- Wine-Bugs - list.winehq.org

[Bug 52554] New: Serbian locale mapping cause crash on multiple installshield wizards
by WineHQ Bugzilla May 6, 2022

May 6, 2022

https://bugs.winehq.org/show_bug.cgi?id=52554 Bug ID: 52554 Summary: Serbian locale mapping cause crash on multiple installshield wizards Product: Wine Version: 7.2 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: major Priority: P2 Component: ntdll Assignee: wine-bugs(a)winehq.org Reporter: lei.pero(a)gmail.com Distribution: --- Since commits: kernelbase: Map LANG_SERBIAN_NEUTRAL in ConvertDefaultLocale(). f51e44c1802338cdd41b38efe2757b642b619d6f ntdll: Map LANG_SERBIAN_NEUTRAL in RtlLocaleNameToLcid(). db3e08770fe9bfb6f06a71761e48a40fe9764058 starting with wine-6.2 - when using locale sr_RS.UTF-8, some (specifically older) installers do crash, for example 3DMark03, preventing installation, passing LANG=en_US.UTF-8 EV works as expected with no crash. Reverting those two specific commits solves the issue. Or changing line in: dlls/ntdll/locale.c from: case MAKELANGID( LANG_SERBIAN, SUBLANG_NEUTRAL ): *lcid = LANG_SERBIAN_NEUTRAL; to: case MAKELANGID( LANG_SERBIAN, SUBLANG_NEUTRAL ): *lcid = MAKELCID( 0x02, SORT_DEFAULT ); solves the issue as well, but that basically renders all those commits useless I assume? On the topic, but unrelated to this bug, I can see changes in dlls/kernel32/nls/srl.nls, with commit: kernel32: Update sr-Latn locale definition. 3b3dfda59951b0f42e297f2b9a31ded04a98d4b4 that doesn't seem right to me, srl.nls should stand for Serbian Latin I assume, srm.nls probably stands for m as Montenegro. So, if I'm correct, addition of srm.nls with commit: kernel32: Add sr-Latn-RS locale definition. db2666e9d20f80968ff6b4b0ea1deae20c3c368b should have changes made to srl.lns. I don't know how this is solved for English for example, since neutral should have a specific code not related to any country where language is used (e.g. US, UK, CA etc.), and the code it seems to be sr_CS as it is changed for sr_RS latin locale, where it shouldn't be changed I assume. -- 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

[Bug 51181] New: d3d10core:d3d10core fails systematically on AMD GPUs
by WineHQ Bugzilla May 6, 2022

May 6, 2022

https://bugs.winehq.org/show_bug.cgi?id=51181 Bug ID: 51181 Summary: d3d10core:d3d10core fails systematically on AMD GPUs Product: Wine Version: 6.8 Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: directx-d3d Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com test_instanced_draw() fails systematically in d3d10core:d3d10core on AMD GPUs (and sometimes there is a crash that follows): d3d10core.c:9275: Test failed: Got 0xfff0f010, expected 0xff80f010 at (160, 0), sub-resource 0. d3d10core.c:9275: Test failed: Got 0xfff0f040, expected 0xff80f040 at (240, 0), sub-resource 0. d3d10core.c:9275: Test failed: Got 0xffaaaacc, expected 0xffbbaacc at (480, 0), sub-resource 0. d3d10core.c:9275: Test failed: Got 0xffaaaa90, expected 0xffbbaa90 at (560, 0), sub-resource 0. Notice how this failure does not happen on the machines that have other GPUs such as cw-gtx560 or QEmu VMs, and how this does not depend on the Windows version or Radeon driver version for that matter: https://test.winehq.org/data/patterns.html#d3d10core:d3d10core So it looks like the test expects fails to account for the Radeon driver results for some reason. The commit that introduced this test is: commit fcd549345d96109179b125d31baf9e9073ecf643 Author: Józef Kucia <jkucia(a)codeweavers.com> AuthorDate: Mon Nov 6 10:55:20 2017 +0100 Commit: Alexandre Julliard <julliard(a)winehq.org> CommitDate: Mon Nov 6 19:37:09 2017 +0100 d3d10core/tests: Add test for SV_InstanceID. Signed-off-by: Józef Kucia <jkucia(a)codeweavers.com> Signed-off-by: Henri Verbeet <hverbeet(a)codeweavers.com> Signed-off-by: Alexandre Julliard <julliard(a)winehq.org> -- 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

[Bug 43995] New: Uplay - Assassin's Creed 4 Black Flag won't start
by wine-bugs＠winehq.org May 6, 2022

May 6, 2022

https://bugs.winehq.org/show_bug.cgi?id=43995 Bug ID: 43995 Summary: Uplay - Assassin's Creed 4 Black Flag won't start Product: Wine-staging Version: 2.20 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: major Priority: P2 Component: -unknown Assignee: wine-bugs(a)winehq.org Reporter: mi.bar.2001(a)gmail.com CC: erich.e.hoover(a)wine-staging.com, michael(a)fds-team.de, sebastian(a)fds-team.de Distribution: --- Created attachment 59652 --> https://bugs.winehq.org/attachment.cgi?id=59652 Terminal log from `wine AC4BFSP.exe` After clicking the "Play" button in Ubisoft Game Launcher (or launching `wine AC4BFSP.exe`), a window pops up informing, that the game is starting, but it disappears and the game doesn't launch. I don't have the Steam version, just the Uplay one. I've set Windows version to XP, wineprefix is 32-bit. Wine version will probably be visible when I post this, but I'm going to mention it just to be safe. Using Wine 2.20 staging -- 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 38

[Bug 52856] New: kernel32:process fails when Wine has no Unix terminal
by WineHQ Bugzilla May 6, 2022

May 6, 2022

https://bugs.winehq.org/show_bug.cgi?id=52856 Bug ID: 52856 Summary: kernel32:process fails when Wine has no Unix terminal Product: Wine Version: unspecified Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: kernel32 Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- kernel32:console fails on the TestBot Wine machines: process.c:1552: Test failed: Opening ConIn process.c:1553: Test failed: Opening ConOut process.c:1556: Test failed: Getting sb info process.c:1557: Test failed: Getting console in mode process.c:1558: Test failed: Getting console out mode process.c:212: 5 failures in child process process.c:1569: Test failed: Getting sb info process.c:1570: Test failed: Getting console in mode process.c:1571: Test failed: Getting console out mode process.c:1582: Test failed: GetLastError: expecting 87 got 12 process.c:1592: Test failed: GetLastError: expecting 87 got 12 process.c:1621: Test failed: Console:SizeX expected 30870, but got 0 process.c:1622: Test failed: Console:SizeY expected 31489, but got 0 process.c:1623: Test failed: Console:CursorX expected 32644, but got 0 process.c:1624: Test failed: Console:CursorY expected 94, but got 0 process.c:1625: Test failed: Console:Attributes expected 0, but got 1 process.c:1626: Test failed: Console:winLeft expected 0, but got 1 process.c:1627: Test failed: Console:winTop expected 4294964424, but got 0 process.c:1628: Test failed: Console:winRight expected 49, but got 0 process.c:1629: Test failed: Console:winBottom expected 30870, but got 0 process.c:1630: Test failed: Console:maxWinWidth expected 31489, but got 0 process.c:1631: Test failed: Console:maxWinHeight expected 11256, but got 0 process.c:1634: Test failed: Console:InputMode expected 0, but got 1 process.c:1635: Test failed: Console:OutputMode expected 7211040, but got 0 process.c:1639: Test failed: Wrong console CP (expected 1252 got 0/0) process.c:1640: Test failed: Wrong console-SB CP (expected 1252 got 0/0) process.c:1645: Test failed: Wrong console mode process.c:1646: Test failed: Wrong console-SB mode https://test.winehq.org/data/patterns.html#kernel32:process The reason is that the TestBot starts Wine from a daemon process which does not have an associated xterm / Konsole / GNOME terminal. This causes Wine to have to allocate a Wine console and that's when things go wrong. A bisect shows that the TestBot is only impacted by this bug since the commit below: commit eff83cd4e00e8ccd4daad05bdd6d76e77664ddd1 Author: Eric Pouech <eric.pouech(a)gmail.com> Date: Mon Apr 11 15:05:09 2022 +0200 kernelbase: Provide a pseudo console environment for initial CUI processes not tied to a Unix tty. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=52771 Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=52761 Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=52743 Signed-off-by: Eric Pouech <eric.pouech(a)gmail.com> Signed-off-by: Jacek Caban <jacek(a)codeweavers.com> Signed-off-by: Alexandre Julliard <julliard(a)winehq.org> But as the commit entry shows this commit also fixed a lot of other issues in the TestBot results (not least of which is getting the wow64 WineTest results in an acceptable shape for test.winehq.org) Also this issue has happened before as evidenced by bugs 25284 and 28220. However those bugs are very old (~2014) so the issue may have been fixed and reappeared (the TestBot has not used a Unix terminal to run the Wine tests since at least 2018-06 (date on local-testagentd.service) and probably 2017 or more (f1d368722b65 tools commit)). Bug 28220 does have a more recent and interesting comment by Esme with regards to handle inheritance: https://bugs.winehq.org/show_bug.cgi?id=28220#c3 -- 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 52559] New: kernel32:resource times out since 2021-12-29 at two win10 testbot systems.
by WineHQ Bugzilla May 6, 2022

May 6, 2022

https://bugs.winehq.org/show_bug.cgi?id=52559 Bug ID: 52559 Summary: kernel32:resource times out since 2021-12-29 at two win10 testbot systems. Product: Wine Version: 7.2 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: kernel32 Assignee: wine-bugs(a)winehq.org Reporter: bernhardu(a)mailbox.org Distribution: --- I wondered because the windows test patterns showed timeouts for the kernel32:resource [1] since some days. I tried a run with verbose timestamps and got [2]. There it shows gaps of ~6 seconds, and it looks like the delay originates from an execution of EndUpdateResourceA [3]. Unfortunately there are a lot calls for this function and therefore the test times out. Searching the net led me to the autohotkey forums where this function was associated with windows defender. Is windows defender active at these VMs? [1] https://test.winehq.org/data/tests/kernel32:resource.html [2] https://testbot.winehq.org/JobDetails.pl?Key=108231&f201=exe32.report#k201 [3] https://source.winehq.org/git/wine.git/blob/781277de62822c2bac23dc892a99665… -- 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 52079] New: oleacc:main crashes randomly on Windows 10 1709+
by WineHQ Bugzilla May 6, 2022

May 6, 2022

https://bugs.winehq.org/show_bug.cgi?id=52079 Bug ID: 52079 Summary: oleacc:main crashes randomly on Windows 10 1709+ Product: Wine Version: unspecified Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: oleacc Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com oleacc:main crashes randomly on Windows 10 1709+ in both 32- and 64-bit: main.c:1041: Test failed: V_VT(&v) = 0 main.c:1048: Test failed: V_VT(&v) = 0 main.c:1049: Test failed: V_DISPATCH(&v) = 0000000000000000 main.c:1049: this is the last test seen before the exception 1050:main: unhandled exception c0000005 at 0000000000401597 oleacc:main:1050 done (-1073741819) in 0s Test failed: crash (c0000005) https://test.winehq.org/data/patterns.html#oleacc:main Unsurprisingly a bisect shows that the tests started failing with the commit that introduced the test: commit 1a3db363c6b4065af532a8db1e6309be753b1781 Author: Connor McAdams <cmcadams(a)codeweavers.com> Date: Mon Sep 20 18:03:31 2021 +0200 oleacc: Add Client_get_accFocus tests. Signed-off-by: Connor McAdams <cmcadams(a)codeweavers.com> Signed-off-by: Piotr Caban <piotr(a)codeweavers.com> Signed-off-by: Alexandre Julliard <julliard(a)winehq.org> At the time the crash happened after line 1037 which points to line 1039 as a likely culprit: 1032 /* Set focus to each child window. */ 1033 SetFocus(btn); 1034 hr = IAccessible_get_accFocus(acc, &v); 1035 ok(hr == S_OK, "hr %#x\n", hr); 1036 ok(V_VT(&v) == VT_DISPATCH, "V_VT(&v) = %d\n", V_VT(&v)); 1037 ok(V_DISPATCH(&v) != NULL, "V_DISPATCH(&v) = %p\n", V_DISPATCH(&v)); 1038 1039 hr = IDispatch_QueryInterface(V_DISPATCH(&v), &IID_IOleWindow, (void**)&ow); 1040 ok(hr == S_OK, "got %x\n", hr); That call has been moved to _check_acc_hwnd() and so the crash now happens in the first QueryInterface() call in that function: hr = IUnknown_QueryInterface(unk, &IID_IOleWindow, (void**)&ow); called from check_acc_hwnd((IUnknown*)V_DISPATCH(&v), btn); -- 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 52904] New: kernel32:locale breaks oleaut32:varformat on Wine in the French locale
by WineHQ Bugzilla May 6, 2022

May 6, 2022

https://bugs.winehq.org/show_bug.cgi?id=52904 Bug ID: 52904 Summary: kernel32:locale breaks oleaut32:varformat on Wine in the French locale Product: Wine Version: unspecified Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: kernel32 Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- kernel32:locale breaks oleaut32:varformat on Wine in the French locale: varformat.c:642: Test failed: Unexpected hr 0x80020005 for L"0,000 EUR" varformat.c:649: Test failed: Unexpected hr 0x80020005 for L"0,000 EUR" https://test.winehq.org/data/patterns.html#oleaut32:varformat When starting from a clean wineprefix one would normally get "0,000 €" for this test. But if one runs kernel32:locale before the Euro symbol is replaced with "EUR" which is then not recognized as a currency symbol. A bisect shows that the failures started with this commit: commit 78a33f968461fa3a01d3eec837b4c24f353bc00b Author: Alexandre Julliard <julliard(a)winehq.org> AuthorDate: Tue Mar 29 22:12:25 2022 +0200 kernelbase: Reimplement SetLocaleInfoW(). Signed-off-by: Alexandre Julliard <julliard(a)winehq.org> The commit modifies both kernel32 and kernel32:locale, and both parts of the patch are needed to reproduce the failure. -- 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 52867] New: System.Globalization.CultureInfo.GetCultureInfo fails for 'en' value
by WineHQ Bugzilla May 6, 2022

May 6, 2022

https://bugs.winehq.org/show_bug.cgi?id=52867 Bug ID: 52867 Summary: System.Globalization.CultureInfo.GetCultureInfo fails for 'en' value Product: Wine Version: 7.6 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: ntdll Assignee: wine-bugs(a)winehq.org Reporter: galtgendo(a)o2.pl Distribution: --- Created attachment 72261 --> https://bugs.winehq.org/attachment.cgi?id=72261 stacktrace (filing under ntdll, as that where relevant files were before the commit series that most likely broke it) After upgrading to 7.6, still occurring in 7.7, a regression was made - the last published dnSpy version had been running relatively well in its basic functionality (I'm actually not any good with it), other than minor display glitches; now it stacktraces immediately. I believe the stacktrace makes it clear what happens and why I suspect that locale commit series are at fault here. That release of dnSpy embeds its mono environment, so it's not a wine-mono problem. -- 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 52858] New: kernel32:file fails in Windows 8.1 on the cw-rx460 machine
by WineHQ Bugzilla May 6, 2022

May 6, 2022

https://bugs.winehq.org/show_bug.cgi?id=52858 Bug ID: 52858 Summary: kernel32:file fails in Windows 8.1 on the cw-rx460 machine 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:file and msvcp140:msvcp140 fail on the cw-rx460 machine: file.c:905: Test failed: Unexpected error 87. file.c:910: Test failed: Unexpected error 87. https://test.winehq.org/data/patterns.html#kernel32:file These are new tests so unsurprisingly the failures are caused by the commit that introduced them: commit 02faaea93b78d5d53976b2955bef7beffa780d02 Author: Paul Gofman <pgofman(a)codeweavers.com> Date: Mon Apr 11 21:44:36 2022 +0300 kernelbase: Reset last error on success in CopyFileExW(). Signed-off-by: Paul Gofman <pgofman(a)codeweavers.com> Signed-off-by: Alexandre Julliard <julliard(a)winehq.org> It's unclear why the test only fails on this configuration. Also note that msvcp140:msvcp140 fails in a very similar way in the exact same test configuration: see bug 52857. -- 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 4

[Bug 51407] New: kernel32:time's test_GetCalendarInfo() fails in Hindi
by WineHQ Bugzilla May 6, 2022

May 6, 2022

https://bugs.winehq.org/show_bug.cgi?id=51407 Bug ID: 51407 Summary: kernel32:time's test_GetCalendarInfo() fails in Hindi Product: Wine Version: 6.10 Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: kernel32 Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- Created attachment 70263 --> https://bugs.winehq.org/attachment.cgi?id=70263 Broken fix and traces Three GetCalendarInfo() checks fail in Hindi: https://test.winehq.org/data/patterns.html#kernel32:time time.c:728: Test failed: GetCalendarInfoA failed err 87 time.c:731: Test failed: GetCalendarInfoA failed err 87 time.c:738: Test failed: got 0, expected 7 The immediate cause for the failure is that Hindi is one of the few languages for which NLS_IsUnicodeOnlyLcid() returns true. In turn this causes the ANSI GetCalendarInfoA() to return ERROR_INVALID_PARAMETER. The last failure just reflects the fact that the two calls to GetCalendarInfoA() should not have failed. Only Hindi fails and only in Wine. But before declaring this a Wine bug we need to look at Windows. * For some reason probably related to SetWinLocale [1] the system locale of w10pro64_hi is still English. * This causes both the ANSI and Unicode GetCalendarInfo() to return English names (Monday). * Manually setting the system locale to Hindi is possible on Windows 10 2004. What's interesting is the disabled and checked-by-default checkbox in that dialog: [X] Beta: Use Unicode UTF-8 for worldwide language support * And indeed doing that and running the test again results in GetCalendarInfoA() returning UTF-8 strings on Windows! time.c:730: ret=19 sdayname1="\xe0\xa4\xb8\xe0\xa5\x8b\xe0\xa4\xae\xe0\xa4\xb5\xe0\xa4\xbe\xe0\xa4\xb0" सोमवार time.c:735: ret2=19 sdayname1="\xe0\xa4\xb8\xe0\xa5\x8b\xe0\xa4\xae\xe0\xa4\xb5\xe0\xa4\xbe\xe0\xa4\xb0" सोमवार time.c:742: ret2=19 sdayname1=L"\0938\094b\092e\0935\093e\0930" * The beta moniker strongly suggests that this is a new option. Indeed selecting Hindi as the system locale was not possible in Windows 8.1 for instance. * This is probably also why SetWinLocale has trouble setting the system locale to Hindi. So the Windows behavior can be summarized as such: * Return localized values for non-Unicode-only locales. * In older Windows versions setting the system locale to a Unicode-only locale is impossible. * In newer Windows versions setting the system locale to a Unicode-only locale is possible and then it uses UTF-8. Commenting out the Hindi line in NLS_IsUnicodeOnlyLcid() fixes the failures in Wine (see attached patch). But this fix is wrong because it causes GetCalendarInfoA() to return question marks instead of intelligible strings. (it also impacts many APIs but that's probably ok based on the Windows behavior) So Wine should do one of the following: 1. Stick to a 'safe' system locale when $LANG is set to a Unicode-only locale. Typically this would be English. 2. Or use UTF-8 when the system locale is a Unicode-only locale. Option 2 would actually make a lot of sense on Unix systems but it is also more likely to break Windows applications. So option 1 may be the safer (and simpler?) option (for now). [1] https://source.winehq.org/git/tools.git/blob/HEAD:/testbot/bin/SetWinLocale… -- 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 7