http://bugs.winehq.org/show_bug.cgi?id=34642
Bug #: 34642 Summary: Premiere Pro 2 not finish loading Product: Wine Version: 1.7.0 Platform: x86-64 OS/Version: Linux Status: UNCONFIRMED Severity: critical Priority: P2 Component: ole AssignedTo: wine-bugs@winehq.org ReportedBy: tom1vuu2011@gmail.com Classification: Unclassified
Last two lines of log:
err:pulse:pulse_contextcallback Context failed: Connection refused err:ole:RevokeDragDrop invalid hwnd (nil)
I don't know if it's failing because pulse isn't installed.
http://bugs.winehq.org/show_bug.cgi?id=34642
Rosanne DiMesio dimesio@earthlink.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Severity|critical |normal
--- Comment #1 from Rosanne DiMesio dimesio@earthlink.net 2013-10-02 14:00:42 CDT --- Not critical. http://bugs.winehq.org/page.cgi?id=fields.html#importance
http://bugs.winehq.org/show_bug.cgi?id=34642
--- Comment #2 from John Franklin tom1vuu2011@gmail.com 2013-10-03 03:20:13 CDT --- Pulse is not the issue.
https://bugs.winehq.org/show_bug.cgi?id=34642
Michael Müller michael@fds-team.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |michael@fds-team.de
--- Comment #3 from Michael Müller michael@fds-team.de --- Hi,
I tried to debug this issue and found a function related to the problem. I was unable to find the exact reason for the problem, maybe someone else has more luck to reveal the last piece of the puzzle using my findings. However, I came up with a workaround to prevent this problem.
When you start premiere pro 2 using +relay as debug channel, the application works as expected but is really slow because of all the debug output. This means that the problem is either a timing issue or related to the redirect coded added by the +relay channel.
To verify my assumption, I added my own debug channel which does the same like relay, but without printing any debug messages at all. The program still worked using the new debug channel, so I limited the dlls affected by my code to find out which one is causing the problem. Adding the relay code only to ntdll.dll was sufficient to make premiere pro happy. I did the same with the ordinals exported by ntdll and was able to pinpoint it to a single function: RtlInterlockedPushEntrySList
Adding the relay code only to ntdll.dll.RtlInterlockedPushEntrySList is sufficient to make the application start. Anyway, I was unable to find a proper fix for this without adding the relay code. My idea that the function may need to be hotpatchable, like the relay code, was wrong since adding DECLSPEC_HOTPATCH to RtlInterlockedPushEntrySList did not solve the problem.
(Tested with wine-1.7.26)
Regards, Michael
https://bugs.winehq.org/show_bug.cgi?id=34642
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download Status|UNCONFIRMED |NEW URL| |http://download.adobe.com/p | |ub/adobe/magic/creativesuit | |e/CS2_EOL/PPRO/PPRO_2.0_Ret | |-NH_UE.zip CC| |focht@gmx.net Component|ole |user32 Summary|Premiere Pro 2 not finish |Adobe Premiere Pro 2.0 |loading |exits silently on startup | |('GetMenuItemInfo' must | |zero out 'cch' if mask | |doesn't specify | |'MIIM_TYPE') Ever confirmed|0 |1
--- Comment #4 from Anastasius Focht focht@gmx.net --- Hello folks,
confirming.
If you are encountering strange issues when using 'relay' it's best to leave it out of the equation completely. It modifies stack layout/values which effects code that uses stack-based data structures with uninitialized data members -> tends to hide/move problem domains.
--- snip --- $ pwd /home/focht/.wine/drive_c/Program Files/Adobe/Adobe Premiere Pro 2.0
$ WINEDEBUG=+tid,+seh,+loaddll,+process wine ./Adobe\ Premiere\ Pro.exe
log.txt 2>&1
... 0009:trace:loaddll:load_native_dll Loaded L"C:\program files\adobe\adobe premiere pro 2.0\plug-ins\en_us\premiere asio.dll" at 0xe3e0000: native 0009:fixme:win:EnumDisplayDevicesW ((null),0,0x33e760,0x00000000), stub! 0009:trace:loaddll:load_builtin_dll Loaded L"C:\windows\system32\dinput.dll" at 0xf2700000: builtin 0009:trace:loaddll:free_modref Unloaded module L"C:\windows\system32\dinput.dll" : builtin 0009:trace:loaddll:load_builtin_dll Loaded L"C:\windows\system32\mmdevapi.dll" at 0xf3060000: builtin 0009:trace:loaddll:load_builtin_dll Loaded L"C:\windows\system32\winealsa.drv" at 0xf25e0000: builtin ... 0009:trace:seh:raise_exception code=e06d7363 flags=1 addr=0x7b83ae8f ip=7b83ae8f tid=0009 0009:trace:seh:raise_exception info[0]=19930520 0009:trace:seh:raise_exception info[1]=0033e868 0009:trace:seh:raise_exception info[2]=7e67ef24 0009:trace:seh:raise_exception eax=7b826d6d ebx=7b8be000 ecx=19930520 edx=0033e794 esi=76a40082 edi=3b520041 0009:trace:seh:raise_exception ebp=0033e7d8 esp=0033e774 cs=0023 ds=002b es=002b fs=0063 gs=006b flags=00200283 0009:trace:seh:call_stack_handlers calling handler at 0x97ecea code=e06d7363 flags=1 0009:trace:seh:call_stack_handlers handler at 0x97ecea returned 1 ... 0009:trace:loaddll:free_modref Unloaded module L"C:\Program Files\Adobe\Adobe Premiere Pro 2.0\Plug-ins\Common\PremiereFiltersMetaPlugin.prmp" : native 0009:trace:loaddll:free_modref Unloaded module L"C:\Program Files\Adobe\Adobe Premiere Pro 2.0\Plug-ins\en_US\TransitionGPUCenterPeel.prm" : native 0009:trace:loaddll:free_modref Unloaded module L"C:\Program Files\Adobe\Adobe Premiere Pro 2.0\Plug-ins\en_US\TransitionGPUPageCurl.prm" : native ... --- snip ---
The C++ exception is the interesting one, it's the result of a memory allocation failure.
Call site:
--- snip --- ... 00974AE4 56 PUSH ESI 00974AE5 8B7424 08 MOV ESI,DWORD PTR SS:[ESP+8] ; alloc size 00974AE9 EB 11 JMP SHORT UIFramew.00974AFC 00974AEB 56 PUSH ESI 00974AEC E8 DF080000 CALL <JMP.&MSVCR71._callnewh> 00974AF1 85C0 TEST EAX,EAX 00974AF3 59 POP ECX 00974AF4 75 06 JNZ SHORT UIFramew.00974AFC 00974AF6 FF15 603DA900 CALL DWORD PTRDS:[<&MSVCP71.?_Nomemory@std@@YAXXZ>] 00974AFC 56 PUSH ESI 00974AFD E8 AA040000 CALL <JMP.&MSVCR71.malloc> 00974B02 85C0 TEST EAX,EAX 00974B04 59 POP ECX 00974B05 74 E4 JE SHORT UIFramew.00974AEB 00974B07 5E POP ESI 00974B08 C3 RETN --- snip ---
Caller parent:
--- snip --- 00826C50 55 PUSH EBP 00826C51 56 PUSH ESI 00826C52 33ED XOR EBP,EBP 00826C54 57 PUSH EDI 00826C55 8B7C24 10 MOV EDI,DWORD PTR SS:[ESP+10] ; #wchars 00826C59 3BFD CMP EDI,EBP 00826C5B 8BF1 MOV ESI,ECX 00826C5D 896E 04 MOV DWORD PTR DS:[ESI+4],EBP 00826C60 896E 08 MOV DWORD PTR DS:[ESI+8],EBP 00826C63 896E 0C MOV DWORD PTR DS:[ESI+C],EBP 00826C66 74 4A JE SHORT UIFramew.00826CB2 00826C68 81FF FFFFFF7F CMP EDI,7FFFFFFF 00826C6E 76 05 JBE SHORT UIFramew.00826C75 00826C70 E8 0D5DF3FF CALL UIFramew.0075C982 00826C75 53 PUSH EBX 00826C76 8D1C3F LEA EBX,DWORD PTR DS:[EDI+EDI] ; bytes = #wchars*2 00826C79 53 PUSH EBX 00826C7A E8 65DE1400 CALL UIFramew.00974AE4 ; allocate 00826C7F 8D0C03 LEA ECX,DWORD PTR DS:[EBX+EAX] --- snip ---
Next caller parent gives the hint:
--- snip -- ... 00826F86 C78424 B0000000... MOV DWORD PTR SS:[ESP+B0],1 00826F91 C74424 68 30000000 MOV DWORD PTR SS:[ESP+68],30 00826F99 C74424 6C 62000000 MOV DWORD PTR SS:[ESP+6C],62 00826FA1 899C24 8C000000 MOV DWORD PTR SS:[ESP+8C],EBX 00826FA8 899C24 88000000 MOV DWORD PTR SS:[ESP+88],EBX 00826FAF FF15 1443A900 CALL DWORD PTR DS:[<&USER32.GetMenuItemInfoW>] 00826FB5 3BC3 CMP EAX,EBX 00826FB7 0F84 15010000 JE UIFramew.008270D2 00826FBD 8B8424 80000000 MOV EAX,DWORD PTR SS:[ESP+80] ; cch 00826FC4 8D4C24 18 LEA ECX,DWORD PTR SS:[ESP+18] 00826FC8 894424 10 MOV DWORD PTR SS:[ESP+10],EAX 00826FCC 51 PUSH ECX 00826FCD 40 INC EAX ; include '\0' (+1) 00826FCE 50 PUSH EAX ; #wchars 00826FCF 8D8C24 90000000 LEA ECX,DWORD PTR SS:[ESP+90] 00826FD6 895C24 20 MOV DWORD PTR SS:[ESP+20],EBX 00826FDA E8 72D1F2FF CALL UIFramew.00754151 ... <loop menu items> ... 00754151 E9 FA2A0D00 JMP UIFramew.00826C50 ... --- snip ---
The code calls 'GetMenuItemInfoW' with only partially initialized stack-based 'MENUITEMINFO' structure ('cbSize' and 'fMask' members).
After return from 'GetMenuItemInfoW':
--- snip --- $+0 00000030 ; cbSize $+4 00000062 ; fMask (MIIM_ID | MIIM_DATA | MIIM_STRING) $+8 7ED98000 ; fType $+C 0033E9E0 ; fState $+10 00000000 ; wID $+14 00000000 ; hSubMenu $+18 3AF80042 ; hbmpChecked $+1C 006D0DB0 ; hbmpUnchecked $+20 00000000 ; dwItemData $+24 00000000 ; dwTypeData $+28 0097E8F2 ; cch ; leftover stack garbage from previous usage $+2C 00000000 ; hbmpItem --- snip ---
Source: http://source.winehq.org/git/wine.git/blob/cb2b15921daeebf763e8eb34fd9fa336d...
MSDN: http://msdn.microsoft.com/en-us/library/windows/desktop/ms647578%28v=vs.85%2...
--- quote --- MIIM_DATA 0x00000020 Retrieves or sets the dwItemData member. ... MIIM_ID 0x00000002 Retrieves or sets the wID member. ... MIIM_STRING 0x00000040 Retrieves or sets the dwTypeData member. --- quote ---
--- quote --- cch
Type: UINT
The length of the menu item text, in characters, when information is received about a menu item of the MFT_STRING type. However, cch is used only if the MIIM_TYPE flag is set in the fMask member and is zero otherwise. Also, cch is ignored when the content of a menu item is set by calling SetMenuItemInfo.
Note that, before calling GetMenuItemInfo, the application must set cch to the length of the buffer pointed to by the dwTypeData member. If the retrieved menu item is of type MFT_STRING (as indicated by the fType member), then GetMenuItemInfo changes cch to the length of the menu item text. If the retrieved menu item is of some other type, GetMenuItemInfo sets the cch field to zero.
The cch member is used when the MIIM_STRING flag is set in the fMask member. --- quote ---
$ sha1sum PPRO_2.0_Ret-NH_UE.zip 24dc9b897b83f7a4e75f7bb5921a6e2a5ba94d36 PPRO_2.0_Ret-NH_UE.zip
$ du -sh PPRO_2.0_Ret-NH_UE.zip 1.1G PPRO_2.0_Ret-NH_UE.zip $ wine --version wine-1.7.26
Tidbit: the Adobe PPro 2.0 installer crashes for me at the end ... no one mentioned this here?
Regards
https://bugs.winehq.org/show_bug.cgi?id=34642
Sebastian Lackner sebastian@fds-team.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |sebastian@fds-team.de
--- Comment #5 from Sebastian Lackner sebastian@fds-team.de --- Thanks for the nice analysis. :) The following patch by Michael Müller should fix this issue:
https://github.com/compholio/wine-compholio/blob/master/patches/user32-GetMe...
https://bugs.winehq.org/show_bug.cgi?id=34642
--- Comment #6 from Anastasius Focht focht@gmx.net --- Hello FDS team,
--- quote --- Thanks for the nice analysis. :) The following patch by Michael Müller should fix this issue: --- quote ---
no problem ;-)
I still wonder ... you didn't encounter a crash with the installer? This is the more interesting bug to me as it's about license manager/activation.
Regards
https://bugs.winehq.org/show_bug.cgi?id=34642
--- Comment #7 from Michael Müller michael@fds-team.de --- Hi AF,
I downloaded the german EOL version some time ago and it does not crash after the installation:
24dc9b897b83f7a4e75f7bb5921a6e2a5ba94d36 PPRO_2.0_Ret-NH_D.zip
However, there are still bugs in the application itself, like you can not import any videos. Premiere pro will simply freeze in a critical section which makes the application quite useless. I am going to open a separate bug report for it.
Regards, Michael
https://bugs.winehq.org/show_bug.cgi?id=34642
--- Comment #8 from Anastasius Focht focht@gmx.net --- Hello folks,
the following link gives some insight: http://www.xyclopsoft.com/blog/2013/01/07/get-free-adobe-cs2/
After EOL'ing CS2 products and making them available for free (at own risk on modern Vista/Win7 OS), Adobe also provided special serials/keys for the product downloads.
Those keys have the online activation scheme *deactivated* by design. I ran into the problem/crash because the old key I used for investigating the installer problem involves activation scheme. This is no longer supported since the activation server for these EOL'd products is dead anyway.
If people encounter this problem of CS2/PPro2 installer crashing with Wine using an old key which involves product activation and the installer works for them on Windows please notify/contact me, so I might be able to investigate further.
Regards
https://bugs.winehq.org/show_bug.cgi?id=34642
Michael Müller michael@fds-team.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |37239
https://bugs.winehq.org/show_bug.cgi?id=34642
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Fixed by SHA1| |997e197f23144fd53af6fceeec4 | |71411996ed6a8 Status|NEW |RESOLVED Resolution|--- |FIXED
--- Comment #9 from Anastasius Focht focht@gmx.net --- Hello folks,
this is fixed by commit http://source.winehq.org/git/wine.git/commitdiff/997e197f23144fd53af6fceeec4...
Thanks Michael
Regards
https://bugs.winehq.org/show_bug.cgi?id=34642
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #10 from Alexandre Julliard julliard@winehq.org --- Closing bugs fixed in 1.7.27.
-
wine-bugs＠winehq.org