execute using hardware. I apologize for spreading misinformation around. Cheers, Kuba

Sept. 17, 2007

Re: old dos/windows program crashing

by Alexander Nicolaysen Soernes

On Monday 17 September 2007 15:10:39 Kuba Ober wrote: > > I obviously tried WINE then, but it didn't work, > > however, dosbox works with it (though slow, obviously) > > Wine will not be any faster. Dosbox doesn't emulate the CPU (at least when > run on x86 machines), it's your old good VM environment, just like if you > ran DOS under win9x. > > Cheers, Kuba Isn't the whole point of DOSBox to emulate an old computer with all the necessar stuff like SoundBlaster 16? Last I checked it had a setting to set the number of emulated cycles. Alexander N. Sørnes

Sept. 17, 2007

Re: old dos/windows program crashing

by Kuba Ober

> I obviously tried WINE then, but it didn't work, > however, dosbox works with it (though slow, obviously) Wine will not be any faster. Dosbox doesn't emulate the CPU (at least when run on x86 machines), it's your old good VM environment, just like if you ran DOS under win9x. Cheers, Kuba

Sept. 17, 2007

Re: old dos/windows program crashing

by Kuba Ober

> I have a program, that I got like ages ago. It > effectively is an atari emulator for dos, by the name > "pacifist", but I originally got it installed under > windows ME. It has an old atari-basic program > "installed" in it, (a simple game) written by a friend > in it, which I wanted to play. > I decided to try to get it to run again under Linux. > > The website it came from originally doesn't seem to > have the program available for download anymore Tried to look at www.archive.org? It'll likely still be there. Cheers, Kuba

Sept. 17, 2007

Re: AppDB: cannot submit a screenshot even when logged in

by Alexander Nicolaysen Sørnes

On Sunday 16. September 2007 20:56:47 Vit Hrachovy wrote: > Hello everybody, > I may missed something, but I cannot submit a screenshot to the > application I maintain in AppDB and I'm logged in. > > There's always a nasty 'Log in to submit screenshots' instead of Submit > button. > > Is it planned downtime? > Regards > Vit > > P.S. It's http://appdb.winehq.org/screenshots.php?iAppId=4025 Do you have cookies enabled? I have looked at the code and I cannot see a reason why you should see that message if you are logged in. If it still does not work, please post the URL of the page which shows the error. Alexander N. Sørnes

Sept. 16, 2007

Re: AppDB: cannot submit a screenshot even when logged in

by Vit Hrachovy

Hi Alexander, it suddenly started working - I blame wrong cache or TTL :-( It was strange, because I could Delete the image, but couldn't post. Anyhow, please forget the issue. Thanks and regards Vit Alexander Nicolaysen Sørnes wrote: > Do you have cookies enabled? I have looked at the code and I cannot see a > reason why you should see that message if you are logged in. If it still > does not work, please post the URL of the page which shows the error. > > > Alexander N. Sørnes

Sept. 16, 2007

AppDB: cannot submit a screenshot even when logged in

by Vit Hrachovy

Hello everybody, I may missed something, but I cannot submit a screenshot to the application I maintain in AppDB and I'm logged in. There's always a nasty 'Log in to submit screenshots' instead of Submit button. Is it planned downtime? Regards Vit P.S. It's http://appdb.winehq.org/screenshots.php?iAppId=4025

Sept. 16, 2007

Re: DirectDraw opengl renderer

by Roderick Colenbrander

> Grand news, and thanks a lot for pointing this out - Zork: Nemesis and > Zork: Grand Inquisitor are now finally playable using these two options > opengl/readdraw. There is a small glitch with screen redrawing, but hey > - it works after YEARS! Thanks, I have had reports of other games suddenly working in this version but I doubt my patches have caused it. There are still a few issues in the code to which fixes have been submitted by Stefan and me. Hopefully after those patches things will still work. I would ask people to monitor git closely this week to see if no regressions will appear. Roderick -- Psssst! Schon vom neuen GMX MultiMessenger gehört? Der kanns mit allen: http://www.gmx.net/de/go/multimessenger

Sept. 16, 2007

Re: DirectDraw opengl renderer

by Vit Hrachovy

Grand news, and thanks a lot for pointing this out - Zork: Nemesis and Zork: Grand Inquisitor are now finally playable using these two options opengl/readdraw. There is a small glitch with screen redrawing, but hey - it works after YEARS! Regards Vit Roderick Colenbrander wrote: > If you are interested in using DirectDrawRenderer=opengl also make sure to > play with the RenderTargetLockMode (default=readdraw). In case of Tiberian > Sun 'readtex' is a lot faster. > > Regards, > Roderick

Sept. 16, 2007

Re: DirectDraw opengl renderer

by Roderick Colenbrander

> As I understand it, there is still a long time before the opengl renderer > will become the default one. With Wine 0.9.45, some bugs were fixed > allowing the opengl renderer to be used for C&C Tiberian Sun, and I must > say the speed improvements are dramatic! Without any FPS tools I would > definitely say performance has increased by several hundred per cent. Now > the game is so fast I may have to turn down the game speed setting. :) If > you move the mouse it still slows down, though, as it does with the gdi > renderer. There are roughly three bottlenecks for 2D games: - GetDC/ReleaseDC which are used to draw to the framebuffer using GDI functions (requires nasty framebuffer downloads/uploads) - Depth conversion (mainly for palette p8 games like starcraft / c&c1 / red alert) - LockRect/UnlockRect which give direct access (a pointer) to framebuffer memory In 0.9.45 I added PBO support for use with LockRect/UnlockRect. PBOs provide us with a pointer to DMA memory or video memory (depending on what the implementation prefers). This greatly helps Tiberian Sun and Red Alert II (the games I tested this code with a lot). There are some other remaining issues (some PBO regressions) but I hope to get those fixed soon. After that I plan to add some more optimizations to the upload/download code (in some cases we have to flip a surface in software which is slow) and I will add an accelerated P8 backend to make C&C1/RedAlert/StarCraft a LOT faster. If you are interested in using DirectDrawRenderer=opengl also make sure to play with the RenderTargetLockMode (default=readdraw). In case of Tiberian Sun 'readtex' is a lot faster. Regards, Roderick

Sept. 16, 2007

DirectDraw opengl renderer

by Alexander Nicolaysen Sørnes

Hello everyone, I know this has been discussed before (and not that long ago), but I think I have some wonderful new information to add to the debate of whether or not to add an option for this in winecfg. As I understand it, there is still a long time before the opengl renderer will become the default one. With Wine 0.9.45, some bugs were fixed allowing the opengl renderer to be used for C&C Tiberian Sun, and I must say the speed improvements are dramatic! Without any FPS tools I would definitely say performance has increased by several hundred per cent. Now the game is so fast I may have to turn down the game speed setting. :) If you move the mouse it still slows down, though, as it does with the gdi renderer. Anyway, I think this should apply to several other games as well (and I will test more). But I was wondering is this is enough to add an option in winecfg? Alexander N. Sørnes

Sept. 16, 2007

Re: RICHED20: additional tests plus RFC

by Alexander Nicolaysen Sørnes

On Sunday 16. September 2007 04:34:17 a_villacis(a)palosanto.com wrote: > I am tracking down a richedit bug in FACETS (bug #1713). > > Apparently, the OCX control used by this app expects a particular > SendMessage() > with the message EM_SETCHARFORMAT to return 1. Currently builtin returns 0 > unconditionally, making this app bail out with "Invalid property value". By > changing the return value to 1 unconditionally, the app gets "fixed", but > this > is not a proper solution, since other apps might break with this change. > > This patch adds additional testing of EM_SETCHARFORMAT. All todo_wine's > succeed > in an installation of WinXP SP2. Please confirm this is so for all of you. > > There are two code blocks commented out. The first one makes the app crash > on WinXP, and is listed for completeness. The other one was supposed to > test what > happens when the riched20 control receives a blanked-out CHARFORMAT2 > structure. > However this triggers an assertion in Wine at style.c line 63. WinXP > doesn't bat an eye, and possibly ignores these structure values. Unless > somebody objects, I will shortly send a patch that replaces the assertion > with a NULL return in style.c, plus a NULL check in editor.c . > > There are more riched20 tests that fail in WinXP SP2, unrelated to > EM_SETCHARFORMAT, but I would like confirmation that (1) somebody else sees > failing tests in native Windows, not just me, and (2) this patch does not > introduce any new failing tests in native Windows. > > Changelog: > * Add additional tests for EM_SETCHARFORMAT message of riched20 control Hey, great to see some work on EM_SETCHARFORMAT. If you're interested, Wine has a builtin WordPad (just type 'wine wordpad') which exposes some problems. Try to select a font in the font list when no text is selected, for instance. Alexander N. Sørnes

Sept. 16, 2007

compile wine from sources - strange messages

by Andrey Gavrilin

Hello. I wish to compile wine from sources. Has received such message. The distribution: Ubuntu 7.04. What headers files does not suffice? PS I used the program the translator from Russian to English. configure: WARNING: jack/jack.h: present but cannot be compiled configure: WARNING: jack/jack.h: check for missing prerequisite headers? configure: WARNING: jack/jack.h: see the Autoconf documentation configure: WARNING: jack/jack.h: section "Present But Cannot Be Compiled" configure: WARNING: jack/jack.h: proceeding with the preprocessor's result configure: WARNING: jack/jack.h: in the future, the compiler will take precedence configure: WARNING: ## ------------------------------------ ## configure: WARNING: ## Report this to wine-devel(a)winehq.org ## configure: WARNING: ## ------------------------------------ ## configure: WARNING: pthread.h: present but cannot be compiled configure: WARNING: pthread.h: check for missing prerequisite headers? configure: WARNING: pthread.h: see the Autoconf documentation configure: WARNING: pthread.h: section "Present But Cannot Be Compiled" configure: WARNING: pthread.h: proceeding with the preprocessor's result configure: WARNING: pthread.h: in the future, the compiler will take precedence

Sept. 16, 2007

Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Reece Dunn

On 16/09/2007, Steven Edwards <winehacker(a)gmail.com> wrote: > On 9/16/07, Reece Dunn <msclrhd(a)googlemail.com> wrote: > > Do you want to file a bug report with Google, or should I do it? > > Please file it. Done. http://groups.google.com/group/google-mail-problem-solving-uk/browse_thread… - Reece

Sept. 16, 2007

Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Steven Edwards

On 9/16/07, Reece Dunn <msclrhd(a)googlemail.com> wrote: > Do you want to file a bug report with Google, or should I do it? Please file it. Thanks -- Steven Edwards "There is one thing stronger than all the armies in the world, and that is an idea whose time has come." - Victor Hugo

Sept. 16, 2007

Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Stefan Dösinger

Am Sonntag, 16. September 2007 11:27:17 schrieb Reece Dunn: > If I select "Show original" from the "Reply" menu, I see: > > User-Agent: Thunderbird 1.5.0.13 (X11/20070824) > > which would explain why the people using Thunderbird can see it > correctly. I also see: I'm using kmail here

Sept. 16, 2007

Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Reece Dunn

On 15/09/2007, Jesse Allen <the3dfxdude(a)gmail.com> wrote: > On 9/15/07, Stefan Dösinger <stefandoesinger(a)gmx.at> wrote: > > Am Samstag, 15. September 2007 18:11:43 schrieb Maarten Lankhorst: > > > Steven Edwards schreef: > > > > Its not all of your files only some come through with junk letters > > > > like this one has > > > > > > > > ¨s1* > > > > > > > > as the filename. Another one had the Eurosymbol Squared....it could > > > > just be gmail being goofy as I noticed some of the patches come > > > > through fine. > > > > > > I haven't tried webmail of gmail, but gmail received through thunderbird > > > works fine for me. Not sure why webmail would be any different then pop3. > > Working fine here too > > It has to be the webmail of gmail then. I see it on gmail. I'm seeing the filenames as being garbled here on gmail as well, so it does look like a gmail issue. I don't know about other webmail, though. If I select "Show original" from the "Reply" menu, I see: User-Agent: Thunderbird 1.5.0.13 (X11/20070824) which would explain why the people using Thunderbird can see it correctly. I also see: Content-Type: text/x-patch; name*0="0003-gdi-Implement-BiDi-classification-of-characters-and.patch" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename*0="0003-gdi-Implement-BiDi-classification-of-characters-and.pat"; filename*1="ch" So the name is fine. I'm wondering if it is the way that gmail is processing the Content-Disposition line. Do you want to file a bug report with Google, or should I do it? - Reece

Sept. 16, 2007

Re: patches to make wine compile on osx

by Roderick Colenbrander

Hi Daniel, A few comments regarding the patches. First of all you should use a single patch file for each patch (this makes tracking patches easier). Now over to the patches. Regarding wineosxtypeerrors.patch: - 'GLenum' should be 'unsigned int' instead of int. I don't understand why the casts are needed as it should just get the type info from gl.h. I think that using '%u' for unsigned data could fix most of your problems. Regarding wineosxglexterrors.patch: - Same comment regarding unsigned int data. (make this a separate patch) - GL_TEXTURE_RECTANGLE_NV should be handled in include/wine/wgl.h, strange that I didn't receive compile failures for linux about it .. perhaps it is already in the mesa headers. (make this a separate patch) Anyway I would define GL_TEXTURE_RECTANGLE_NV like this: #ifndef GL_NV_texture_rectangle #define GL_TEXTURE_RECTANGLE_NV 0x84F5 #define GL_TEXTURE_BINDING_RECTANGLE_NV 0x84F6 #define GL_PROXY_TEXTURE_RECTANGLE_NV 0x84F7 #define GL_MAX_RECTANGLE_TEXTURE_SIZE_NV 0x84F8 #endif The extension is fully compatible with GL_ARB_texture_rectangle which osx should have and GL_EXT_texture_rectangle is likely the same. (the constants between the ARB and NV one are also the same, I expect the same for EXT and ARB) Regards, Roderick Colenbrander

Sept. 16, 2007

[OT] Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Kari Hurtta

"Jesse Allen" <the3dfxdude(a)gmail.com> writes in gmane.comp.emulators.wine.devel: > On 9/15/07, Stefan Dösinger <stefandoesinger(a)gmx.at> wrote: > > Am Samstag, 15. September 2007 18:11:43 schrieb Maarten Lankhorst: > > > Steven Edwards schreef: > > > > Its not all of your files only some come through with junk letters > > > > like this one has > > > > > > > > ¨s1* > > > > > > > > as the filename. Another one had the Eurosymbol Squared....it could > > > > just be gmail being goofy as I noticed some of the patches come > > > > through fine. > > > > > > I haven't tried webmail of gmail, but gmail received through thunderbird > > > works fine for me. Not sure why webmail would be any different then pop3. > > Working fine here too > > > > > > It has to be the webmail of gmail then. I see it on gmail. ---------------- Content-Type: text/x-patch; name*0="0003-gdi-Implement-BiDi-classification-of-characters-and.patch" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename*0="0003-gdi-Implement-BiDi-classification-of-characters-and.pat"; filename*1="ch" ----------------- Looks like RFC 2231: "MIME Parameter Value and Encoded Word Extensions: Character Sets, Languages, and Continuations" Perhaps there is some bug in RFC 2231 parser on gmail. / Kari Hurtta

Sept. 16, 2007

Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Jesse Allen

On 9/15/07, Stefan Dösinger <stefandoesinger(a)gmx.at> wrote: > Am Samstag, 15. September 2007 18:11:43 schrieb Maarten Lankhorst: > > Steven Edwards schreef: > > > Its not all of your files only some come through with junk letters > > > like this one has > > > > > > ¨s1* > > > > > > as the filename. Another one had the Eurosymbol Squared....it could > > > just be gmail being goofy as I noticed some of the patches come > > > through fine. > > > > I haven't tried webmail of gmail, but gmail received through thunderbird > > works fine for me. Not sure why webmail would be any different then pop3. > Working fine here too > > It has to be the webmail of gmail then. I see it on gmail.

Sept. 15, 2007

Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Stefan Dösinger

Am Samstag, 15. September 2007 18:11:43 schrieb Maarten Lankhorst: > Steven Edwards schreef: > > Its not all of your files only some come through with junk letters > > like this one has > > > > ¨s1* > > > > as the filename. Another one had the Eurosymbol Squared....it could > > just be gmail being goofy as I noticed some of the patches come > > through fine. > > I haven't tried webmail of gmail, but gmail received through thunderbird > works fine for me. Not sure why webmail would be any different then pop3. Working fine here too

Sept. 15, 2007

Re: Bug#441070: Dfw5 draws graphs black (regression)

by Chris Morgan

Yes, please file a bug report, http://bugs.winehq.org Chris On 9/15/07, Willi Mann <foss-ml(a)wm1.at> wrote: > > > Please let me know if I can do anything to help fixing this bug. > > As nobody answered so far, should I file a bug report? > > Willi > > Plase CC: me on replies. > > > >

Sept. 15, 2007

Re: Bug#441070: Dfw5 draws graphs black (regression)

by Willi Mann

> Please let me know if I can do anything to help fixing this bug. As nobody answered so far, should I file a bug report? Willi Plase CC: me on replies.

Sept. 15, 2007

Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Maarten Lankhorst

Steven Edwards schreef: > Its not all of your files only some come through with junk letters > like this one has > > ¨s1* > > as the filename. Another one had the Eurosymbol Squared....it could > just be gmail being goofy as I noticed some of the patches come > through fine. > I haven't tried webmail of gmail, but gmail received through thunderbird works fine for me. Not sure why webmail would be any different then pop3. Cheers, Maarten

Sept. 15, 2007

Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Steven Edwards

On 9/15/07, Maarten Lankhorst <m.b.lankhorst(a)gmail.com> wrote: > Steven Edwards schreef: > > Hi Maarten, > > > > Could you configure your git scripts to output different filenames > > when generating patches? Somehow they get managed for me in my gmail > > inbox. It makes it quite a pain for me to review. > > > Hi Steven, > > What kind of different filenames? .txt instead of .patch? Its not all of your files only some come through with junk letters like this one has ¨s1* as the filename. Another one had the Eurosymbol Squared....it could just be gmail being goofy as I noticed some of the patches come through fine. -- Steven Edwards "There is one thing stronger than all the armies in the world, and that is an idea whose time has come." - Victor Hugo

Sept. 15, 2007

Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Maarten Lankhorst

Steven Edwards schreef: > Hi Maarten, > > Could you configure your git scripts to output different filenames > when generating patches? Somehow they get managed for me in my gmail > inbox. It makes it quite a pain for me to review. > Hi Steven, What kind of different filenames? .txt instead of .patch? Cheers, Maarten

Sept. 15, 2007

Re: [PATCH 2/7] gdi: Implement BiDi classification of characters

by Steven Edwards

Hi Maarten, Could you configure your git scripts to output different filenames when generating patches? Somehow they get managed for me in my gmail inbox. It makes it quite a pain for me to review. Thanks -- Steven Edwards "There is one thing stronger than all the armies in the world, and that is an idea whose time has come." - Victor Hugo

Sept. 15, 2007

old dos/windows program crashing

by Joris Huizer

Hello, I have a program, that I got like ages ago. It effectively is an atari emulator for dos, by the name "pacifist", but I originally got it installed under windows ME. It has an old atari-basic program "installed" in it, (a simple game) written by a friend in it, which I wanted to play. I decided to try to get it to run again under Linux. The website it came from originally doesn't seem to have the program available for download anymore - so I just copied the files from the windows ME disc. I obviously tried WINE then, but it didn't work, however, dosbox works with it (though slow, obviously) so I know I'm not missing required files (though, perhaps, registry settings - not sure how likely that is, though) Anyway. This isn't such a big deal, though I'd like to look at what's going on in WINE, and possibly fix it - though I need help here. This program is, among things, a "PMODE W/DOS Extender" - so that means it probably does some obscure real-mode/protected-mode stuff. Console output is as follows, Warning: unprotecting memory to allow real-mode calls. NULL pointer accesses will no longer be caught. err:int:__wine_emulate_instruction mov cr0,eax at 0x00000544 err:int:DOSVM_Int33Handler int33: unknown/not implemented parameters: int33: AX 0015, BX 0003, CX 11f7, DX 1165, SI 0055, DI 03ac, DS 11df, ES 11e7 fixme:int:DOSVM_Int10Handler Select vertical resolution - not supported fixme:int:DOSVM_Int10Handler Load ROM 8x8 Double Dot Patterns - Not Supported fixme:int31:DOSVM_Int31Handler physical address mapping (0x00000000) - unimplemented wine: Unhandled page fault on read access to 0xffffffff at address 0x101f:0x0003ed71 (thread 002d), starting debugger... fixme:dbghelp:addr_to_linear Failed to linearize address 08f0:00000002 (mode 0) fixme:dbghelp:addr_to_linear Failed to linearize address 08f0:00000000 (mode 0) Also I attached the visual display the "winevdm" shows as the program crashes. (It seems it crashes somewhere in generated code in kernel32.dll - in DPMI_PendingEventCheck() - written by the winebuild tool) I know some things about real-mode code, but not too much. Could you give me pointers on what to look at next? BTW, please CC me, I'm not on the mailing list currently. Thanks, Joris ____________________________________________________________________________________ Check out the hottest 2008 models today at Yahoo! Autos. http://autos.yahoo.com/new_cars.html

Sept. 15, 2007

Re: add shell32.755

by Maarten Lankhorst

Louis. Lenders schreef: > Hi, i'm sending in 2 patches from Anastasius Focht for review, and > hopefully for inclusion in git. Otherwise they might get "lost" and > forgotten. > > They're helpful for testing apps with native comdlg32.dll. So although > they do not really fix any real life app (not that i know of I'm missing a few tests to see if this behavior is actually what it should be like in windows. You should always add them to implementations of new functions. Cheers, Maarten

Sept. 15, 2007

Re: msi(1/3): Don't crash if rec is NULL

by James Hawkins

On 9/14/07, Jeremy White <jwhite(a)winehq.org> wrote: > Your patches also cure the symptom, Juan; thanks. > > I agree with the concerns James expressed, though: I didn't feel that I > understood the whole story. (Heck, I really have *no* clue what I'm > talking about <grin>). > > However, the call to MSI_FormatMessageW is not an internal call, I think; > it looks to me as though it's called as a direct result of an inbound > call to MsiFormatRecordA(): > 0014:Call msi.MsiFormatRecordA(00000002,00000005,1000e768,7b7ea55c) ret=10003064 > (in http://bugs.winehq.org/attachment.cgi?id=8060 about 10-15 lines from bottom) > > So the question in my mind is: > Can a record still be a legitimate record if MSI_RecordGetFieldCount returns 0? > > If the answer is yes, then I think the patches are correct. > I don't know off the top of my head, but tests should definitely be added before a fix goes in. -- James Hawkins

Sept. 14, 2007

Re: msi(1/3): Don't crash if rec is NULL

by Jeremy White

Your patches also cure the symptom, Juan; thanks. I agree with the concerns James expressed, though: I didn't feel that I understood the whole story. (Heck, I really have *no* clue what I'm talking about <grin>). However, the call to MSI_FormatMessageW is not an internal call, I think; it looks to me as though it's called as a direct result of an inbound call to MsiFormatRecordA(): 0014:Call msi.MsiFormatRecordA(00000002,00000005,1000e768,7b7ea55c) ret=10003064 (in http://bugs.winehq.org/attachment.cgi?id=8060 about 10-15 lines from bottom) So the question in my mind is: Can a record still be a legitimate record if MSI_RecordGetFieldCount returns 0? If the answer is yes, then I think the patches are correct. Cheers, Jeremy > The public APIs check for bad records and return ERROR_INVALID_HANDLE, > so they will never send in a NULL rec. Something internally is > sending in a NULL rec, and that needs to be fixed. >

Sept. 14, 2007

Re: fnt2fon: janitorial IgnoredReturnValues

by Alexandre Julliard

Lionel_Debroux <lionel_debroux(a)yahoo.fr> writes: > * the current input file (fp) is not always closed when exiting; Closing the input on exit is useless. > * the atexit() and signal() calls are placed too late in the program, > and their return values are not checked; Their return value is irrelevant, they won't fail, and if they do you can't do anything useful about it anyway. > * I've just seen on Michael Stefaniuc's potential issues page that > Smatch reports a memory leak, and it looks like a legitimate error. Memory leaks in a short lived tool like fnt2fon don't really matter. I'm not saying these things can't be fixed, but the effort has to be proportional to the results, and in the case of something like fnt2fon it's most likely not worth it. It's better to spend time fixing bugs in the dlls. -- Alexandre Julliard julliard(a)winehq.org

Sept. 14, 2007

Re: fnt2fon: janitorial IgnoredReturnValues

by Lionel_Debroux

> > tools/fnt2fon is flagged in the janitorial page "IgnoredReturnValues". > > This page is outdated, but its content is still valid for fnt2fon, > > which ignores a number of return values. > I don't think that's worth fixing, the warnings are essentially > useless, it just makes the code uglier. OK, I won't do that again :-) > There may be a few legitimate errors in there, like checking the > fopen() result, and those could be fixed; There are indeed several legitimate errors, such as: * not checking the fopen() result; * the current input file (fp) is not always closed when exiting; * the atexit() and signal() calls are placed too late in the program, and their return values are not checked; * I've just seen on Michael Stefaniuc's potential issues page that Smatch reports a memory leak, and it looks like a legitimate error. > but please don't add checking to every call that gcc complains about. OK. Actually, it's not GCC, but splint, since the headers on my SimplyMEPIS 6.x/Debian box don't contain __attribute__((warn_unused_result)). Vincent Béron has certainly used another distro (Fedora-type ?) to get GCC to complain about those calls. Regards, Lionel Debroux.

Sept. 14, 2007

Re: msi(1/3): Don't crash if rec is NULL

by James Hawkins

On 9/14/07, Juan Lang <juan.lang(a)gmail.com> wrote: > > MSI_FormatRecordW is an internal function and we should be crashing if > > rec is NULL. > > Are you sure? This is causing the iTunes installer to crash for a > couple people. Besides, I don't think the code is written with this > assumption in mind. In deformat_string_internal: > > static DWORD deformat_string_internal(MSIPACKAGE *package, LPCWSTR ptr, > WCHAR** data, DWORD len, MSIRECORD* record, > INT* failcount) > ... > if (ptr==NULL) > { > TRACE("Deformatting NULL string\n"); > *data = NULL; > return 0; > } > > So deformat_string_internal handles NULL correctly. This patch > changes one of the call sites to avoid dereferencing a NULL pointer > before passing it to a function that deals with a NULL input. > The public APIs check for bad records and return ERROR_INVALID_HANDLE, so they will never send in a NULL rec. Something internally is sending in a NULL rec, and that needs to be fixed. -- James Hawkins

Sept. 14, 2007

Re: msi(1/3): Don't crash if rec is NULL

by Juan Lang

> MSI_FormatRecordW is an internal function and we should be crashing if > rec is NULL. Are you sure? This is causing the iTunes installer to crash for a couple people. Besides, I don't think the code is written with this assumption in mind. In deformat_string_internal: static DWORD deformat_string_internal(MSIPACKAGE *package, LPCWSTR ptr, WCHAR** data, DWORD len, MSIRECORD* record, INT* failcount) ... if (ptr==NULL) { TRACE("Deformatting NULL string\n"); *data = NULL; return 0; } So deformat_string_internal handles NULL correctly. This patch changes one of the call sites to avoid dereferencing a NULL pointer before passing it to a function that deals with a NULL input. --Juan

Sept. 14, 2007

Re: fnt2fon: janitorial IgnoredReturnValues (1/5)

by Alexandre Julliard

Lionel_Debroux <lionel_debroux(a)yahoo.fr> writes: > tools/fnt2fon is flagged in the janitorial page "IgnoredReturnValues". > This page is outdated, but its content is still valid for fnt2fon, which > ignores a number of return values. I don't think that's worth fixing, the warnings are essentially useless, it just makes the code uglier. There may be a few legitimate errors in there, like checking the fopen() result, and those could be fixed; but please don't add checking to every call that gcc complains about. -- Alexandre Julliard julliard(a)winehq.org

Sept. 14, 2007

Re: msi(1/3): Don't crash if rec is NULL

by James Hawkins

On 9/14/07, Juan Lang <juan.lang(a)gmail.com> wrote: > This series of patches is in response to Jeremy's troubles with > iTunes. They may or may not fix the installer for him, but they > prevent potential crashes. > --Juan > MSI_FormatRecordW is an internal function and we should be crashing if rec is NULL. -- James Hawkins

Sept. 14, 2007

Problem with TCP Option 0x00000008

by Christian Mahnke

Hello, i have a problem with a program that uses a network license manager. I've installed native DCOM using the winetricks script. When i start the program i got the message "fixme:winsock:WS_setsockopt Unknown IPPROTO_TCP optname 0x00000008" after the loading of native rpcltccm.dll. To work around this i've compiled a modified wine 0.9.44 with an empty case statement for 0x00000008 in socket.c. It doesn't help much, the program just fails a bit later. The TCP Option 0x00000008 is defined as TCP_ATMARK in ws2_ipdef.h in the Windows SDK. Does anyone know how this should be mapped ? Cheers, Christian

Sept. 14, 2007

Problem with TCP Option 0x00000008

by Christian Mahnke

Hello, i have a problem with a program that uses a network license manager. I've installed native DCOM using the winetricks script. When i start the program i got the message "fixme:winsock:WS_setsockopt Unknown IPPROTO_TCP optname 0x00000008" after the loading of native rpcltccm.dll. To work around this i've compiled a modified wine 0.9.44 with an empty case statement for 0x00000008 in socket.c. It doesn't help much, the program just fails a bit later. The TCP Option 0x00000008 is defined as TCP_ATMARK in ws2_ipdef.h in the Windows SDK. Does anyone know how this should be mapped ? Cheers, Christian

Sept. 14, 2007

Re: winex11.drv/wintab32[1/5]: store categories field-by-field (resend)

by Alexandre Julliard

"Mikolaj Zalewski" <mikolajz(a)google.com> writes: > There is a problem to have the same prototype and make winex11.drv > call handle only Unicode as some of the WTInfo output are binary data > and some are strings. Should I duplicate the knowledge of which are > which in wintab32? The W->A translation in wintab32 of course needs to know the format of the data to map strings correctly, that's not a problem. -- Alexandre Julliard julliard(a)winehq.org

Sept. 14, 2007

An enumeration of network resources

by Konstantin Kondratyuk

Hello! I try to organize an enumeration of network resources by means of functions WNetOpenEnum and WNetEnumResource. Function WNetOpenEnumW contains a line: " ret = providerTable->table[index].openEnum( dwScope, dwType, dwUsage, lpNet, &handle); " which addresses to function in dll. The information about loaded dll is absent in the wine registry. Whether it means, what there are no preparations of DLL's for work with a network (similar ntlanman.dll in Windows)? -- Best regards, Konstantin Kondratyuk.

Sept. 14, 2007

Re: carbon poker under wine

by Trent Waddington

On 9/13/07, Trent Waddington <trent.waddington(a)gmail.com> wrote: > wget http://fpdownload.macromedia.com/get/flashplayer/current/install_flash_play… > ./wine install_flash_player.exe Forget I said that. Grab: http://rtfm.insomnia.org/~qg/Flash9b.ocx and put it in ~/.wine/drive_c/windows/system32/Macromed/Flash, cd to that directory and do: ~/wine/wine ../../regsvr32.exe Flash9b.ocx which should give you: Successfully registered DLL Flash9b.ocx And that's all you need. Trent

Sept. 13, 2007

Re: First song purchased on iTunes 7 with Wine...

by Jeremy White

Juan Lang wrote: >> 7.4.2? I tried that last night, and didn't end up getting >> any files installed. Am I missing any tricks? I had a virgin .wine... > > Did you try it with my patches installed? Last night's didn't have > them, this morning's did. Try again with today's .git. Hopefully > something didn't break between 7.1.x and 7.4.x. No joy :-(. fresh git, make distclean, clean ~/.wine. Install just fails; Last 2 lines of console output is: err:msi:custom_get_thread_return Invalid Return Code 1723 err:msi:ITERATE_Actions Execution halted, action L"InstallPackages" returned 1603 But this is now the wrong place for this; sorry for the noise. I've posted a report in this bug: http://bugs.winehq.org/show_bug.cgi?id=8705 Cheers, Jeremy

Sept. 13, 2007

Re: winex11.drv/wintab32[1/5]: store categories field-by-field (resend)

by Mikolaj Zalewski

> You shouldn't change the prototype. Driver functions should follow the > signature of the corresponding APIs as far as possible. There is a problem to have the same prototype and make winex11.drv call handle only Unicode as some of the WTInfo output are binary data and some are strings. Should I duplicate the knowledge of which are which in wintab32? Mikolaj Zalewski

Sept. 13, 2007

Re: First song purchased on iTunes 7 with Wine...

by Juan Lang

> 7.4.2? I tried that last night, and didn't end up getting > any files installed. Am I missing any tricks? I had a virgin .wine... Did you try it with my patches installed? Last night's didn't have them, this morning's did. Try again with today's .git. Hopefully something didn't break between 7.1.x and 7.4.x. --Juan

Sept. 13, 2007