September 2013 - Wine-bugs - list.winehq.org

[Bug 29792] New: Gothic 2 (JoWood Productions) installer fails due to media validation tool failing (don't add FILE_ATTRIBUTE_ARCHIVE by default to file entries)
by wine-bugs＠winehq.org 07 Sep '19

07 Sep '19

http://bugs.winehq.org/show_bug.cgi?id=29792 Bug #: 29792 Summary: Gothic 2 (JoWood Productions) installer fails due to media validation tool failing (don't add FILE_ATTRIBUTE_ARCHIVE by default to file entries) Product: Wine Version: 1.4-rc1 Platform: x86 OS/Version: Linux Status: NEW Severity: normal Priority: P2 Component: ntdll AssignedTo: wine-bugs(a)winehq.org ReportedBy: focht(a)gmx.net Classification: Unclassified Hello, I first wanted to investigate bug 22515 hence i bought the game for a few bucks and looked into it... Well, it seems there exist different editions of the game and the problems might not be connected at all. Although this is related to copy protection it's not the SecuROM 4.x tests failing here - that part of protection works fine. I have the 3 CD set by JoWood Productions released on November 29, 2002. Near the end of installation process there is a tool "g2setup.exe" launched which verifies the authenticity of install media. That process returns a specific exit code (0x777 -> 1911) which basically says: "all checks passed". Any different exit code will roll back the whole installation! Current workaround: kill the installer from another terminal when this verification process starts. Everything has already been copied up to this point so just wait until the disc spins up and execute 'wineserver -k'. --- snip --- 0024:Starting process L"C:\\Program Files\\Jowood\\Gothic II\\g2setup.exe" (entryproc=0x4c6929) --- snip --- --- snip --- -=[ ProtectionID v0.6.4.0 JULY]=- (c) 2003-2010 CDKiLLER & TippeX Build 07/08/10-17:57:05 ... Scanning -> Z:\home\focht\.wine\drive_c\Program Files\Jowood\Gothic II\g2setup.dll File Type : 32-Bit Dll (Subsystem : Win GUI / 2), Size : 229376 (038000h) Byte(s) [File Heuristics] -> Flag : 00000000000000001100001100000001 (0x0000C301) [!] Armadillo v2.xx - v3.xx detected ! [i] Splash Setting (0x0) -> NONE / Duration : 00 second(s) [CompilerDetect] -> Visual C/C++ - Scan Took : 0.560 Second(s) --- snip --- Various anti-debugging/hacking trickery which of course can be bypassed ;-) Additionally to physical media analysis of GOTHIC2_CD3 (part of SecuROM) there are "custom" checks present. Most of them are simple ones like comparing disc serial against hard coded value. The directory structure is read from disc and CRC32 checksums are calculated for specific files. --- snip --- ATTRIBS SIZE CRC32 NAME ____________________________________________________________ A R 362,789,836 5f4cb098 Gothic2-Setup.W03 A R 123 JoWooD Homepage.url A R 14,816 Readme.txt A R 289 Register.url A R 5,569 eula.txt A R 766 gothic2.ico (DIR) 168,146,318 ArxDemo A R 31,744 1d31e5c6 ArxDemo\Setup.exe A R 877 ArxDemo\Setup.ini A R 1,304,064 ArxDemo\Setup.msi A R 101,581,207 ArxDemo\Setup1.cab A R 23,103,162 ArxDemo\Setup2.cab A R 31,150,491 ArxDemo\Setup3.cab A R 1,309,184 ArxDemo\Setup_Deutsch.msi A R 1,708,856 34ec21c0 ArxDemo\instmsi.exe A R 1,822,520 e5563f20 ArxDemo\instmsiw.exe ... Scan time.............[ 551.80] Seconds (551801 ms) Total files...........[ 35] Files Total size............[ 558,468,227] Bytes Dirlist Text CRC32....[ 0032ED44] ____________________________________________________________ ... --- snip --- Forget the scan time, I was debugging at this place ;-) What makes things complicated is that additionally to per-file CRC32 checksums there is also a CRC32 for the whole in-memory folder listing content generated (see snippet) -> "Dirlist Text CRC32" (in snippet the address is printed, not the value itself). That overall checksum didn't match the internal hard-coded one. One problem is that Wine adds FILE_ATTRIBUTE_ARCHIVE by default to each file info entry. Source: http://source.winehq.org/git/wine.git/blob/c7cb3e6cb21598281cf2b00b2ccd8323… This is breaks the overall checksum (file entries get "A" added). The file entries have to have only FILE_ATTRIBUTE_READONLY set -> "R" (reside on CDROM). If FILE_ATTRIBUTE_ARCHIVE is omitted (and a second bug is worked around) the verification process exits with exit code 0x777 which allows the main installer to successfully finish. --- There is another problem, a bug in the verification tool that unfortunately prevents a "full" fix for this installer. One of the file CRC32 on CD3 gets miscalculated ... It took me some hours to find out this brain damage after successfully pinning down the first problem and wondering why the overall checksum still didn't match. --- snip --- ... 0024:Call KERNEL32.CreateFileA(0032d734 "D:\\arxdemo\\setup.exe",80000000,00000001,00000000,00000003,00000080,00000000) ret=00d86dfe 0024:Ret KERNEL32.CreateFileA() retval=000000a0 ret=00d86dfe 0024:Call KERNEL32.GetFileSize(000000a0,00000000) ret=00d87385 0024:Ret KERNEL32.GetFileSize() retval=00007c00 ret=00d87385 0024:Call ntdll.RtlAllocateHeap(00fc0000,00000000,00010010) ret=00be379b 0024:Ret ntdll.RtlAllocateHeap() retval=00fc5238 ret=00be379b 0024:Call KERNEL32.ReadFile(000000a0,00fc5238,00010000,00bfd464,00000000) ret=00d8725a 0024:Ret KERNEL32.ReadFile() retval=00000001 ret=00d8725a 0024:Call ntdll.RtlFreeHeap(00fc0000,00000000,00fc5238) ret=00be3cea 0024:Ret ntdll.RtlFreeHeap() retval=00000001 ret=00be3cea 0024:Call KERNEL32.CloseHandle(000000a0) ret=00d8743a 0024:Ret KERNEL32.CloseHandle() retval=00000001 ret=00d8743a --- snip --- "D:\\arxdemo\\setup.exe" is the culprit. The file size is 31744 bytes. The tool allocates a fixed 65552 byte buffer to read file content in and calculate checksum on (actually 65536 bytes are used for checksumming). Prior to calculating file CRC32 for "D:\\arxdemo\\setup.exe", the 350 MiB "D:\\gothic2-setup.w03" was checksummed in 0x10000 byte chunks. It had the same heap chunk address 0x00fc5238 which was later reused for "D:\\arxdemo\\setup.exe" checksum calculation. --- snip --- 0024:Call KERNEL32.CreateFileA(0032dd20 "D:\\gothic2-setup.w03",80000000,00000001,00000000,00000003,00000080,00000000) ret=00d86dfe 0024:Ret KERNEL32.CreateFileA() retval=0000009c ret=00d86dfe 0024:Call KERNEL32.GetFileSize(0000009c,00000000) ret=00d87385 0024:Ret KERNEL32.GetFileSize() retval=159fbbcc ret=00d87385 0024:Call ntdll.RtlAllocateHeap(00fc0000,00000000,00010010) ret=00be379b 0024:Ret ntdll.RtlAllocateHeap() retval=00fc5238 ret=00be379b 0024:Call KERNEL32.ReadFile(0000009c,00fc5238,00010000,00bfd464,00000000) ret=00d8725a 0024:Ret KERNEL32.ReadFile() retval=00000001 ret=00d8725a ... 0024:Call KERNEL32.ReadFile(0000009c,00fc5238,00010000,00bfd464,00000000) ret=00d8725a 0024:Ret KERNEL32.ReadFile() retval=00000001 ret=00d8725a 0024:Call KERNEL32.ReadFile(0000009c,00fc5238,00008480,00bfd464,00000000) ret=00d8725a 0024:Ret KERNEL32.ReadFile() retval=00000001 ret=00d8725a 0024:Call KERNEL32.ReadFile(0000009c,00fc5238,00008480,00bfd464,00000000) ret=00d8725a 0024:Ret KERNEL32.ReadFile() retval=00000001 ret=00d8725a 0024:Call ntdll.RtlFreeHeap(00fc0000,00000000,00fc5238) ret=00be3cea 0024:Ret ntdll.RtlFreeHeap() retval=00000001 ret=00be3cea ... --- snip --- Because the heap chunk was reused and the buffer was fully filled from previous run, the read file operation for "D:\\arxdemo\\setup.exe" only initialized the first 31744 bytes. The checksum is done on 65536 bytes of buffer which includes garbage/leftover already present. Well, the developers didn't bother to pass HEAP_ZERO_MEMORY to RtlAllocateHeap() or explicit memset(ptr, 0, nbytes). The tool "works" in Windows most likely due to different heap manager implementation/usage or maybe application shims. They put some effort into anti-debugging/hacking/copying and rambled about piracy in hidden messages (found while debugging) ... and yet managed to put in such bugs. You can hack/patch ntdll.dll RtlAllocateHeap() -> forcing flags |= HEAP_ZERO_MEMORY to verify the installer really works after fixing the first problem. Regards -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 5

[Bug 30421] New: obscure GDI Message Queue + Focus bug
by wine-bugs＠winehq.org 05 Sep '19

05 Sep '19

http://bugs.winehq.org/show_bug.cgi?id=30421 Bug #: 30421 Summary: obscure GDI Message Queue + Focus bug Product: Wine Version: 1.3.37 Platform: x86 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: lkcl(a)lkcl.net Classification: Unclassified ok, this is a rather obscure bug that i've been experiencing for quite some time, but hadn't got a handle on any of the functions being used, so couldn't really provide a useful report. as i am now running pyjamas-desktop which uses python ctypes bindings to gdi32 (!!) i can now provide useful feedback. please see the following file, the GetMessage loop towards the end: http://pyjs.org/pygit/#file=pyjd/mshtml.py&id=0d4b6787d01c3d90f2c8801c5c4c4… basically, the repro is as follows: * run wine python.exe Hello.py (an example) * move the mouse out of the GDI32 (IWebBrowser2 OLE control) window * move the mouse into the xterm where wine was fired up * press ctrl-c now, _normally_, under xp, nothing happens... and nothing _should_ happen... ever. ctrl-c should be *completely* ignored. but, continue: * move the mouse _back_ into the GDI32 window * the program exits the exit should *not* be happening when the mouse is moved back. which brings me on to the various experiences that i've had with focus issues, especially dialog boxes. maybe this is because i'm using fvwm2, but the behaviour of focus on dialog boxes and especially inputboxes within dialog boxes is _very_ erratic. the mouse is needed to be moved into the parent window in order sometimes for keyboard focus to arrive on the inputbox within the dialog box! sometimes that doesn't work, and the mouse is required to be moved onto the bar at the top of the parent window (the one with the title - the x-windows / fvwm window bar!) all very odd, and probably a different bug, but hey :) -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 9

[Bug 19182] New: Allow to completely disable MIME-type and application integration
by wine-bugs＠winehq.org 01 Sep '19

01 Sep '19

http://bugs.winehq.org/show_bug.cgi?id=19182 Summary: Allow to completely disable MIME-type and application integration Product: Wine Version: 1.1.25 Platform: PC OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: diafero(a)arcor.de Note: I could not find a component "winemenubilder" which is why I selected "-unknown". It would be great to have a way to tell wine not to create a whole bunch of files across the .config and .local directories. I don't want any mime-type which is registered by a windows application run through wine to reflect on my system - if I really want to use wine to open a certain filetype (which never happened so far) I will do that manually. Nor do I want all the unnecessary start menu entries to be added to my KDE menu - they end up in the wrong category anyway. I like to have the control of creating these entries myself. For the menu it worked fine to simply remove the "Wine" menu directory with the menu editor. The annoying files still got created, but they were ignored. However now wine started to spam my mime directory, too, and it constantly re-creates the files in there - a really annoying behaviour. This not only leads to the danger of me accidentally starting wine without wanting to do so, it also drives some applications crazy (for example the KDE screenshot application no longer recognizes ".jpg" to be a JPEG extension since wine added .jfif). I understand that many people want wine to just magically integrate into their Linux environment, but there are also people like me who switched to Linux because they like to have control, and Windows often prefers its own preferences over the user's. Wine should really not bring that to Linux, or it will get as annoying as Windows is. If I could not patch wine myself to no longer create these files, I had no way to stop all these mime types from being added, no configuration panel to remove them, nothing. I do not even know which of my many WINEPREFIXes created the files. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 102

[Bug 22383] New: Dragon Age: Origins - massive flickering when StrictDrawOrdering is disabled
by wine-bugs＠winehq.org 30 Aug '19

30 Aug '19

http://bugs.winehq.org/show_bug.cgi?id=22383 Summary: Dragon Age: Origins - massive flickering when StrictDrawOrdering is disabled Product: Wine Version: 1.1.43 Platform: x86-64 OS/Version: Linux Status: NEW Keywords: regression Severity: minor Priority: P2 Component: directx-d3d AssignedTo: wine-bugs(a)winehq.org ReportedBy: austinenglish(a)gmail.com CC: hverbeet(a)gmail.com Using wine-1.1.42-390-g4fc6512 + the patch from bug 22202, there is massive flickering of vertical black lines across the intro video. Disabling StrictDrawOrdering fixes it. Can't seem to get a screenshot, but if you really need one, let me know. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 33

[Bug 33834] New: user32: multiline edit control renders tabs not correctly using usp10
by wine-bugs＠winehq.org 30 Aug '19

30 Aug '19

http://bugs.winehq.org/show_bug.cgi?id=33834 Bug #: 33834 Summary: user32: multiline edit control renders tabs not correctly using usp10 Product: Wine Version: 1.6-rc2 Platform: x86 OS/Version: Linux Status: UNCONFIRMED Severity: minor Priority: P3 Component: user32 AssignedTo: wine-bugs(a)winehq.org ReportedBy: bernhardu(a)vr-web.de Classification: Unclassified Created attachment 44856 --> http://bugs.winehq.org/attachment.cgi?id=44856 minimal test program An multiline edit control from user32 with custom tab widths set by window message EM_SETTABSTOPS gets the tabs not rendered correctly. Only the first and the last "column" is visible. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 10

[Bug 13610] New: DragonRiders: main character, dragon and sky appear black
by wine-bugs＠winehq.org 28 Aug '19

28 Aug '19

http://bugs.winehq.org/show_bug.cgi?id=13610 Summary: DragonRiders: main character, dragon and sky appear black Product: Wine Version: 1.0-rc3 Platform: Other URL: http://appdb.winehq.org/objectManager.php?sClass=version &iId=8946 OS/Version: other Status: UNCONFIRMED Severity: enhancement Priority: P2 Component: directx-d3d AssignedTo: wine-bugs(a)winehq.org ReportedBy: rainmaker(a)icgag.it Created an attachment (id=13574) --> (http://bugs.winehq.org/attachment.cgi?id=13574) How it looks in Wine The main character and the dragon in the DragonRiders game appear completely black except for their eyes, as if they had no texture applied (check the attached screenshots). From a bird's view of the village at the beginning, it looks like the sky suffers from the same problem. I couldn't verify whether this applies to other characters in the game as well or not, since I only played for a few minutes. The same bug happens in Cedega as well. I can only guess it's some kind of chroma key problem (as the one in Ultima IX), but the game cannot be configured with respect to that. The only hint is a debug error that floods the console (check attached log): err:d3d_surface:IWineD3DBaseSurfaceImpl_BltFast DXTC decompression not supported by now but I don't know if it might be related. There's also a message concerning an unsupported Stippled Alpha feature. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 37

[Bug 26617] New: Need for Speed Shift 2: large black squares appear in cockpit, obscuring much of the screen
by wine-bugs＠winehq.org 28 Aug '19

28 Aug '19

http://bugs.winehq.org/show_bug.cgi?id=26617 Summary: Need for Speed Shift 2: large black squares appear in cockpit, obscuring much of the screen Product: Wine Version: 1.3.16 Platform: x86 OS/Version: Linux Status: NEW Severity: normal Priority: P2 Component: directx-d3d AssignedTo: wine-bugs(a)winehq.org ReportedBy: dank(a)kegel.com It's as if part of the screen is censored. The black blocks usually show up after about twenty or forty seconds of racing. They're there even on lowest settings. Makes game unplayable. There are quite a few messages like this: fixme:d3d_surface:surface_load_location Downloading RGB surface 0x79f9eb50 to reload it as sRGB. fixme:d3d_surface:IWineD3DBaseSurfaceImpl_Blt Filter WINED3DTEXF_LINEAR not supported in software blit. fixme:d3d_surface:IWineD3DBaseSurfaceImpl_Blt Stretched blit not implemented for bpp 64! -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 46

[Bug 23081] New: Dragon Age: Origins - shadow under character flickers
by wine-bugs＠winehq.org 28 Aug '19

28 Aug '19

http://bugs.winehq.org/show_bug.cgi?id=23081 Summary: Dragon Age: Origins - shadow under character flickers Product: Wine Version: 1.2-rc2 Platform: x86 OS/Version: Linux Status: NEW Severity: normal Priority: P2 Component: directx-d3d AssignedTo: wine-bugs(a)winehq.org ReportedBy: dank(a)kegel.com The shadow under the character flickers badly, especially in the character selector before the game launches. (This is with winetricks strictdrawordering=enabled as per bug 22383 ) It is present at least as far back as wine-1.1.36, but it gets harder to tell as you go further back, because of other problems. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 14

[Bug 23730] New: Dragon Age: Origins, black flickering in opening menu screen when mouse moves
by wine-bugs＠winehq.org 28 Aug '19

28 Aug '19

http://bugs.winehq.org/show_bug.cgi?id=23730 Summary: Dragon Age: Origins, black flickering in opening menu screen when mouse moves Product: Wine Version: 1.2 Platform: x86 OS/Version: Linux Status: NEW Severity: normal Priority: P2 Component: msvcrt AssignedTo: wine-bugs(a)winehq.org ReportedBy: dank(a)kegel.com Now that bug 22200 is fixed, Dragon Age Online no longer needs native msvcr80. But when you move the mouse, there are black bars in the game's opening menu page. You can work around this with WINEDEBUG=-msvcrt. The messages this suppresses look like this: fixme:msvcrt:_controlfp_s (0x518e8f8 655391 196608) semi-stub fixme:msvcrt:__crtGetStringTypeW (unk 0, type 1, wstr 0x518e58c(1), 0x518e590) partial stub This seems to be a concrete case of a verbose FIXME causing suboptimal performance. Should we a) declare this to be not a problem because the user is supposed to read the manual or appdb and know to use WINEDEBUG=-all when running this app? b) declare this to be not a problem because the distro is supposed to patch or wrap wine to always run apps with WINEDEBUG=-all? c) temporarily mark the FIXMEs that cause the performance problem to be displayed once only? or d) something else? -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 10

[Bug 29397] New: Flickering black boxes over models in Orcs Must Die!
by wine-bugs＠winehq.org 28 Aug '19

28 Aug '19

http://bugs.winehq.org/show_bug.cgi?id=29397 Bug #: 29397 Summary: Flickering black boxes over models in Orcs Must Die! Product: Wine Version: 1.3.35 Platform: x86 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: jeff(a)deserettechnology.com Classification: Unclassified Created attachment 38054 --> http://bugs.winehq.org/attachment.cgi?id=38054 Black boxes covering player models While playing Orcs Must Die!, flickering black boxes appear on the edges of many models, including background models that are part of the arena as well as character models. These black boxes rarely stay in place long enough to capture with a screenshot, but I have one that shows two such boxes. They change position as the player moves and animations execute, but the boxes are present in almost every frame. Tested with 1.1.35 and 962230064d and got identical results. I have never tried to play with earlier versions so I don't know if this is a regression or not. The attached screenshot shows a black box over the archer's head and the player's elbow. The box on the archer is permanent, the box on the players elbow flickers as movement animations occur. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 14