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

[Bug 50298] New: w1064v1909 shows the "create a Windows account" fullscreen dialog
by WineHQ Bugzilla 19 Dec '22

19 Dec '22

https://bugs.winehq.org/show_bug.cgi?id=50298 Bug ID: 50298 Summary: w1064v1909 shows the "create a Windows account" fullscreen dialog Product: Wine-Testbot Version: unspecified Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- Created attachment 68868 --> https://bugs.winehq.org/attachment.cgi?id=68868 w1064v1909 "Let's finish setting up your device" dialog w1064v1909 corresponds to the 1909-live snapshot of the w1064 VM. That snapshot is created automatically by the TestBot. Currently this snapshot opens on the "create a Windows account" fullscreen dialog, the one with the following title: Let's finish setting up your device It seems this dialog 'pops up' automatically on boot if one does not provide a Microsoft account when installing Windows. When that happens is hard to predict, Windows seems to wait a few days or reboots before doing. Furthermore something WineTest does seems to make this dialog go away because it's not visible in the final WineTest screenshot. However it is visible if one runs a single test such as ntdll:exception. It's not entirely clear if this dialog is interfering with some tests but its presence is definitely unwanted. -- 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 47847] New: Disable Windows 10's "privacy wizard" after locale change
by WineHQ Bugzilla 19 Dec '22

19 Dec '22

https://bugs.winehq.org/show_bug.cgi?id=47847 Bug ID: 47847 Summary: Disable Windows 10's "privacy wizard" after locale change Product: Wine-Testbot Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- The TestBot can automatically set up locale-specific live snapshots (after the locale has been manually installed, see point 1 of bug 47804). However on Windows 10 the mandatory reboot is sometimes followed by a wizard asking a bunch of questions like whether to use audio speech recognition, what kind of reports to send Microsoft, etc. I think this is the "Choose your Privacy Settings" wizard. Of course that breaks the snapshot creation since the TestBot cannot answer these questions, fails to detect they are even being asked (because TestAgentd is already running in the background), and creates a snapshot of the half-baked state. So one question is: how can one prevent Windows from putting up this wizard? But then, while I had this wizard for the French locale, I did not get it for Arabic. Would it perchance only happen for European locales? If so the pain may be bearable. -- 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 51411] New: Add a mixed locale Windows test configuration.
by WineHQ Bugzilla 19 Dec '22

19 Dec '22

https://bugs.winehq.org/show_bug.cgi?id=51411 Bug ID: 51411 Summary: Add a mixed locale Windows test configuration. Product: Wine-Testbot Version: unspecified Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- The TestBot's locale test configurations usually have a matching set of locales for SystemDefaultLCID, USystemPreferredUILanguages, UserDefaultUILanguage and ThreadUILanguage. For instance for French we have: SystemDefaultLCID 040c UserDefaultLCID 0409 ThreadLocale 0409 SystemPreferredUILanguages 040C,0409 UserDefaultUILanguage 040c ThreadUILanguage 040c As a result tests can substitute GetSystemDefaultLCID() for GetThreadUILanguage() and get away with it. In turn this allows Wine to substitute these too without causing the tests to fail (see for instance mlang:mlang, bug 51410). To avoid this the TestBot needs a test configuration where all the important locale settings are different so these issues are detected immediately. -- 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 51451] New: Many Windows 10 VMs are incompletely localized
by WineHQ Bugzilla 19 Dec '22

19 Dec '22

https://bugs.winehq.org/show_bug.cgi?id=51451 Bug ID: 51451 Summary: Many Windows 10 VMs are incompletely localized Product: Wine-Testbot Version: unspecified Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com The localized w10pro64* VMs don't have the right UserDefaultLCID and are sometimes missing even more: w10pro64 fr_FR SystemDefaultLCID 040c UserDefaultLCID 0409 <--- should be 040c ThreadLocale 0409 ??? SystemPreferredUILanguages 040C,0409 UserDefaultUILanguage 040c ThreadUILanguage 040c w10pro64 ar_MA SystemDefaultLCID 0c01 UserDefaultLCID 0409 <--- should be 0c01 ThreadLocale 0409 ??? SystemPreferredUILanguages 0409 <--- should contain 0c01 UserDefaultUILanguage 0409 <--- should be 0c01 ThreadUILanguage 0409 <--- should be 0c01 w10pro64 he_IL UserDefaultLCID 0409 <--- should be 40d w10pro64 hi_IN Being a Unicode-only language means setting the system locale to Hindi was not possible in older Windows versions (up to early Windows 10). Now it is but it's considered Beta and uses the UTF-8 encoding in ANSI APIs. So it's not clear if it should be changed (and how to do it programmatically). SystemDefaultLCID 0409 <--- Unicode-only language UserDefaultLCID 0409 <--- should be 0439 ThreadLocale 0409 ??? SystemPreferredUILanguages 0439,0409 UserDefaultUILanguage 0439 ThreadUILanguage 0439 w10pro64 ja_JP UserDefaultLCID 0409 <--- should be 411 w10pro64 ko_KO UserDefaultLCID 0409 <--- should be 412 w10pro64 pt_BR UserDefaultLCID 0409 <--- should be 416 w10pro64 ru_RU UserDefaultLCID 0409 <--- should be 419 w10pro64 zh_CN UserDefaultLCID 0409 <--- should be 804 This prevents testing how Windows behaves when formatting and parsing their currencies in oleaut32:varformat for instance. Strangely enough the w7u* VMs don't have this issue: w7u de_DE SystemDefaultLCID 0407 UserDefaultLCID 0407 ThreadLocale 0407 SystemPreferredUILanguages 0407 UserDefaultUILanguage 0407 ThreadUILanguage 0407 -- 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 2

[Bug 12159] New: Alt-click moves the windows instead of giving access to alternate functionalities in Photoshop Cs2
by wine-bugs＠winehq.org 18 Dec '22

18 Dec '22

http://bugs.winehq.org/show_bug.cgi?id=12159 Summary: Alt-click moves the windows instead of giving access to alternate functionalities in Photoshop Cs2 Product: Wine Version: 0.9.58. Platform: PC OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P1 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: mariejoaile(a)gmail.com Steps to reproduce: 1- Open Photoshop 2- Create a new document or open one 3- Pick a selection tool or clone or any of the tool making use of an alternate function using alt 4- Try to use this alt-click option and instead of for instance for the select tool, deselecting, it will move the window. This action is very important because there is no other way to use certain of Photoshop's tool, making tools like clone and healing useless. -- 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 4

[Bug 53250] New: ole32:compobj - test_CoGetClassObject() sometimes fails on Windows 7 and 8
by WineHQ Bugzilla 17 Dec '22

17 Dec '22

https://bugs.winehq.org/show_bug.cgi?id=53250 Bug ID: 53250 Summary: ole32:compobj - test_CoGetClassObject() sometimes fails on Windows 7 and 8 Product: Wine Version: unspecified Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: ole32 Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com ole32:compobj - test_CoGetClassObject() sometimes fails on Windows 7 and 8: compobj.c:831: Test failed: Section not found. compobj.c:841: Test failed: Unexpected return value. compobj.c:850: Test failed: Unexpected return value. compobj.c:856: Test failed: Unexpected module name L"testlib.dll". https://test.winehq.org/data/patterns.html#ole32:compobj The failures are rare enough to cause false positives. -- 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 2

[Bug 53222] New: shell32:shlfolder - test_SHChangeNotify() gets unexpected WM_USER_NOTIFY messages on Windows 10
by WineHQ Bugzilla 16 Dec '22

16 Dec '22

https://bugs.winehq.org/show_bug.cgi?id=53222 Bug ID: 53222 Summary: shell32:shlfolder - test_SHChangeNotify() gets unexpected WM_USER_NOTIFY messages on Windows 10 Product: Wine Version: unspecified Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: shell32 Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com shell32:shlfolder - test_SHChangeNotify() gets unexpected WM_USER_NOTIFY messages on Windows 10. The exact failures vary a bit though: shlfolder.c:5027: SHChangeNotify tests (0) ... shlfolder.c:5014: RMDIR: took 2 tries shlfolder.c:4964: verifying pidls for: RMDIR shlfolder.c:4977: Test failed: Didn't expect a WM_USER_NOTIFY message (event: 10) or shlfolder.c:4978: SHChangeNotify tests (0) shlfolder.c:4915: verifying pidls for: MKDIR shlfolder.c:4928: Test failed: Didn't expect a WM_USER_NOTIFY message (event: 40000) or shlfolder.c:5027: SHChangeNotify tests (0) shlfolder.c:4964: verifying pidls for: MKDIR shlfolder.c:5014: MKDIR: took 2 tries shlfolder.c:4964: verifying pidls for: MKDIR shlfolder.c:4977: Test failed: Didn't expect a WM_USER_NOTIFY message (event: 8) or shlfolder.c:4978: SHChangeNotify tests (1) ... shlfolder.c:4965: CREATE: took 0 tries shlfolder.c:4915: verifying pidls for: RMDIR shlfolder.c:4928: Test failed: Didn't expect a WM_USER_NOTIFY message (event: 3) or shlfolder.c:5014: CREATE: took 3 tries shlfolder.c:4960: Test failed: CREATE: expected notification type 2, got: 40000 shlfolder.c:4964: verifying pidls for: CREATE shlfolder.c:4967: Test failed: GetDisplayNameOf failed: 0x80070057 shlfolder.c:4977: Test failed: Didn't expect a WM_USER_NOTIFY message (event: 3) https://test.winehq.org/data/patterns.html#shell32:shlfolder The failures have been seen on Windows 10 1507, 21H1 and 21H2 but it is likely that the other Windows 10 versions behave the same way. -- 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 41784] New: Wine error under LinuxMint 17.3 after running PDF24 Creator
by wine-bugs＠winehq.org 16 Dec '22

16 Dec '22

https://bugs.winehq.org/show_bug.cgi?id=41784 Bug ID: 41784 Summary: Wine error under LinuxMint 17.3 after running PDF24 Creator Product: Wine Version: unspecified Hardware: x86 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs(a)winehq.org Reporter: marialvez(a)gmail.com Distribution: --- Created attachment 56219 --> https://bugs.winehq.org/attachment.cgi?id=56219 Using the PDF24 Creator program, in Wine mode and after close the application, my Pc show these bug.- Y work with Linux Mint 17.3 Cinnamon 64-bit Version 2.8.8 Using the PDF24 Creator program, in Wine mode and after close the application, my Pc show these bug.- Y work with Linux Mint 17.3 Cinnamon 64-bit Version 2.8.8 -- 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 7

[Bug 52930] New: d3dcompiler_47:hlsl_d3d9 fails randomly on Windows
by WineHQ Bugzilla 14 Dec '22

14 Dec '22

https://bugs.winehq.org/show_bug.cgi?id=52930 Bug ID: 52930 Summary: d3dcompiler_47:hlsl_d3d9 fails randomly on Windows Product: Wine Version: unspecified Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: directx-d3d-util Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Created attachment 72305 --> https://bugs.winehq.org/attachment.cgi?id=72305 Small patch to apply on top of 58d20aae3772^ to get failures d3dcompiler_47:hlsl_d3d9 fails randomly on Windows 8.1 to 21H2: hlsl_d3d9.c:1691: Test failed: Got unexpected hr 0x88760b59. hlsl_d3d9.c:1692: Test failed: Got unexpected blob. hlsl_d3d9.c:1693: Test failed: Got unexpected errors. https://test.winehq.org/data/patterns.html#d3dcompiler_47:hlsl_d3d9 A bisect against Windows 10 21H1 on cw-gtx560 shows that the failures started with the commit below: commit 58d20aae3772473a90e43ee236b793f0421eecb1 Author: Eric Pouech <eric.pouech(a)gmail.com> Date: Tue Feb 8 19:48:44 2022 +0100 d3dcompiler/tests: Build without -DWINE_NO_LONG_TYPES. Signed-off-by: Eric Pouech <eric.pouech(a)gmail.com> Signed-off-by: Henri Verbeet <hverbeet(a)codeweavers.com> Signed-off-by: Alexandre Julliard <julliard(a)winehq.org> The patch does not actually make any change that should matter but I have confirmed the bisect result manually: 58d20aae3772^ -> 200 consecutive runs with no failure 58d20aae3772 -> fails after 9 to 33 runs - Applying just the Makefile.in parts and last 4 chunks of hlsl_d3d9.c of that commit is sufficient to get the failures. - Patching only the makefiles is not enough. - I did not try to identify which of the 4 chunks are needed to get failures. They all patch code that's executed after the failure anyway. So there must be some memory corruption that's sensitive to memory layout or some subtle issue with the generated code. I also if(0)ed-out all the other test_xxx() functions, and also if(0)ed-out the tests that preceded the failing one in test_include() but I still get the failure so it's not a case where an ill-advised NULL-behavior test causes memory corruption. -- 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 53402] New: fake dlls such as OPENGL32.dll won't load when their unix lib is linked by LLVM lld
by WineHQ Bugzilla 13 Dec '22

13 Dec '22

https://bugs.winehq.org/show_bug.cgi?id=53402 Bug ID: 53402 Summary: fake dlls such as OPENGL32.dll won't load when their unix lib is linked by LLVM lld Product: Wine Version: 7.13 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: ntdll Assignee: wine-bugs(a)winehq.org Reporter: w12101111(a)outlook.com Distribution: --- I build wine 7.13 using Gentoo Linux ebuild (app-emulation/wine-vanilla-7.13) with clang & ld.lld and x86_64-w64-mingw32. In wine 7.13, opengl32.dll is a wine fake dll, and I found that any program require opengl failed to launch (take https://github.com/gkv311/wglinfo as an example): 0024:err:module:LdrInitializeThunk "OPENGL32.dll" failed to initialize, aborting 0024:err:module:LdrInitializeThunk Initializing dlls for L"Z:\\home\\han\\Downloads\\wglinfo64.exe" failed, status c0000005 Use gdb to debug this, I found that the real address of DllMain is 0x7a9be8e0, but ntdll try to call 0x17a9be8e0: #0 0x000000017a9be8e0 in ?? () #1 0x0000000170032ac7 in call_dll_entry_point (proc=0x17a9be8e0, module=0x7a9d0000 <__wine_spec_pe_header+65274>, reason=1, reserved=0x11fb00) at /tmp/portage/app-emulation/wine-vanilla-7.13/work/wine-7.13/dlls/ntdll/loader.c:305 (gdb) x /6i 0x000000017a9be8e0 => 0x17a9be8e0: Cannot access memory at address 0x17a9be8e0 (gdb) x /6i 0x000000007a9be8e0 0x7a9be8e0 <DllMain>: push %rbp 0x7a9be8e1 <DllMain+1>: mov %rsp,%rbp 0x7a9be8e4 <DllMain+4>: and $0xfffffffffffffff8,%rsp 0x7a9be8e8 <DllMain+8>: sub $0x30,%rsp 0x7a9be8ec <DllMain+12>: mov %rcx,0x18(%rsp) 0x7a9be8f1 <DllMain+17>: mov %edx,0x14(%rsp) The address of `module` (0x7a9d0000) is calculated by dlopen_dll ( dlls/ntdll/unix/loader.c:1294 ): nt = dlsym( handle, "__wine_spec_nt_header" ) module = (HMODULE)((nt->OptionalHeader.ImageBase + 0xffff) & ~0xffff); map_so_dll( nt, module ) And in function static NTSTATUS map_so_dll( const IMAGE_NT_HEADERS *nt_descr, HMODULE module ): BYTE *addr = (BYTE *)module; fixup_rva_ptrs( &nt->OptionalHeader.AddressOfEntryPoint, addr, 1 ); static inline void fixup_rva_ptrs( void *array, BYTE *base, unsigned int count ) { BYTE **src = array; DWORD *dst = array; for ( ; count; count--, src++, dst++) *dst = *src ? *src - base : 0; } This code is equal to nt->OptionalHeader.AddressOfEntryPoint = (DWORD)((BYTE*)nt->OptionalHeader.AddressOfEntryPoint - (BYTE *)module); Note that nt->OptionalHeader.AddressOfEntryPoint is DWORD, this code implies that nt->OptionalHeader.AddressOfEntryPoint is greater than (nt->OptionalHeader.ImageBase + 0xffff) & ~0xffff), otherwise it will overflow. This header is write by output_module ( tools/winebuild/spec32.c:699 ), nt->OptionalHeader.AddressOfEntryPoint is the address of DllMain, and nt->OptionalHeader.ImageBase is the address of __wine_spec_pe_header Read opengl32.dll.so using readelf: > readelf -S /usr/lib/wine-vanilla-7.13/wine/x86_64-unix/opengl32.dll.so There are 31 section headers, starting at offset 0x386b20: Section Headers: [Nr] Name Type Address Off Size ES Flg Lk Inf Al [ 0] NULL 0000000000000000 000000 000000 00 0 0 0 [ 1] .dynsym DYNSYM 000000007a800238 000238 002370 18 A 4 1 8 [ 2] .gnu.hash GNU_HASH 000000007a8025a8 0025a8 000b30 00 A 1 0 8 [ 3] .hash HASH 000000007a8030d8 0030d8 000bd8 04 A 1 0 4 [ 4] .dynstr STRTAB 000000007a803cb0 003cb0 00133a 00 A 0 0 1 [ 5] .rela.dyn RELA 000000007a804ff0 004ff0 043890 18 A 1 0 8 [ 6] .rela.plt RELA 000000007a848880 048880 000120 18 AI 1 20 8 [ 7] .rodata PROGBITS 000000007a8489a0 0489a0 0140d0 00 AMS 0 0 16 [ 8] .eh_frame_hdr PROGBITS 000000007a85ca70 05ca70 00cc7c 00 A 0 0 4 [ 9] .eh_frame PROGBITS 000000007a8696f0 0696f0 045d24 00 A 0 0 8 [10] .text PROGBITS 000000007a8b0420 0af420 10fce0 00 AX 0 0 32 [11] .init PROGBITS 000000007a9c0100 1bf100 011008 00 AX 0 0 1 [12] .fini PROGBITS 000000007a9d1108 1d0108 000003 00 AX 0 0 1 [13] .plt PROGBITS 000000007a9d1110 1d0110 0000d0 00 AX 0 0 16 [14] .init_array INIT_ARRAY 000000007a9d21e0 1d01e0 000008 00 WA 0 0 8 [15] .fini_array FINI_ARRAY 000000007a9d21e8 1d01e8 000010 00 WA 0 0 8 [16] .data.rel.ro PROGBITS 000000007a9d2200 1d0200 00fcb0 00 WA 0 0 16 [17] .dynamic DYNAMIC 000000007a9e1eb0 1dfeb0 000170 10 WA 4 0 8 [18] .got PROGBITS 000000007a9e2020 1e0020 000018 00 WA 0 0 8 [19] .data PROGBITS 000000007a9e3038 1e0038 009120 00 WA 0 0 8 [20] .got.plt PROGBITS 000000007a9ec158 1e9158 000078 00 WA 0 0 8 [21] .bss NOBITS 000000007a9ec1d0 1e91d0 006090 00 WA 0 0 16 [22] .debug_line PROGBITS 0000000000000000 1e91d0 039046 00 0 0 1 [23] .debug_line_str PROGBITS 0000000000000000 222216 00004e 01 MS 0 0 1 [24] .comment PROGBITS 0000000000000000 222264 00003e 01 MS 0 0 1 [25] .debug_abbrev PROGBITS 0000000000000000 2222a2 000c02 00 0 0 1 [26] .debug_info PROGBITS 0000000000000000 222ea4 0ed62d 00 0 0 1 [27] .debug_str PROGBITS 0000000000000000 3104d1 028720 01 MS 0 0 1 [28] .symtab SYMTAB 0000000000000000 338bf8 026d30 18 30 6249 8 [29] .shstrtab STRTAB 0000000000000000 35f928 000117 00 0 0 1 [30] .strtab STRTAB 0000000000000000 35fa3f 0270de 00 0 0 1 Key to Flags: W (write), A (alloc), X (execute), M (merge), S (strings), I (info), L (link order), O (extra OS processing required), G (group), T (TLS), C (compressed), x (unknown), o (OS specific), E (exclude), R (retain), l (large), p (processor specific) > readelf -s /usr/lib/wine-vanilla-7.13/wine/x86_64-unix/opengl32.dll.so | grep __wine_spec_pe_header 1: 000000007a9c0106 0 NOTYPE LOCAL DEFAULT 11 __wine_spec_pe_header > readelf -s /usr/lib/wine-vanilla-7.13/wine/x86_64-unix/opengl32.dll.so | grep DllMain 6227: 000000007a9be8e0 147 FUNC LOCAL HIDDEN 10 DllMain In output_module, __wine_spec_pe_header is put into .init section: .section ".init","ax" jmp 1f __wine_spec_pe_header: .skip 69632 1: DllMain is in .text section But LLVM lld put .text before .init, so the assumption in map_so_dll is not valid, and the address has a extra 2^32 because of the subtraction overflow and cast to DWORD. If I change function output_module, put __wine_spec_pe_header in .text just like PLATFORM_SOLARIS, then OPENGL32.DLL will load without any issue. -- 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