I have the same issues with 1.1.7 during gameplay: missing objects/textures

fixme:d3d_shader:IWineD3DPixelShaderImpl_CompileShader Shader 0x207d2730 recompiled more than 50 times
fixme:d3d_shader:IWineD3DVertexShaderImpl_CompileShader Vertexshader 0x81d4470 recompiled more than 50 times
fixme:d3d_shader:IWineD3DVertexShaderImpl_CompileShader Vertexshader 0x81d4470 recompiled more than 50 times
fixme:d3d_shader:IWineD3DPixelShaderImpl_CompileShader Shader 0x207d2730 recompiled more than 50 times
fixme:d3d_shader:IWineD3DVertexShaderImpl_CompileShader Vertexshader 0x81d4470 recompiled more than 50 times
fixme:d3d_shader:IWineD3DVertexShaderImpl_CompileShader Vertexshader 0x81d4470 recompiled more than 50 times
fixme:d3d_shader:IWineD3DPixelShaderImpl_CompileShader Shader 0x207d2730 recompiled more than 50 times
fixme:d3d_shader:IWineD3DVertexShaderImpl_CompileShader Vertexshader 0x81d4470 recompiled more than 50 times
fixme:d3d_shader:IWineD3DVertexShaderImpl_CompileShader Vertexshader 0x81d4470 recompiled more than 50 times

I can confirm. 
Same on my machine.
OpenGL renderer string: GeForce 8800 GTS/PCI/SSE2
OpenGL version string: 2.1.2 NVIDIA 177.80
Ubuntu 8.10 Intrepid Ibex

This bug also occurs for me. I believe there is an opening video which I was able to skip (I think), however, the graphical artifacts and distortion still remain seen, and the game is completely unplayable. Note: There seems to be a green square in the bottom right hand corner that never seems to leave, even if you skip the cutscene. This is a huge letdown, seeing how the other "Next-Generation" Tomb Raider games (Legend and Anniversary) ran excellently. These new problems are probably due to the complete overhaul that Crystal Dynamics/Eidos performed on the engine of the game. It would be awesome to see the demo running before the full release of the game, and knowing the wine devs, I'm 100% sure this is possible.

I checked the gpu"s temperature during the game.
It seems there is no any GPU rendering during the game because my gpu's temp stays on 56 degrees. It is very strange.
I use Nvidia 8600M GT with the latest beta driver (177.61.02)

(In reply to comment #4)
> I checked the gpu"s temperature during the game.
> It seems there is no any GPU rendering during the game because my gpu's temp
> stays on 56 degrees. It is very strange.
> I use Nvidia 8600M GT with the latest beta driver (177.61.02)
> 

Nice discovery. By the way, when you try to run the game, does it slow down your computer at all? Because when I try running the demo, it makes everything extremely slow, although this could be due to the fact that I have a single-core computer.

Yes, after loading the demo, my cpus are on 100%.
I simply use ALT+F4 :-D
Can both of you create regression test for TRU? Cause I cant, and developers need this thing to debug the problem and solve it.

*** Bug 15880 has been marked as a duplicate of this bug. ***

From the attached log, I noticed that the 2 first big errors are :

> fixme:d3d:debug_d3dformat Unrecognized 1515474505 (as fourcc: INTZ) WINED3DFORMAT!
> err:d3d:CheckTextureCapability Unhandled format=unrecognized
> fixme:d3d:debug_d3dformat Unrecognized 1515667794 (as fourcc: RAWZ) WINED3DFORMAT!
> err:d3d:CheckTextureCapability Unhandled format=unrecognized

Video card : GeForce 7600 GT
Drivers : 173.14.09

This bug may be related with bug #12816, since they seem to have the same kinds of errors. Thoughts ?

with 1.1.8

wine: Unhandled page fault on read access to 0x00000004 at address 0x455010 (thread 0014), starting debugger...
Unhandled exception: page fault on read access to 0x00000004 in 32-bit code (0x00455010).
Register dump:
 CS:0023 SS:002b DS:002b ES:002b FS:0063 GS:006b
 EIP:00455010 ESP:7ebfb6cc EBP:00455000 EFLAGS:00010206(   - 00      - RIP1)
 EAX:00000000 EBX:7ee91ff4 ECX:7ebfb964 EDX:00000000
 ESI:004500c8 EDI:00451ee0
Stack dump:
0x7ebfb6cc:  00451ee0 004500c8 7ebfb988 7ebfb6ec
0x7ebfb6dc:  7ee91ff4 00000000 7ebfb964 7ee92500
0x7ebfb6ec:  7ee8ff37 7ee92500 7ebfb964 001140d0
0x7ebfb6fc:  00000000 00114070 7ebfb974 7ebfb7bc
0x7ebfb70c:  f7e76c61 00000000 7ee92500 00340014
0x7ebfb71c:  7b8b6ff4 00010000 7ebfb96c 7ebfb964
Backtrace:
=>1 0x00455010 in glaide32.sys (+0x5010) (0x00455000)
  2 0x835d0000 (0x0000e860)
  3 0x00000000 (0x00000000)
0x00455010: movl	0x4(%eax),%eax
Modules:
Module	Address			Debug info	Name (29 modules)
PE	  450000-  45f180	Export          glaide32.sys
ELF	7b800000-7b93e000	Deferred        kernel32<elf>
  \-PE	7b820000-7b93e000	\               kernel32
ELF	7bc00000-7bcab000	Deferred        ntdll<elf>
  \-PE	7bc10000-7bcab000	\               ntdll
ELF	7bf00000-7bf04000	Deferred        <wine-loader>
ELF	7ebfd000-7ec10000	Deferred        libresolv.so.2
ELF	7ec1e000-7ec34000	Deferred        hal<elf>
  \-PE	7ec20000-7ec34000	\               hal
ELF	7ec34000-7ec54000	Deferred        iphlpapi<elf>
  \-PE	7ec40000-7ec54000	\               iphlpapi
ELF	7ec54000-7ecbb000	Deferred        rpcrt4<elf>
  \-PE	7ec60000-7ecbb000	\               rpcrt4
ELF	7edcc000-7ee06000	Deferred        ntoskrnl<elf>
  \-PE	7edd0000-7ee06000	\               ntoskrnl
ELF	7ee06000-7ee5b000	Deferred        advapi32<elf>
  \-PE	7ee10000-7ee5b000	\               advapi32
ELF	7ee5b000-7ee66000	Deferred        libnss_files.so.2
ELF	7ee66000-7ee6f000	Deferred        libnss_compat.so.2
ELF	7ee7e000-7ee93000	Deferred        winedevice<elf>
  \-PE	7ee80000-7ee93000	\               winedevice
ELF	7efb6000-7efdc000	Deferred        libm.so.6
ELF	7efdf000-7efe9000	Deferred        libnss_nis.so.2
ELF	7efe9000-7f000000	Deferred        libnsl.so.1
ELF	f7d25000-f7d29000	Deferred        libdl.so.2
ELF	f7d29000-f7e6c000	Deferred        libc.so.6
ELF	f7e6c000-f7e84000	Deferred        libpthread.so.0
ELF	f7ea8000-f7fdf000	Deferred        libwine.so.1
ELF	f7fe0000-f7ffd000	Deferred        ld-linux.so.2
Threads:
process  tid      prio (all id:s are in hex)
00000008 
	00000009    0
0000000a 
	0000000b    0
0000000c 
	00000015    0
	00000013    0
	00000012    0
	0000000e    0
	0000000d    0
0000000f (D) C:\windows\system32\winedevice.exe
	00000014    0 <==
	00000011    0
	00000010    0
Backtrace:
=>1 0x00455010 in glaide32.sys (+0x5010) (0x00455000)
  2 0x835d0000 (0x0000e860)
  3 0x00000000 (0x00000000)
ALSA lib pcm.c:2104:(snd_pcm_open_conf) Cannot open shared library /usr/lib/alsa-lib/libasound_module_pcm_pulse.so
ALSA lib pcm.c:2104:(snd_pcm_open_conf) Cannot open shared library /usr/lib/alsa-lib/libasound_module_pcm_pulse.so
fixme:ntdll:NtQueryInformationProcess (0xffffffff,info_class=34,0x20baa0c,0x00000004,0x20baa08) Unknown information class
fixme:ntdll:NtQuerySystemInformation info_class SYSTEM_HANDLE_INFORMATION
fixme:ntdll:NtQueryObject Unsupported information class 3
fixme:debugstr:CheckRemoteDebuggerPresent (0xffffffff)->(0x20b97f8): Stub!
fixme:ntdll:NtQueryInformationProcess (process=0xffffffff) Unimplemented information class: ProcessDebugFlags
fixme:d3d:IWineD3DImpl_FillGLCaps OpenGL implementation supports 32 vertex samplers and 32 total samplers
fixme:d3d:IWineD3DImpl_FillGLCaps Expected vertex samplers + MAX_TEXTURES(=8) > combined_samplers
fixme:win:EnumDisplayDevicesW ((null),0,0x20b9744,0x00000000), stub!
fixme:ntdll:server_ioctl_file Unsupported ioctl 24000 (device=2 access=1 func=0 method=0)
fixme:psapi:EnumPageFilesA (0x1498780, 0x2098adc) stub
fixme:psapi:EnumPageFilesA (0x1498780, 0x20633f8) stub
fixme:ntdll:NtQueryInformationProcess (process=0xffffffff) Unimplemented information class: ProcessDeviceMap
fixme:ntdll:NtQuerySystemInformation info_class SYSTEM_HANDLE_INFORMATION
fixme:ntdll:NtQueryObject Unsupported information class 3
fixme:debugstr:CheckRemoteDebuggerPresent (0xffffffff)->(0x20cf894): Stub!
err:file:GetLongPathNameW UNC pathname L"\\\\.\\pipe\\svcctl"
fixme:ntdll:NtQueryInformationProcess (process=0xffffffff) Unimplemented information class: ProcessDebugFlags
fixme:ole:CoInitializeSecurity ((nil),-1,(nil),(nil),0,3,(nil),0,(nil)) - stub!
err:ole:CoGetClassObject class {9a5ea990-3034-4d6f-9128-01f3c61022bc} not registered
err:ole:CoGetClassObject class {9a5ea990-3034-4d6f-9128-01f3c61022bc} not registered
err:ole:create_server class {9a5ea990-3034-4d6f-9128-01f3c61022bc} not registered
fixme:ole:CoGetClassObject CLSCTX_REMOTE_SERVER not supported
err:ole:CoGetClassObject no class object {9a5ea990-3034-4d6f-9128-01f3c61022bc} could be created for context 0x17
fixme:d3d:debug_d3dformat Unrecognized 1515474505 (as fourcc: INTZ) WINED3DFORMAT!
err:d3d:CheckTextureCapability Unhandled format=unrecognized
fixme:d3d:debug_d3dformat Unrecognized 1515667794 (as fourcc: RAWZ) WINED3DFORMAT!
err:d3d:CheckTextureCapability Unhandled format=unrecognized
wine: Unhandled page fault on execute access to 0x02660035 at address 0x2660035 (thread 0009), starting debugger...

I've run a regression test and submitted a new bug (bug 15955) for the crash. So we have to wait until bug 15955 is fixed.

(In reply to comment #10)
> I've run a regression test and submitted a new bug (bug 15955) for the crash.
> So we have to wait until bug 15955 is fixed.
> 

I just upgraded to 1.1.8, wine doesn't crash for me when the demo start up.

Created attachment 17372 [details]
tru terminal output errors

 Using the latest snapshot as of 11-19-2008 , let's call it 1.1.8b , Underworld renders in raw blocks and triangles of off-grey for the world textures and blue for the water. On loading the level, I get the neon green to orange box in the lower right hand side but it disappears after a minute or so. Oddly enough I do get a water sprites(?) when splashing and the menus run perfectly. No character models are rendered. This is with all settings/Effects(FS Effects,Shadows,etc) to Off and textures on Low. 
attaching a small file of terminal output.

OpenGL version string: 2.1.2 NVIDIA 177.82
OpenGL shading language version string: 1.20 NVIDIA via Cg compiler

Created attachment 17389 [details]
dirty hack which disables the vertexshader recompilation

This hack disables the vertexshader recompilation (Probably we could use this extension http://www.opengl.org/registry/specs/EXT/vertex_array_bgra.txt for doing this for us. But I have no glue where to put it in.). Now the game speed is fine. 

But there are several other glitches like bug 9717 (Using fbo makes the screen grey for me).

Off course when you jump into the water the game looks fine.

But there are tree parts off the screen ...the upper half, the bottom left and the bottom right.

I tried with NVIDIA 180.08 - geforce 8800GTS, kernel 2.6.27.5 on fedora9, ORM=backbuffer

It is very interesting, i use 1.1.9, nvidia 177.82, and the game's rendering is the same gray field as before. I tried it with fbo, pbo, bbo too.
:-(((

I also receive this message thousends time:

fixme:d3d_shader:IWineD3DPixelShaderImpl_CompileShader Shader 0x207d2730
recompiled more than 50 times

I also receive this message thousends time:

fixme:d3d_shader:IWineD3DPixelShaderImpl_CompileShader Shader 0x207d2730
recompiled more than 50 times

fixme:d3d_shader:shader_glsl_color_correction Add color fixup for vertex texture WINED3DSIO_TEXLDL

fixme:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_OUT_OF_MEMORY (0x505) from glDrawElements @ drawprim.c / 266

fixme:d3d:context_check_fbo_status FBO status GL_FRAMEBUFFER_UNSUPPORTED_EXT (0x8cdd)

fixme:d3d:context_check_fbo_status 	Color attachment 0: (0x1645e8) WINED3DFMT_A8R8G8B8 4096x3072
fixme:d3d:context_check_fbo_status 	Depth attachment: (0x164c80) WINED3DFMT_D16 4096x3072
fixme:d3d:IWineD3DDeviceImpl_ClearSurface >>>>>>>>>>>>>>>>> GL_INVALID_FRAMEBUFFER_OPERATION_EXT (0x506) from glClear @ device.c / 5129
fixme:d3d_draw:drawStridedFast >>>>>>>>>>>>>>>>> GL_OUT_OF_MEMORY (0x505) from glDrawElements @ drawprim.c / 266

(In reply to comment #14)
> Created an attachment (id=17389) [details]
> dirty hack which disables the vertexshader recompilation
> 
Using 1.19 with the patch improves 'performance' but polygonal edges are still projected off into infinity and textures are absent except for water effects(splashing,etc). 
In the registry under \hkeycu\software\crystaldynamics\tombraider\graphics\AdapterIdentifier is {00000000-0000-0000-0000-000000000000} while in TR:Legend it's {AEB2CDD4-6E41-43EA-941C-8361CC760781}. Just found it interesting. It reverts back to 0's if you set it differently. The DisableOptions (32bit textures,etc) have limited effect or cause more problems/instability.

Created attachment 17463 [details]
tru render bug screenshots

Created attachment 17464 [details]
tru render bug screenshots

Created attachment 17465 [details]
tru render bug screenshots

I can confirm foregoing buggy rendering screenshots under wine1.0.1 - and nv178.x driver. As mentioned elsewhere, TRU employs shader 3 model (according to Eidos), whilst TRL and TRA employes shader 2 model. However, I do not know whether this has implications on Wine.

Created attachment 17485 [details]
Picture with ORM=backbuffer

This is a picture with OffscreenRenderingMode = backbuffer. There is the problem with the "shadow on the upper half of the screen" visible. If I am not out of the water the hole upper half of the screen is grey.

(In reply to comment #22)
> I can confirm foregoing buggy rendering screenshots under wine1.0.1 - and
> nv178.x driver. As mentioned elsewhere, TRU employs shader 3 model (according
> to Eidos), whilst TRL and TRA employes shader 2 model. However, I do not know
> whether this has implications on Wine.
> 

In my copy of Legend UseShader30 is set to true(1) and UseShader20 is Off(0) ,without any problems, in the registry. Or maybe it just ignores such settings? I tried adding those DWORDS to the TRU reg w/o success.

(In reply to comment #24)
> 
> 
> In my copy of Legend UseShader30 is set to true(1) and UseShader20 is Off(0)
> ,without any problems, in the registry. Or maybe it just ignores such settings?
> I tried adding those DWORDS to the TRU reg w/o success.
> 
You are right. Both TRL and TRA use those DWORDS in the reg with values you report. You can , however, set both dwords to zero or one , or set UseShader20=1
and UseShader30=0 w/o impact - for me visible at least. From what I read from the tombraider.com forums a cup of windozzed users are pretty angry because they cannot use their 'old' cards , because of missing shader 2 support in TRU. Those 'old' cards were, however, working for them in TRA & TRL. 

(In reply to comment #25)
> (In reply to comment #24)
> > 
> > 
> From what I read from
> the tombraider.com forums a cup of windozzed users are pretty angry because
> they cannot use their 'old' cards , because of missing shader 2 support in TRU.
> Those 'old' cards were, however, working for them in TRA & TRL. 
> 

They might as well break something if it isn't broken , right? :) I vote for Rico's idea in Comment #14.

*** Bug 16717 has been marked as a duplicate of this bug. ***

*** Bug 16719 has been marked as a duplicate of this bug. ***

I have wine 1.1.12

I intalled new nvidia 180.12

Tomb RAider Anniversay - AMAZNING RENDERING SPEED
Tomb RAider LEgend - with and without NEXTGENCONT AMAZING RENDERING SPEED!!!!
Tomb Raider Underworld - stil unuseable

will it be ever fixed?
Am wait from December, am on the first level, and lara is always burned. :-)))
Because I cant see anything! :-D

I can help with d3d rendering, but i dont know who works with this bug, witch developer of wine.

The speed problem is fixed in current git.

If you have some time, you could search the source of the grey wall (probably something todo with fog?) which takes all the screen here. Only the osd is shown (ORM=fbo).

(In reply to comment #23)
> Created an attachment (id=17485) [details]
> Picture with ORM=backbuffer
> 
> This is a picture with OffscreenRenderingMode = backbuffer. There is the
> problem with the "shadow on the upper half of the screen" visible. If I am not
> out of the water the hole upper half of the screen is grey.
> 

Strange. No matter what I set as OffscreenRenderingMode (backbuffer or fbo) I get a total grey screen and the game crashes after a few seconds.
Menu looks fine but after the mission is loaded screen turns all grey and it crashes after some seconds.
I'm using wine-1.1.13 and an Nvidia 8600GTS with the 180.11.02 driver.

(In reply to comment #31)
> (In reply to comment #23)
> > Created an attachment (id=17485) [details] [details]
> > Picture with ORM=backbuffer
> > 
> > This is a picture with OffscreenRenderingMode = backbuffer. There is the
> > problem with the "shadow on the upper half of the screen" visible. If I am not
> > out of the water the hole upper half of the screen is grey.
> > 
> 
> Strange. No matter what I set as OffscreenRenderingMode (backbuffer or fbo) I
> get a total grey screen and the game crashes after a few seconds.
> Menu looks fine but after the mission is loaded screen turns all grey and it
> crashes after some seconds.
> I'm using wine-1.1.13 and an Nvidia 8600GTS with the 180.11.02 driver.
> 

Yes, but I found something!

Start a new game. Press backward (S) when its loaded. You hit the wall, the game is accelerated!!!! Move you mouse, its sooo fast on that point!!!!!
Strange!

Created attachment 18905 [details]
hack for geforce8 on x86_64

@Florian
You could try this patch, it's an updated version of the one for Gothic3. It should let you run Tombraider without an early crash. At least it does it for me (driver 180.18).

(In reply to comment #33)
> Created an attachment (id=18905) [details]
> hack for geforce8 on x86_64
> 
> @Florian
> You could try this patch, it's an updated version of the one for Gothic3. It
> should let you run Tombraider without an early crash. At least it does it for
> me (driver 180.18).
> 

woooooooooooooooooooooooooooooooooooooooooooooo

with NVIDIA 182.25 driver, the rendering speed is amazing faaaaaaaaaaaaaaaaaaaaassssssssssssttttttttttttttttttt!!!!!!!!!!!!!!!

We are sooo close!!!!!!!!!!!!!!!!

:-)))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))

(In reply to comment #34)
> (In reply to comment #33)
> > Created an attachment (id=18905) [details] [details]
> > hack for geforce8 on x86_64
> > 
> > @Florian
> > You could try this patch, it's an updated version of the one for Gothic3. It
> > should let you run Tombraider without an early crash. At least it does it for
> > me (driver 180.18).
> > 
> 
> woooooooooooooooooooooooooooooooooooooooooooooo
> 
> with NVIDIA 182.25 driver, the rendering speed is amazing
> faaaaaaaaaaaaaaaaaaaaassssssssssssttttttttttttttttttt!!!!!!!!!!!!!!!
> 
> We are sooo close!!!!!!!!!!!!!!!!
> 
> :-)))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))
> 

sorry!!!
NVIDIA Linux driver 180.25

http://www.nvnews.net/vbulletin/showthread.php?p=1914053

I use fbo, sometimes the whole screen is orange, but sometimes its ok! :-)

Created attachment 19023 [details]
orm=backbuffer

(In reply to comment #33)
> Created an attachment (id=18905) [details]
> hack for geforce8 on x86_64
> 
> @Florian
> You could try this patch, it's an updated version of the one for Gothic3. It
> should let you run Tombraider without an early crash. At least it does it for
> me (driver 180.18).
> 

The patch fixes the crashes, indeed. Thank you for that.
If I use backbuffer as ORM it's like in the attached picture. If I use fbo the screen is just grey. If I dive (at least I assume I did) the screen turns into some kind of turquoise.

From what I experienced with backbuffer the performance seemed to be okay, though I did not really test it.

I attached photos:

1. the strange orange filter that hides textures
2. some almost perfect rendeing (the speed is faster than under Vista shi*t!!!!)

I never use backbuffer or pixcelbuffer because you will get an unreal rendering. only framebuffer can provide the real rendering. (I meanx textures and effects etc...)

With Nvidia boosted driver the game is still unplayable because of that orange thing....??????????????? In demo version at Thailand level it is grey. Should I use a rubber or sg? :-)

Is it silly to ask Nvidia and Eidos about this texture rendering problem?
Cause no one cares about it here, and this strange rendering thing is existed from 2006 when Tomb Raider Legend with PS3.0 appeared!

Created attachment 19025 [details]
almost_perfect_1.jpg

Created attachment 19026 [details]
almost_perect_2.jpg

Created attachment 19027 [details]
almost_perfect_3.jpg

Created attachment 19028 [details]
strange_area_like_a_filter_hides_textures.jpg

Looks like depth buffer precision issues. I probably won't have time to look into this soon. The improved speed is probably due to EXT_vertex_array_bgra support in the newer drivers, which should prevent the vertex shader recompilation.

Actually, it's probably not so much about precision as about the depth buffer still having old data.

(In reply to comment #43)
> Actually, it's probably not so much about precision as about the depth buffer
> still having old data.
> 

thanks!

What can I do? I would like to help if you dont have time. I dont know what is old data az depth buffer. Is it in wine or in nvidia driver?

What should i check?

:-)

I'm pretty sure the issue is in Wine. Does the log contain any FIXMEs from surface_load_ds_location()? Typically you'd start with looking at how the frame is rendered (ie, dumping the backbuffer and depth buffer for each subframe) and then matching that against the log output.

(In reply to comment #45)
> I'm pretty sure the issue is in Wine. Does the log contain any FIXMEs from
> surface_load_ds_location()? Typically you'd start with looking at how the frame
> is rendered (ie, dumping the backbuffer and depth buffer for each subframe) and
> then matching that against the log output.
> 

Am at work, after work, i will check this and will report here! Thanks your help! :)

(In reply to comment #45)
> I'm pretty sure the issue is in Wine. Does the log contain any FIXMEs from
> surface_load_ds_location()? Typically you'd start with looking at how the frame
> is rendered (ie, dumping the backbuffer and depth buffer for each subframe) and
> then matching that against the log output.
> 

Another (maybe) interesting thning:

You can set at setup Tripple buffering
when its on, you can see the game textures sometimes (i attached the screenshost via tripplebuffering)

Now, i turned it off, in this case you only can see the orange or gray screen!


To find that youve explained about surface_load_ds_location(), should i run this command:

WINEDEBUG=d3d wine tru.exe ???

Here is a videofile about this problem:

http://www.youtube.com/watch?v=fhIZElEWz68


When level is loading, the spinning ring is very slow, you recieve this fixme message:
fixme:d3d_shader:IWineD3DPixelShaderImpl_CompileShader Shader 0x207d2730
recompiled more than 50 times

You can fix this by modify registry:

HKCU\Software\Crytal Dynamics\TombRaider: Underworld\Graphics

"DontDeferShaderCreation" "0"


I dont know what is.

Same issue here (similar results) with lots of
--
err:d3d:getColorBits Unsupported format: WINED3DFMT_R16G16_FLOAT
err:d3d:IWineD3DImpl_IsPixelFormatCompatibleWithRenderFmt Unable to check compatibility for Format=WINED3DFMT_R16G16_FLOAT
---

messages.

Using the 185.19 NV drivers on a GTX 285.
wine 1.1.18

Still present in wine-1.1.29

"DontDeferShaderCreation" "0"
This does really speedup the spinning of the ring while game loading. But still my display is all orange. Have tried all in-game options to turn on/off. Nothing helps. :(

And Anti-Aliasing is off here and cannot be set on. :(

I have made some progress:

You need to press W/S or some other movement keys and try to rotate the view with your mouse. Then you will see some rendered pictures and all over organge (new game, of course). While you rotate the view around Lara, the camera sometimes when away from her or closes up to her.

Once the camera is a bit more far away, the screen turns orange all over. Once the camera closes up to Lara, you see her.

But still with 'flickering' orange polygons in all textures.

Some typos in last comment, here is the corrected:

--------------------
I have made some progress:

You need to press W/S or some other movement keys and try to rotate the view with your mouse. Then you will see some rendered pictures and not all over orange (you need start a new game, of course). While you rotate the view around Lara, the  camera sometimes went away from her or closes up to her.

Once the camera is a bit more far away, the screen turns orange all over the screen. When the camera closes up to Lara, you see her. But still you have 'flickering' orange polygons in all textures.
--------------------
Sorry, it is 2:00am here... :/

Orange rendering: I noticed that when ever I switch which ever display\ grapchic detail setting then the orange disappears and screen is rendered correctly until camera moves or Lara moves.

Hi,

I came across the GTA4 bugzilla entry http://bugs.winehq.org/show_bug.cgi?id=16328. GTA4 had (or still has) problems because of missing pixelformat called INTZ. There is a patch attached which fixes this (http://bugs2.winehq.org/attachment.cgi?id=26516), and as it turns out, it also fixes the orange/grey screen filler bug.

After applying that patch onto today's git the game is mostly playable (C2D@3Ghz, 8800GT, 1920x1200). There are still a few places where, depending on the camera angle, every 3d object disappears or 3d objects leave smudges on the background.

1 install wine 1.1.42
2.copy the patched files (wine_1.1.42_patched.7z) to /usr/lib/wine
3. start the game :-)

http://www.youtube.com/watch?v=zGs-54s_HI0

Totally playable game! Finally! :-)))))))))))))

There is an attached patch for it (tru.patch) I do hope this will be included in the next wine version.

*BUG******************************************************
I have found a bug, with mmdevapi.dll.so, in 1.1.42 there is no sfx during the gameplay. I use the 1.1.35 version of this file, that was the last point when there were sfx under TRU. How can I report this bug?
**********************************************************

Created attachment 27207 [details]
the patch file

patched binary files

http://www.4shared.com/file/256995773/1fe5a1bb/wine_1142_patched.html

(In reply to comment #55)

Thank you!

Hi Lazlo,
please create a new bug report in bugzilla for the 
regression that your patch fixes.

(In reply to comment #56)
> 1 install wine 1.1.42
> 2.copy the patched files (wine_1.1.42_patched.7z) to /usr/lib/wine
> 3. start the game :-)
> 
> Totally playable game! Finally! :-)))))))))))))

Does not work at all for me. I have always à lot of GL_INVALID_OPERATION and shader errors coming from d3d_shader. My screen is mostly gray with some black areas when I move (like screenshots from comments #19 #20 #21 in fact).
I'm using the demo from URL with a GeForce 7600 GT and nvidia drivers 190.53.
Wine 1.1.42 is built with the patch from comment #55.

(In reply to comment #56)

Works for me. Thank you guys!

gentoo 2.6.33
nvidia 195.36.03 9800GT
wine 1.1.38 
  + tru.patch
  + 1.1.35's mmdevapi.dll.so (for sounds)

(In reply to comment #57)
> Created an attachment (id=27207) [details]
> the patch file

Ok. I applied this patch now. When I set the ORM to backbuffer the game will run smoothly but I have a big grey square in the left bottom corner of the screen. which makes it unplayable, though the performance looks good.
With OffscreenRenderingMode set to fbo the game runs very slow and the picture is kind of distorted.

ArchLinux 64bit
AMD II X3 435
Nvidia gts250 (195.36.15 driver)
4GB Ram

I will try with an ATI System later and I can add Screenshots if necessary.

(In reply to comment #63)
> (In reply to comment #57)
> > Created an attachment (id=27207) [details] [details]
> > the patch file
> 
> Ok. I applied this patch now. When I set the ORM to backbuffer the game will
> run smoothly but I have a big grey square in the left bottom corner of the
> screen. which makes it unplayable, though the performance looks good

I did some further testings on this by disabling all graphic settings.
The square is caused by the option "volumetric effects". Not sure if it's called like this in English.
It's listed above Shadows.
Tested on the system mentioned above with patched wine-1.1.44

With wine-1.2-rc1/2 (updated today) the sound is only available if you stay in menu and did not load a save-game or start a new one. So it is still an issue (I have retail version here). And I think you have the same with the demo... Can someone with the demo try it?

Created attachment 29915 [details]
updated patch for latest GIT

I have updated the patch in attachment http://bugs.winehq.org/attachment.cgi?id=27207 to fit with latest wine GIT (updated some minutes ago). There was a failed hunk which I was able to fix (by putting the code from the rej-file in the right place).

(In reply to comment #45)
> I'm pretty sure the issue is in Wine. Does the log contain any FIXMEs from
> surface_load_ds_location()? Typically you'd start with looking at how the frame
> is rendered (ie, dumping the backbuffer and depth buffer for each subframe) and
> then matching that against the log output.

There are a lot of these and the fire (when you start a new game) is not displayed correctly.

So how do I do this? I'm no programmer in C/C++ but I like to help.

I got this:
fixme:d3d_surface:surface_load_ds_location No up to date depth stencil location.

This screenshot was taken from the retail (starting map) just before I applied the updated attached patch:
http://www.mxchange.org/downloads/wine/tru/screenshot-20100730_0427.png

This screenshot was taken after the patch was applied:
http://www.mxchange.org/downloads/wine/tru/screenshot-20100730_0454.png

As you can see, the fire I was talking about vanished, because the fire fighters did a good and really quick job? ;-) No, when I move Lara closer to the (still existing but invisible) fire, and turned around so I can see her face, the face got illuminated by the fire.

I think the vanishing sound after starting up a game is related to this line:
-------------------
err:mmdevapi:MMDevEnum_Create OpenAL support not compiled in, application will not find sound devices
-------------------
I have Debian Unstable/Squeeze AMD64 and the configure script told me that there are no 32-bit libraries, still Debian has not released a lib32openal (or so?) package for Unstable. :(

I found a workaround for the missing sound:

I have downloaded the 32-bit version of libopenal and libopenal-dev, extracted both manually, put all from /usr/lib into /usr/lib32, set a symbolic link from libopenal.so.1 to libopenal.so (very important step!) in /usr/lib32, edited the package information file openal.pkg (libdir=/usr/lib32) and configure was no longer complaining about missing libopenal 32-bit development files. :-)

So let's hope this works...

Created attachment 30439 [details]
updated patch for latest GIT

Created attachment 30489 [details]
updated patch for latest GIT

The previous patch included modification in configure from recent automake tools.

Finally TOMB RAIDER UNDERWOLRD works almost perfectly!

1. on the first level, the fire is unvisible!

2. audio driver is crackling: change current  mmdevapi.dll.so to 1.1.43 wine one

Fire is still not visible in wine-1.3.9
Later in the croft manor this issue makes the game almost impossible to be completed. You have to make some shadows match the silhouettes of two angels. As the fire is not displayed there are also no shadows.

Hi, 
since this is my first thread here...thanks a lot for this great work. 
Im so glad that wine and opengl porting games to linux where it actually belongs to, because linux is fun! 

game is playable here also, but have 2 issues:

when diving in first level there is no background, just grey...
I know from winxp that there is background.
see the attached image

sound is not working on videos...

my config:
- gentoo x86_64 kernel 2.6.36-gentoo-r5
- nvidia 260.19.29
- xorg-server 1.9.4
- wine 1.3.14 unpatched
- TRU with Patch v1.1

the err messages:
err:d3d_surface:surface_set_texture_name Surface has SFLAG_INTEXTURE set, but no texture name.
err:d3d_surface:surface_modify_location Surface 0x215f08 does not have any up to date location.
fixme:d3d_surface:surface_load_ds_location No up to date depth stencil location.
err:d3d_surface:surface_modify_location Surface 0x215f08 does not have any up to date location.
fixme:d3d_surface:surface_load_ds_location No up to date depth stencil location.
...

nearly all the time i get: 
err:mmdevapi:ACR_ReleaseBuffer Starting from 1014

Is there anything I can try?

Created attachment 33507 [details]
first level- no background, only playground

Does winetricks d3d9 help?  I don't expect it to, but...

> Does winetricks d3d9 help?  I don't expect it to, but...
no it didn't. 

i rechecked at xp again, it looks like a sort of cloudiness is somehow missing.
Thats why I see the background under water.
Also there where sun beams at xp when she is diving near to the surface...
I like the realistic scenery of the game. Hm...
next thing I will try is to go backwards in wine revisions.

Did anyone see the cloudiness and sunbeams under water yet...?

> Did anyone see the cloudiness and sunbeams under water yet...?
As far as I recall my playing experiences with TRU, I cannot recall them. Maybe you can make a screenshot from the game under your XP?

Then it can be easily compared.

Created attachment 33585 [details]
diff backgrounds left xp, right wine

good idea...

This bug appears to affect Deus Ex Human Revolution (beta) as well. Game is filled with orange when graphical settings are turned up (like "strange area" attachment), and looks like the "almost perfect 1" attachment when graphical settings are turned down.

Created attachment 35699 [details]
Output during failure to initialize Direct3D

Game error box: "Failed to initialize Direct3D with these settings."

TRU, full version. Happens with or without patch, and no matter what settings I choose in the config program.

Wine 1.3.24, Ubuntu Lucid, Geforce 6200, driver 275.21.

The game runs correctly under Win XP on the same machine.

I assume the game doesn't recognize Shader Model 3.0, although my card supports it.

(In reply to comment #82)
> Game error box: "Failed to initialize Direct3D with these settings."
I get this when disabling -support for vertex shader- or deactivate -pixel shader- in winecfg.

also compare:
[HKEY_USERS\S-1-5-21-0-0-0-1000\Software\Wine\Direct3D]
"DirectDrawRenderer"="opengl"
"Multisampling"="enabled"
"OffscreenRenderingMode"="fbo"
"PixelShaderMode"="enabled"
"VertexShaderMode"="hardware"
"VideoMemorySize"="1024"


The game runs here with 1.3.24 like before...
I've the impression that the detail level of the ground under water has increased a bit, but the ugly background is still there.

(In reply to comment #83)

I have all these options set, except I only have 512 MB video ram.

Created attachment 35708 [details]
tru getting through setup without glsl, then crashing

I noticed when I set UseGLSL to disabled, the setup program accepts my settings and apparently the game tries to start. However, I get an unhandled page fault then instead.

I see a %eax in the log, can it have to do with sound? I use OSS 4, not alsa. Fmod problem?

I'd like to reflect on comment #76. That visual glitch is still present in Wine-1.3.28, but only when anti-aliasing is turned off.
Now Wine supports multisampling, and I noticed that water appears correctly when anti-aliasing is enabled in the game.

Nvidia 250 / driver 280.13

(In reply to comment #86)
> I'd like to reflect on comment #76. That visual glitch is still present in
> Wine-1.3.28, but only when anti-aliasing is turned off.
> Now Wine supports multisampling, and I noticed that water appears correctly
> when anti-aliasing is enabled in the game.
> 
If that's the case, setting "AlwaysOffscreen" to "enabled" may also work as a workaround.

(In reply to comment #87)
> (In reply to comment #86)
> > I'd like to reflect on comment #76. That visual glitch is still present in
> > Wine-1.3.28, but only when anti-aliasing is turned off.
> > Now Wine supports multisampling, and I noticed that water appears correctly
> > when anti-aliasing is enabled in the game.
> > 
> If that's the case, setting "AlwaysOffscreen" to "enabled" may also work as a
> workaround.

Yes, with 'AlwaysOffscreen' set to 'enabled', water looks fine to me when Lara is underwater, whether AA is turned on or off.

confirmed, 
with wine 1.3.27 here, I had to simply switch on AA and it worked...
even after removing "AlwaysOffscreen" "enabled" again.

I remember playing around earlier with AA settings and it never fixed anything.
But now the game runs really fine. I see the sunbeams and all the beauty cloudiness in the background...and this in a nice wine environment! 

Also the sound in Videos is working fine now!

So, ThankS guys from my side...really great work! 
I did'nt expect a solution anymore...but now I can start to go on with the game...

and for me the best thing is...

Today I can remove this Windows Xp shit from my computer again!!!


config:
- gentoo x86_64 2.6.39-gentoo-r5
- nvidia 270.41.19
- xorg-server 1.10.4
- wine 1.3.27 unpatched
- TRU with Patch v1.1

The problem with the "orange" plain color instead of polygon textures still exist with Wine 1.5.5 - but it definitely seems to be a bug in the graphics driver and not wine, because Underworld runs perfectly on a 64 Bit Ubuntu 12.04 Wine 1.5.5 computer with a nVidia GT430 graphicsa card (and driver 295.40). BUt these graphics problems show on a 64 Ubuntu 12.04 Wine 1.5.5 computer with Radeon HD 6550D graphics (and driver Catalyst 11.8)! I tried most graphics settings (all on, all off, etc.) but the missing texture problem persists.

Reported FIXED by the last comment.

Downloaded demo from http://www.fileplanet.com/193841/190000/fileinfo/Tomb-Raider:-Underworld-Demo

works without any troubles on 1.7.6

Confirmed fixed by 3 different users.

Closing bugs fixed in 1.7.7.