June 2022 - Wine-bugs - list.winehq.org

[Bug 52995] New: shell32:shelllink crashes in Wine cause shell32:shellpath's test_PathResolve() to fail (test.bat file)
by WineHQ Bugzilla 17 Jun '22

17 Jun '22

https://bugs.winehq.org/show_bug.cgi?id=52995 Bug ID: 52995 Summary: shell32:shelllink crashes in Wine cause shell32:shellpath's test_PathResolve() to fail (test.bat file) Product: Wine Version: unspecified Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: shell32 Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- shell32:shellpath's test_PathResolve() fails seemingly randomly: shellpath.c:3049: Test failed: test 4: expected L"C:\\users\\winetest\\Temp\\test.lnk", got L"C:\\users\\winetest\\Temp\\test.bat" shellpath.c:3049: Test failed: test 8: expected L"C:\\users\\winetest\\Temp\\test.lnk", got L"C:\\users\\winetest\\Temp\\test.bat" shellpath.c:3048: Test failed: test 14: expected 0, got 1 shellpath.c:3049: Test failed: test 14: expected L"C:\\users\\winetest\\Temp\\test", got L"C:\\users\\winetest\\Temp\\test.bat" shellpath.c:3049: Test failed: test 16: expected L"C:\\users\\winetest\\Temp\\test.lnk", got L"C:\\users\\winetest\\Temp\\test.bat" shellpath.c:3049: Test failed: test 19: expected L"C:\\users\\winetest\\Temp\\test.lnk", got L"C:\\users\\winetest\\Temp\\test.bat" shellpath.c:3049: Test failed: test 20: expected L"C:\\users\\winetest\\Temp\\test.lnk", got L"C:\\users\\winetest\\Temp\\test.bat" shellpath.c:3049: Test failed: test 24: expected L"C:\\users\\winetest\\Temp\\test.lnk", got L"C:\\users\\winetest\\Temp\\test.bat" shellpath.c:3048: Test failed: test 26: expected 0, got 1 shellpath.c:3049: Test failed: test 26: expected L"C:\\users\\winetest\\Temp\\test", got L"C:\\users\\winetest\\Temp\\test.bat" https://test.winehq.org/data/patterns.html#shell32:shellpath But it turns out the dates when shell32:shellpath failures all match dates when shell32:shelllink crashed: https://test.winehq.org/data/patterns.html#shell32:shelllink What happens is that shell32:shellpath looks for a "test.lnk" file. But if one creates a "c:\users\...\Temp\test.bat" file before starting the test, that's what gets returned instead, thus causing the failures. shell32:shelllink just happens to use a file by that name. Normally it gets cleaned up but that just does not happen when shell32:shelllink crashes! So I think the right fix would be to make shell32:shellpath more resilient: * Delete "test.*" before checking for test.lnk. But that runs the risk of deleting files dear to the person running the test (they are in a Temp directory but still). * Use some GetTempFileName() or some similar API to generate a unique filename. * Or just use a name that's less likely to be a leftover from a previous test, "shellpath.lnk" for instance. Very basic but maybe sufficient. -- 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 6

[Bug 52998] New: xaudio2_7:xaudio2 fails on Windows 1909+
by WineHQ Bugzilla 17 Jun '22

17 Jun '22

https://bugs.winehq.org/show_bug.cgi?id=52998 Bug ID: 52998 Summary: xaudio2_7:xaudio2 fails on Windows 1909+ Product: Wine Version: unspecified Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: xaudio2 Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com xaudio2_7:xaudio2 has the same set of systematic failures on all Windows 1909+ versions: xaudio2.c:1096: Test failed: xaudio2_8.dll: CreateFX({a90bc001-e897-e897-7439-435500000000}) failed: 8007000e xaudio2.c:1096: Test failed: xaudio2_8.dll: CreateFX({a90bc001-e897-e897-7439-435500000001}) failed: 8007000e xaudio2.c:1096: Test failed: xaudio2_8.dll: CreateFX({a90bc001-e897-e897-7439-435500000002}) failed: 8007000e xaudio2.c:1096: Test failed: xaudio2_8.dll: CreateFX({a90bc001-e897-e897-7439-435500000003}) failed: 8007000e xaudio2.c:1116: Test failed: xaudio2_8.dll: CreateFX(CLSID_AudioVolumeMeter) failed: 8007000e https://test.winehq.org/data/patterns.html#xaudio2_7:xaudio2 The error code is E_OUTOFMEMORY and the failures correspond to the following formats: CLSID_FXEQ27 CLSID_FXMasteringLimiter27 CLSID_FXReverb27 CLSID_AudioVolumeMeter Strangely enough the test succeeds on Windows 10 1507, 1607, 1709 and 1809 without any visible skip. So it's as if support for these formats was removed (but that's a strange error code to say so), or something more must be done to get them initialized properly. -- 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 34176] New: Kernel32:CompareStringW strange behavior ?
by wine-bugs＠winehq.org 17 Jun '22

17 Jun '22

http://bugs.winehq.org/show_bug.cgi?id=34176 Bug #: 34176 Summary: Kernel32:CompareStringW strange behavior ? Product: Wine Version: 1.6 Platform: x86-64 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: kernel32 AssignedTo: wine-bugs(a)winehq.org ReportedBy: wine.manette(a)manette.be Classification: Unclassified Created attachment 45460 --> http://bugs.winehq.org/attachment.cgi?id=45460 Sorce and compiled executable for wineTest.exe + winedbg log.txt Hi all, I'm running wine 1.6 (i386) under Debian 7.1 (x86_64). When running a console program of mine (Delphi XE2 compiled under Windows Vista) I'm observing an important behavior difference under wine than nativevely. I narrowed down the symptom to a string comparison of some sort. Source file below is a code snippet to reproduce the problem (source is UTF-8, Delphi is utf-8 / WideString, WideChar aware). Here is the (correct/expected) result when program is run natively under Windows Vista: G:\__Project-2\Code\wineTest>wineTest * TEST-001: TRUE * TEST-002: TRUE * TEST-003: TRUE * TEST-004: TRUE TRUE * TEST-005: TRUE TRUE * TEST-003: TRUE TRUE Here is the same program output when run from Wine: root(a)Bilbo-100:~# wine wineTest.exe Wine cannot find the ncurses library (libncurses.so.5). fixme:thread:GetThreadUILanguage : stub, returning default language. fixme:thread:GetThreadPreferredUILanguages 56, 0x33d8f8, (nil) 0x33d920 fixme:thread:GetThreadPreferredUILanguages 56, 0x33d8f8, (nil) 0x33d920 * TEST-001: TRUE * TEST-002: TRUE * TEST-003: TRUE * TEST-004: TRUE FALSE * TEST-005: TRUE FALSE * TEST-003: TRUE FALSE For what I understood the TStringList IndexOf method may use Windows API's CompareStringW function (http://source.winehq.org/WineAPI/CompareStringW.html) Is this a configuration problem (locale ?) May this be a bug or implementation problem ? I can provide the corrsponding binary wineTest.exe ////////////////////////////////////////////////// Using winedbg I could crosscheck that CompareStringW for two different strings result in equality (CSTR_LESS_THAN = 1, CSTR_EQUAL = 2, CSTR_GREATER_THAN = 3). Maybe this is because of locale setting or flags setup ? root(a)Bilbo-100:~# WINEDEBUG=+relay,+snoop wine wineTest.exe &> log.txt root(a)Bilbo-100:~# grep CompareStringW log.txt 0009:Call KERNEL32.CompareStringW(00000400,00000001,00549cac L"m\00b2",00000002,00449890 L"m\00b2",00000002) ret=004187d0 0009:Ret KERNEL32.CompareStringW() retval=00000002 ret=004187d0 0009:Call KERNEL32.CompareStringW(00000400,00000001,00549cac L"m\00b2",00000002,004498bc L"m2",00000002) ret=004187d0 0009:Ret KERNEL32.CompareStringW() retval=00000002 ret=004187d0 0009:Call KERNEL32.CompareStringW(00000400,00000001,00549cac L"m\00b3",00000002,004498f8 L"m\00b3",00000002) ret=004187d0 0009:Ret KERNEL32.CompareStringW() retval=00000002 ret=004187d0 0009:Call KERNEL32.CompareStringW(00000400,00000001,00549cac L"m\00b3",00000002,0044990c L"m3",00000002) ret=004187d0 0009:Ret KERNEL32.CompareStringW() retval=00000002 ret=004187d0 0009:Call KERNEL32.CompareStringW(00000400,00000001,00549cac L"\00e9\00e9\00e9",00000003,00449920 L"\00e9\00e9\00e9",00000003) ret=004187d0 0009:Ret KERNEL32.CompareStringW() retval=00000002 ret=004187d0 0009:Call KERNEL32.CompareStringW(00000400,00000001,00549cac L"\00e9\00e9\00e9",00000003,00449934 L"eee",00000003) ret=004187d0 0009:Ret KERNEL32.CompareStringW() retval=00000002 ret=004187d0 /////////////////////////////////////////////////// program wineTest; {$APPTYPE CONSOLE} Uses Classes; var sl : TStringList; begin Write(ErrOutput, '* TEST-001: '); if not('m²'='m2') then Writeln(ErrOutput, 'TRUE') else Writeln(ErrOutput, 'FALSE'); Write(ErrOutput, '* TEST-002: '); if not('m³'='m3') then Writeln(ErrOutput, 'TRUE') else Writeln(ErrOutput, 'FALSE'); Write(ErrOutput, '* TEST-003: '); if not('ééé'='eee') then Writeln(ErrOutput, 'TRUE') else Writeln(ErrOutput, 'FALSE'); sl := TStringList.Create; Write(ErrOutput, '* TEST-004: '); sl.Clear; sl.Add('m²'); if not(sl.IndexOf('m²')<0) then Write(ErrOutput, 'TRUE') else Write(ErrOutput, 'FALSE'); Write(ErrOutput, ' '); if sl.IndexOf('m2')<0 then Writeln(ErrOutput, 'TRUE') else Writeln(ErrOutput, 'FALSE'); Write(ErrOutput, '* TEST-005: '); sl.Clear; sl.Add('m³'); if not(sl.IndexOf('m³')<0) then Write(ErrOutput, 'TRUE') else Write(ErrOutput, 'FALSE'); Write(ErrOutput, ' '); if sl.IndexOf('m3')<0 then Writeln(ErrOutput, 'TRUE') else Writeln(ErrOutput, 'FALSE'); Write(ErrOutput, '* TEST-003: '); sl.Clear; sl.Add('ééé'); if not(sl.IndexOf('ééé')<0) then Write(ErrOutput, 'TRUE') else Write(ErrOutput, 'FALSE'); Write(ErrOutput, ' '); if sl.IndexOf('eee')<0 then Writeln(ErrOutput, 'TRUE') else Writeln(ErrOutput, 'FALSE'); end. -- 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.

2 14

[Bug 49232] New: Saints Row The Third Remastered shows floating black rectangles
by WineHQ Bugzilla 17 Jun '22

17 Jun '22

https://bugs.winehq.org/show_bug.cgi?id=49232 Bug ID: 49232 Summary: Saints Row The Third Remastered shows floating black rectangles Product: Wine Version: 5.8 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: directx-d3d Assignee: wine-bugs(a)winehq.org Reporter: rca(a)psy-q.ch Distribution: --- Created attachment 67236 --> https://bugs.winehq.org/attachment.cgi?id=67236 Screenshot showing black rectangles in the scene The game shows floating black rectangles in your view. They don't appear when looking straight down at the ground and also not in indoor scenes. Attached screenshot shows the effect I am guessing it might have something to do with street lamp halos or an effect like that. The log mentions: 0170:fixme:d3d:wined3d_device_update_sub_resource Ignoring flags 0x2. 00d0:err:d3d:wined3d_debug_callback 0x3a1967e0: "GL_INVALID_OPERATION in glCopyImageSubData(internalFormat mismatch)". -- 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 53168] New: kernel32:debugger - test_ExitCode() has a rare timeout condition running dbg,attach,process,event,detach
by WineHQ Bugzilla 17 Jun '22

17 Jun '22

https://bugs.winehq.org/show_bug.cgi?id=53168 Bug ID: 53168 Summary: kernel32:debugger - test_ExitCode() has a rare timeout condition running dbg,attach,process,event,detach Product: Wine Version: unspecified Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: kernel32 Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com kernel32:debugger - test_ExitCode() has a rare timeout condition running dbg,attach,process,event,detach: debugger.c:610: running dbg,attach,process,event,detach... debugger.c:621: waiting for child exit... debugger.c:198: child: crashing... debugger.c:659: waiting for the debugger... debugger.c:166: debugger traces: processing debug messages setting event done debugging... kernel32:debugger:0dcc done (258) in 125s https://test.winehq.org/data/patterns.html#kernel32:debugger The failures are very rare (~0.6%) but always happen in the same location: in the (last) crash_and_debug(... "dbg,attach,process,event,detach") call in test_ExitCode(). This failure has mostly been seen on Windows 10 21H1 but it happened at least once in 1507. So it is likely that all Windows 10 versions are impacted. It also happens in both 32- and 64-bit code. -- 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 1

[Bug 32224] New: LTspice: selected text/component after Ctrl+F not highlighted
by wine-bugs＠winehq.org 16 Jun '22

16 Jun '22

http://bugs.winehq.org/show_bug.cgi?id=32224 Bug #: 32224 Summary: LTspice: selected text/component after Ctrl+F not highlighted Product: Wine Version: unspecified Platform: x86-64 OS/Version: Linux Status: UNCONFIRMED Severity: minor Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: testelor(a)yahoo.com Classification: Unclassified Steps to reproduce: 1) Download and install LTspice - http://ltspice.linear.com/software/LTspiceIV.exe 2) Run LTspice using: $ wine scad3.exe -nowine 3) Create a new schematic, press 'R' (resistor), for example, and place it twice by clicking. There should be R1 and R2 on the schematic. 4) Press Ctrl+F and search for "r2", no quotes, it should be highlighted in yellow, but they're not. Side note: - create a new schematic, press F2 to bring up the device list. - select current, place it on the schematic - right-click on it and type '1', no quotes, then OK - press 'R', place it next to the current source, in parallel - right click on it and replace "R" with '1', no quotes, then OK - press 'G' and place ground on the lower pins of both components - press F3 and draw a net by clicking on the unconnected current source's pin, then on the resistor's - press 'S' and write ".tran 1", no quotes, don't forget the dot, place it on the schematic - in the toolbar, press the 'running man' button, or right-click on the schematic and choose "run" - it should be almost instant, a new empty window will pop up, above - move the mouse cursor over the resistor until the cursor changes into a current-clamp (something black with a red arrow), then click once to display the current through the resistor - in the newly created window, there will be a trace displayed and a label at the top, reading I(R1) Alt+left-clicking on the label will highlight the resistor. That's how it should look like after Ctrl+F. Alternatively, you can just run "wine scad3.exe", no "-nowine" switch, and repeat the 4 steps in the beginning. No screenshot as it doesn't make sense to show what doesn't happen... -- 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.

2 8

[Bug 28282] New: Sound constantly crackling in StarCraft II
by wine-bugs＠winehq.org 15 Jun '22

15 Jun '22

http://bugs.winehq.org/show_bug.cgi?id=28282 Summary: Sound constantly crackling in StarCraft II Product: Wine Version: 1.3.26 Platform: x86-64 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: pitlicek(a)gmail.com Sound is crackling since 1.3.26 (and still does in 27) and console is spammed with "err:alsa:wine_snd_pcm_recover underrun occurred". Using ALSA in Wine (PulseAudio in Ubuntu 11.04), full acceleration (changing to emulation doesn't fix it), 48 kHz (my SoundBlaster can handle it, and again, change makes no difference), 16 bps. Using OSS and padsp results in no sound at all. But it's still better than no sound in 1.3.25 after mmdevapi rewrite/implementation. ;) -- 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.

2 111

[Bug 52951] New: gdi32:font - test_GetGlyphOutline() fails on "VL Gothic" in Wine
by WineHQ Bugzilla 15 Jun '22

15 Jun '22

https://bugs.winehq.org/show_bug.cgi?id=52951 Bug ID: 52951 Summary: gdi32:font - test_GetGlyphOutline() fails on "VL Gothic" in Wine Product: Wine Version: unspecified Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: gdi32 Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- gdi32:font - test_GetGlyphOutline() fails on "VL Gothic" in Wine: font.c:4885: Test failed: expected 36, got 19 (VL Gothic:136) font.c:4903: Test failed: expected 36, got 19 (VL Gothic:136) https://test.winehq.org/data/patterns.html#gdi32:font This begs two questions: 1. Is this a bug in the font, or a bug in the test? 2. Why does this happen on the debian11 (and debiant) VMs, but not on fg-deb64 (my development machine)? On Debian VL Gothic is brought in by this package dependency chain: libsane -> libpoppler-glib8 -> libpoppler102 -> poppler-data -(suggest)-> fonts-vlgothic Interestingly poppler-data has alternatives to fonts-vlgothic one of which is already installed: fonts-ipaexfont-gothic fonts-ipfont-gothic (installed) fonts-ipafont-nonfree-jisx0208 fonts-mona fonts-takao-gothic fonts-umeplus-cl fonts-vlgothic (installed) But the reason for the difference is that fg-deb64 has the AR PL UMing CN font (from the fonts-arphic-uming package) and not the TestBot VMs. The lack of UMing also causes a test to be skipped on the TestBot VMs: 1: (IPAGothic:128) 2: (NanumGothicCoding:129) font.c:4867: Tests skipped: Fixed-pitch TrueType font for charset 134 is not available 4: (VL Gothic:136) So I'm going to install fonts-arphic-uming on the TestBot VMs which should swipe these two failures under the rug. Note: I don't have information on how much this will increase the runtime (see bug 51367) but it seems like it shouldn't be too bad. -- 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 3

[Bug 51433] New: Battle.Net: Installer terminates with BLZBNTBTS0000005C error screen
by WineHQ Bugzilla 15 Jun '22

15 Jun '22

https://bugs.winehq.org/show_bug.cgi?id=51433 Bug ID: 51433 Summary: Battle.Net: Installer terminates with BLZBNTBTS0000005C error screen Product: Wine Version: 6.12 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs(a)winehq.org Reporter: pep(a)bouah.net Distribution: --- Created attachment 70285 --> https://bugs.winehq.org/attachment.cgi?id=70285 Wine logs Hi there, Trying to install Heartstone (Blizzard), it failed when updating the battle.net updater. I reproduced with the Battle.Net Installer directly, downloaded from here: https://www.blizzard.com/en-us/download/confirmation?platform=windows&local… (shasum seems to give me something different each time I download it again) The error dialog the installer shows says "More help: BLZBNTBTS0000005C". Logs from the installer in attachments. I get the same error with wine and wine and wine-staging 6.12 with a fresh profile. -- 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 53160] New: Ancestral Quest 16.0.11: Data from linked FamilySearch shrinks and then disappears
by WineHQ Bugzilla 14 Jun '22

14 Jun '22

https://bugs.winehq.org/show_bug.cgi?id=53160 Bug ID: 53160 Summary: Ancestral Quest 16.0.11: Data from linked FamilySearch shrinks and then disappears Product: Wine Version: 7.6 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: user32 Assignee: wine-bugs(a)winehq.org Reporter: drjk011(a)live.com Distribution: --- Ancestral Quest is software designed for storing and managing personal family history data. When a person is linked to FamilySearch (free account) and 'Review with FamilySearch" is selected, data from FamilySearch is retrieved and the textbox is supposed to be populated so local data can be compared with FamilySearch data. The data momentarily is presented, then shrinks smaller and smaller until it disappears from view. With text data, there are no error messages, just disappearing text. If Ancestral Quest is a registered version, then photos may be compared with FamilySearch as well. In this case, the same action occurs until the there is a divide-by-zero error and the program crashes. Hardware: PclinuxOS 2022 KDE Plasma 5.24.5 Ryzen 5-3600, 23.5 GiB Ram, NVIDIA GTX 1060 Wine 7.6 Software: Ancestral Quest 16.0.11 by Incline Software http://www.ancestralquestonline.com/updates/getlatest.asp?demo=1 Error demonstrated with the “Scrapbook Sample Database” freely downloadable from the publisher http://www.ancquest.com/download.htm Connecting to FamilySearch by the Church of Jesus Christ of Latter Day Saints. One must be registered, but it is a free account. https://www.familysearch.org/en/ Attachments: Four (4) screenshots from Windows of how it is supposed to look: Screenshot_main.png; Screenshot_sub-details.png; Screenshot_sources.png; and Screenshot_sub-photos.png Two (2) mpg files showing what is actually seen: "RealTime.mpg" shows accessing the above mentioned 4 areas, and "Clips slowed 90%.mpg" showing what is happening slow enough to actually evaluate. One (1) backtrace file "backtrace.txt" showing information when the photos cause the program failure. Comment: >From my working with this issue and attempting to find a resolution, to me it seems that this is not exactly a failure of Wine, but rather a procedure or function dealing with size and placement (located in User32.dll) that is not yet fully implemented. I could not find any patches or work that addresses this issue. One backtrace (that I cannot yet duplicate) referenced a WINPOS_SysCommandSizeMove procedure followed by a WINPROC_call_window procedure; indicating that there were missing parameters. I would be happy to provide more details and/or logs if that would be helpful. Thanks... -- 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 8