http://bugs.winehq.org/show_bug.cgi?id=25512
Summary: The program crashes after splash screen Product: Wine Version: 1.3.9 Platform: x86 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs@winehq.org ReportedBy: newilkman@yahoo.com
This is a new version of WTLibrary and a main change i of MS Visual C++ 2008 to MS Visual C++ 2009. The program starts an te Splask screen is shown and then a second screen is displayed which says that the program can't continue due to a program error.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #1 from Igor Tarasov tarasov.igor@gmail.com 2010-12-13 14:23:22 CST --- Created an attachment (id=32483) --> (http://bugs.winehq.org/attachment.cgi?id=32483) Crash log
Crash log as reported by one of the users.
http://bugs.winehq.org/show_bug.cgi?id=25512
Paul b01@westhouseit.co.uk changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW Ever Confirmed|0 |1
--- Comment #2 from Paul b01@westhouseit.co.uk 2010-12-13 15:16:19 CST --- *** This bug has been confirmed by popular vote. ***
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #3 from Jeff Zaroyko jeffz@jeffz.name 2010-12-13 15:27:07 CST --- Please update the summary to include the name of the program.
The crash log is missing debug symbols from the backtrace, please compile from source and generate a new one. http://wiki.winehq.org/Backtraces
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #4 from nille newilkman@yahoo.com 2010-12-14 09:44:19 CST --- The program that crashes is Watchtower Library 2010.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #5 from nille newilkman@yahoo.com 2010-12-14 10:15:15 CST --- Created an attachment (id=32492) --> (http://bugs.winehq.org/attachment.cgi?id=32492) WINEDEBUG=+iphlpapi wine start WTLibrary.exe &>debug.txt
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #6 from Robert Backhaus robbak@robbak.com 2010-12-14 19:10:04 CST --- Created an attachment (id=32496) --> (http://bugs.winehq.org/attachment.cgi?id=32496) Crash log and backtrace with debug symbols, WINEDEBUG=+iphlpapi
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #7 from Juan Lang juan_lang@yahoo.com 2010-12-14 19:24:30 CST --- Why are you including +iphlpapi, when the crash is in msvcr90?
What happens if you use builtin msvcr90 instead?
http://bugs.winehq.org/show_bug.cgi?id=25512
Robert Backhaus robbak@robbak.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |robbak@robbak.com
--- Comment #8 from Robert Backhaus robbak@robbak.com 2010-12-14 19:31:05 CST --- (In reply to comment #7)
Why are you including +iphlpapi, when the crash is in msvcr90?
What happens if you use builtin msvcr90 instead?
Another user, who claims to have a working copy using native iphlpapi, suggested it. doesn't work for me, though.
Is there a builtin msvcr90? I have not put any overrides to load native. The installer sideloads the msvc runtime.
http://bugs.winehq.org/show_bug.cgi?id=25512
Dmitry Timoshkov dmitry@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|The program crashes after |Watchtower Library 2010 |splash screen |crashes after splash screen
http://bugs.winehq.org/show_bug.cgi?id=25512
business.kid@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |business.kid@gmail.com
--- Comment #9 from business.kid@gmail.com 2010-12-16 03:57:07 CST --- (In reply to comment #8)
(In reply to comment #7)
Why are you including +iphlpapi, when the crash is in msvcr90?
What happens if you use builtin msvcr90 instead?
Another user, who claims to have a working copy using native iphlpapi, suggested it. doesn't work for me, though.
Is there a builtin msvcr90? I have not put any overrides to load native. The installer sideloads the msvc runtime.
At JHB's suggestion, I have tried this with msvcr90 set to 'builtin' only on a 32 bit box, with clean install. I don't even get the splash screen. I'm a noob with bugzilla, but I'll try to attach the full log. Business_kid
bash-4.1$ grep -n unimplemented WarnHeapTraceNtdllLoaddll 47900:wine: Call from 0x7eda6b83 to unimplemented function msvcr90.dll._ui64tow_s, aborting 52038:Unhandled exception: unimplemented function msvcr90.dll._ui64tow_s called in 32-bit code (0x7eda6b83). 52118: 1 0x7e5e65cf __wine_spec_unimplemented_stub+0x2e(module="msvcr90.dll", function="ÿ ã'") [/home/dec/wine-git/dlls/winecrt0/stub.c:34] in msvcr90 (0x0033e7e8) 52301: 1 0x7e5e65cf __wine_spec_unimplemented_stub+0x2e(module="msvcr90.dll", function="ÿ ã'") [/home/dec/wine-git/dlls/winecrt0/stub.c:34] in msvcr90 (0x0033e7e8) 52310:wine: Call from 0x7eda6b83 to unimplemented function msvcr90.dll._ui64tow_s, aborting 52311:wine: Call from 0x7eda6b83 to unimplemented function msvcr90.dll._ui64tow_s, aborting
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #10 from business.kid@gmail.com 2010-12-16 04:01:08 CST --- Created an attachment (id=32508) --> (http://bugs.winehq.org/attachment.cgi?id=32508) Debug output.
This is from a clean install on wine-git-2010-12-11 with msvcr90 set to builtin (only & )WINEDEBUG=+warn,+heap,+trace,+msvcr90,+ntdll,+loaddll Gets interesting around line 47,900
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #11 from Juan Lang juan_lang@yahoo.com 2010-12-16 11:56:48 CST --- Created an attachment (id=32515) --> (http://bugs.winehq.org/attachment.cgi?id=32515) Patch: Implement _ui64tow_s
Does this (totally untested, copy and paste) patch help with native msvcr90?
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #12 from Juan Lang juan_lang@yahoo.com 2010-12-16 11:58:25 CST --- (In reply to comment #8)
Another user, who claims to have a working copy using native iphlpapi, suggested it. doesn't work for me, though.
I be very surprised if that claim were true. Native iphlpapi is likely to depend on unimplemented, and probably undocumented, Windows features, and I wouldn't expect it to work at all.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #13 from business.kid@gmail.com 2010-12-16 17:21:31 CST --- Created an attachment (id=32520) --> (http://bugs.winehq.org/attachment.cgi?id=32520) Patched log
OK. Rebuilt wine-git-2010-12-11 with the patch applied. Ran with native msvcr90 only. Fresh ~/.wine and fresh install of WTLibrary 2010. Haven't opened it to look - I don't want a divorce!
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #14 from Juan Lang juan_lang@yahoo.com 2010-12-16 17:27:46 CST --- (In reply to comment #13)
Created an attachment (id=32520)
--> (http://bugs.winehq.org/attachment.cgi?id=32520) [details]
Patched log
Could you attach one without any WINEDEBUG settings?
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #15 from business.kid@gmail.com 2010-12-17 04:24:16 CST --- Created an attachment (id=32528) --> (http://bugs.winehq.org/attachment.cgi?id=32528) 2 debug outputs
Well, this is strange. The first attached trace is with no debug symbols, and a builtin msvcr90 & msvcrt (That was patched with your msvcr90 patch as well). Then I added a native iphlpapi.dll from dll-files.com and it sat uo and ran, on your patched version(32 bits) with msvcr90 & msvcrt set as builtin, and iphlpapi set as native. It also runs on the 64 bit box, (without your patch) with iphlpapi set as native, and no other options. I ran WINEDEBUG=+loaddll on the 64 bit to check, and it's using builtin (unpatched) msvcr90 & msvcrt files. It is not happy in itself (err:ole:CoGetClassObjects {blah} not registered on 64 bit & err:fixme:__clean_info_type_names_internal (blah) stub on the 32 bit. Nevertheless, iphlpapi is the difference between using it and not using it. I'll continue try stuff as long as you need that done here, of course.
http://bugs.winehq.org/show_bug.cgi?id=25512
business.kid@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #32508|0 |1 is obsolete| | Attachment #32520|0 |1 is obsolete| |
--- Comment #16 from business.kid@gmail.com 2010-12-18 08:05:00 CST --- Created an attachment (id=32535) --> (http://bugs.winehq.org/attachment.cgi?id=32535) wine WTLibrary.exe > file 2>&1
I'm up and running here. Thanks to all who helped. Here's the current log with the 2010 version wine-git-2010-12-11 msvcr90 patch override to native on iphlpapi.dll override to Builtin on msvcr90.dll override to Builtin on msvcrt.dll mfc90u.dll from dll-files.com (as also iphlpapi.dll)
Now the nappy evidently isn't 100% clean, the program interface is ok. Tested on 32 & 64 bit Slackware-13.1. The latter needs Alien Bob's multilib packages, as they provide 32 bit libs.
http://bugs.winehq.org/show_bug.cgi?id=25512
SSB silas-wine@flatline.org.uk changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |silas-wine@flatline.org.uk
--- Comment #17 from SSB silas-wine@flatline.org.uk 2010-12-19 15:21:25 CST --- If you are having trouble getting Watchtower Library 2010 working on WINE, but you have Watchtower Library 2008, then you could first install Watchtower Library 2008 and then copy the rs_data folder from the Watchtower Library 2010 CD to .wine/drive_c/Program\ Files/Watchtower/Watchtower\ Library\ 2008/E (replacing E with whatever language you're installing). You'll then get the Watchtower Library 2008 program but with the Watchtower Library 2010 data, and it seems to work.
The 2009 program will NOT accept 2010's rs_data, but the 2008 program WILL. I believe 2007, 2006 and 2005 will also accept more recent rs_data versions. So this method can also be used if you're stuck with an old version of WINE (on an old distro on old hardware) which can't run the recent years' programs. For that matter this method can sometimes be used to update the data of broken Windows boxes that for whatever reason won't install a newer version of the program. Always rename (don't delete) the existing rs_data directory, in case its replacement does not work and you need to back off the change.
Of course it would be ideal to fix it so the latest program runs by itself, because we don't know how many minor hidden bugs might be introduced by using a different year's rs_data (and also it means you have to get hold of an old CD, which not everybody has). But the rs_data hack seems to be a good temporary fix.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #18 from business.kid@gmail.com 2010-12-20 03:43:36 CST --- Created an attachment (id=32550) --> (http://bugs.winehq.org/attachment.cgi?id=32550) Output with mcvcr90 patch & no other mods.
It's been pointed out I should have done this quicker - apologies. This patch seems good.
This is from git 2010-1220, with the 0003-Implement-_ui64tow_s.patch applied, & a fresh install of wine & wt library 2010. Changes from defaults: winetricks gecko corefonts msvcr90 set builtin only The usual symlinks to humour XFCE(MEPS\ Platform\ 2.3 symlinked into system32)
msvcr90 no longer gets a mention. It still crashes out, unfortunately. FWIW, the access address (0x7efcd356) is suspiciously high on this box, which has only 1 gig of memory.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #19 from Nat nat.bugzilla@gmail.com 2010-12-22 04:24:11 CST --- Created an attachment (id=32587) --> (http://bugs.winehq.org/attachment.cgi?id=32587) wine-iphlpapi-fix.patch
I got this from a friend.
As the comment in there explains: "MSVC runtime requires DnsSuffix to be valid pointer make it an empty string."
Would be nice have this fix included for wine-1.3.10 and wine.1.2.next.
Thanks!
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #20 from business.kid@gmail.com 2010-12-22 07:55:44 CST --- Created an attachment (id=32589) --> (http://bugs.winehq.org/attachment.cgi?id=32589) ~/wine/drive_c/Program Files/Watchtower/Watchtower Library 2010/E/output
THAT sorts it out! This is a good one!
Built git of 2010-12-22 with both patches on this bug, msvcr90.dll & iphlpapi.dll set as Builtin, packaged it on Slackware-13.1 & fresh ~/.wine and fresh install of WT Library. Not a native dll in sight (except for what came with the wtlib, of course).
This program now loads and runs as it should
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #21 from business.kid@gmail.com 2010-12-22 09:48:39 CST ---
Built git of 2010-12-22 with both patches on this bug, msvcr90.dll & iphlpapi.dll set as Builtin, packaged it on Slackware-13.1 & fresh ~/.wine and fresh install of WT Library. Not a native dll in sight (except for what came with the wtlib, of course).
Same setup _without_ the 0003-Implement-_ui64tow_s.patch (unpatched msvcr90.dll and msvcrt.dll) also works. Same overrides as above. So the iphlpapi.dll patch makes the difference, for this program at least. I am having trouble with reversing the 0003-Implement-_ui64tow_s.patch, so I used dlls from 2010-12-13. HTH.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #22 from Austin English austinenglish@gmail.com 2010-12-22 13:05:26 CST --- (In reply to comment #19)
Created an attachment (id=32587)
--> (http://bugs.winehq.org/attachment.cgi?id=32587) [details]
wine-iphlpapi-fix.patch
I got this from a friend.
As the comment in there explains: "MSVC runtime requires DnsSuffix to be valid pointer make it an empty string."
Would be nice have this fix included for wine-1.3.10 and wine.1.2.next.
Thanks!
Patches should be sent to wine-patches@winehq.org, by the original author. Patches are not picked up from bugzilla.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #23 from Robert Backhaus robbak@robbak.com 2010-12-22 21:29:53 CST --- Created an attachment (id=32594) --> (http://bugs.winehq.org/attachment.cgi?id=32594) archive containg patched iphlpapi.dll.so and iphlpapi.dll.fake
Here are the patched copies of the dll.so file. Installing the iphlpapi.dll.so file to /usr/lib32/wine/ and iphlpapi.dll.fake to /usr/lib32/wine/fakeddlls/iphlpapi.dll got it working for me. On 32 bit systems, use /lib/ instead of /lib32/. I do not know if the updated faked dll is required. This should allow us to have a working program without native dlls and without building from git until the patch is released in stable.
http://bugs.winehq.org/show_bug.cgi?id=25512
Igor Tarasov tarasov.igor@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |tarasov.igor@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=25512
Arno Teigseth arno@teigseth.no changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |arno@teigseth.no
--- Comment #24 from Arno Teigseth arno@teigseth.no 2010-12-23 13:26:09 CST --- (In reply to comment #21)
Same setup _without_ the 0003-Implement-_ui64tow_s.patch (unpatched msvcr90.dll and msvcrt.dll) also works.
Yes, the iphlpapi GetAdaptersAddresses fix is the only fix needed to kill this bug.
http://bugs.winehq.org/show_bug.cgi?id=25512
Nat nat.bugzilla@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |nat.bugzilla@gmail.com
--- Comment #25 from Nat nat.bugzilla@gmail.com 2010-12-25 04:33:42 CST --- (In reply to comment #22)
(In reply to comment #19)
Created an attachment (id=32587)
--> (http://bugs.winehq.org/attachment.cgi?id=32587) [details] [details]
wine-iphlpapi-fix.patch
...
Would be nice have this fix included for wine-1.3.10 and wine.1.2.next.
...
Patches are not picked up from bugzilla.
Seems like someone posted a similar fix to wine-patches the other day, that solves same problem.
http://bugs.winehq.org/show_bug.cgi?id=25512
Dan Kegel dank@kegel.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |dank@kegel.com
--- Comment #26 from Dan Kegel dank@kegel.com 2010-12-26 12:28:05 CST --- http://source.winehq.org/patches/ says http://www.winehq.org/pipermail/wine-patches/2010-December/097143.html needs tests before it can be committed.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #27 from Robert Backhaus robbak@robbak.com 2010-12-26 21:54:02 CST --- (In reply to comment #26)
http://source.winehq.org/patches/ says http://www.winehq.org/pipermail/wine-patches/2010-December/097143.html needs tests before it can be committed.
Patch needs rebasing. Manually applied, the changes do fix the issue.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #28 from Robert Backhaus robbak@robbak.com 2010-12-26 21:54:46 CST --- (In reply to comment #26)
http://source.winehq.org/patches/ says http://www.winehq.org/pipermail/wine-patches/2010-December/097143.html needs tests before it can be committed.
Patch needs rebasing. Manually applied, the changes do fix the issue.
How should we test it, and where should we post the results?
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #29 from Dan Kegel dank@kegel.com 2010-12-26 22:06:48 CST --- He means we need test cases added to the conformance test suite. See http://www.winehq.org/docs/winedev-guide/testing
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #30 from business.kid@gmail.com 2010-12-27 10:57:15 CST --- Created an attachment (id=32638) --> (http://bugs.winehq.org/attachment.cgi?id=32638) iphlpapi patch against git
I applied the patch manually to git, diffed with the a clean git, and attach the output. It runs the Watchtower Library 2010, but I wouldn't rush this upstream. There's an extra 'if' condition in that routine and I would like to see a Wine coder pass his eyes over it first. Better do it once and do it right.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #31 from Arno Teigseth arno@teigseth.no 2010-12-27 12:13:41 CST --- (In reply to comment #30)
... and I would like to see a Wine coder pass his eyes over it first. Better do it once and do it right.
Well if you subscribe to the wine-patches mailing list you can see that Timo's already working on it :)
Search for "Timo" on http://source.winehq.org/patches/
and you will see that the wine testbot says "OK" to those tests.
http://bugs.winehq.org/show_bug.cgi?id=25512
Michael Stefaniuc mstefani@redhat.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |mstefani@redhat.com Target Milestone|--- |1.2.x
http://bugs.winehq.org/show_bug.cgi?id=25512
Francisco infoerudito@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |infoerudito@gmail.com
--- Comment #32 from Francisco infoerudito@gmail.com 2010-12-27 15:30:18 CST --- fixme:exec:SHELL_execute flags ignored: 0x00000100 wine: cannot find L"C:\windows\system32\wtlibrary.exe" ebeldur@ebeldur-desktop:~$ fixme:appbar:handle_appbarmessage SHAppBarMessage(ABM_GETAUTOHIDEBAR, hwnd=(nil), edge=3): stub fixme:appbar:handle_appbarmessage SHAppBarMessage(ABM_GETAUTOHIDEBAR, hwnd=(nil), edge=1): stub fixme:appbar:handle_appbarmessage SHAppBarMessage(ABM_GETAUTOHIDEBAR, hwnd=(nil), edge=0): stub fixme:appbar:handle_appbarmessage SHAppBarMessage(ABM_GETAUTOHIDEBAR, hwnd=(nil), edge=2): stub wine: Unhandled page fault on read access to 0x00000000 at address 0x78556ff2 (thread 001c), starting debugger... err:seh:setup_exception_record nested exception on signal stack in thread 001c eip 7bc74470 esp 7ffdbc7c stack 0x242000-0x340000
My system is Ubuntu 10.04 64bits
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #33 from Arno Teigseth arno@teigseth.no 2010-12-27 20:09:26 CST --- (In reply to comment #32)
My system is Ubuntu 10.04 64bits
is this from "wine start WTLibrary.exe" OUTSIDE of .wine/drive_c/Program Files/Watchtower/Watchtower Library 2010/S/ ?
what is your wine version? (wine --version)
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #34 from Robert Backhaus robbak@robbak.com 2010-12-27 21:53:59 CST --- iphlpapi patch has been committed ( http://source.winehq.org/patches/data/69753 )
Current git is now working. Thanks to all!
http://bugs.winehq.org/show_bug.cgi?id=25512
Dan Kegel dank@kegel.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED
--- Comment #35 from Dan Kegel dank@kegel.com 2010-12-27 22:06:42 CST --- Reported fixed.
http://bugs.winehq.org/show_bug.cgi?id=25512
--- Comment #36 from Jacek Caban jacek@codeweavers.com 2011-01-03 08:16:35 CST --- commit bfc8832ccb633ea1e8dbdcd31c24d7fbae3c5315
http://bugs.winehq.org/show_bug.cgi?id=25512
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #37 from Alexandre Julliard julliard@winehq.org 2011-01-07 12:39:43 CST --- Closing bugs fixed in 1.3.11.
http://bugs.winehq.org/show_bug.cgi?id=25512
Simon Tretter simon@tretter.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |simon@tretter.net
http://bugs.winehq.org/show_bug.cgi?id=25512
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Target Milestone|1.2.x |---
https://bugs.winehq.org/show_bug.cgi?id=25512
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Fixed by SHA1| |bfc8832ccb633ea1e8dbdcd31c2 | |4d7fbae3c5315 CC| |focht@gmx.net Component|-unknown |iphlpapi