March 2020 - Wine-Bugs - list.winehq.org

[Bug 48861] x64dbg fails to set debuggee command line via File -> Change Command Line menu

by WineHQ Bugzilla

https://bugs.winehq.org/show_bug.cgi?id=48861 Anastasius Focht <focht(a)gmx.net> changed: What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download URL| |https://github.com/x64dbg/x | |64dbg/releases/download/sna | |pshot/snapshot_2020-02-23_1 | |6-38.zip --- Comment #1 from Anastasius Focht <focht(a)gmx.net> --- Hello folks, I've created a snapshot of a snapshot *kekw* via Internet Archive from: https://github.com/x64dbg/x64dbg/releases/download/snapshot/snapshot_2020-0… -> https://web.archive.org/web/20200330230227/https://github-production-releas… $ sha1sum snapshot_2020-02-23_16-38.zip b068b7bbfab1987baf4e10c8df1b9431d88e1e28 snapshot_2020-02-23_16-38.zip $ du -sh snapshot_2020-02-23_16-38.zip 31M snapshot_2020-02-23_16-38.zip Regards -- 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.

March 30, 2020

[Bug 48861] New: x64dbg fails to set debuggee command line via File -> Change Command Line menu

by WineHQ Bugzilla

https://bugs.winehq.org/show_bug.cgi?id=48861 Bug ID: 48861 Summary: x64dbg fails to set debuggee command line via File -> Change Command Line menu Product: Wine Version: 5.5 Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: kernelbase Assignee: wine-bugs(a)winehq.org Reporter: focht(a)gmx.net Distribution: --- Hello folks, https://github.com/x64dbg/x64dbg/wiki/Frequently-Asked-Questions --- quote --- Q: Help, how do I pass arguments to the program I want to debug? A: Use the File -> Change Command Line option: --- quote --- This doesn't work under Wine. One has to start the debuggee using an undocumented way. https://github.com/x64dbg/x64dbg/issues/210 The documentation task https://github.com/x64dbg/docs/issues/26 is still open since 2016. Examples: --- snip -- # starts x64dbg with 64-bit debuggee and args $ wine ./x96dbg.exe "c:\\windows\\system32\\cmd.exe" /k echo "%PROCESSOR_ARCHITECTURE%" # starts x32dbg with 32-bit debuggee and args $ wine ./x96dbg.exe "c:\\windows\\syswow64\\cmd.exe" /k echo "%PROCESSOR_ARCHITECTURE%" --- snip --- Anyway, the problem here is how Wine implements 'kernel32.GetCommandLine{A,W}' functions. Starting with x64dbg sources... https://github.com/x64dbg/x64dbg/blob/d2c62e69c2cd4d42cc4509612ad3cfd34e914… --- snip --- bool dbgsetcmdline(const char* cmd_line, cmdline_error_t* cmd_line_error) { // Make sure cmd_line_error is a valid pointer cmdline_error_t cmd_line_error_aux; if(cmd_line_error == NULL) cmd_line_error = &cmd_line_error_aux; // Get the command line address if(!getcommandlineaddr(&cmd_line_error->addr, cmd_line_error)) return false; auto command_line_addr = cmd_line_error->addr; // Convert the string to UTF-16 auto command_linewstr = StringUtils::Utf8ToUtf16(cmd_line); if(command_linewstr.length() >= 32766) //32766 is maximum character count for a null-terminated UNICODE_STRING command_linewstr.resize(32766); // Convert the UTF-16 string to ANSI auto command_linestr = Utf16ToAnsi(command_linewstr.c_str()); // Fill the UNICODE_STRING to be set in the debuggee UNICODE_STRING new_command_line; new_command_line.Length = USHORT(command_linewstr.length() * sizeof(WCHAR)); //max value: 32766 * 2 = 65532 new_command_line.MaximumLength = new_command_line.Length + sizeof(WCHAR); //max value: 65532 + 2 = 65534 new_command_line.Buffer = PWSTR(command_linewstr.c_str()); //allow cast from const because the UNICODE_STRING will not be used locally // Allocate remote memory for both the UNICODE_STRING.Buffer and the (null terminated) ANSI buffer duint mem = MemAllocRemote(0, new_command_line.MaximumLength + command_linestr.size()); if(!mem) { cmd_line_error->type = CMDL_ERR_ALLOC_UNICODEANSI_COMMANDLINE; return false; } // Write the UNICODE_STRING.Buffer to the debuggee (UNICODE_STRING.Length is used because the remote memory is zeroed) if(!MemWrite(mem, new_command_line.Buffer, new_command_line.Length)) { cmd_line_error->addr = mem; cmd_line_error->type = CMDL_ERR_WRITE_UNICODE_COMMANDLINE; return false; } // Write the (null-terminated) ANSI buffer to the debuggee if(!MemWrite(mem + new_command_line.MaximumLength, command_linestr.data(), command_linestr.size())) { cmd_line_error->addr = mem + new_command_line.MaximumLength; cmd_line_error->type = CMDL_ERR_WRITE_ANSI_COMMANDLINE; return false; } // Change the pointers to the command line if(!fixgetcommandlinesbase(mem, mem + new_command_line.MaximumLength, cmd_line_error)) return false; // Put the remote buffer address in the UNICODE_STRING and write it to the PEB new_command_line.Buffer = PWSTR(mem); if(!MemWrite(command_line_addr, &new_command_line, sizeof(new_command_line))) { cmd_line_error->addr = command_line_addr; cmd_line_error->type = CMDL_ERR_WRITE_PEBUNICODE_COMMANDLINE; return false; } // Copy command line copyCommandLine(cmd_line); return true; } --- snip --- Corresponding Wine trace log snippet: --- snip --- ... 0009:Call KERNEL32.VirtualAllocEx(0000024c,00000000,00000087,00003000,00000040) ret=01dd445c 0009:Call ntdll.NtAllocateVirtualMemory(0000024c,0032ce3c,00000000,0032ce58,00003000,00000040) ret=7b022f18 0009: queue_apc( handle=024c, call={APC_VIRTUAL_ALLOC,addr==00000000,size=00000087,zero_bits_64=0,op_type=3000,prot=40} ) 0045: *wakeup* signaled=256 0009: queue_apc() = 0 { handle=02e8, self=0 } 0045: select( flags=2, cookie=7ffdb10c, timeout=infinite, prev_apc=0000, result={APC_VIRTUAL_QUERY,status=0,base=00330000,alloc_base=00330000,size=00001000,state=1,prot=2,alloc_prot=80,alloc_type=100}, data={WAIT,handles={0048}} ) 0045: select() = KERNEL_APC { timeout=infinite, call={APC_VIRTUAL_ALLOC,addr==00000000,size=00000087,zero_bits_64=0,op_type=3000,prot=40}, apc_handle=004c } 0009: select( flags=2, cookie=0032c9ac, timeout=infinite, prev_apc=0000, result={}, data={WAIT_ALL,handles={02e8}} ) 0009: select() = PENDING { timeout=infinite, call={APC_NONE}, apc_handle=0000 } 0045: select( flags=2, cookie=7ffdb10c, timeout=infinite, prev_apc=004c, result={APC_VIRTUAL_ALLOC,status=0,addr=00380000,size=00001000}, data={WAIT,handles={0048}} ) 0009: *wakeup* signaled=0 0045: select() = PENDING { timeout=infinite, call={APC_NONE}, apc_handle=0000 } 0009: get_apc_result( handle=02e8 ) 0009: get_apc_result() = 0 { result={APC_VIRTUAL_ALLOC,status=0,addr=00380000,size=00001000} } 0009:Ret ntdll.NtAllocateVirtualMemory() retval=00000000 ret=7b022f18 0009:Ret KERNEL32.VirtualAllocEx() retval=00380000 ret=01dd445c 0009:Call KERNEL32.WriteProcessMemory(0000024c,00380000,05ab7250,00000058,0032cea8) ret=0211b7b8 0009:Call ntdll.NtWriteVirtualMemory(0000024c,00380000,05ab7250,00000058,0032cea8) ret=7b023199 0009: write_process_memory( handle=024c, addr=00380000, data={22,00,43,00,3a,00,5c,00,77,00,69,00,6e,00,64,00,6f,00,77,00,73,00,5c,00,73,00,79,00,73,00,77,00,6f,00,77,00,36,00,34,00,5c,00,77,00,69,00,6e,00,65,00,6d,00,69,00,6e,00,65,00,2e,00,65,00,78,00,65,00,22,00,20,00,2d,00,74,00,65,00,73,00,74,00,61,00,72,00,67,00,31,00} ) 0045: *signal* signal=19 0009: write_process_memory() = 0 0009:Ret ntdll.NtWriteVirtualMemory() retval=00000000 ret=7b023199 0009:Ret KERNEL32.WriteProcessMemory() retval=00000001 ret=0211b7b8 0009:Call KERNEL32.WriteProcessMemory(0000024c,0038005a,05a989f8,0000002d,0032cea8) ret=0211b7b8 0009:Call ntdll.NtWriteVirtualMemory(0000024c,0038005a,05a989f8,0000002d,0032cea8) ret=7b023199 0009: write_process_memory( handle=024c, addr=0038005a, data={22,43,3a,5c,77,69,6e,64,6f,77,73,5c,73,79,73,77,6f,77,36,34,5c,77,69,6e,65,6d,69,6e,65,2e,65,78,65,22,20,2d,74,65,73,74,61,72,67,31,00} ) 0045: *signal* signal=19 0009: write_process_memory() = 0 0009:Ret ntdll.NtWriteVirtualMemory() retval=00000000 ret=7b023199 0009:Ret KERNEL32.WriteProcessMemory() retval=00000001 ret=0211b7b8 --- snip --- Both, ANSI and UNICODE buffers are allocated and filled in debuggee process space. https://github.com/x64dbg/x64dbg/blob/d2c62e69c2cd4d42cc4509612ad3cfd34e914… --- snip --- static bool fixgetcommandlinesbase(duint new_command_line_unicode, duint new_command_line_ascii, cmdline_error_t* cmd_line_error) { duint getcommandline; if(!valfromstring("kernelBase:GetCommandLineA", &getcommandline)) { if(!valfromstring("kernel32:GetCommandLineA", &getcommandline)) { cmd_line_error->type = CMDL_ERR_GET_GETCOMMANDLINE; return false; } } if(!patchcmdline(getcommandline, new_command_line_ascii, cmd_line_error)) return false; if(!valfromstring("kernelbase:GetCommandLineW", &getcommandline)) { if(!valfromstring("kernel32:GetCommandLineW", &getcommandline)) { cmd_line_error->type = CMDL_ERR_GET_GETCOMMANDLINE; return false; } } if(!patchcmdline(getcommandline, new_command_line_unicode, cmd_line_error)) return false; return true; } --- snip --- Corresponding Wine trace log snippet: --- snip --- ... 0009:Call KERNEL32.LoadLibraryExW(05ab78d8 L"Z:\\home\\focht\\projects\\wine\\mainline-install-x86_64\\lib\\wine\\kernelbase.dll",00000000,00000001) ret=01e0dbdf ... 0009:Ret KERNEL32.LoadLibraryExW() retval=7b000000 ret=01e0dbdf ... 0009:Call KERNEL32.GetProcAddress(7b000000,05ab78bb "GetCommandLineA") ret=01e0b046 0009:Ret KERNEL32.GetProcAddress() retval=7b0036ec ret=01e0b046 0009:Call KERNEL32.FreeLibrary(7b000000) ret=01e0ddf9 0009:Ret KERNEL32.FreeLibrary() retval=00000001 ret=01e0ddf9 ... 0009:Call KERNEL32.ReadProcessMemory(0000024c,7b0036ec,0032ce2c,00000064,0032ce08) ret=0211b6bc 0009:Call ntdll.NtReadVirtualMemory(0000024c,7b0036ec,0032ce2c,00000064,0032ce08) ret=7b022cf9 0009: read_process_memory( handle=024c, addr=7b0036ec ) 0045: *signal* signal=19 0009: read_process_memory() = 0 { data={8b,ff,55,8b,ec,5d,68,36,01,0c,00,b8,00,50,07,7b,50,ff,50,04,c2,00,00,90,90,90,90,90,90,90,90,90,8b,ff,55,8b,ec,5d,68,37,01,0c,00,b8,00,50,07,7b,50,ff,50,04,c2,00,00,90,90,90,90,90,90,90,90,90,8b,ff,55,8b,ec,5d,68,38,01,0a,00,b8,00,50,07,7b,50,ff,50,04,c2,08,00,90,90,90,90,90,90,90,90,90,8b,ff,55,8b} } 0009:Ret ntdll.NtReadVirtualMemory() retval=00000000 ret=7b022cf9 0009:Ret KERNEL32.ReadProcessMemory() retval=00000001 ret=0211b6bc ... --- snip --- x32dbg reads the opcodes of 32-bit 'kernelbase.GetCommandLineA' and is not happy about it. --- snip --- static bool patchcmdline(duint getcommandline, duint new_command_line, cmdline_error_t* cmd_line_error) { duint command_line_stored = 0; unsigned char data[100]; cmd_line_error->addr = getcommandline; if(!MemRead(cmd_line_error->addr, & data, sizeof(data))) { cmd_line_error->type = CMDL_ERR_READ_GETCOMMANDLINEBASE; return false; } #ifdef _WIN64 /* 00007FFC5B91E3C8 | 48 8B 05 19 1D 0E 00 | mov rax,qword ptr ds:[7FFC5BA000E8] 00007FFC5B91E3CF | C3 | ret | This is a relative offset then to get the symbol: next instruction of getmodulehandle (+7 bytes) + offset to symbol (the last 4 bytes of the instruction) */ if(data[0] != 0x48 || data[1] != 0x8B || data[2] != 0x05 || data[7] != 0xC3) { cmd_line_error->type = CMDL_ERR_CHECK_GETCOMMANDLINESTORED; return false; } DWORD offset = * ((DWORD*) & data[3]); command_line_stored = getcommandline + 7 + offset; #else //x86 /* 750FE9CA | A1 CC DB 1A 75 | mov eax,dword ptr ds:[751ADBCC] | 750FE9CF | C3 | ret | */ if(data[0] != 0xA1 || data[5] != 0xC3) { cmd_line_error->type = CMDL_ERR_CHECK_GETCOMMANDLINESTORED; return false; } command_line_stored = * ((duint*) & data[1]); #endif //update the pointer in the debuggee if(!MemWrite(command_line_stored, &new_command_line, sizeof(new_command_line))) { cmd_line_error->addr = command_line_stored; cmd_line_error->type = CMDL_ERR_WRITE_GETCOMMANDLINESTORED; return false; } return true; } --- snip --- Wine source: https://source.winehq.org/git/wine.git/blob/9bcb1f5195d1e65e0e7afb288d36fee… --- snip --- 1008 static STARTUPINFOW startup_infoW; 1009 static char *command_lineA; 1010 static WCHAR *command_lineW; --- snip --- https://source.winehq.org/git/wine.git/blob/9bcb1f5195d1e65e0e7afb288d36fee… --- snip --- 1043 /*********************************************************************** 1044 * GetCommandLineA (kernelbase.@) 1045 */ 1046 LPSTR WINAPI DECLSPEC_HOTPATCH GetCommandLineA(void) 1047 { 1048 return command_lineA; 1049 } 1050 1051 1052 /*********************************************************************** 1053 * GetCommandLineW (kernelbase.@) 1054 */ 1055 LPWSTR WINAPI DECLSPEC_HOTPATCH GetCommandLineW(void) 1056 { 1057 return NtCurrentTeb()->Peb->ProcessParameters->CommandLine.Buffer; 1058 } --- snip --- Depending on which compiler (GCC or LLVM/MingW) is used to build Wine, the following code is generated: === GCC === --- snip -- $ gcc -v Using built-in specs. COLLECT_GCC=/usr/bin/gcc COLLECT_LTO_WRAPPER=/usr/libexec/gcc/x86_64-redhat-linux/8/lto-wrapper OFFLOAD_TARGET_NAMES=nvptx-none OFFLOAD_TARGET_DEFAULT=1 Target: x86_64-redhat-linux Configured with: ../configure --enable-bootstrap --enable-languages=c,c++,fortran,objc,obj-c++,ada,go,lto --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-bugurl=http://bugzilla.redhat.com/bugzilla --enable-shared --enable-threads=posix --enable-checking=release --enable-multilib --with-system-zlib --enable-__cxa_atexit --disable-libunwind-exceptions --enable-gnu-unique-object --enable-linker-build-id --with-gcc-major-version-only --with-linker-hash-style=gnu --enable-plugin --enable-initfini-array --with-isl --enable-libmpx --enable-offload-targets=nvptx-none --without-cuda-driver --enable-gnu-indirect-function --enable-cet --with-tune=generic --with-arch_32=i686 --build=x86_64-redhat-linux Thread model: posix gcc version 8.3.1 20190223 (Red Hat 8.3.1-2) (GCC) --- snip --- --- snip --- $ gdb mainline-install-x86_64/lib64/wine/kernelbase.dll.so -batch -ex 'disassemble GetCommandLineA' Dump of assembler code for function GetCommandLineA: 0x000000007b05fd18 <+0>: push %rbp 0x000000007b05fd19 <+1>: mov 0x738a0(%rip),%rax # 0x7b0d35c0 <command_lineA> 0x000000007b05fd20 <+8>: mov %rsp,%rbp 0x000000007b05fd23 <+11>: and $0xfffffffffffffff0,%rsp 0x000000007b05fd27 <+15>: leaveq 0x000000007b05fd28 <+16>: retq End of assembler dump. --- snip --- --- snip --- $ gdb mainline-install-x86_64/lib64/wine/kernelbase.dll.so -batch -ex 'disassemble GetCommandLineW' Dump of assembler code for function GetCommandLineW: 0x000000007b05fd58 <+0>: push %rbp 0x000000007b05fd59 <+1>: mov %gs:0x30,%rax 0x000000007b05fd62 <+10>: mov 0x60(%rax),%rax 0x000000007b05fd66 <+14>: mov %rsp,%rbp 0x000000007b05fd69 <+17>: and $0xfffffffffffffff0,%rsp 0x000000007b05fd6d <+21>: mov 0x20(%rax),%rax 0x000000007b05fd71 <+25>: mov 0x78(%rax),%rax 0x000000007b05fd75 <+29>: leaveq 0x000000007b05fd76 <+30>: retq End of assembler dump. --- snip --- NOTE: offset-0x4 to work around dwarf symbol processing bug. --- snip --- $ gdb mainline-install-x86_64/lib/wine/kernelbase.dll.so -batch -ex 'disassemble GetCommandLineA-0x4' Dump of assembler code for function GetCommandLineA: 0x7b060620 <+0>: mov %edi,%edi 0x7b060622 <+2>: push %ebp 0x7b060623 <+3>: mov %esp,%ebp 0x7b060625 <+0>: pop %ebp 0x7b060626 <+1>: mov 0x7b0d4a20,%eax 0x7b06062b <+6>: ret End of assembler dump. --- snip --- --- snip --- $ gdb mainline-install-x86_64/lib/wine/kernelbase.dll.so -batch -ex 'disassemble GetCommandLineW-0x4' Dump of assembler code for function GetCommandLineW: 0x7b060640 <+0>: mov %edi,%edi 0x7b060642 <+2>: push %ebp 0x7b060643 <+3>: mov %esp,%ebp 0x7b060645 <+0>: mov %fs:0x18,%eax 0x7b06064b <+6>: mov 0x30(%eax),%eax 0x7b06064e <+9>: pop %ebp 0x7b06064f <+10>: mov 0x10(%eax),%eax 0x7b060652 <+13>: mov 0x44(%eax),%eax 0x7b060655 <+16>: ret End of assembler dump. --- snip --- === LLVM MinGW === --- snip --- $ clang -v clang version 10.0.0 (https://github.com/llvm/llvm-project.git c49194969430f0ee817498a7000a979a7a0ded03) Target: x86_64-unknown-linux-gnu Thread model: posix InstalledDir: /home/focht/projects/llvm-mingw-20191230-ubuntu-16.04/bin Found candidate GCC installation: /usr/lib/gcc/x86_64-redhat-linux/8 Selected GCC installation: /usr/lib/gcc/x86_64-redhat-linux/8 Candidate multilib: .;@m64 Candidate multilib: 32;@m32 Selected multilib: .;@m64 --- snip --- --- snip --- $ gdb mainline-install-x86_64/lib/wine/kernelbase.dll -batch -ex 'disassemble GetCommandLineA' Dump of assembler code for function _GetCommandLineA(a)0: 0x7b0313e0 <+0>: push %ebp 0x7b0313e1 <+1>: mov %esp,%ebp 0x7b0313e3 <+3>: mov 0x7b076bf8,%eax 0x7b0313e8 <+8>: pop %ebp 0x7b0313e9 <+9>: ret End of assembler dump. --- snip --- --- snip --- $ gdb mainline-install-x86_64/lib/wine/kernelbase.dll -batch -ex 'disassemble GetCommandLineW' Dump of assembler code for function _GetCommandLineW(a)0: 0x7b0313f0 <+0>: push %ebp 0x7b0313f1 <+1>: mov %esp,%ebp 0x7b0313f3 <+3>: mov %fs:0x18,%eax 0x7b0313f9 <+9>: mov 0x30(%eax),%eax 0x7b0313fc <+12>: mov 0x10(%eax),%eax 0x7b0313ff <+15>: mov 0x44(%eax),%eax 0x7b031402 <+18>: pop %ebp 0x7b031403 <+19>: ret End of assembler dump. --- snip --- --- snip --- $ gdb mainline-install-x86_64/lib64/wine/kernelbase.dll -batch -ex 'disassemble GetCommandLineA' Dump of assembler code for function GetCommandLineA: 0x000000007b0389b0 <+0>: mov 0x4de49(%rip),%rax # 0x7b086800 0x000000007b0389b7 <+7>: retq End of assembler dump. --- snip --- --- snip --- $ gdb mainline-install-x86_64/lib64/wine/kernelbase.dll -batch -ex 'disassemble GetCommandLineW' Dump of assembler code for function GetCommandLineW: 0x000000007b0389c0 <+0>: mov %gs:0x30,%rax 0x000000007b0389c9 <+9>: mov 0x60(%rax),%rax 0x000000007b0389cd <+13>: mov 0x20(%rax),%rax 0x000000007b0389d1 <+17>: mov 0x78(%rax),%rax 0x000000007b0389d5 <+21>: retq End of assembler dump. --- snip --- ===== General problems: (1) asymmetric design of kernelbase GetCommandLineA/W commmand buffer handling The ANSI variant returns the command line buffer address directly while the UNICODE variant goes through the PEB. The symmetric design was changed in 2003 with https://source.winehq.org/git/wine.git/commitdiff/b53b5bcb50e75263e8bbffed3… ("- fixed a couple of bugs in ntdll...") to be asymmetrical. A big squash-like commit, doing major refactoring/various bugfixing = several things in one go. No mentioning of this specific design change. Part of Wine 0.9 release. (2) HOTPATCH entry Commit https://source.winehq.org/git/wine.git/commitdiff/911e50849a9bc05ab11d896a3… ("kernel32: Move process startup information functions to kernelbase.") added it. Part of Wine 4.16 release. I couldn't find a bug report about hooking problems of GetCommandLineA/W which requires those to be hotpatch-compatible entries. I guess this was added for no special reason :| (3) compiler choice/optimization settings, influencing generated code --- Fixing above problems is not that easy by just returning 'command_lineW' in 'W' variant and removing DECLSPEC_HOTPATCH because of (3). Generated code: --- snip --- $ gdb mainline-install-x86_64/lib/wine/kernelbase.dll -batch -ex 'disassemble GetCommandLineA' Dump of assembler code for function _GetCommandLineA(a)0: 0x7b0313e0 <+0>: push %ebp 0x7b0313e1 <+1>: mov %esp,%ebp 0x7b0313e3 <+3>: mov 0x7b076bf8,%eax 0x7b0313e8 <+8>: pop %ebp 0x7b0313e9 <+9>: ret End of assembler dump. --- snip --- --- snip --- $ gdb mainline-install-x86_64/lib/wine/kernelbase.dll -batch -ex 'disassemble GetCommandLineW' Dump of assembler code for function _GetCommandLineW(a)0: 0x7b0313f0 <+0>: push %ebp 0x7b0313f1 <+1>: mov %esp,%ebp 0x7b0313f3 <+3>: mov 0x7b076bf4,%eax 0x7b0313f8 <+8>: pop %ebp 0x7b0313f9 <+9>: ret End of assembler dump. --- snip --- --- snip --- $ gdb mainline-install-x86_64/lib64/wine/kernelbase.dll -batch -ex 'disassemble GetCommandLineA' Dump of assembler code for function GetCommandLineA: 0x000000007b0389b0 <+0>: mov 0x4de49(%rip),%rax # 0x7b086800 0x000000007b0389b7 <+7>: retq End of assembler dump. --- snip --- --- snip --- $ gdb mainline-install-x86_64/lib64/wine/kernelbase.dll -batch -ex 'disassemble GetCommandLineW' Dump of assembler code for function GetCommandLineW: 0x000000007b0389c0 <+0>: mov 0x4de31(%rip),%rax # 0x7b0867f8 0x000000007b0389c7 <+7>: retq End of assembler dump. --- snip --- The 64-bit variant looks good, but 32-bit is no-go. To have full control over the generated code, small __i386__ and __x86_64__ assembly implementations for A and W variant each are sufficient. NOTE: the assembly for __x86_64__ needs to use RIP-relative addressing (PIE) -> movq command_lineW(%rip), %rax * encoding: [0x48,0x8b,0x05,A,A,A,A] * kind: reloc_riprel_4byte_movq_load With the small assembly functions in place, x64dbg and x32dbg were able to set the debuggee command line via menu. If assembly is not an option, consider changing A/W variant to be symmetrical to make this bug report somewhat useful in the end (but still WONTFIX). $ sha1sum snapshot_2020-02-11_01-07.zip 03fa18ffb1cbdd46a6b3c9b7e19332e7e2024edc snapshot_2020-02-11_01-07.zip $ du -sh snapshot_2020-02-11_01-07.zip 31M snapshot_2020-02-11_01-07.zip $ wine --version wine-5.5 Regards -- 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.

March 30, 2020

[Bug 48860] RegiStax 6: mouse does not work in open file dialog

by WineHQ Bugzilla

https://bugs.winehq.org/show_bug.cgi?id=48860 Zebediah Figura <z.figura12(a)gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Regression SHA1|62ed79ae1ad14d8c46309e93d52 | |1cef5797625de | -- 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.

March 30, 2020

[Bug 48860] RegiStax 6: mouse does not work in open file dialog

by WineHQ Bugzilla

https://bugs.winehq.org/show_bug.cgi?id=48860 Zebediah Figura <z.figura12(a)gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Keywords| |regression Regression SHA1|62ed79ae1ad14d8c46309e93d52 |62ed79ae1ad14d8c46309e93d52 |1cef5797625de |1cef5797625de |setupregistax6.exe | -- 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.

March 30, 2020

[Bug 48860] RegiStax 6: mouse does not work in open file dialog

by WineHQ Bugzilla

https://bugs.winehq.org/show_bug.cgi?id=48860 eserradi(a)gmx.com changed: What |Removed |Added ---------------------------------------------------------------------------- Hardware|x86 |x86-64 -- 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.

March 30, 2020

[Bug 48860] New: RegiStax 6: mouse does not work in open file dialog

by WineHQ Bugzilla

https://bugs.winehq.org/show_bug.cgi?id=48860 Bug ID: 48860 Summary: RegiStax 6: mouse does not work in open file dialog Product: Wine Version: 5.0 Hardware: x86 URL: http://www.astronomie.be/registax/setupregistax6.exe OS: Linux Status: UNCONFIRMED Severity: minor Priority: P2 Component: -unknown Assignee: wine-bugs(a)winehq.org Reporter: eserradi(a)gmx.com Regression SHA1: 62ed79ae1ad14d8c46309e93d521cef5797625de setupregistax6.exe Distribution: Debian Mouse does not work in open file dialog in RegiStax V6. A workaround to restore mouse functionality is to press the tab key until the combobox with label "Look in:", then press down key to expand the item list and move mouse over one item to make the mouse work again. It worked very well in wine version 1.8.7, then switched to wine v4 and v5 and stop working. It looks like wine opens two instances of RegiStax as in the title bar appears "RegiStax V6 <2>", it might be related to the mouse problem. Previous versions of the program are not affected. -- 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.

March 30, 2020

[Bug 48802] Logitech Extreme 3D Pro joystick works fine in joy.cpl but buttons wrongly mapped in Battlefield 4

by WineHQ Bugzilla

https://bugs.winehq.org/show_bug.cgi?id=48802 --- Comment #32 from Luca Boccassi <luca.boccassi(a)gmail.com> --- (In reply to Zebediah Figura from comment #31) > I do not have access to Battlefield 4, no. If you are up for more debugging I'd be happy to donate a copy, if it can help? -- 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.

March 30, 2020

[Bug 48855] Borderlands 3: Unable to save

by WineHQ Bugzilla

https://bugs.winehq.org/show_bug.cgi?id=48855 --- Comment #2 from mrs11teen(a)gmail.com --- I think I managed to apply the patch to file.c correctly in v5.5 (I couldn't get the test patch to apply though, my tests/file.c file looks different for some reason.) Unfortunately, the error still occurs and terminal output still throws that same fixme. -- 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.

March 30, 2020

[Bug 48858] Passmark PerformanceTest fails to run CPU test

by WineHQ Bugzilla

https://bugs.winehq.org/show_bug.cgi?id=48858 --- Comment #1 from Austin English <austinenglish(a)gmail.com> --- Anything in the terminal? -- 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.

March 30, 2020

[Bug 48851] WinSCP file transfer "The specified data could not be decrypted"

by WineHQ Bugzilla

https://bugs.winehq.org/show_bug.cgi?id=48851 --- Comment #2 from Austin English <austinenglish(a)gmail.com> --- (In reply to Timothy C. Quinn from comment #1) > If there are any pre-built or wine-ready linux distributions that can run a > later version of wine, I'd be glad to test on a VM. However I cannot upgrade > my current version of my production desktop at this time so I'm stuck with > 4.0.3 at this time. WineHQ doesn't have any, but it does have binaries prebuilt for debian/fedora/ubuntu, so you should be able to boot a livecd of one of those in a vm and install the wine package: https://wiki.winehq.org/Download -- 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.

March 30, 2020