[Bug 40334] New: AION - Wine /Unhandled exception: page fault on read access to 0x00000000 in 64-bit code (0x0000000000000000).
https://bugs.winehq.org/show_bug.cgi?id=40334
Bug ID: 40334 Summary: AION - Wine /Unhandled exception: page fault on read access to 0x00000000 in 64-bit code (0x0000000000000000). Product: Wine Version: 1.7.55 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs@winehq.org Reporter: ryancasanovalv@gmail.com Distribution: ---
Created attachment 54000 --> https://bugs.winehq.org/attachment.cgi?id=54000 Got this error message after installing AION on wine/ubuntu 15.10
Unhandled exception: page fault on read access to 0x00000000 in 64-bit code (0x0000000000000000).
https://bugs.winehq.org/show_bug.cgi?id=40334
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Version|1.7.55 |1.7.36
--- Comment #1 from Austin English austinenglish@gmail.com --- Please retest in 1.8.x/1.9.x, 1.7 series is unsupported.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #2 from Ryan ryancasanovalv@gmail.com --- Created attachment 54006 --> https://bugs.winehq.org/attachment.cgi?id=54006 Attachment2
Same thing with latest version.
https://bugs.winehq.org/show_bug.cgi?id=40334
fjfrackiewicz@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |fjfrackiewicz@gmail.com
--- Comment #3 from fjfrackiewicz@gmail.com --- (In reply to Ryan from comment #2)
Created attachment 54006 [details] Attachment2 [details]
Same thing with latest version.
Have you tried running this application in Windows XP mode or does it not support anything lower than Windows 7?
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #4 from Ryan ryancasanovalv@gmail.com --- Created attachment 54020 --> https://bugs.winehq.org/attachment.cgi?id=54020 Same thing with Win XP.
https://bugs.winehq.org/show_bug.cgi?id=40334
LeonMaxx leonmaxx@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |leonmaxx@gmail.com
--- Comment #5 from LeonMaxx leonmaxx@gmail.com --- Created attachment 63884 --> https://bugs.winehq.org/attachment.cgi?id=63884 aion64_crash_log.txt
Still an issue. Tested with vanilla Wine 4.3. Will try to get some debug information.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #6 from LeonMaxx leonmaxx@gmail.com --- Sorry, attached wrong log. After testing with vanilla wine I tested with Proton/Wine 3.16 and saved crash log from this launch. But logs seems to be identical.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #7 from LeonMaxx leonmaxx@gmail.com --- Created attachment 63889 --> https://bugs.winehq.org/attachment.cgi?id=63889 aion64_winedebug_log.txt.gz
This is log from starting 64-bit Aion client on freshly built Wine 4.4 with "WINEDEBUG=+all".
Command line: wine explorer /desktop=AION,1920x1080 ./bin64/aion.bin -ip:79.110.83.80 -noweb -noauthgg -st -charnamemenu -ingamebrowser -webshopevent:6 -f2p -lbox -litelauncher -64 -ncping -nosatab -aiontv -nobs -60f2p -n20 /SessKey: /CompanyID:11 /ChannelGroupIndex:-1 -lang:ENG -litestep:9
Client download link: https://en.aion.gameforge.com/website/download/
https://bugs.winehq.org/show_bug.cgi?id=40334
Louis Lenders xerox.xerox2000x@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |xerox.xerox2000x@gmail.com
--- Comment #8 from Louis Lenders xerox.xerox2000x@gmail.com --- Maybe it chokes because of missing registry keys for VideoBiosVersion and SystemBiosVersion
See below.
Could you try add them (look in windows how they look like), does it help anything?
advapi32.RegOpenKeyA(80000002,1403dd7d2 "Hardware\description\System",1403dd811) ret=1403e20a3
19454.454:002b:002c:Call advapi32.RegQueryValueExA(000000a8,1403dd7ee "SystemBiosVersion",00000000,00000000,1403dd819,1403dda19) ret=1403e21d3 19454.454:002b:002c:trace:reg:RegQueryValueExA (0xa8,"SystemBiosVersion",(nil),(nil),0x1403dd819,0x1403dda19=512) 19454.454:002b:002c:trace:heap:RtlAllocateHeap (0x10000,70000062,00000024): returning 0x760e0 19454.454:002b:002c:trace:reg:NtQueryValueKey (0xa8,L"SystemBiosVersion",2,0x23fa90,256) 002c: get_key_value( hkey=00a8, name=L"SystemBiosVersion" ) 002c: get_key_value() = OBJECT_NAME_NOT_FOUND { type=-1, total=0, data={} } 19454.454:000c:000d:trace:heap:RtlFreeHeap (0x10000,70000062,0x107e0): returning TRUE 19454.454:000c:000d:Ret advapi32.RegQueryValueExA() retval=00000000 ret=7fddc150a076 19454.454:000c:000d:Call advapi32.RegCloseKey(00000080) ret=7fddc150a083 000d: close_handle( handle=0080 ) 000d: close_handle() = 0 19454.455:000c:000d:Ret advapi32.RegCloseKey() retval=00000000 ret=7fddc150a083 19454.455:000c:000d:trace:shell:HCR_GetClassNameA -- (Query file associations) 19454.455:000c:000d:trace:shell:DllGetClassObject CLSID:{a07034fd-6caa-4954-ac3f-97a27216f98a} (Query file associations),IID:{00000001-0000-0000-c000-000000000046} (IID_ICla ssFactory) 19454.455:000c:000d:trace:shell:DllGetClassObject index[10] 19454.455:000c:000d:Call ntdll.RtlAllocateHeap(00010000,00000000,00000030) ret=7fddc15524fd 19454.455:002b:002c:trace:heap:RtlFreeHeap (0x10000,70000062,0x760e0): returning TRUE 19454.455:002b:002c:Ret advapi32.RegQueryValueExA() retval=00000002 ret=1403e21d3
19454.455:002b:002c:Call advapi32.RegQueryValueExA(000000a8,1403dd800 "VideoBiosVersion",00000000,00000000,1403dd819,1403dda19) ret=1403e22ef 19454.455:002b:002c:trace:reg:RegQueryValueExA (0xa8,"VideoBiosVersion",(nil),(nil),0x1403dd819,0x1403dda19=512) 19454.455:002b:002c:trace:heap:RtlAllocateHeap (0x10000,70000062,00000022): returning 0x76130 19454.455:002b:002c:trace:reg:NtQueryValueKey (0xa8,L"VideoBiosVersion",2,0x23fa90,256) 002c: get_key_value( hkey=00a8, name=L"VideoBiosVersion" ) 002c: get_key_value() = OBJECT_NAME_NOT_FOUND { type=-1, total=0, data={} } 19454.455:002b:002c:trace:heap:RtlFreeHeap (0x10000,70000062,0x76130): returning TRUE 19454.455:002b:002c:Ret advapi32.RegQueryValueExA() retval=00000002 ret=1403e22ef
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #9 from Louis Lenders xerox.xerox2000x@gmail.com --- Created attachment 63891 --> https://bugs.winehq.org/attachment.cgi?id=63891 reg keys
Here`s how they look like in my virtualbox
You could try ``wine regedit vidbios.reg``
Maybe it helps, maybe not. But worth a try i`d say
https://bugs.winehq.org/show_bug.cgi?id=40334
Louis Lenders xerox.xerox2000x@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #63891|0 |1 is obsolete| |
--- Comment #10 from Louis Lenders xerox.xerox2000x@gmail.com --- Created attachment 63893 --> https://bugs.winehq.org/attachment.cgi?id=63893 reg keys
This is a regfile with only the two keys.
Could you try ``regedit vidb.reg``, see if it changes anything?
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #11 from LeonMaxx leonmaxx@gmail.com --- Thanks for suggestion. Unfortunately nothing changed.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #12 from Louis Lenders xerox.xerox2000x@gmail.com --- Created attachment 63900 --> https://bugs.winehq.org/attachment.cgi?id=63900 part of +relay,+seh log
Attached (hopefully relevant part of) +relay log.
I can confirm this bug (the missing registrykeys appear red herring, i created them in wineboot but nothing changed)
I haven`t got a clue why it doesn`t start, tried a few things, no luck...
Note: for me, rather then crashing, it end quickly with 0030:err:seh:setup_exception stack overflow 3584 bytes in thread 0030 eip 000000007bcba95c esp 0000000000150810 stack 0x150000-0x152000-0x250000
https://bugs.winehq.org/show_bug.cgi?id=40334
Louis Lenders xerox.xerox2000x@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Ever confirmed|0 |1 Keywords| |download URL| |https://en.aion.gameforge.c | |om/website/download/ Status|UNCONFIRMED |NEW
--- Comment #13 from Louis Lenders xerox.xerox2000x@gmail.com --- confirming sha1sum 20180103_AION_Setup.exe 138133c8eba9ff6cb805d8bda372919653036f97 20180103_AION_Setup.exe
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #14 from LeonMaxx leonmaxx@gmail.com --- Louis, thanks for your attention to this bug report.
Things that I forgot to mention: 1. 32-bit client do work, but with bad performance when there is many activity on screen e.g. many characters in main city location, etc. (even with ESYNC and Gallium-nine enabled).
2. 32-bit client is slow on Windows too, that is why I wanted to try 64-bit version with hope for better performance.
3. 64-bit client on Windows have much better performance than 32-bit client in activity heavy scenes.
Command line to run 32-bit client: wine explorer /desktop=AION,1920x1080 ./bin32/aion.bin -ip:79.110.83.80 -noweb -noauthgg -st -charnamemenu -ingamebrowser -webshopevent:6 -f2p -lbox -litelauncher -ncping -nosatab -aiontv -nobs -60f2p -n20 /SessKey: /CompanyID:11 /ChannelGroupIndex:-1 -lang:ENG -litestep:9
https://bugs.winehq.org/show_bug.cgi?id=40334
Fred dracoanarion@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |dracoanarion@gmail.com
--- Comment #15 from Fred dracoanarion@gmail.com --- Is there anyway we could help to get this bug solved ? It still valid for wine-staging 4.19.
https://bugs.winehq.org/show_bug.cgi?id=40334
Paul Gofman gofmanp@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |gofmanp@gmail.com
--- Comment #16 from Paul Gofman gofmanp@gmail.com --- Created attachment 65650 --> https://bugs.winehq.org/attachment.cgi?id=65650 Export missing function from ntdll
I've tested the game.
64 bit version wants an undocumented function in ntdll which is likely very similar to NtQuerySystemInformation(). The patch is fixing that. No logs could help here, the application seems to get some API functions by directly reading import tables from loaded dll images.
Then there is the crash after loading the data. I will put a proof of concept hack which fixes that in the next comment, though it definitely deserves a separate bug report and is likely going to be a long standing issue.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #17 from Paul Gofman gofmanp@gmail.com --- Created attachment 65651 --> https://bugs.winehq.org/attachment.cgi?id=65651 hack
After fixing the crash at the very start (see Comment #16) the game loads it data for a while and then crashes again during the transition screen before showing login screen.
The crash happens in CrySystem.dll in CryFree(). It does not like really high memory pointers, it probably indexes some part of them in some table and that exceeds the bounds of the table. It looks like it expects the memory returned by HeapAlloc to have limited address range. I am attaching the hack which allows the game to run (by providing zero bits mask in heap virtual memory allocations). To allow this work well I also included the patch by Rémi Bernon which is currently hanging in development patches list waiting for review. Anyway, this patch is not something I would recommend to include in any more or less universal build.
The game relies on the addresses allocated in (relatively) low space and it is not guaranteed on Windows (though probably usually the case unless the lower space is exhausted). I checked that there are no special suspicious flags on heaps creation used.
I found an open source for Cry system on github. I looked at the code very briefly but it looks like the present version does not have this bug, so maybe the game is using some old version and the bug might go away as soon as the game updates Cry system.
Besides this problem and the one related to this bug (Comment #16), there is also a crash in wined3d right after logging in to server. It is the same for 32 and 64 bit version. I hope this one will be fixed rather sooner than later.
With all these issues bypassed, the game starts and runs seemingly flawless (I created a character, walked quite a bit around the initial area and even managed to kill some big fly near the gates).
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #18 from LeonMaxx leonmaxx@gmail.com --- Thank you Paul for debugging this issue. I always thought that early versions of CryEngine is terrible.
What debugging tools do you use?
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #19 from Louis Lenders xerox.xerox2000x@gmail.com --- Created attachment 65695 --> https://bugs.winehq.org/attachment.cgi?id=65695 patch to work around bug 46568
Hi Paul,
After fixing the crash at the very start (see Comment #16) the game loads it >data for a while and then crashes again during the transition screen before showing >login screen.
The crash happens in CrySystem.dll in CryFree(). It does not like really high >memory pointers, it probably indexes some part of them in some table and that >exceeds the bounds of the table.
I gave the game a try with your first patch (add RtlGetNativeSystemInfo) and i can get to the loginscreen with another patch i apply always to work around bug 46568; So might be same issue. Patches are from Derek Lesho (see https://bugs.winehq.org/show_bug.cgi?id=46568). Does this patch help for you too?
Unfortenately i`ve not been able to login, i get a message the server is not reachable :(
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #20 from Paul Gofman gofmanp@gmail.com --- (In reply to Louis Lenders from comment #19)
Created attachment 65695 [details] patch to work around bug 46568
Patches are from Derek Lesho (see https://bugs.winehq.org/show_bug.cgi?id=46568). Does this patch help for you too?
Those patches are indeed about somewhat similar issues, though native (old) msxml and Crysystem32.dll may have different assumptions about the address range, I am not sure. FWIW I think both Derek's and my patch are absolutely no go for upstream, at least in their current form. I don't know, maybe something more limited in scope or based on some configuration setting maybe worked out.
Unfortenately i`ve not been able to login, i get a message the server is not reachable :(
That's strange, I had absolutely no issues with login. Here is my string for running aion.bin with the current directory being game install, C:\aion in my case (I got that by running from launcher first): wine 'C:\aion\bin64\aion.bin' -ip:79.110.83.80 -noweb -noauthgg -st -charnamemenu -ingamebrowser -webshopevent:6 -f2p -lbox -litelauncher -64 -ncping -nosatab -aiontv -nobs -60f2p -n20 /SessKey: /CompanyID:11 /ChannelGroupIndex:-1 -lang:eng -litestep:9
If that doesn't work, maybe you can start that from launcher (it works for me also), which might also download some updates if neeeded.
FWIW, after the successful login it will crash unless you apply one of:
1. 2 patches: https://source.winehq.org/patches/data/173737 https://source.winehq.org/patches/data/173738
or
2. A patch attached here: https://www.winehq.org/pipermail/wine-devel/2019-November/154702.html
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #21 from Louis Lenders xerox.xerox2000x@gmail.com --- (In reply to Paul Gofman from comment #20)
I don't know, maybe something more limited in scope or based on some >configuration setting maybe worked out.
That would be nice if something along these lines could be included for example in Staging (maybe set some env. variable to trigger the patches like Staging did already for few other "features")
Just for the info, i apply patches from msxml6-bug since few weeks standard and haven`t seen any strange crashes / errors that i know of
That's strange, I had absolutely no issues with login.
Maybe something is wrong at my side; It says either ID and password do not match (with your long options added), or authorizarion server is unreachable.
I see in console:
0091:err:winediag:IcmpCreateFile Failed to use ICMP (network ping), this requires special permissions.
Any idea how to configure to get rid of this error?
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #22 from Fred dracoanarion@gmail.com --- Thanks for your work on this Paul and Louis, though I'm afraid I don't get it working, and I still don't know why.
I wonder if that can be due to my setup, the app get stuck quite soon with the following error logs:
002a:err:ntdll:RtlpWaitForCriticalSection section 0x7becbb20 "loader.c: loader_section" wait timed out in thread 002a, blocked by 0009, retrying (60 sec) 002b:err:ntdll:RtlpWaitForCriticalSection section 0x7becbb20 "loader.c: loader_section" wait timed out in thread 002b, blocked by 0009, retrying (60 sec)
(repeating for for different threads)
I wonder if that can be due to my wine build somehow, though I tried several ways with the same result: following the wiki instructions, using the ubuntu official debian build files, or using the winehq debian build files. Same result.
I used the current head as a base (190fb605c3 "quartz/tests: Fix some tests on w7u." for my last trial), plus the 2 following patches: ntdll: Export RtlGetNativeSystemInformation() function. ntdll: Fix find_free_area outside of reserved areas with zero_bits != 0. (v2)
I also tried with 4.20 tag as a base. Same result.
Please note that I tried my own build without the patches to double check that i could reproduce the original issue.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #23 from LeonMaxx leonmaxx@gmail.com --- (In reply to Fred from comment #22)
002a:err:ntdll:RtlpWaitForCriticalSection section 0x7becbb20 "loader.c: loader_section" wait timed out in thread 002a, blocked by 0009, retrying (60 sec) 002b:err:ntdll:RtlpWaitForCriticalSection section 0x7becbb20 "loader.c: loader_section" wait timed out in thread 002b, blocked by 0009, retrying (60 sec)
Same errors in my build too.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #24 from Louis Lenders xerox.xerox2000x@gmail.com --- (In reply to LeonMaxx from comment #23)
(In reply to Fred from comment #22)
002a:err:ntdll:RtlpWaitForCriticalSection section 0x7becbb20 "loader.c: loader_section" wait timed out in thread 002a, blocked by 0009, retrying (60 sec) 002b:err:ntdll:RtlpWaitForCriticalSection section 0x7becbb20 "loader.c: loader_section" wait timed out in thread 002b, blocked by 0009, retrying (60 sec)
Same errors in my build too.
Is that after logging in or just after trying to start the program?
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #25 from LeonMaxx leonmaxx@gmail.com --- (In reply to Louis Lenders from comment #24)
Is that after logging in or just after trying to start the program?
After start, even before windows is created.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #26 from Louis Lenders xerox.xerox2000x@gmail.com --- (In reply to LeonMaxx from comment #25)
(In reply to Louis Lenders from comment #24)
Is that after logging in or just after trying to start the program?
After start, even before windows is created.
Did you give this one a try as well? https://bugs.winehq.org/attachment.cgi?id=65695&action=diff
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #27 from LeonMaxx leonmaxx@gmail.com --- (In reply to Louis Lenders from comment #26)
Did you give this one a try as well? https://bugs.winehq.org/attachment.cgi?id=65695&action=diff
Sorry for not mentioning, I used exactly this patch.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #28 from Louis Lenders xerox.xerox2000x@gmail.com --- (In reply to LeonMaxx from comment #27)
(In reply to Louis Lenders from comment #26)
Did you give this one a try as well? https://bugs.winehq.org/attachment.cgi?id=65695&action=diff
Sorry for not mentioning, I used exactly this patch.
And with Paul`s attached patches , does it work then?
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #29 from Paul Gofman gofmanp@gmail.com --- (In reply to Fred from comment #22)
ntdll: Export RtlGetNativeSystemInformation() function. ntdll: Fix find_free_area outside of reserved areas with zero_bits != 0. (v2)
Which one exactly is the second one and how do you decide that it is sufficient? If this is that you got from wine-devel patches, it of course does not include the hack, the game will crash before you see the login screen.
Can you try exactly with the patches attached to comments #16 and #17 and not anything else (neither in addition nor instead)? On top of the current Wine git.
The additional patches referenced in Comment #20 are not needed now, the fix for that issue is there in the current git.
If you still can't start it, can you please attach the full output (not some selected or last lines), or even better with WINEDEBUG=+seh,+loaddll?
Just to note, it still works fine here with the referenced patches applied on top of current git. How exactly do you start the game? Do you have the latest version of the game client updated?
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #30 from Paul Gofman gofmanp@gmail.com --- (In reply to Louis Lenders from comment #26)
(In reply to LeonMaxx from comment #25)
(In reply to Louis Lenders from comment #24)
Is that after logging in or just after trying to start the program?
After start, even before windows is created.
Did you give this one a try as well? https://bugs.winehq.org/attachment.cgi?id=65695&action=diff
I wonder how some seh stack handlers search hack which is in the patch referenced here is related to the high memory range problem in this game? Stacking unrelated app specific hacks together is not necessarily supposed to improve things, I would strongly advise against stacking unrelated hacks together :)
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #31 from Fred dracoanarion@gmail.com --- Created attachment 65785 --> https://bugs.winehq.org/attachment.cgi?id=65785 stuck logs after patch from #16 + #17
(In reply to Paul Gofman from comment #29)
(In reply to Fred from comment #22)
ntdll: Export RtlGetNativeSystemInformation() function. ntdll: Fix find_free_area outside of reserved areas with zero_bits != 0. (v2)
Which one exactly is the second one and how do you decide that it is sufficient? If this is that you got from wine-devel patches, it of course does not include the hack, the game will crash before you see the login screen.
Can you try exactly with the patches attached to comments #16 and #17 and not anything else (neither in addition nor instead)? On top of the current Wine git.
Don't get me wrong there, I wanted to have the status described in comment #16. Not a working game. But I'm not even reaching this state. The unhandled fault is passed through (likely thanks to the patch from comment #16), but I'm not getting to the point that the game starts loading (not even getting to the splash screen).
The attached log is 5.2Gb big uncompressed. Likely an infinite loop that I had to stop. The code base is head (190fb605c3 (origin/master, origin/HEAD, master) quartz/tests: Fix some tests on w7u.) + comment #16 and #17 patches (exactly).
I'm using the following to start the game: /mnt/Data/Jeux/.wine64_aion/drive_c$ env WINEPREFIX="/mnt/Data/Jeux/.wine64_aion" /opt/wine-aion/bin/wine64 'C:\AION\bin64\AION.bin' -ip:79.110.83.80 -noweb -noauthgg -st -charnamemenu -ingamebrowser -webshopevent:6 -f2p -lbox -litelauncher -64 -ncping -nosatab -aiontv -nobs -60f2p -n20 /SessKey: /CompanyID:11 /ChannelGroupIndex:-1 -lang:fra -litestep:9
The game is up to date, I asked the launcher to run a check to confirm it. Side note: the launcher doesn't start the 64bit version by default on my setup, which is rather unexpected from my side. Again, I'm not quite sure to have a valid environment there :(
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #32 from Paul Gofman gofmanp@gmail.com --- (In reply to Fred from comment #31)
The attached log is 5.2Gb big uncompressed. Likely an infinite loop that I had to stop. The code base is head (190fb605c3 (origin/master, origin/HEAD, master) quartz/tests: Fix some tests on w7u.) + comment #16 and #17 patches (exactly).
Thanks for the logs. I've got some guess of the possible reason for your problem here. Unless I am missing something, it seems like you are building all binaries as .so, not as PE (e. g., you have kernelbase.dll.so in your built / installed Wine binaries), right? I would start from making a default Wine build which is using mingw and produces PE binaries (.dll files) for the most of the Wine libs. I didn't yet bother to make a non-PE build to verify that I will reproduce your crash (yes, it is the crash, just looping through the same handler), but it looks quite possible that the might be here.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #33 from Paul Gofman gofmanp@gmail.com --- I've checked with --without-mingw build, and I could reproduce exactly the same crash. So I hope as soon as you mange to make a normal build with mingw, this problem will be solved.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #34 from Fred dracoanarion@gmail.com --- Created attachment 65787 --> https://bugs.winehq.org/attachment.cgi?id=65787 Login screenshot
Thanks Paul, I didn't have the config option, but I was missing the mingw-gcc on my 64bit system, and config was "silently" disabling it. I got build-dep this time, seems I missed this step in the 64bits env.
With this fixed, I could reach the login !
Now next problem, the graphisms are broken. It's unplayable :D.
Yet another problem in my configuration ?
https://bugs.winehq.org/show_bug.cgi?id=40334
Fred dracoanarion@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #65785|0 |1 is obsolete| |
--- Comment #35 from Fred dracoanarion@gmail.com --- Created attachment 65788 --> https://bugs.winehq.org/attachment.cgi?id=65788 Login texture issue (+seh,+loaddll) logs
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #36 from Paul Gofman gofmanp@gmail.com --- (In reply to Fred from comment #34)
Now next problem, the graphisms are broken. It's unplayable :D.
‘winetricks d3dcompiler_47’, maybe use Wine virtual desktop. I think it should’ve been the same with 32 bit version. Also, if you alredy run it without native d3dcompiler and get to login screen without long loading screen, it may keep some glitches, you should force it to reliad the data. I always see it reloading data after it did not terminate normally (e.g stop it by control+c from terminal).
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #37 from Fred dracoanarion@gmail.com --- I do have a couple of winetricks to get Aion working on wine-stable, but not this one.
With d3dcompiler, I get the texture back (with some glitches even with data load), but I'm not passing the character selection anymore.
I believe I didn't need any winetricks on staging 4.17 with a 32bit prefix, though I'm not able to reproduce this anymore with my builds (keep trying).
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #38 from Paul Gofman gofmanp@gmail.com --- (In reply to Fred from comment #37)
I do have a couple of winetricks to get Aion working on wine-stable, but not this one.
With d3dcompiler, I get the texture back (with some glitches even with data load), but I'm not passing the character selection anymore.
Could you please specify "not passing"? If it crashes, could you please attach new "WINEDEBUG=+loaddll,+seh" log?
Maybe also try with native d3dx9_38.dll, d3dx9_31.dll. I am not sure I needed that though.
I believe I didn't need any winetricks on staging 4.17 with a 32bit prefix, though I'm not able to reproduce this anymore with my builds (keep trying).
Not possible without d3dcompiler or d3dx9 native override, unless the game client changed significantly between those tries. But most likely you had some overrides. It wants to compiler some shaders, that is not implemented in builtin Wine dlls.
https://bugs.winehq.org/show_bug.cgi?id=40334
Fred dracoanarion@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #65787|0 |1 is obsolete| | Attachment #65788|0 |1 is obsolete| |
--- Comment #39 from Fred dracoanarion@gmail.com --- Created attachment 65790 --> https://bugs.winehq.org/attachment.cgi?id=65790 Stuck after character selection WINEDEBUG=+loaddll,+seh
I don't know if that can be qualified as a crash. The game stays blocked either just after pressing the "Ok" button, or in the loading splash screen after selecting the character.
I added both winetricks. If that may help, I can also provide a log of my 32 bit setup (with wine-stable), that includes some few winetricks.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #40 from Paul Gofman gofmanp@gmail.com --- Created attachment 65793 --> https://bugs.winehq.org/attachment.cgi?id=65793 hack
Absolutely the same crash as I observed initially and which I worked around by the hack in comment #17.
1. Can you please double check that you have the hack from comment #17 applied (exactly that, which consists of 2 patches inside), and only that plus the one from Comment #16? And no other patches / "improvements"?
2. If you had the hack from comment #17, but it is crashing with it, can you please try to replace the hack from Comment #17 with the one attached here? It is mostly the same but limits the allocation range more.
3. If it is still crashing, can you please attach WINEDEBUG=+seh,+loaddll,+virtual log? And if not crashing right away, but does it later during gameplay, the same log recorded from start might be also interesting.
https://bugs.winehq.org/show_bug.cgi?id=40334
Fred dracoanarion@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #65790|0 |1 is obsolete| |
--- Comment #41 from Fred dracoanarion@gmail.com --- Created attachment 65796 --> https://bugs.winehq.org/attachment.cgi?id=65796 Crash after character selection WINEDEBUG=+seh,+loaddll,+virtual log
Ok, looks I did mess my build between comment #34 and comment #37 (I had the patch of Leon instead of your hack). Though the crash is still here, just happens few seconds later. I tried with comment #16 and comment #17 patches. The logs here are with comment #16 and comment #40 patches.
I also recreated my prefix, and now have the following wine tricks (coming from my original aion32 prefix): d3dx9_26 d3dx9_28 d3dx9_31 d3dx9_35 d3dx9_36 d3dx9_39 d3dx9_42 d3dx9_43 d3dx9
I don't have glitches anymore.
Crash occurred while moving the camera with other (relatively lot of) players in the area. The crash didn't occur with a secondary character (with no other player around), though I didn't try to play more than a couple of minutes.
https://bugs.winehq.org/show_bug.cgi?id=40334
Paul Gofman gofmanp@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #65651|0 |1 is obsolete| | Attachment #65793|0 |1 is obsolete| |
--- Comment #42 from Paul Gofman gofmanp@gmail.com --- Created attachment 65797 --> https://bugs.winehq.org/attachment.cgi?id=65797 hack
(In reply to Fred from comment #41)
I don't have glitches anymore.
Crash occurred while moving the camera with other (relatively lot of) players in the area. The crash didn't occur with a secondary character (with no other player around), though I didn't try to play more than a couple of minutes.
Thanks, now it is interesting and looks like my hack is missing something rather than there are some build / setup problems remaining. Could you please try that with the new version attached here (i. e., the one from Comment #16 plus this one)? I would hope it will avoid similar crashes / hangs now. If it crashes / hangs again, the same log with the new patch would be useful.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #43 from Fred dracoanarion@gmail.com --- Created attachment 65802 --> https://bugs.winehq.org/attachment.cgi?id=65802 Crash during play WINEDEBUG=+seh,+loaddll,+virtual log
I'm sorry, it seems it's not yet fixed :)
I put 2 logs in the tarball. One I could play for 2 or 3 minutes, though that's a bit early here to have a lot of players around. Then the application detected a problem, wine complained about a wrong DLL installed (never seen that before) and license issue, and then the app crashed. I named it wine_aion_crash_wrong_dll.log
On the other log, I could only play few seconds, moved the camera a bit fast, and got stuck. I killed the app using wineserver -k.
https://bugs.winehq.org/show_bug.cgi?id=40334
Paul Gofman gofmanp@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #65797|0 |1 is obsolete| |
--- Comment #44 from Paul Gofman gofmanp@gmail.com --- Created attachment 65803 --> https://bugs.winehq.org/attachment.cgi?id=65803 Force down top allocation order
(In reply to Fred from comment #43)
Created attachment 65802 [details] Crash during play WINEDEBUG=+seh,+loaddll,+virtual log
I'm sorry, it seems it's not yet fixed :)
Thanks for testing this, it is really helpful as I cannot reproduce the crashes, no matter how quick I rotate the camera.
These crashes are initially all about the same, the strange message is just the consequence of trying to generate a crash report by the game.
I am attaching a broader fix, could you please try with this one? This one is something I see as a bit less of a hack already, not ready for inclusion anywhere, but maybe somewhere in the direction how it could be done in principle.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #45 from Paul Gofman gofmanp@gmail.com --- Also, it still looks a bit strange how quick you can get this crash even with the earlier versions of the patch. Just in case, can you please double check that you appling the latest version of the patch and rebuilding the affected parts correctly? What you actually need to rebuild and install is 64 bit ntdll (dlls/ntdll) after applying the latest patch. I use to it by entering dlls/ntdll in build dir, doing 'make install' and later double check that I the newly build dll is used:
lsof -p <pid> | grep ntdll.dll (where <pid> is process id of aion.bin).
Unfortunately it is a bit hard to guess from the logs with the previous patches if the patch is actually there or not. I should've added smth for that, will do for the next attempt if the problem is not solved yet.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #46 from Fred dracoanarion@gmail.com --- This one looks to be a bit more stable. I'll test it deeper later today.
Considering your last comment, I'm afraid I added one little patch on top of your latest, I hope you won't mind too much (and that it will not affect the behavior too much neither):
diff --git a/dlls/ntdll/virtual.c b/dlls/ntdll/virtual.c index 4af11b0a5e..18896ee211 100644 --- a/dlls/ntdll/virtual.c +++ b/dlls/ntdll/virtual.c @@ -1202,6 +1202,7 @@ static NTSTATUS map_view( struct file_view **view_ret, void *base, size_t size, goto done; }
+ FIXME("Fred to Paul: Forcing top down allocation order\n"); if (!(ptr = find_free_area( address_space_start, alloc.limit, size, mask, top_down, TRUE, VIRTUAL_GetUnixProt(vprot) )))
I also added a similar one to my previous build to verify that the changes were really in. My build environment seems now to be working as expected (or it fakes it pretty well !).
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #47 from Paul Gofman gofmanp@gmail.com --- (In reply to Fred from comment #46)
FIXME("Fred to Paul: Forcing top down allocation order\n"); if (!(ptr = find_free_area( address_space_start, alloc.limit, size, mask, top_down, TRUE,VIRTUAL_GetUnixProt(vprot) )))
Great idea.
This one looks to be a bit more stable. I'll test it deeper later today.
So with this last one it did not crash for you yet? Looking forward for your deeper test results. Thanks for testing this. If it will crashes again eventually, it is good to have the same log as before, so probably it worth running the game with the same debug flags for now. The log may grow bigger, but I hope it won't become impossibly big since it is compressed very well.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #48 from Fred dracoanarion@gmail.com --- So far good news. I'm playing for about 4h now, without a crash ! And the 64bit version of Aion is keeping its promises, way better fps in crowded environment, and I've got no more issues with water :D
I'll keep playing with it for a while, but that's really nice so far !
The only drawback I noticed is that I can't get the window switching (alt+tab) to work without virtual desktop, but that's minor and may not be related at all.
Thanks a lot Paul for your work. If you want me to test another version of this for a release, I'm here.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #49 from Paul Gofman gofmanp@gmail.com --- (In reply to Fred from comment #48)
So far good news. I'm playing for about 4h now, without a crash ! And the 64bit version of Aion is keeping its promises, way better fps in crowded environment, and I've got no more issues with water :D
Great news!
Let's keep this bug about the original issue now (the crash due to missing function), we will need to split the high memory addresses issue off to another bug so it can be referenced if we will be adding something like this fix to Staging.
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #50 from LeonMaxx leonmaxx@gmail.com --- When you'll create new bug for high memory, please post a link here, for me to track progress too. Thanks.
https://bugs.winehq.org/show_bug.cgi?id=40334
Paul Gofman gofmanp@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |48175
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #51 from Paul Gofman gofmanp@gmail.com --- I've created a separate Bug #48175 for high memory issue. This bug stays about initial crash which is due to missing function (see Comment #16).
https://bugs.winehq.org/show_bug.cgi?id=40334
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|AION - Wine /Unhandled |Multiple games and |exception: page fault on |applications crash on |read access to 0x00000000 |unimplemented function |in 64-bit code |ntdll.dll.RtlGetNativeSyste |(0x0000000000000000). |mInformation (AION, AVG | |Free Edition 2012-2014, | |TuneUp Utilities 2014) Component|-unknown |ntdll CC| |focht@gmx.net
--- Comment #52 from Anastasius Focht focht@gmx.net --- Hello folks,
refining summary to collect more apps/games that need this native API.
While revisiting bug 30499 ("Multiple Avira AVG product installers crash due to access of undocumented PEB field "UnicodeCaseTableData" (AVG Free Edition 2012-2014, TuneUp Utilities 2014)") it turns out more apps need 'ntdll.RtlGetNativeSystemInformation'.
--- snip --- $ pwd /home/focht/.wine/drive_c/Program Files (x86)/TuneUp Utilities 2014
$ wine ./SystemControl.exe 0021:fixme:win:RegisterPowerSettingNotification (0x9cc60,{31f9f286-5084-42fe-b720-2b0264993763},1): stub ... wine: Call from 0x7bc7006c to unimplemented function ntdll.dll.RtlGetNativeSystemInformation, aborting --- snip ---
More references/imports of that native API in WINEPREFIX:
--- snip --- $ grep -ralZ RtlGetNativeSystemInformation .wine/drive_c | xargs -r0i bash -c "echo "{}" && winedump -j import "{}" | grep RtlGetNativeSystemInformation"
.wine/drive_c/windows/Installer/ae9b.msi .wine/drive_c/Program Files (x86)/TuneUp Utilities 2014/tuavgx.dll .wine/drive_c/Program Files (x86)/TuneUp Utilities 2014/avgrepliba.dll 0007d298 647 RtlGetNativeSystemInformation .wine/drive_c/Program Files (x86)/TuneUp Utilities 2014/avgdumpa.exe 000a9360 647 RtlGetNativeSystemInformation .wine/drive_c/Program Files (x86)/TuneUp Utilities 2014/avgreplibx.dll .wine/drive_c/Program Files (x86)/TuneUp Utilities 2014/avgdiagex.exe .wine/drive_c/Program Files (x86)/TuneUp Utilities 2014/tuavga.dll 00043348 647 RtlGetNativeSystemInformation .wine/drive_c/Program Files (x86)/TuneUp Utilities 2014/avgdumpx.exe .wine/drive_c/ProgramData/{FE8D473A-6F06-4F99-B5F4-BED72B2A038C}/{D3742F82-1C1A-4DCC-ABBD-0E831C0185CC}.msi --- snip ---
http://down.filepuma.com/files/system-utilities/tuneup-utilities-2014/TuneUp...
Snapshot of installer via Internet Archive:
https://web.archive.org/web/20191127225631/http://down.filepuma.com/files/sy...
$ sha1sum TuneUp_Utilities_2014_v14.0.1000.340.exe 43b17e17c855e83b57f73ea5c8c13aebe805e0da TuneUp_Utilities_2014_v14.0.1000.340.exe
$ du -sh TuneUp_Utilities_2014_v14.0.1000.340.exe 28M TuneUp_Utilities_2014_v14.0.1000.340.exe
$ wine --version wine-4.20-285-g4ccdf3e58a
Regards
https://bugs.winehq.org/show_bug.cgi?id=40334
Alistair Leslie-Hughes leslie_alistair@hotmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |STAGED Staged patchset| |https://github.com/wine-sta | |ging/wine-staging/tree/mast | |er/patches/ntdll-RtlGetNati | |veSystemInformation
https://bugs.winehq.org/show_bug.cgi?id=40334
--- Comment #53 from Paul Gofman gofmanp@gmail.com --- Should be fixed upstream by https://source.winehq.org/git/wine.git/commit/c6b501acdcabe7d3ebcf5ce13729fd...
https://bugs.winehq.org/show_bug.cgi?id=40334
Alistair Leslie-Hughes leslie_alistair@hotmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|STAGED |RESOLVED Fixed by SHA1| |c6b501acdcabe7d3ebcf5ce1372 | |9fd9f4d9176e9 Resolution|--- |FIXED
--- Comment #54 from Alistair Leslie-Hughes leslie_alistair@hotmail.com --- Fixed by https://source.winehq.org/git/wine.git/?a=commit;h=c6b501acdcabe7d3ebcf5ce13...
https://bugs.winehq.org/show_bug.cgi?id=40334
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #55 from Alexandre Julliard julliard@winehq.org --- Closing bugs fixed in 5.4.
https://bugs.winehq.org/show_bug.cgi?id=40334
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- URL|https://en.aion.gameforge.c |https://web.archive.org/web |om/website/download/ |/20140414143844/http://aa-d | |ownload.avg.com/filedir/ins | |t/avg_tuht_stf_all_2014_380 | |.exe
-
wine-bugs＠winehq.org -
WineHQ Bugzilla