February 2010 - Wine-Bugs - list.winehq.org

[Bug 21741] New: Sonic & Knuckles Collection Demo installer produces (non-fatal) WINNLS32 errors
by wine-bugs＠winehq.org Aug. 14, 2021

Aug. 14, 2021

http://bugs.winehq.org/show_bug.cgi?id=21741 Summary: Sonic & Knuckles Collection Demo installer produces (non-fatal) WINNLS32 errors Product: Wine Version: 1.1.38 Platform: x86-64 URL: http://www2.sega.com/support/pcdownloads/sonic_knuckle s_demo/ OS/Version: Linux Status: UNCONFIRMED Severity: minor Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: RandomAccountName(a)mail.com CC: julliard(a)winehq.org Created an attachment (id=26268) --> (http://bugs.winehq.org/attachment.cgi?id=26268) Terminal output from current git When installing the demo for Sonic & Knuckles Collection, the user is present with the following error message four times: Can't GetProcAddress() for WINNLS32.DLL WINNLS32EnableIME Installation succeeds in spite of these messages. This is a regression which I tracked down to the following commit: 9bf9eb094102c8e1d3661c88d6cf5e8129bc4dc4 is first bad commit commit 9bf9eb094102c8e1d3661c88d6cf5e8129bc4dc4 Author: Alexandre Julliard <julliard(a)winehq.org> Date: Tue Mar 24 09:31:03 2009 +0100 winnls32: Make winnls.dll into a stand-alone 16-bit module. :100644 100644 3619da5d499922e7b8929fbfeffff56aafeaf08a b2c44c2518de7640e14dd2b80f3c6a3db97b5ee5 M .gitignore :100755 100755 2b50abf47065900234add54b9d702259dd67ce31 e11f83c7ed3cafb6db7481d8597e07f040405eb1 M configure :100644 100644 c3f7008e81206748c1ca1866ad3fb6d051896766 9ee0e88331867ed2f32d9199acbce4a7ec0cd5fb M configure.ac :040000 040000 3a39aa36af68064495b4d06e575d7e0be41b5496 576951279d2974c31f209eec20e953f69807e1e3 M dlls This commit can no longer be reverted cleanly from current git, but the problem goes away if 1.1.18 is compiled with the patch reverted. -- 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 7

[Bug 21501] New: Visual C++ 2010 beta 2 web installer can't replace corrupted downloads (sharing violation due to wintrust handle leak)
by wine-bugs＠winehq.org Aug. 10, 2021

Aug. 10, 2021

http://bugs.winehq.org/show_bug.cgi?id=21501 Summary: Visual C++ 2010 beta 2 web installer can't replace corrupted downloads (sharing violation due to wintrust handle leak) Product: Wine Version: 1.1.37 Platform: x86 URL: http://www.microsoft.com/visualstudio/en-us/try/defaul t.mspx#download OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: wintrust AssignedTo: wine-bugs(a)winehq.org ReportedBy: focht(a)gmx.net Hello, the web installer verifies the integrity of already downloaded files using hashing and wintrust signature checks. In case of a corrupted file, WinVerifyTrust -> WINTRUST_DefaultVerify -> SOFTPUB_LoadFileMessage fails: --- snip --- ... 0031:trace:wintrust:SoftpubInitialize (0x1004c230) 0031:trace:wintrust:SoftpubInitialize returning 00000000 0031:trace:wintrust:SoftpubLoadMessage (0x1004c230) 0031: create_file( access=80000000, attributes=00000040, sharing=00000001, create=1, options=00000050, attrs=00000080, objattr={rootdir=0000,sd={},name=L""}, filename="/home/focht/.wine/dosdevices/c:/windows/temp/21006.01/1033/VS_PRO/wcu/vc90sp1/vc_red.cab" ) 0031: create_file() = 0 { handle=0178 } 0031: get_handle_fd( handle=0178 ) 0031: *fd* 0178 -> 145 0031: get_handle_fd() = 0 { type=1, removable=0, access=00120089, options=00000050 } 0031:trace:wintrust:SOFTPUB_OpenFile returning 1 0031:trace:wintrust:SOFTPUB_GetFileSubject returning 1 0031:trace:wintrust:SOFTPUB_GetSIP returning 1 0031:trace:wintrust:CryptSIPGetSignedDataMsg (0x38ddf90 0x1004c268 0 0x9bf677c (nil)) 0031:trace:wintrust:WINTRUST_GetSignedMsgFromCabFile (0x38ddf90 0x1004c268 0 0x9bf677c (nil)) 0031:trace:wintrust:WINTRUST_GetSignedMsgFromCabFile cert_offset: 3823372 0031:trace:wintrust:WINTRUST_GetSignedMsgFromCabFile cert_size: 5944 0031:warn:wintrust:WINTRUST_GetSignedMsgFromCabFile offset beyond file, not attempting to read 0031:trace:wintrust:CryptSIPGetSignedDataMsg returning 0 0031:trace:wintrust:SoftpubLoadMessage returning 1 (800b0100) 0031:trace:wintrust:WINTRUST_DefaultVerify returning 800b0100 0031:trace:wintrust:WinVerifyTrust returning 800b0100 ... --- snip --- It seems if WinVerifyTrust fails that way, the caller is not supposed to clean up, e.g. calling WinVerifyTrust with WTD_STATEACTION_CLOSE -> WINTRUST_DefaultClose. The web installer never does - hence the handle + additional data is leaked. This leaked file handle later leads to a sharing violation when urlmon tries to copy the new file from download cache to destination (DownloadBSC_OnStopBinding): --- snip --- 0031: create_file( access=c0010000, attributes=00000040, sharing=00000007, create=1, options=00001040, attrs=00000000, objattr={rootdir=0000,sd={},name=L""}, filename="/home/focht/.wine/dosdevices/c:/windows/temp/21006.01/1033/VS_PRO/wcu/vc90sp1/vc_red.cab" ) 0031: create_file() = SHARING_VIOLATION { handle=0000 } ... --- snip --- After 3 download + replacement tries, it gives up. Properly internally cleaning up (closing the file handle ...) helps the installer/urlmon to replace the file. 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.

2 9

[Bug 18980] New: Canon's CONSOLE Image Control & Storage Software fails on administrator privs check
by wine-bugs＠winehq.org Aug. 10, 2021

Aug. 10, 2021

http://bugs.winehq.org/show_bug.cgi?id=18980 Summary: Canon's CONSOLE Image Control & Storage Software fails on administrator privs check Product: Wine Version: 1.1.23 Platform: PC URL: http://www.usa.canon.com/consumer/controller?act=Model InfoAct&tabact=DownloadDetailTabAct&fcategoryid=326&mo delid=10350 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: advapi32 AssignedTo: wine-bugs(a)winehq.org ReportedBy: focht(a)gmx.net Hello, a user in #winehq came up with this problem, so I looked into it knowing the app won't work at all due to low level hardware driver dependency. Anyway it revealed at least a valid bug ;-) --- snip --- ... 0028:Call advapi32.OpenProcessToken(ffffffff,00000018,0032fce4) ret=0046fa9e 0028:Ret advapi32.OpenProcessToken() retval=00000001 ret=0046fa9e 0028:Call advapi32.GetTokenInformation(000000a8,00000002,00000000,00000000,0032fce0) ret=0046fac3 0028:trace:advapi:GetTokenInformation (0xa8, TokenGroups, (nil), 0, 0x32fce0): 0028:Ret advapi32.GetTokenInformation() retval=00000000 ret=0046fac3 0028:Call msvcrt.malloc(000000a0) ret=5f8038d6 0028:Call ntdll.RtlAllocateHeap(00110000,00000000,000000a0) ret=7ed2d867 0028:Ret ntdll.RtlAllocateHeap() retval=00159260 ret=7ed2d867 0028:Ret msvcrt.malloc() retval=00159260 ret=5f8038d6 0028:Call advapi32.GetTokenInformation(000000a8,00000002,00159260,000000a0,0032fce0) ret=0046faef 0028:trace:advapi:GetTokenInformation (0xa8, TokenGroups, 0x159260, 160, 0x32fce0): 0028:Ret advapi32.GetTokenInformation() retval=00000001 ret=0046faef 0028:Call advapi32.GetSidSubAuthorityCount(0015929c) ret=0046fb10 0028:Ret advapi32.GetSidSubAuthorityCount() retval=0015929d ret=0046fb10 0028:Call KERNEL32.GetLastError() ret=0046fb20 0028:Ret KERNEL32.GetLastError() retval=0000007a ret=0046fb20 0028:Call advapi32.GetSidSubAuthorityCount(001592a8) ret=0046fb10 0028:Ret advapi32.GetSidSubAuthorityCount() retval=001592a9 ret=0046fb10 0028:Call KERNEL32.GetLastError() ret=0046fb20 0028:Ret KERNEL32.GetLastError() retval=0000007a ret=0046fb20 0028:Call advapi32.GetSidSubAuthorityCount(001592b4) ret=0046fb10 0028:Ret advapi32.GetSidSubAuthorityCount() retval=001592b5 ret=0046fb10 0028:Call KERNEL32.GetLastError() ret=0046fb20 0028:Ret KERNEL32.GetLastError() retval=0000007a ret=0046fb20 0028:Call advapi32.GetSidSubAuthorityCount(001592c0) ret=0046fb10 0028:Ret advapi32.GetSidSubAuthorityCount() retval=001592c1 ret=0046fb10 0028:Call KERNEL32.GetLastError() ret=0046fb20 0028:Ret KERNEL32.GetLastError() retval=0000007a ret=0046fb20 0028:Call advapi32.GetSidSubAuthorityCount(001592cc) ret=0046fb10 0028:Ret advapi32.GetSidSubAuthorityCount() retval=001592cd ret=0046fb10 0028:Call KERNEL32.GetLastError() ret=0046fb20 0028:Ret KERNEL32.GetLastError() retval=0000007a ret=0046fb20 0028:Call advapi32.GetSidSubAuthorityCount(001592dc) ret=0046fb10 0028:Ret advapi32.GetSidSubAuthorityCount() retval=001592dd ret=0046fb10 0028:Call KERNEL32.GetLastError() ret=0046fb20 0028:Ret KERNEL32.GetLastError() retval=0000007a ret=0046fb20 0028:Call advapi32.GetSidSubAuthorityCount(001592ec) ret=0046fb10 0028:Ret advapi32.GetSidSubAuthorityCount() retval=001592ed ret=0046fb10 0028:Call KERNEL32.GetLastError() ret=0046fb20 0028:Ret KERNEL32.GetLastError() retval=0000007a ret=0046fb20 0028:Call msvcrt.free(00159260) ret=5f8048cf 0028:Call ntdll.RtlFreeHeap(00110000,00000000,00159260) ret=7ed2da36 0028:Ret ntdll.RtlFreeHeap() retval=00000001 ret=7ed2da36 0028:Ret msvcrt.free() retval=00000001 ret=5f8048cf ... 0028:Call user32.MessageBoxW(00000000,00159560 L"This program requires administrator privileges.",00157d48 L"CONSOLE",00000040) ret=5f86881e --- snip --- With the first call to GetTokenInformation( ... TokenGroups ... ) the app queries the buffer size to hold result data -> lasterror = ERROR_INSUFFICIENT_BUFFER. The app allocates the required buffer and then calls GetTokenInformation() to actually fetch the data. Next, the app loops through sub authorities using GetSidSubAuthorityCount(). Normally it would look for DOMAIN_ALIAS_RID_ADMINS but the problem is the GetLastError() call _before_ the actual comparison. If last error is != 0, the app won't look at data returned by GetSidSubAuthorityCount(). The last successful call to GetTokenInformation() didn't reset last error at all - it's still at ERROR_INSUFFICIENT_BUFFER: --- snip dlls/advapi32/security.c --- BOOL WINAPI GetTokenInformation( HANDLE token, TOKEN_INFORMATION_CLASS tokeninfoclass, LPVOID tokeninfo, DWORD tokeninfolength, LPDWORD retlen ) { ... return set_ntstatus( NtQueryInformationToken( token, tokeninfoclass, tokeninfo, tokeninfolength, retlen)); } --- snip dlls/advapi32/security.c --- --- snip dlls/advapi32/security.c --- /* set last error code from NT status and get the proper boolean return value */ /* used for functions that are a simple wrapper around the corresponding ntdll API */ static inline BOOL set_ntstatus( NTSTATUS status ) { if (status) SetLastError( RtlNtStatusToDosError( status )); return !status; } --- snip dlls/advapi32/security.c --- This is probably an oversight. I don't know what Windows does but the app seems to expect the last error being reset to zero (NTSTATUS success). A conformance test should reveal this. 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.

2 9

[Bug 18717] New: .NET 3.x needs ntdll.dll.RtlEnumerateGenericTableWithoutSplaying (WPF + MILCore driver)
by wine-bugs＠winehq.org Aug. 10, 2021

Aug. 10, 2021

http://bugs.winehq.org/show_bug.cgi?id=18717 Summary: .NET 3.x needs ntdll.dll.RtlEnumerateGenericTableWithoutSplaying (WPF + MILCore driver) Product: Wine Version: 1.1.22 Platform: PC OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: ntdll AssignedTo: wine-bugs(a)winehq.org ReportedBy: focht(a)gmx.net Hello, affects every .NET 3.x GUI app which makes use of WPF. Both, the managed part of WPF and the (unmanaged) MILCore driver make use of the generic table API. I'm still unsure how much of the generic table API needs to be functional and what role such tables play in communication between Framework and MILCore driver. Currently, it seems we can get away with little effort. Returning NULL from stub seems sufficient to prevent the crash and allows apps to proceed further. --- snip --- fixme:shell:URL_ParseUrl failed to parse L"PresentationFramework" fixme:shell:URL_ParseUrl failed to parse L"PresentationCore" fixme:shell:URL_ParseUrl failed to parse L"System" fixme:shell:URL_ParseUrl failed to parse L"WindowsBase" fixme:advapi:RegisterTraceGuidsW 0x6a053a (nil) 0x32dfd0 1 0x32dfac (null) (null) 0x32dfb4 fixme:shell:URL_ParseUrl failed to parse L"PresentationCore.resources" fixme:advapi:RegisterTraceGuidsW 0x532dc234 (nil) 0x53218780 27 0x32ce8c (null) (null) 0x53355f40 fixme:seh:RtlCaptureStackBackTrace (2, 3, 0x533546cc, (nil)) stub! fixme:ntdll:RtlInitializeGenericTable (0x16cae4,0x5327d3ce,0x5327d3a0,0x5327d3b7,(nil)) stub! fixme:ntdll:RtlInitializeGenericTable (0x16cd4c,0x5328ec37,0x5327d3a0,0x5327d3b7,(nil)) stub! fixme:seh:RtlCaptureStackBackTrace (2, 3, 0x533546e4, (nil)) stub! fixme:seh:RtlCaptureStackBackTrace (2, 3, 0x533546fc, (nil)) stub! fixme:win:EnumDisplayDevicesW ((null),0,0x2d5e568,0x00000000), stub! fixme:win:EnumDisplayDevicesW ((null),1,0x2d5e568,0x00000000), stub! wine: Call from 0x7bc48066 to unimplemented function ntdll.dll.RtlEnumerateGenericTableWithoutSplaying, aborting ... fixme:advapi:RegisterEventSourceW ((null),L".NET Runtime 2.0 Error Reporting"): stub fixme:advapi:ReportEventW (0xcafe4242,0x0001,0x0000,0x000003e8,(nil),0x0008,0x000000e0,0x3009a1b4,0x6cc5c0): stub err:eventlog:ReportEventW L"edgermreader.exe" err:eventlog:ReportEventW L"1.0.0.0" err:eventlog:ReportEventW L"47c0d0da" err:eventlog:ReportEventW L"unknown" err:eventlog:ReportEventW L"0.0.0.0" err:eventlog:ReportEventW L"00000000" err:eventlog:ReportEventW L"0" err:eventlog:ReportEventW L"7bc48066" fixme:advapi:DeregisterEventSource (0xcafe4242) stub --- snip --- While you're at it, you might also want to fix the prototype of RtlInitializeGenericTable(). --- snip dlls/ntdll/rtl.c --- PVOID WINAPI RtlInitializeGenericTable(PVOID pTable, PVOID arg2, PVOID arg3, PVOID arg4, PVOID arg5) --- snip dlls/ntdll/rtl.c --- See: http://msdn.microsoft.com/en-us/library/ms796842.aspx Return type is VOID, additionally the parameters might be enhanced with MSDN info. 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.

2 5

[Bug 18716] New: .NET 3.0 WPF requires SystemParametersInfoW( SPI_GETMOUSEVANISH) handled
by wine-bugs＠winehq.org Aug. 10, 2021

Aug. 10, 2021

http://bugs.winehq.org/show_bug.cgi?id=18716 Summary: .NET 3.0 WPF requires SystemParametersInfoW( SPI_GETMOUSEVANISH) handled Product: Wine Version: 1.1.22 Platform: PC URL: http://infogenium.com/proj_reader.php OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: user32 AssignedTo: wine-bugs(a)winehq.org ReportedBy: focht(a)gmx.net Hello, another NET 3.0 app dies due to SystemParametersInfoW( SPI_GETMOUSEVANISH) not handled (triggered by "search" -> "type text to find" action, mouse hovering) --- snip --- fixme:system:SystemParametersInfoW Unimplemented action: 4128 (SPI_GETMOUSEVANISH) First chance exception generated: (0xe6a41c) <System.ComponentModel.Win32Exception> Unhandled exception generated: (0xe6a41c) <System.ComponentModel.Win32Exception> nativeErrorCode=1439 _className=<null> _exceptionMethod=<null> _exceptionMethodString=<null> _message=(0xe6a6d4) "Invalid SPI parameter" _data=(0xe6ad40) <System.Collections.ListDictionaryInternal> _innerException=<null> _helpURL=<null> _stackTrace=(0xe6ad6c) <System.SByte[]> _stackTraceString=<null> _remoteStackTraceString=<null> _remoteStackIndex=0 _dynamicMethods=<null> _HResult=-2147467259 _source=<null> _xptrs=0 _xcode=-532459699 Exception is called:UNHANDLED ... Thread 0x37 Current State:GCUnsafe spot 0)* System.Windows.SystemParameters::get_MouseVanish +0155[native] +0064[IL] in <Unknown File Name>:<Unknown Line Number> 1) System.Windows.Documents.TextEditorTyping::HideCursor +0046[native] +0012[IL] in <Unknown File Name>:<Unknown Line Number> 2) System.Windows.Documents.TextEditorTyping::DoTextInput +0079[native] +0006[IL] in <Unknown File Name>:<Unknown Line Number> 3) System.Windows.Documents.TextEditorTyping.TextInputItem::Do +0079[native] +0020[IL] in <Unknown File Name>:<Unknown Line Number> 4) System.Windows.Documents.TextEditorTyping::ScheduleInput +0069[native] +0028[IL] in <Unknown File Name>:<Unknown Line Number> 5) System.Windows.Documents.TextEditorTyping::OnTextInput +0403[native] +0146[IL] in <Unknown File Name>:<Unknown Line Number> 6) System.Windows.Documents.TextEditor::OnTextInput +0031[native] +0012[IL] in <Unknown File Name>:<Unknown Line Number> 7) System.Windows.Controls.Primitives.TextBoxBase::OnTextInput +0076[native] +0036[IL] in <Unknown File Name>:<Unknown Line Number> 8) System.Windows.UIElement::OnTextInputThunk +0095[native] +0031[IL] in <Unknown File Name>:<Unknown Line Number> 9) System.Windows.Input.TextCompositionEventArgs::InvokeEventHandler +0072[native] +0007[IL] in <Unknown File Name>:<Unknown Line Number> --- snip --- (also suffers from bug 18660) 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.

2 6

[Bug 12718] New: Typing the username in MSN Messenger 7.5 is really slow
by wine-bugs＠winehq.org Aug. 10, 2021

Aug. 10, 2021

http://bugs.winehq.org/show_bug.cgi?id=12718 Summary: Typing the username in MSN Messenger 7.5 is really slow Product: Wine Version: 0.9.60 Platform: PC URL: http://www.filehippo.com/download_msn_messenger/751/ OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: wininet AssignedTo: wine-bugs(a)winehq.org ReportedBy: jaimerave(a)gmail.com When i start to write my email in msn messenger, the letters will take a long time to appear. In the console I'm only watchinh this message every time I type a letter: fixme:wininet:InternetSetOptionW INTERNET_OPTION_SEND/RECEIVE_TIMEOUT fixme:wininet:InternetSetOptionW INTERNET_OPTION_SEND/RECEIVE_TIMEOUT fixme:wininet:InternetSetOptionW Option INTERNET_OPTION_CONNECT_TIMEOUT (60000): STUB -- 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 22

[Bug 20188] New: MSN Messenger 7.5 hangs while signing in
by wine-bugs＠winehq.org Aug. 10, 2021

Aug. 10, 2021

http://bugs.winehq.org/show_bug.cgi?id=20188 Summary: MSN Messenger 7.5 hangs while signing in Product: Wine Version: 1.1.30 Platform: PC URL: http://www.oldapps.com/msn_messenger.php OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: jaimerave(a)gmail.com Created an attachment (id=23793) --> (http://bugs.winehq.org/attachment.cgi?id=23793) Console output after signing in and when it starts loading the contacts the the screen just hangs. -- 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 10059] New: Windows Live Messenger 8. 5 - You can not type into login boxes
by wine-bugs＠winehq.org Aug. 10, 2021

Aug. 10, 2021

http://bugs.winehq.org/show_bug.cgi?id=10059 Summary: Windows Live Messenger 8.5 - You can not type into login boxes Product: Wine Version: CVS/GIT Platform: PC OS/Version: Linux Status: UNCONFIRMED Severity: major Priority: P2 Component: wine-richedit AssignedTo: wine-bugs(a)winehq.org ReportedBy: philip.kent(a)ntlworld.com When trying to run Windows Live Messenger 8.5 (install instructions in AppDB - http://appdb.winehq.org/objectManager.php?sClass=version&iId=8146) you will find you can not log in. This is because it is impossible to type into the textboxes requesting usernames and passwords. They are too small and do not accept input. A log file is attached. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are watching all bug changes.

2 16

[Bug 13777] New: Can't write in login and password fields in MSN8
by wine-bugs＠winehq.org Aug. 10, 2021

Aug. 10, 2021

http://bugs.winehq.org/show_bug.cgi?id=13777 Summary: Can't write in login and password fields in MSN8 Product: Wine Version: 1.0-rc4 Platform: Other OS/Version: other Status: UNCONFIRMED Severity: normal Priority: P2 Component: richedit AssignedTo: wine-bugs(a)winehq.org ReportedBy: igamads(a)gmail.com Created an attachment (id=13810) --> (http://bugs.winehq.org/attachment.cgi?id=13810) Terminal output on msn8 and clean wineprefix Can't right on the fields for Login and password. -- 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 14358] New: fusion: installation .NET assemblies might fail due to incorrect parsing of CLR metadata structures and tables
by wine-bugs＠winehq.org Aug. 10, 2021

Aug. 10, 2021

http://bugs.winehq.org/show_bug.cgi?id=14358 Summary: fusion: installation .NET assemblies might fail due to incorrect parsing of CLR metadata structures and tables Product: Wine Version: CVS/GIT Platform: PC URL: http://www.microsoft.com/downloads/details.aspx?familyid =3c9a983a-ac14-4125-8ba0-d36d67e0f4ad OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: focht(a)gmx.net Hello, a bug for James Hawkins ... low priority. Encountered this when I accidentally forgot the .NET 2.0 prerequisite before Office 2007 installation. Setup dies in wine's fusion due to bugs in CLR metadata processing while installing primary interop assemblies for Office into GAC. --- snip --- 003f:trace:fusion:CreateAssemblyCache (0x32fe9c, 0) 003f:Call ntdll.RtlAllocateHeap(00110000,00000000,00000008) ret=79655027 003f:Ret ntdll.RtlAllocateHeap() retval=00114ec0 ret=79655027 003f:Ret fusion.CreateAssemblyCache() retval=00000000 ret=694077de 003f:trace:fusion:IAssemblyCacheImpl_InstallAssembly (0x114ec0, 0, L"Microsoft.mshtml.dll", (nil)) .. 003f:Call KERNEL32.CreateFileW(00115168 L"Microsoft.mshtml.dll",80000000,00000001,00000000,00000003,00000080,00000000) ret=7965724f 003f:Ret KERNEL32.CreateFileW() retval=00000038 ret=7965724f 003f:Call KERNEL32.CreateFileMappingW(00000038,00000000,00000002,00000000,00000000,00000000) ret=7965728e 003f:Ret KERNEL32.CreateFileMappingW() retval=0000003c ret=7965728e 003f:Call KERNEL32.MapViewOfFile(0000003c,00000004,00000000,00000000,00000000) ret=796572c4 003f:Ret KERNEL32.MapViewOfFile() retval=00440000 ret=796572c4 003f:Call ntdll.RtlImageNtHeader(00440000) ret=796572d6 003f:Ret ntdll.RtlImageNtHeader() retval=00440080 ret=796572d6 003f:Call ntdll.RtlImageRvaToVa(00440080,00440000,00002008,00000000) ret=7965736d 003f:Ret ntdll.RtlImageRvaToVa() retval=00441008 ret=7965736d 003f:Call ntdll.RtlImageRvaToVa(00440080,00440000,001237c4,00000000) ret=7965739f 003f:Ret ntdll.RtlImageRvaToVa() retval=005627c4 ret=7965739f 003f:Call ntdll.RtlImageRvaToVa(00440080,00440000,001237e4,00000000) ret=796573d1 003f:Ret ntdll.RtlImageRvaToVa() retval=005627e4 ret=796573d1 003f:trace:seh:raise_exception code=c0000005 flags=0 addr=0x602636b3 003f:trace:seh:raise_exception info[0]=00000000 003f:trace:seh:raise_exception info[1]=00000000 003f:trace:seh:raise_exception eax=00000000 ebx=79659168 ecx=00000000 edx=0012c8c0 esi=80131018 edi=00000000 003f:trace:seh:raise_exception ebp=0032fc08 esp=0032fbec cs=0073 ds=007b es=007b fs=0033 gs=003b flags=00010246 003f:trace:seh:call_stack_handlers calling handler at 0x7b877bc0 code=c0000005 flags=0 wine: Unhandled page fault on read access to 0x00000000 at address 0x602636b3 (thread 003f), starting debugger... 003f:trace:seh:start_debugger Starting debugger "winedbg --auto 62 68" 003f:trace:seh:call_stack_handlers handler at 0x7b877bc0 returned 1 Unhandled exception: page fault on read access to 0x00000000 in 32-bit code (0x602636b3). --- snip --- CLR metadata section header is incorrectly processed (offset calculation). The version length is rounded to DWORD boundary - this needs to be taken into account (crash on some assemblies). Also the CLR metadata tables offset calculation is not correct. A good crash candidate is "Microsoft.mshtml.dll", the primary interop assembly (PIA) for mshtml.dll A good source of information about PE CLR data structures is here: http://www.ntcore.com/Files/dotnetformat.htm You might also want to have a look at CFF explorer, one of the rare PE tools which support .NET/CLR metadata: http://www.ntcore.com/exsuite.php If you use it on assemblies with large amount of CLR metadata, like Microsoft.mshtml.dll PIA, it might take lots of time to parse all tables. It gives you all the info (metadata table offsets/RVA's) for verification of calculations. I wrote my own little test which just calls fusion's IAssemblyCache_InstallAssembly on assemblies to verify my findings/patches. I'll leave out any patch snippets to give potential patch implementor full freedom. For testing purposes you can use the "Office 2003 Update: Redistributable Primary Interop Assemblies" from Micro$oft support which is available for download (see URL). Extract and run 'msiexec -x O2003PIA.msi'. 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.

2 10