[i945] (Jaunty) Eschalon Book I runs only in windowed mode and it's terribly slow with Intel video cards

[Lele83x]

Binary package hint: xserver-xorg-video-intel

Eschalon Book I is a very good native commercial Linux game that should run well on computers with poor video cards and poor overall specs. This is the homepage: http://basiliskgames.com/book1.html

My computer is a Thinkpad x60s with Intel L2400 cpu, 1G of ram, intel 945 chipset with intel 950 video card.
On my computer running a live of jaunty (a daily build: 2 feb 2009) from a usb stick, with the game loaded on my regular hard drive (with intrepid of course), the game runs ONLY in windowed mode. When it runs it is slow to the level you cannot even play it, the game does not crash but you are forced to quit because it become unresponsive.
If I try to run it in full screen all I get is a black screen with no music and no loading in the background.

In this other bug https://bugs.launchpad.net/bugs/321401 I reported the logs I got out of it.

Of course the game runned well on Hardy, didn't run at all on Intrepid, and now runs poorly on jaunty.

[lspci]
00:00.0 Host bridge [0600]: Intel Corporation Mobile 945GM/PM/GMS, 943/940GML and 945GT Express Memory Controller Hub [8086:27a0] (rev 03)
 Subsystem: Lenovo Device [17aa:2017]
00:02.0 VGA compatible controller [0300]: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller [8086:27a2] (rev 03)
 Subsystem: Lenovo Device [17aa:201a]

[Geir Ove Myhr (gomyhr)]

Hi again Axx83,

I attach the files from bug 321401 here as well for convenience. I saw
you got a reply from the Eschalon developers at
http://basiliskgames.com/forums/viewtopic.php?f=11&t=2080#p16840 :

"I'm afraid it's a known issue. Intel's current linux drivers do not
support the texture compression which the game needs. Basically,
unlike nVidia and ATI, Intel decided to put the texture compression in
software. But because the algorithm is patented, it can't be
implemented in free drivers."

I'm a bit puzzled by this. Especially since it ran well in Hardy.
While something goes wrong in mesa in the case of Intrepid, it looks
like it somewhat works for Jaunty although not very well. In this post
http://basiliskgames.com/forums/viewtopic.php?f=11&t=1951#p15771
it seems that one of the Intel driver developers, Eric Anholt, would
be happy to talk to the Eschalon developers about how they can improve
the performance. If this also results in a patch for the intel driver,
I guess that can also be included in Ubuntu.

I think the only way to get something done about this if the Echelon
developers get on the boat. It would be nice if they could comment
here, but as Eric said that it was not working working very well with
his developer's stack, this will really have to be solved upstream. I
searched for Eschalon on bugs.freedesktop.org which is where bugs for
xorg (including the Intel driver) are registered, but it didn't give
any hits. Maybe you can try and get the Eschalon developers to talk to
the Intel driver developers, and if they already did, give some
details about what the problems were and if they can be solved?

Geir Ove

[Lele83x (lele83x)]

Thanks for the help on this, I'm a newbie regarding bugs.

The reply I got here: http://basiliskgames.com/forums/viewtopic.php?f=11&t=2080#p16840
was from the developers at Hothead, the creators of Penny Arcade's "On the Rain-Slick Precipice of Darkness", and I guess their problem is quite different since their game is full 3d and all that stuff while Eschalon is totally 2d. I emailed them because neither that game runs on my intel card.

But again I'm no expert on drivers and 3d stuff, I just know a lot about patents and did a quick search.

Actually I'm a bit puzzled about Eschalon running VERY FINE on hardy while it runs so bad on Jaunty (and not at all on Intrepid) but so many things are involved in this that things like this can happen I guess.

I'll email the developers of Eschalon about this bug and the possibility for them to solve it by collaboration with the drivers programmers.

Thanks again for the help.

[Geir Ove Myhr (gomyhr)]

Sorry, I assumed Hothead was the company behind Eschalon (well, it says basiliskgames in the url, so I should have known better).

I hope this can be sorted out. I will leave this bug as confirmed for now.

[Geir Ove Myhr (gomyhr)]

It seems that it is not only Eschalon which has problems with fullscreen on 945GM. See bug 326524.

[Lele83x (lele83x)]

I sent an email to the developers of Eschalon but they said it was not their fault, but the programming language's, so I sent an email to the developers of Blitzmax the programming language used for Eschalon and this was the banswer I received:

EMAIL I REICEIVED FROM BLITZMAX DEVELOPERS

Thanks, I've forwarded your email onto our developers. I'm not sure there's
much we'll be able to do, as it sounds like it's a driver-specific problem,
but it's good for us to be aware of these things all the same. If we find
out anything we'll let you know.

Blitz Research.

----- Original Message -----
From: <...>
To: <[e-mail]>
Sent: Friday, February 06, 2009 9:01 PM
Subject: [SUPPORT] blitzmax and issue with new linux intel drivers

> Dear developers
>
> I tried to contact the developers of Eschalon book 1, a game created with
> your language, about a bug with the game and the latest version of the
> intel graphics driver for linux.
>
> This is the bug i noticed them https://bugs.launchpad.net/bugs/324292
>
> The bug is a serious one and it makes games created with your language
> preatty unplayable if you own intel video cards and run linux.
>
> In the email I told them to contact the intel linux driver developers to
> solve this bug because they where directly interested in collaborating
> with other programmers to solve the bug.
>
> The developers at Eschalon answered me tough that they can't do anything
> and the programming language deals with the graphics on its own, and that
> this problem should be turned to you at blitzmax.
>
> Could it be possible for you to collaborate with intel linux driver
> developers to solve this bug ?
>
> Thanks for the attention
>
> THIS IS THE EMAIL I SENT TO BASILISK GAMES:
>
> Dear developers I'm a happy owner of your game linux versione since it
> came out.
>
> Lately I've been a lot less happy bacause with the new kernel, and the new
> versions of Ubuntu the game does not run anymore.
> I filed a bug about this and a very competent person has given me a lot of
> help on the matter.
>
> The bug is this one :
> https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/324292
>
> Please read the bug very carefully and the answers I received.
>
> He suggested the developers of the game to collaborate with the intel
> driver developers as Eric Anholt (one of the intel drivers developers)
> himself suggested on your game forums
> http://basiliskgames.com/forums/viewtopic.php?f=11&t=1951#p15771 it's all
> on the bug page anyway.
>
> This is the contact page of the driver developers
> http://intellinuxgraphics.org/feedback.html
>
> Please solve this bug, if it's not solved in the next driver release me
> and all the linux owners of intel cards won't be able to play book 2.
>
> Thanks for the attention
>
> THIS IS THE ANSWER I REICEVED FROM BASILISK GAME
>
> Thanks for the email, Lele.
>
> I hate to be someone who passes the blame around, but we really have
> nothing to do with the implementation of the OpenGL driver on Linux. You
> may want to forward this information back to the Linux guys:
>
> We use a development package called "BlitzMax". You can find it at
> http://bli...

[Geir Ove Myhr (gomyhr)]

Thank you for the update, Axx83. You've done a lot of good research.

It seems that the fullscreen issue has been fixed for many other program by the latest driver update (xserver-xorg-video-intel 2:2.6.1-1ubuntu2). This was supposed to fix bug 324913 (which it probably did), but it is reported that it also fixed bug 326524 where the problem was that many games would not run fullscreen.

Could you test the updated version once it hits the archives? (it is not on my mirror yet, but it made it into the 20090210 daily-live at http://cdimage.ubuntu.com/daily-live/20090210/, you can see it in the files jaunty-desktop-*.manifest).

The performance issue is probably the same as for many other OpenGL applications. We really hope this can be sorted out for Jaunty. Setting
        Option "AccelMethod" "UXA
in the Device section of your /etc/X11/xorg.conf may improve performance, but unfortunately it is also not as stable as is required for it to be the default for Jaunty. See https://lists.ubuntu.com/archives/ubuntu-x/2009-February/000395.html .

[Geir Ove Myhr (gomyhr)]

I'll just add that that I downloaded and tested the Eschalon Book I demo, and it seemed to work quite well on my Thinkpad X61 Tablet (965GM chipset) with an updated Jaunty.

Short instructions if others want to test:
1. Download the tar.gz from one of the links at http://basiliskgames.com/book1.html
2. Unpack the tar.gz-file to folder (e.g. in your home directory)
3. Install the package libstdc++5 if not already installed (`sudo apt-get install libstdc++5`)
4. Run the file eschalon_book_1_demo in the eschalon_book_1_demo folder.

[Lele83x (lele83x)]

I'll test it right away !

[Lele83x (lele83x)]

I tested it today with a daily live (14 feb) and even tough now the game starts in full screen, it keeps been terribly slow.

[Geir Ove Myhr (gomyhr)]

Great that the fullscreen issue has been solved. I suspect the general performance issue is pretty general and not specifically related to Eschalon. The ubuntu X developers are looking at ways to optimize the performance, but unfortunately it looks like it's a choice between speed and stability/corruption in general. On some chipsets, though, the problems may not appear, while the performance may increase.

You can try to set
        Option "AccelMethod" "UXA"
in the device section of your xorg.conf and see if it improves performance. There is some ongoing reporting at https://wiki.ubuntu.com/X/UxaTesting , and the current results for 945GM is not overly positive. It would be nice if you could add your experience there as well.

[Bryce Harrington (bryce)]

[This is an automatic notification.]

A new major version of the -intel driver is now available in Karmic.

This version includes a major reworking of the acceleration
architecture, which resolves a huge number of issues. We do not know
whether it resolves the issue you reported.

Would you mind testing Karmic Alpha-2 and seeing if it is still a
problem? CD ISO images are available here:

  http://cdimages.ubuntu.com/releases/karmic/

If the issue can still be reproduced on karmic, please report here with
your findings, and attach a fresh Xorg.0.log from your test, and we will
be able to forward the bug upstream.

Otherwise, if the bug no longer exists in Karmic, let us know that as
well.

In the off chance you encounter different bugs while attempting to test
Karmic, please report those as new bug reports.

[Lele83x (lele83x)]

The game runs perfectly fine with 2.6.30 kernel and newest intel graphic driver

[Geir Ove Myhr (gomyhr)]

Thank you for letting us know that this has now been fixed. Closing this bug.

[Thomas Trummer (truetom)]

Sorry, but the game is as slow as ever.

[Thomas Trummer (truetom)]

I'll take that back. For some reason HW acceleration doesn't work with the latest xorg drivers. I'll try again when they do.

[Thomas Trummer (truetom)]

Latest updates brought HW acceleration back, works fine now. Sorry for the noise.

00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03)