fixme:d3d:resource_check_usage Unhandled usage flags 0x8 is probably harmless,can you attach a log with +d3d debug?

Yes, I'll get to that sometime after work.  I didn't think the console messages were harmful either, but I don't think it was ever spamming them that frequently before.  I normally only run on a 5000 line buffer and it fills that pretty rapidly.

well, you can use grep -E -v to get rid of the unwantd ones ;) pass the unwated one as parameter, ie "foo|bar|foobar"

Created attachment 45330 [details]
d3d debug output

Last relevant lines from winedebug using +d3d relay

Created attachment 45331 [details]
Screenshot of issue

Strangely the trees get texture but buildings, characters, and mines seem not to get textured.

Created attachment 45332 [details]
Running with system monitor.

It probably doesn't matter, but it is interesting the affect it is having on my normally unused cores.

Problem still exists in Wine 1.6 RC5

(In reply to comment #4)
> Created attachment 45330 [details]
> d3d debug output
> 
> Last relevant lines from winedebug using +d3d relay

Lots of textures seem to be set to (nil) - interesting.

can you test running it with GDI? if after that its no longer black reverse it to OpenGL and turn of GLSL, i had it often on my older laptopt that turning on GLSL turned my textures black. looking at the debug now ;)

Please run a regression test:
http://wiki.winehq.org/RegressionTesting

Building Wine on 64bit systems scares me, but according to AppDB, it was fine with 1.5.28, and is still present in 1.7

As an extra, it's also running extremely slowly (1-2fps) but i guess that's for another bug report

I can confirm this bug with Wine 1.7.6 and 1.5.30. Running with the "-nostartup" launch option doesn't make any difference.

I'm on a Debian Sid box with a Radeon HD 6470M, using the open source driver.

(In reply to comment #12)
> I can confirm this bug with Wine 1.7.6 and 1.5.30. Running with the
> "-nostartup" launch option doesn't make any difference.
> 
> I'm on a Debian Sid box with a Radeon HD 6470M, using the open source driver.

Sorry, I forgot to mention it is 64bit OS and Wine compiled as 32bit.

For sure, it's a regression. I've tested with Wine 1.5.28 and all textures are correct! Are there any information that I can provide to help fixing this bug? Thanks.

The fatest way to see this bug fixed is to perform a regression test:
http://wiki.winehq.org/RegressionTesting

So, here it is:

"e0494afa007a1d1ba2f4b649beedca27cb20808e is the first bad commit
commit e0494afa007a1d1ba2f4b649beedca27cb20808e
Author: Henri Verbeet <hverbeet@codeweavers.com>
Date:   Tue Jun 11 10:21:28 2013 +0200

    wined3d: Introduce a helper function for printing floats in GLSL shaders.

:040000 040000 c4b9e3a37188f45aff9cb35d00ef45aa6242e0a4 4a7a6eadaf6ebfe495915b798111dd49d1ed8768 M	dlls"

Is that what you need?

Thanks.

(In reply to comment #16)
> So, here it is:
> 
> "e0494afa007a1d1ba2f4b649beedca27cb20808e is the first bad commit
> commit e0494afa007a1d1ba2f4b649beedca27cb20808e
> Author: Henri Verbeet <hverbeet@codeweavers.com>
> Date:   Tue Jun 11 10:21:28 2013 +0200
> 
>     wined3d: Introduce a helper function for printing floats in GLSL shaders.
> 
> :040000 040000 c4b9e3a37188f45aff9cb35d00ef45aa6242e0a4
> 4a7a6eadaf6ebfe495915b798111dd49d1ed8768 M    dlls"
> 
> Is that what you need?
> 
> Thanks.

Yes, thanks.

Does this have any kind of demo that can reproduce the issue? The commit you found looks somewhat unlikely, given that there's no debug output with GLSL compiler errors. Does reverting it fix the issue?

I don't think there's a demo. I've searched and I don't see any: http://store.steampowered.com/search/?snr=1_5_9__12&term=age+of+empires

By reverting, do you mean, checkout a single commit before e0494afa007a1d1ba2f4b649beedca27cb20808e and try again (compile wine, test ...) ?

Thank you very much.

Checking out to the previous commit (b02951dfa7ff416f54205dab17ae3bb98e53ce7c, immediately before e0494afa007a1d1ba2f4b649beedca27cb20808e) does work. All textures are all right.

The commit pointed out by git bisect seems to be the problem. Maybe I did something wrong, I'm not sure.

What else can I make to help?

Thank you.

Created attachment 46583 [details]
debug hack

(In reply to comment #19)
> I don't think there's a demo. I've searched and I don't see any:
> http://store.steampowered.com/search/?snr=1_5_9__12&term=age+of+empires
> 
That's unfortunate. Does the attached hack also fix the issue?

Henri Verbeet, your patch fixes the problem!!!

Steps:

- git checkout master
- git pull origin master
- patch -p1 < /tmp/debug-hack.patch
- clear; CC="ccache gcc -m32" ./configure --verbose --disable-tests; make -j 4;
- make install

And then, I tested the game.

Thank you.

Created attachment 46591 [details]
another hack

How about this one?

I'll give it a try Henri Verbeet.
Should I checkout master again or test both together?
Thank you!

(In reply to comment #24)
> I'll give it a try Henri Verbeet.
> Should I checkout master again or test both together?
> Thank you!

Just that patch on top of current git.

It's still working with both patches applied.

Are you online on IRC, Henri?

Thanks.

(In reply to comment #26)
> Are you online on IRC, Henri?
> 
Yeah, #winehackers.

Created attachment 46593 [details]
patch

The attached patch should be more of a proper fix.

It works as expected. Will it be integrated into Wine source code?
Thank you very much, Henri Verbeet.

Patch commited:
http://source.winehq.org/git/wine.git/?a=commit;h=12f16f2127ffce9a05eedf008c3aa3732751a7e3

Should be fixed by commit 12f16f2127ffce9a05eedf008c3aa3732751a7e3.

Closing bugs fixed in 1.7.7.

Removing 1.6.x milestone from bugs included in 1.6.2.