[BUG] R6025 - pure virtual function call

[americantabloid]

Hi!

When I go thru my media files, I sometimes experience this bug,

R6025 - pure virtual function call

Here is a screenshot of the errortext,

http://i475.photobucket.com/albums/rr117/americantabloid/Purevirtualfunction.jpg

I was scrolling thru my media files, first my video directory, then my series directory, doing some fanart updates and thumbs updates on my network share when it happened.

System specs, XBMC Atlantis 4, Vista x64, Radeon x1950 pro, Catalyst 8.8.

I knew I read on the forum about a similar problem with the pure virtual function error, but I have searched on pure, pure virtual, r6025 and didn't find it.

Xbmc debug log
http://pastebin.com/f400e62fc

I also know that there where a similar ticket created on trac, but I can't seam to find that either...I know that you asked for more info in that ticket and thats why I created this thread.

Please let me know if you want any additional information or help with testing done.

Best regards
at

[jmarshall]

The key is to get it reproducible. It appears as though it's only in some circumstances, which is why I can't reproduce it. Once I can reproduce it (reliably) then I can investigate.

The ticket is here:

http://xbmc.org/trac/ticket/4520

Feel free to reopen it and add your experience.

Cheers,
Jonathan

[americantabloid]

Hi Jonathan!
Will try some more during the week and try to see if I can find any common causes for the bug to show up, so I can get it reproducible for you.

Thanks for the link.

Regards
at

[xeonicxpression]

The key is to get it reproducible. It appears as though it's only in some circumstances, which is why I can't reproduce it. Once I can reproduce it (reliably) then I can investigate.

The ticket is here:

http://xbmc.org/trac/ticket/4520

Feel free to reopen it and add your experience.

Cheers,
Jonathan

Sorry that I've kind of been MIA since creating that ticket. Hopefully I'll be able to tackle it again this week. My computer has been out of commission for quite awhile now. Life got in the way of getting it up and going again. I'll upgrade to the latest atlantis release and let you know if it still happens.

[elupus]

http://support.microsoft.com/kb/125749

might be of help.

[jmarshall]

elupus:

It appears to be due to two simultaneous accesses to the imagelib dll, but I'm not sure how it's occuring. The grabbing of the dll is critical sectioned off, but perhaps it's due to the ResolveExports() call not being protected, and the dll.Load() function being called almost simulataneously from different threads??

As can be seen from the debuglog, the exception occurs when the dll is being unloaded.

Wish I could reproduce it so that I can get a callstack!

Cheers,
Jonathan

[americantabloid]

Hi Jonathan!

I have been doing some additional testing.

I have run XBMC on my laptop, connected to the same share as on the Vista64 machine that has resulted in the R6025 error.

Much to my surprise I can't get the error to reproduce on my laptop.

The spec of the laptop is HP nx7000 Pentium M 1.5GHz, Radeon 9200 graphics (Hp supplied driver) and Windows Xp Professional SP3.

What I did to try reproduce the bug is move around my series libary, where I have the series sorted in file structure like series name\season\episodes.

(I think the bug has something to do about going thru the libary and let XBMC generate screenshots from the episodes.)

Then I did the same with my Vista machine and after a short while I run into the R6025 error. To bad I didn't start XBMC with the "Run as an Administrator" option that I need to get an XBMC debug log written.

Anyway I started again and after awhile XBMC hang again. When I Ctrl+Alt+Del to get back to the desktop, it said that XBMC had stopped working. (I did not get the prompt about the R6025, but I took a quick look at the new debug log and seamed similar like the previous posted log, so I decided to post it anyway.)

http://pastebin.com/f3ca61368

Could the bug have something to do about running on dualcore vs single core machines?

Best regards
at

[jmarshall]

It could be affected by number of cores, and it is certainly the same problem.

[americantabloid]

Hi again!

Would it help if you could have Remote Desktop access to my machine and use it for debug purpose?

I don't know how to get you any more useful information.

Regards
at

[jmarshall]

Nope. The only thing that will make a difference is finding out how to reproduce it reliably (i.e. every time, or at least close to it).

[americantabloid]

Did a test to run XBMC on my computer that I use for fileshare duties. It has a Pentium E2160 processor (dual core), Windows Xp Pro SP2 and tried usual procedure, updating information of the media files, making XBMC generate thumbs when there is non to download for series. Just doing the same as on the Vista machine, same files since it is on the file sharing machine.

Can't get it to reproduce on this machine. Will do clean install on the Vista machine and see if it helps.

//at

[Freddo]

I'm having a problem with my new skin, I think this is related to the use or the largeimage control, and ImageLib.dll, basically any time an image is being loaded using largeimage theres a chance XBMC will crash, it seems to be quite random what'll set it off, sometimes it takes hours sometimes it happens right away, but the more largeimage loads the more likely it is to happen.

http://pastebin.com/f70acab6d

There's the debug log, you can see at the very end it freaks out due to msvcr71.dll.

Build being used is atlantis beta1, although I've also encountered the same problem on the latest SVN and various builds in between.

Platform is windows vista ultimate, specs are well above requirements, bar the intel gma945 graphics but to my knowledge that shouldn't affect gui usage (I'm just using this machine as a skinning tool basically).

So, can a Dev take a look at the log and tell me, is this something I'm doing wrong with my skin or is it an actual bug? if so I'll post a ticket for it.

[WiSo]

I don't think it's related to MSVCP71.dll as we don't need since we emulate it.
I assume it's something in imagelib maybe jmarshall can give a hint here?

[jmarshall]

There's a trac report already for this. We believe it's due to the imagelib being loaded/unloaded in a not-threadsafe manner (I suspect due to the resolving of imports/exports)

http://xbmc.org/trac/ticket/4520

Please add your findings to that. The key is to get it reproducible (easily) for a dev so we can hit the callstack to see what is possibly going wrong.

The log there is slightly different, but it gave the same error, so may well be another manifestation of the same problem.

Cheers,
Jonathan

[Freddo]

appended my findings but I don't have permissions to re-open the ticket. I believe this can be made reproducable by simply making extensive use of the largeimage control in a skin, since it seems to have started occurring as soon as I started using them.

[americantabloid]

Oki with the release of the Beta 1, I decided to do a complete reinstall of XBMC on my Vista Ultimate x64 machine. (See previous post in this thread for complete spec.)

Anyway, removed all xbmc content in AppData folder before installing the new version.

Started to rescan my files from scratch. Tv series worked. Started with the movies directory, bang XBMC hangs after a short while. Ctrl-Alt-Del back to desktop and got the Pure Virtual Function error. Tried again with a rescan of the movies folder, same error after a short while.

Tried to post the debug log to pastebin, but pastebin complains about anti spam reacted to the debug log.

So I don't get this errror on two of my XP machines, but it always shows on my Vista Ultimate x64 machine. Scanning is done on the same content on my fileserver.

Regards
at

[americantabloid]

This is another thread about the same problem.

http://xbmc.org/forum/showthread.php?t=36618

//at

[americantabloid]

Added the information to the closed trac id. Couldn't not reopen the ticket.

//at

[Freddo]

thanks, hopefuly the devs will be able to reproduce this one soon, my HTPC runs XP so thats not a problem but it's bugging me like crazy on my vista machine where I'm trying to develop my skin. jrm if you're reading has this additional information enabled you to reproduce or do you still need more info?

[jmarshall]

I don't have Vista, so unless it's reproducible on XP, I'm not your man I'm afraid!

[WiSo]

I have vista but only 32 bit and no decent test library. I can try a scan but doubt to find something useful.
Does it crash on a specific movie, file or something from looking in the logfile?
Otherwise I dunno on how to reproduce.

[Freddo]

as far as I can tell its nothing to do with scanning movies or the library, it's just a case of the large image control being used in a skin, for anything, the more its used the faster it'll crash.

[WiSo]

Ah true should have read before post ;)
Which skin do you use?
I'm not good in this skinning stuff so need something ready for testing.

[Freddo]

I've been making my own skin that uses a lot of largeimage controls.

[americantabloid]

Here is my latest debug log. I run PM.HD, but get the error in PM3 also.

(Sorry for it did take awhile for posting the latest debug log. I had trouble fining a pastebin that accepted the debug log. Pastebin.com did not accept it.)

http://pastebin.ca/1211410

I can't either give an exact example on how to reproduce the error. When the error shows, it is when I move around my files structure or go to one big video folder (with lot of sub folders with movies) and right click and trying to set content for the movies.

Regards
at

[barshad]

I've created a CD to Install Full Window XP on USB Flash Drive (not PE), everything is working fine (boot time from power on to XBMC is 45 seconds, in XMBC Live its more than 2 and half minutes) except when i shutdown the program terminates with error something like 'Pure virtual function call!' in error window,
Some times i also receive this error on my normal installtion when i
use "Scan for new contents"
Can some one help me how to fix this error?

[WiSo]

merging threads.

Please post a logfile. So it looks like that this issue isn't Vista specific

[DragonFly]

These errors are mostly generated by the microsoft visual c++ runtime libraries. Maybe you don't have the "Msvcp71.dll" or the "Msvcr71.dll". Also try installing the "Microsoft Visual C++ 2005 SP1 Redistributable Package".

Quick question though. Why would you install XP on a Flash drive? Boot time from a Flash medium is much slower than from a real HD.

Also, the XBMC Live CD is intended to run on various machines. Your version (if not PE) can only run on the machine it's made on.

[Freddo]

I'm led to understand msvcp71.dll isn't required or a possible cause since it's emulated and not used directly?

[DragonFly]

Have a look in the Windows forum regarding issues with the mentioned libraries. http://xbmc.org/forum/showthread.php?t=37684 (reply #7 and #8). Mostly related to scripts and plugins not working, but who knows?

[WiSo]

I'm led to understand msvcp71.dll isn't required or a possible cause since it's emulated and not used directly?

Yes and no. It seems so that the XBMC binary needs no runtime dll that is a yes, even though I don't know why.
All dlls loaded via our dllloader won't need it either.
But the vis plugins and libcurl are loaded via the windows loadlibrary method because it crashes with ours. Therefore you need for those libraries the msvcp71.dll.

[americantabloid]

Here is another debug log with the same error from a different machine.

This was run on a HP 6710b, Core 2 Duo T7250, 1,5 Gb memory and Intel 965 graphics with Vista Business SP1 32 bit version.

Same video share as my previous posts in this thread. It did take almost an hour before it appaired this time. The share is around 400 GB and includes movies in subfolders.

What I did was, right clicked on my video share in XBMC, choosed Set Content with option Run Automated Scan, Use folder Name For Lookup and Scan Recursively.

Here is the debug log.
http://www.pastebin.ca/1213795

Regards
at

[Freddo]

what revision was that at? I think this was fixed @revision:15692, at least I haven't had it occur since that fix.

[americantabloid]

Beta 1. Will look at the ticket.

//at

[americantabloid]

Freddo, how do you know that it is fixed?

I can't find any new information in the ticket.

//at

[Freddo]

the changelog read:

@15692 wiso

fixed: memory for pixels in TextureManager?.cpp wasn't freed. Please tell if this was per purpose or if it is freed somewhere else.

since that fix I haven't had a crash, it needs more people to build and confirm the crash isn't happening before the ticket can be closed though.

[americantabloid]

Thanks freddo!

Will get around to compile my own version and test tomorrow.

I have just been keeping track of this tickets information, not compiled my own versions.

http://xbmc.org/trac/ticket/4520

Regards
at

[jmarshall]

I very much doubt that that would have fixed the issue, given that all that did is free up some memory when loading animated gifs.

[Freddo]

...how inexplicable, maybe the fix..fixed more than you thought? I've still yet to experience this crash since that fix went in, and I was getting it at least once an hour before.

[Freddo]

This might be related to the way imagelib.dll seems to leak memory like a sieve, I can sit and scroll a list that has a largeimage with listitem.icon as the source and memory usage will climb until it's all used (~1.6gb) I'm sure that's not normal behavior.

[americantabloid]

Quick update,

Still trying to get all the source fram svn, so I haven't been able to some additinal testing with the latest versions.

Will hopefully get some done tomorrow or in the weekend.

Regards
at

[jmarshall]

Freddo: All textures are released after n seconds. I think n is currently 2.

It doesn't leak at all to my knowledge. Just because memory usage as referenced by the OS seems to be high, it may well just be that it's kept them available for the app as it appears to want to reuse it (which will ofcourse happen when scrolling through lists).

Cheers,
Jonathan

[americantabloid]

Hi Jonathan and Freddo!

I finally got all the source from svn and compiled it. I still get the error. Same method as before. Set Content for a video share.

http://www.pastebin.ca/1220120.

This was run on the HP laptop with Vista previously mentioned in the thread. Same share as always.

Svn revision around 15799.

Regards
At

[jmarshall]

Unfortunately I can't reproduce it, so all I have is guesses. Please add your log, and details to the trac report so that it's always available to us.

Thanks,
Jonathan

[americantabloid]

Hi Jonathan!

I got a background in programming, but I been out of it the last couple of years and are not that familiar in the VS2008 environment, so if you would give me instructions, I would try and get you the information you need. But if you think it will be too hard, we will just have to wait for someone else to get it reproducible.

(Where I should place suitable breakpoints in the code or something.)

Anyway thanks for showing the interest.

Regards
at

[jmarshall]

Hi there,

Basically, do a debug build and run it from visual studio. You may have to set the XBMC_HOME environment variable, or change the output path of the executeable so it dumps it down into SVN.

Then just try and reproduce while running from the debugger. When it asserts you should <hopefully> get dumped down into the debugger, which should give you a callstack for the threads and so on. That's the info we're interested in. We need to know the state of all threads when it asserts, so that we can determine how this race condition is popping up. If it doesn't assert during the unload of the dll, you could try adding a try {} catch (...) block to the dll unloader and set a breakpoint within that.

Cheers,
Jonathan

[americantabloid]

Oki Jonathan, I will give it a shot at debugging during the week.

Thanks
at

[americantabloid]

This is really strange. I have been running in debug mode for 4-5 hours. Same svn 15799 as used in building the app with the win build script yesterday. Been doing the same as has triggered the error before...and no error.

I'am really puzzled now. Could the debug enviroment handle this kind of errors, so I don't get same error text?

Regards
at

[jmarshall]

Try running in release mode - you may still get some information if it breaks.

Cheers,
Jonathan

[americantabloid]

Since there has been an release of beta2 since my last post I decided to give that a run(The downloaded xbmc for windows released by xbmc that is).

I tried both my Vista machines (see previous posts for detaíled specs) that has shown the virtual error message.

Same procedure as before...setting content, move around xbmc.

I did take about close to 15 minutes to get the error on my Vista x64 machine this time.

http://www.pastebin.ca/1222661

I'am still running on my x32-bit Vista laptop for about an hour now, setting content for the video directory. I will report back on the result for that machine later and if I get an error, I will start again in the Visual Studio enviroment and try to get a call stack.

Regards
at

[americantabloid]

The laptop run for 1 hour and about 20 minutes before the crash. I will attach this xbmc.log to the trac ticket.

Will get the source for beta 2 from svn and give a try in VS later today or during the week. I did try a different video library this time for the Vista x64 machine so I can't relate the crash to any specific movie or action.

Am I the only one still expericing this error?

Regards
at

[Freddo]

could this be related and possibly explain why some vista users are experiencing this and some aern't, apparently the msvcp71.dll which I am seeing errors for in the debug log isn't included with vista by default, it's down to users to procure it or developers to include it, so I must have picked up one version somewhere, other users have a different version and they are fine? perhaps XBMC should include a reference msvcp71.dll?

http://forums.microsoft.com/MSDN/ShowPost.aspx?PostID=1151823&SiteID=1

this link explains in more detail, especially post 8.

[jmarshall]

No, that is nothing to do with it. Said dll should certainly not be loaded or required.

[Swwethip1]

I just got this error yesterday.. i have been using xbmc for about a month and a 1/2..

Was using IRSS with xbmc.. yesterday i finally tried out evenghost, so i uninstalled IRSS installed EG, was on the default HD skin and went to switch mediastream(from the menu) and then got the error..

dont have the log atm, but can post it when i get home tonight

[americantabloid]

I did run under VS for over an hour trying to get error message, but no luck. This time the release version.

The only thing that might by of interest is a couple of this,

"First-chance exception at 0x76ff42eb in XBMC.exe: Microsoft C++ exception: CL_EXCEPCION_FINAL_FUNCION at memory location 0x0375e264..
The thread 'Win32 Thread' (0x1084) has exited with code 123 (0x7b)"

But the they could just be a planned thrown exception, I am not that familiar with the code. Maybe Jonathan can decide if this is relevant.

Regards
at

(Very strange I can't get this error VS. I have absolutly no problem reproducing the error with the releases.)

[americantabloid]

Swwethip1, can you please post your specs also

Regards
at

[jmarshall]

Where did you get the above error - in the same situation (i.e. when unloading the imagelib.dll?)

Cheers,
Jonathan

[americantabloid]

Where did you get the above error - in the same situation (i.e. when unloading the imagelib.dll?)

Cheers,
Jonathan

Sorry, I can't really tell were I got them. I just run the app under VS and did what I usually do to provoke the bug, updating cover art and moving around in XBMC.

Since I can't get it to reproduce, when run under VS, I just checked the log in VS after exiting the app. These exceptions was the only thing that stood out from VS log file.

(I didn't set any catch statements in the code. I need to familarize myself with the code more first.)

Can we reopen the bug again? I don't seam to have the credentials for this.

Regards
at

[jmarshall]

Have reopened the ticket and attached an experimental release build for you to play with (exe only). It adds exception handling to the unload of the dll - hopefully it'll at least provide some information in the log if an exception occurs, and may even "fix" the issue - who knows.

http://xbmc.org/trac/ticket/4520

Cheers,
Jonathan

[americantabloid]

Have reopened the ticket and attached an experimental release build for you to play with (exe only). It adds exception handling to the unload of the dll - hopefully it'll at least provide some information in the log if an exception occurs, and may even "fix" the issue - who knows.

http://xbmc.org/trac/ticket/4520

Cheers,
Jonathan

Thanks Jonathan, will give it a try later today.

//at

[americantabloid]

Hi Jonathan!

Tried the exe file on my Vista x64 machine and got a crash after a short while. Attached the log file to the ticket. Will try it later in the week under VS and on my other Vista machine.

Regards
at

[jmarshall]

Have updated with a new attempt (better logging) to hopefully try and track down the source of the problem.

[americantabloid]

New debug log using the lastest xbmc jonathan uploaded ticket. My Hp Vista x32 laptop used.

Will get into VS now and try and get some useful information.

Regards
at

[americantabloid]

Updated ticket with a log file from VS2008 running Jonathans latest exe. No error message shown during a lengthly run. Same laptop was used as in the post above.

Now I'am starting to run out of ideas.

Regards
at

[paco]

I also get this error from time to time. I will download the new files and give the logging a try. I mainly have it happen when I'm playing music in the background while browsing my pictures folders. The last time this error appeared, I was playing a song and swapping between home and the artist's album directory. I hope this bug can be tracked down. At the moment, I try to be cautious when doing too much in XBMC.

If it's any help:

Vista SP1
Intel C2D 2.66
Asus P5K-VM
Nvidia 8500GT (Latest drivers)
Imon remote software emulating XBMC key presses within XBMC

[hgumus]

Hi,

Here is the log collected with the exe downloaded from the ticket.
http://pastebin.com/m718211da

I've tried scanning movie folder and it ended with pure function call error.

I've downloaded latest from svn and tried with VS2008 but couldnt reproduce the error yet.

[jmarshall]

Log is truncated, but I suspect it ends the same as the others.

I'm not sure exactly where it's going wrong at this point. Will add even more logging and post a new exe at some point.

Thanks for your continued patience.

[WiSo]

@jmarshall: maybe not related to this topic but can we somehow increase the delay to unload a dll? It makes no sense that we continuously reload imagelib I think.

[hgumus]

This time I've put msvcp71.dll in XBMC\system\ folder, and it didnt crush but
VideoInfoScanner ended with exception:

11:00:46 T:7304 M:867835904 DEBUG: SECTION:UnloadDelayed(DLL: Q:\system\ImageLib.dll)
11:00:46 T:7304 M:867835904 DEBUG: Q:\system\MSVCP71.dll (base 03D10000): Detected memory leaks: 1 leaks
11:00:46 T:7304 M:867831808 DEBUG: leak caller address 3d12d3c, size 128, counter 1
11:00:46 T:7304 M:867831808 DEBUG: MSVCP71.dll: Total bytes leaked: 128
11:00:46 T:7304 M:868335616 DEBUG: Q:\system\ImageLib.dll (base 04DD0000): Detected memory leaks: 1 leaks
11:00:46 T:7304 M:868335616 DEBUG: leak caller address 4e98ed8, size 128, counter 1
11:00:46 T:7304 M:868335616 DEBUG: ImageLib.dll: Total bytes leaked: 128
11:00:46 T:7304 M:869412864 DEBUG: SECTION:UnloadDelayed(DLL: ) DONE
11:00:52 T:7036 M:869421056 ERROR: Recv failed: 10060
11:00:52 T:7036 M:869384192 ERROR: VideoInfoScanner: Exception while scanning.
11:00:52 T:7036 M:869384192 DEBUG: CThread::staticThread, deleting thread graphic context
11:00:52 T:7036 M:869384192 DEBUG: Thread 7036 terminating


Log:
http://pastebin.com/m7237c909

Here is another try ending with exception in very short time
http://pastebin.com/m10c543d8

[hgumus]

Here is new logs collected with new exe uploaded by jmarshall

With mscvp71.dll in XBMC/system

http://pastebin.com/m5596a33b

Without mscvp71.dll in XBMC/system
http://pastebin.com/m4bae1399

Both didnt crash, ended with exception; but crashed after "end" button pressed giving continuous error log: Some other thread trying to leave our critical section.

[WiSo]

Please don't do the msvcp libraries into the system folder. They may get read and will overwrite our own functionality and may also crash the system.
This message is unfortunately a little bit misleading as we export the needed functions of msvcp on our own.

[americantabloid]

Updated ticket with new crash report, running the new uploaded xbmc.exe.

Regards
at

[jmarshall]

hgumus: Your problem is I believe unrelated. You do not get the "pure virtual function call" error, do you?

[paco]

Well, it took me a while with this new executable (almost 25 minutes), but I finally got that damn error to pop up. As usual, I had music playing in the background while browsing through my pictures.

Here's the log:

http://pastebin.com/f1f65de29

I hope this helps.

[hgumus]

hgumus: Your problem is I believe unrelated. You do not get the "pure virtual function call" error, do you?

The only log with "pure vital function call" which I posted is http://xbmc.org/forum/showpost.php?p=233867&postcount=66. The others gave only exception in log.

[americantabloid]

Updated with new crash running Jonathans latest exe.

Regards
at

[americantabloid]

Oki, here we go.

Thought I try to summarize my findings so far.

I have four systems at my disposal,

1. Hp nx7000
(Pentium M 1,5GHz, Radeon 9200, Hp supplied driver, WinX Sp3, 1 Gb memory).

2. Home built
(Pentium E2160 (Dual core), Ati x1300, Catalyst 8.9, WinXp Sp2, 2 GB memory.) (Primary used as file server.)

3. Hp 6710p
(T7250 Core 2 Duo 2.0GHz, GMA965, Intel Driver 7.14.10.1437, Vista Business Sp1 x32, 2 GB memory.)

4. Home built
(Core 2 Duo 3.0GHz, Ati x1950 pro, Catalyst 8.9, Vista Ultimate Sp1 x64, 4 GB memory.)

With the Vista systems (3 and 4) the error is highly reproducible running XBMC (from before Beta1 to svn 15932(081019)). The error occurs from within a couple of minutes to little bit over an hour on this two systems. I have run both PM3 and PM.HD and received this error.

To reproduce the error, this is what I do.

Right clicked on my video share in XBMC, choose Set Content with option Run Automated Scan, Use folder Name For Lookup and Scan Recursively.

The share is around 400 GB and includes movies in subfolders. I have also tried a different share with high resolution movies and the same error shows for this local share.

It has also occurred when I was in the information screen in XBMC, where you see processor information and hard drive information, though I had previously been setting some content for movies and series.

The only setting I change in XBMC is language (swedish), resolution (full screen) and sometimes automatic generate thumbs for content without cover art and turn on debug mode.

(I have also tried to run in debug and release mode under VS2008, but here I can’t get error message to appear in Windows. I haven’t found anything special in the log in VS2008 either, but I’am no expert here. But I have run for several hours this way and no error message under VS 2008. This on my system 3, HP 6710b.)

Best regards
At
Sweden

[jmarshall]

Thanks americantabloid + others.

I've posted another exe - this one the logging is a little more explicit, but is essentially the same. I really don't get what's going on here - it appears the linked list we use to store the dll's exports appears to be getting corrupt, though how this is happening is a mystery at this point. It'll be interesting to see whether it's getting corrupt at the same point or not each time.

I've also uploaded a second executable that simply never unloads the imagelib - this should "fix" the situation - please let me know how it goes. If it's memory corruption, then other bad things may well result!

Cheers,
Jonathan

[americantabloid]

Hi, I have done some additional testing using the new files. Updated the ticket with the new information.

I'am a bit short on time during the week, so I may not be able to do some new testing until late in the week. Will keep an eye on the thread and the ticket and will test anything you need testing as soon as times permits.

Thanks for otherwise wonderful app by the way!!!

Regards
at

[hastarin]

I believe an issue I opened may be related to this. I've commented as such on both tickets, but thought I'd post here as well for good measure.

http://xbmc.org/trac/ticket/5213

[Incriminator]

Just installed on a Vista32 Box - Quad Core, 4GB ram.
I get the error while trying to scan music folders into the library. ImageLib.dll seems to be the issue.
What happens is the scanning will crash on this R6025 error, I will close XBMC, start it back up with auto scanning on, and it will resume scanning.
If I do this enough, it eventually gets through each directory.
My music folder has about 30 directories.
Scanning seems to freeze irrespective of directory content size.

Here is the dialog box that pops up:
Runtime Error!
Program C:\Program Files\XBMC\XBMC.exe
R6025
pure virtual function call

Strange as this happened with the XBOX as well trying to scan items to the library- I have 35416 files representing 320GB of music. With the XBOX these were SMB shares, but now the files are local. Strange.

Problem Event Name: APPCRASH
Application Name: XBMC.exe
Application Version: 2.1.0.0
Application Timestamp: 48ea4922
Fault Module Name: XBMC.exe
Fault Module Version: 2.1.0.0
Fault Module Timestamp: 48ea4922
Exception Code: 40000015
Exception Offset: 0006218e
OS Version: 6.0.6001.2.1.0.256.1
Locale ID: 4105
Additional Information 1: 3187
Additional Information 2: d9727574d1e3f3d82eb83012990f381f
Additional Information 3: 965e
Additional Information 4: 88eafd7c728faeca6e3e37686e8c0857


Edited debug log...


(http://pastebin.com/m3cb55c40)
http://pastebin.com/m3cb55c40

10:32:12 T:30916 M:1932173312 DEBUG: SECTION:UnloadDelayed(DLL: Q:\system\ImageLib.dll)
10:32:12 T:30916 M:1933230080 ERROR: exception in CApplication::Process()
10:32:13 T:30916 M:1933176832 DEBUG: SECTION:UnloadDelayed(DLL: Q:\system\ImageLib.dll)

[jmarshall]

There's already a workaround for the Atlantis release.

[WiSo]

Tested again on Vista 32Bit with fanart enabled in library mode. No crash visible (during debug) and the video plays fine.
It's only a small videolib though.

[NineT9mustang]

I wont be of much help here since i dont have a log and cant reproduce it right now... But I just got the Pure Virtual error

Vista X64, Was browsing my movies (1tb if that helps?) and changing thumbnails/dvd posters... then it crashed

Later when I have time I'll try reproducing

Edit: Built Nov 4th

[jmarshall]

The workaround is only in the Atlantis branch. It is not in the main linuxport branch.

[cheetah05]

The workaround is only in the Atlantis branch. It is not in the main linuxport branch.

I can confirm that this is working for me. Vista x64 (64bits).

I used to get the error, I installed the latest build from here: http://www.mediafire.com/xbmc.

I don't get this error anymore (so far).

[WiSo]

I can confirm that this is working for me. Vista x64 (64bits).

I used to get the error, I installed the latest build from here: http://www.mediafire.com/xbmc.


afaik this is the linuxbranch without the patch.

[cheetah05]

afaik this is the linuxbranch without the patch.

Lol...:s.

Well it installed on windows fine?....so surely that means it must compiled/built for windows.

Also, I got that link from the top of this forum XBMC for Windows - Help and Support (http://xbmc.org/forum/forumdisplay.php?f=59) in this thread: Third-party SVN builds for the community (unofficial and unsupported) (http://xbmc.org/forum/showthread.php?t=35452).

...but come to think of it, it did include some stuff that was supposed to be linux branch only....like the "Programs in Home menu" setting/functionality.

[jmarshall]

Linux branch == main trunk. It's for all non-xbox platforms.

The Atlantis branch is for the 8.10 release for all non-xbox platforms. The workaround is in this branch only.

[jmarshall]

cheetah05: Please try a build from the win32dllloader branch - ikons has done one up (see the sticky in these forums).

[americantabloid]

cheetah05: Please try a build from the win32dllloader branch - ikons has done one up (see the sticky in these forums).

For everyone that has this problem, try Jonathans suggestion. It worked for my two Vista system.

Updated the ticket with this information.

Regards
at

[azido]

i've just installed ikons version.

now i cannot fire up xbmc in windowed mode? both links (xbmc and xbmc (windowed)) fire up xbmc in full screen (thus changing my screen resolution)

strange side behaviour: using xbmc fullscreen and switching to windowed mode with "\" works nice. using "xbmc windowed" also starts up full screen, when switching back to windowed, it only uses a small window.

screenshot:

http://www.250kb.de/u/081201/j/bf21aa2f.jpg

regarding to the pure virtual function call, i'm testing now if this version elimiates the problem. still have lots of movies to add to the database ;)

as far as i can see, vista users (especially those on 64bit) have this error the most.

lemme update some movies in db and either you get my log soon (in case of error) or tomorrow get the info you can put this "fix" in a current svn version where my displaying probs also were fixed :PP

edit: after populating the database, browsing it using fanart, playing some movies for about 4+ hours now i might say: the problem seems to be resolved with this fix - at least i got no single crash at all. pvc issue seems to be eliminated.

[azido]

as promised, here i am again with my report.

for me, after installing that patched version of xbmc with the different windll-loader, the issue is resolved.

i did not get any single crash since i've installed this version yesterday - and did a lot of database inserting and displaying with and without fanart enabled.

@jmarshall: whatever you've done here, add it to the main branch ;)

cheers,
azido

[jmarshall]

Thanks all for testing. I'll merge it back to the main branch when I'm back from vacation.

Cheers,
Jonathan