[Bug 38216] New: Wrong screen in Civilization V with new engines. Good with old 1.7.18.

https://bugs.winehq.org/show_bug.cgi?id=38216

Bug ID: 38216
           Summary: Wrong screen in Civilization V with new engines. Good
                    with old 1.7.18.
           Product: Wine
           Version: 1.7.38
          Hardware: x86
                OS: Mac OS X
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: directx-d3d
          Assignee: [e-mail]
          Reporter: [e-mail]

Created attachment 51009
  --> https://bugs.winehq.org/attachment.cgi?id=51009
Screen with engine 1.7.36

I got Civilization V working with engines up to 1.7.18.
Screenshot attached.
But with new engines 1.7.36, 1.7.38 initial screen is broken.
Screenshot attached too.
Log differences are in those lines:
1.7.36
fixme:d3dx:D3DXLoadSurfaceFromMemory Format conversion missing 0x33545844 ->
0x33545844
fixme:d3dx:D3DXLoadSurfaceFromMemory Format conversion missing 0x35545844 ->
0x35545844
fixme:d3dx:D3DXLoadSurfaceFromMemory Format conversion missing 0x35545844 ->
0x35545844
while not in 1.7.18

[Bug 38216] Wrong screen in Civilization V with new engines. Good with old 1.7.18.

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #1 from Sergey  ---
Created attachment 51010
  --> https://bugs.winehq.org/attachment.cgi?id=51010
Screen with engine 1.7.2

[Bug 38216] Wrong screen in Civilization V with new engines. Good with old 1.7.18.

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #2 from Sergey  ---
Hi,

I narrowed the search.
Good screen up to engine 1.7.23
Bad screen with all engines >= 1.7.24
With engine 1.7.38 no image at all. Just contours.

Why there is global regression? Is there any good news?

[Bug 38216] Wrong screen in Civilization V with new engines. Good with old 1.7.23.

https://bugs.winehq.org/show_bug.cgi?id=38216

Sergey  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Wrong screen in             |Wrong screen in
                   |Civilization V with new     |Civilization V with new
                   |engines. Good with old      |engines. Good with old
                   |1.7.18.                     |1.7.23.

[Bug 38216] Wrong screen in Civilization V with new engines. Good with old 1.7.23.

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #3 from Sergey  ---
one of the possible difference between these versions
     {"GL_ARB_timer_query",                  ARB_TIMER_QUERY               },
+    {"GL_ARB_uniform_buffer_object",        ARB_UNIFORM_BUFFER_OBJECT     },
     {"GL_ARB_vertex_array_bgra",            ARB_VERTEX_ARRAY_BGRA         },
AFAIK this extension OpenGL3.1 is not supported by my Videocard.
Is it a possible reason of the problem?

[Bug 38216] Wrong screen in Civilization V with new engines. Good with old 1.7.23.

https://bugs.winehq.org/show_bug.cgi?id=38216

Michael Müller  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #4 from Michael Müller  ---
Can you reproduce the problem using the demo? For me the demo works flawless on
Linux with Wine 1.7.38.

[Bug 38216] Wrong screen in Civilization V with new engines. Good with old 1.7.23.

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #5 from Sergey  ---
Can you give me a link on the demo?

I am using Radeon HD6670 in MacOS 10.9.5 if this does matter.
I can confirm my observations with 1.7.38 as well as with many custom engines.
The key difference is 1.7.24.diff

[Bug 38216] Wrong screen in Civilization V with new engines. Good with old 1.7.23.

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #6 from Michael Müller  ---
You can download the demo via steam. There are also some standalone installers,
but they all require you to install steam, so it doesn't make any difference
and you can directly use it to download the demo.

[Bug 38216] Wrong screen in Civilization V with new engines. Good with old 1.7.23.

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #7 from Sergey  ---
I have no steam. 
Is it good demo?
http://www.fileplanet.com/216393/210000/fileinfo/Sid-Meier%27s-Civilization-...

[Bug 38216] Wrong screen in Civilization V with new engines

https://bugs.winehq.org/show_bug.cgi?id=38216

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |regression
            Summary|Wrong screen in             |Wrong screen in
                   |Civilization V with new     |Civilization V with new
                   |engines. Good with old      |engines
                   |1.7.23.                     |

--- Comment #8 from Austin English  ---
Please run a regression test:
http://wiki.winehq.org/RegressionTesting

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Wrong screen in             |Wrong screen in
                   |Civilization V with new     |Civilization V
                   |engines                     |

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #9 from Sergey  ---
Not easy for me but I will to try. It takes me a time.

$ git bisect start dlls/wined3d
$ git bisect good wine-1.7.23
$ git bisect bad wine-1.7.24
Bisecting: 4 revisions left to test after this (roughly 2 steps)
[6a1a18e7e2cf3fce8455ae86f0693e3e5df13004] d3d10core: Implement
d3d10_device_CopyResource().

:(

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #10 from Austin English  ---
(In reply to Sergey from comment #9)
> Not easy for me but I will to try. It takes me a time.
>
> $ git bisect start dlls/wined3d
> $ git bisect good wine-1.7.23
> $ git bisect bad wine-1.7.24
> Bisecting: 4 revisions left to test after this (roughly 2 steps)
> [6a1a18e7e2cf3fce8455ae86f0693e3e5df13004] d3d10core: Implement
> d3d10_device_CopyResource().
>
> :(

You need to compile wine and retest for your issue, then run 'git bisect
good/bad' and repeat until you have identified the faulty commit.

Please use the forums if you need more help.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #11 from Sergey  ---
OK, I did this.
----
e9b0a0e181daa98c645a2c1527b6434624d83ec4 is the first bad commit
commit e9b0a0e181daa98c645a2c1527b6434624d83ec4
Author: Andrei Slăvoiu 
Date:   Wed Aug 6 23:13:09 2014 +0300

wined3d: Use capabilities from the shader backend or the fragment pipeline
to determine the D3D level.

:040000 040000 5c1cba93697ab212f596c7028f91eb1c4cf11086
ab11f49dee29b4a823d7d9ccc5d8319fa2baceb2 M    dlls

----
But I don't know why this commit is bad. Someone more understanding should look
in.

Anyway, I made one more test. I created most recent engine 1.7.38-2a58869 and
set here some more dump to see a reason of messages I mentions early
Format conversion missing 0x33545844 -> 0x33545844

This message appears due to the different sizes of dst and src.
dst_size.width = 0x80 src_size.width = 0x48
dst_size.height = 0x20 src_size.height = 0x18
So dst sizes are aligned while src no. This is a real reason for broken screen.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #12 from Sergey  ---
I see 1.7.24.diff contains these lines
----
+    else
+    {
+        /* The drawable size of an FBO target is the OpenGL texture size,
+         * which is the power of two size. */
+        *width = context->current_rt->pow2Width;
+        *height = context->current_rt->pow2Height;
+    }
----

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #13 from Sergey Isakov  ---
Latest engine 1.7.38--b436b4b still the same messy screen.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

Michael Müller  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #14 from Andrei Slăvoiu  ---
(In reply to Sergey Isakov from comment #5)
> I am using Radeon HD6670 in MacOS 10.9.5 if this does matter.

(In reply to Sergey Isakov from comment #11)
> OK, I did this.
>
> e9b0a0e181daa98c645a2c1527b6434624d83ec4 is the first bad commit
> commit e9b0a0e181daa98c645a2c1527b6434624d83ec4
> Author: Andrei Slăvoiu 
> Date:   Wed Aug 6 23:13:09 2014 +0300
> wined3d: Use capabilities from the shader backend or the fragment
>
> pipeline to determine the D3D level.
>
> :040000 040000 5c1cba93697ab212f596c7028f91eb1c4cf11086
> ab11f49dee29b4a823d7d9ccc5d8319fa2baceb2 M	dlls
>
> But I don't know why this commit is bad. Someone more understanding should
> look in.

That commit only changes the fallback in case the video card is not explicitly
recognized. But your card should be recognized so the fallback changes should
be irrelevant. Can you post the output of glxinfo or whatever equivalent there
exists for MacOS?

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #15 from Sergey Isakov  ---
Created attachment 51062
  --> https://bugs.winehq.org/attachment.cgi?id=51062
OpenGL Viewer report

I think you mean OpenGL info. 
It is in the attachment.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #16 from Andrei Slăvoiu  ---
<Renderer>ATI Radeon Turks XT Prototype OpenGL Engine</Renderer>

That explains why it's not recognized, so indeed you are hitting the fallback
path. The old fallback code was able to recognize your card as capable of at
least DX10 based on the presence of EXT_GPU_SHADER4, but the new code that uses
the shader capabilities no longer has access to the list of OpenGL extensions.

And since the problem will be solved the right way by using core contexts (work
in progress) I doubt anybody will be willing to workaround this cornercase
until then.

The only solution for now would be to add the renderer string for your card to
cards_amd_binary array in dlls/wined3d/directx.c It should like:
    {"Turks",                     CARD_AMD_RADEON_HD6600},

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #17 from Sergey Isakov  ---
Great! I wrote this line into the recent sources and compile.
YES IT WORKS!
The bug may be closed if some patch will be committed so that I will not patch
it every times.

Thank you!

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #18 from Sergey Isakov  ---
(In reply to Andrei Slăvoiu from comment #16)
> <Renderer>ATI Radeon Turks XT Prototype OpenGL Engine</Renderer>
>
> That explains why it's not recognized, so indeed you are hitting the
> fallback path. The old fallback code was able to recognize your card as
> capable of at least DX10 based on the presence of EXT_GPU_SHADER4, but the
> new code that uses the shader capabilities no longer has access to the list
> of OpenGL extensions.
>
> And since the problem will be solved the right way by using core contexts
> (work in progress) I doubt anybody will be willing to workaround this
> cornercase until then.
>
> The only solution for now would be to add the renderer string for your card
> to cards_amd_binary array in dlls/wined3d/directx.c It should like:
>     {"Turks",                     CARD_AMD_RADEON_HD6600},

Why not add this line into main sources? Is it dangerous for anybody?

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #19 from Andrei Slăvoiu  ---
You can try to submit it yourself following the guidelines from
http://wiki.winehq.org/SubmittingPatches
BTW, is the driver you are using the standard Apple driver? From the rendering
string it sounds like it's some sort of experimental driver.
And until now the assumption inside the wine code was that the MacOS driver
reports the same rendering strings as the binary Linux driver. And the binary
linux driver renderer string has not changed AFAIK.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #20 from Sergey Isakov  ---
No, I am using clear OSX 10.9.5 with official drivers.
I think the name of the renderer taken from the card BIOS. In this case it will
be the same as in linux. I don't know how to check this.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #21 from Sergey Isakov  ---
Hey, this is systematic error
---
err:winediag:init_driver_info Invalid GPU override 1002:6738 specified,
ignoring.
----
This guy needed a line
{"Barts",                     CARD_AMD_RADEON_HD6800},

May be other AMD  families needed such definitions.
This is the bug!

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #22 from Sergey Isakov  ---
Created attachment 51310
  --> https://bugs.winehq.org/attachment.cgi?id=51310
Different AMD card identification in Apple OSX

There are other families of AMD videocard and MacOSX reported them by family
name as it can be seen from this screenshot.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #23 from Andrei Slăvoiu  ---
You might have more success raising this subject on the wine-devel list. I
never touched an Apple device so I can not help you further. All I can tell you
is that the current code treats MacOS as if it should have the same strings as
the Linux Catalyst driver and that is obviously not true (anymore).

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #24 from Sergey Isakov  ---
Created attachment 51312
  --> https://bugs.winehq.org/attachment.cgi?id=51312
Add recent AMD families

If this patch will not be accepted to mainstream than I wish it can be
accessible for all who want it.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #25 from Henri Verbeet  ---
(In reply to Andrei Slăvoiu from comment #23)
> never touched an Apple device so I can not help you further. All I can tell
> you is that the current code treats MacOS as if it should have the same
> strings as the Linux Catalyst driver and that is obviously not true
> (anymore).

It's not that obvious, actually. I'm not necessarily opposed to adding more
strings to match against, but I'd like some more background information. E.g.,
if this is something that changed in the OS X drivers I'd like to know at least
around what version that happened. So far I've seen exactly one person report
this issue. Granted, that may just be because OS X users almost by definition
don't care about Free Software, but "Prototype OpenGL engine" does sound
slightly suspicious.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #26 from Sergey Isakov  ---
Same issue from other user
---
<?xml version="1.0" encoding="UTF-8" ?>
<!-- http://www.realtech-vr.com/glview -->
<GLView version="4.3.4"><System>Mac OS X 10.10.3 (14D136)</System>
<Device>MacBookPro8,2</Device>
<CPU>2.0 GHz       Intel(R) Core(TM) i7-2630QM CPU @ 2.00GHz x 8</CPU>
<Vendor>Intel Inc.</Vendor>
<Version>3.3 INTEL-10.0.31</Version>
<GLSLVersion>3.30</GLSLVersion>
<Renderer>Intel HD Graphics 3000 OpenGL Engine</Renderer>
<Extensions>GL_ARB_blend_func_extended GL_ARB_ES2_compatibility
GL_ARB_explicit_attrib_location GL_ARB_instanced_arrays
GL_ARB_internalformat_query GL_ARB_occlusion_query2 GL_ARB_sampler_objects
GL_ARB_separate_shader_objects GL_ARB_shader_bit_encoding
GL_ARB_shading_language_include GL_ARB_texture_rgb10_a2ui
GL_ARB_texture_storage GL_ARB_texture_swizzle GL_ARB_timer_query
GL_ARB_vertex_type_2_10_10_10_rev GL_EXT_debug_label GL_EXT_debug_marker
GL_EXT_texture_compression_s3tc GL_EXT_texture_filter_anisotropic
GL_EXT_texture_sRGB_decode GL_APPLE_client_storage
GL_APPLE_container_object_shareable GL_APPLE_flush_render
GL_APPLE_object_purgeable GL_APPLE_rgb_422 GL_APPLE_row_bytes
GL_APPLE_texture_range GL_ATI_texture_mirror_once GL_NV_texture_barrier
</Extensions>
---

I just see you don't want to do anything for Mac.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #27 from Henri Verbeet  ---
(In reply to Sergey Isakov from comment #26)
> <Renderer>Intel HD Graphics 3000 OpenGL Engine</Renderer>

How is that the same issue?

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #28 from Sergey Isakov  ---
(In reply to Henri Verbeet from comment #27)
> (In reply to Sergey Isakov from comment #26)
> > <Renderer>Intel HD Graphics 3000 OpenGL Engine</Renderer>
>
> How is that the same issue?

The issue is select_card_handler will not correctly detemine card which may
lead to different render bugs. My case with the game Civilization V is shown in
screenshot attached.
In my case the bug corrected by just adding the line
{"Turks",                       CARD_AMD_RADEON_HD6600},
corresponding to my card identification by Apple's drivers.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #29 from Ken Thomases  ---
(In reply to Sergey Isakov from comment #20)
> No, I am using clear OSX 10.9.5 with official drivers.

This is contradicted by your own attachment 51062 from comment 15:

<System>Mac OS X 10.7.5 (11G63)</System>
> I think the name of the renderer taken from the card BIOS. In this case it
> will be the same as in linux. I don't know how to check this.

I have never seen nor heard of a renderer name like yours ("ATI Radeon Turks XT
Prototype OpenGL Engine"), which is why your patch is being questioned.  A
search of the web for that string shows very few hits, as opposed to, say, "ATI
Radeon HD 4850 OpenGL Engine" which comes from my 27-inch iMac, Late 2009
running 10.9.5.

Your attachment says the model is "iMac12,2".  According to my searches, that's
the 27-inch iMac, 2011. See
http://www.everymac.com/ultimate-mac-lookup/?search_keywords=iMac12,2 and
https://en.wikipedia.org/wiki/IMac_%28Intel-based%29.  According to those
links, those have an AMD Radeon HD 6750M, AMD Radeon HD 6770M, or AMD Radeon HD
6970M.  According to what I can find, those are codenamed Barts LE (67xx
series) or Cayman (69xx series), not Turks.

Is this a hackintosh or have you replaced the GPU somehow?

Anyway, we're just trying to understand the discrepancies between what you're
reporting and everything we've ever seen before from OS X and Mac hardware.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #30 from Sergey Isakov  ---
OK, I have hackintosh although it is not excluded a case with real Mac with
videocard replaced.
Anyway, Barts and Cayman are also absent in your list.
Be sure I am using native Apple's drivers. No custom or third party drivers are
present anywhere.

At last. I say that card identification listed in cards_amd_binary[] and
cards_intel[] doesn't correspond current Apple's drivers.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #31 from Sergey Isakov  ---
Can be closed "WONT FIX".
The patch attached.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

Ken Sharp  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |WONTFIX

--- Comment #32 from Ken Sharp  ---
(In reply to Sergey Isakov from comment #31)
> Can be closed "WONT FIX".

You can close bugs yourself!

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

Sebastian Lackner  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |patch
             Status|RESOLVED                    |REOPENED
                 CC|                            |[e-mail]
         Resolution|WONTFIX                     |---
     Ever confirmed|0                           |1
    Regression SHA1|                            |e9b0a0e181daa98c645a2c1527b
                   |                            |6434624d83ec4

--- Comment #33 from Sebastian Lackner  ---
(In reply to Ken Sharp from comment #32)
> (In reply to Sergey Isakov from comment #31)
> > Can be closed "WONT FIX".
>
> You can close bugs yourself!

There never was a decision that this is a WONTFIX, Sergey was just frustrated
that the patch was not accepted on the first attempt. I would say we should
leave it open until there is a final decision from the d3d devs, or until the
patch was accepted.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

[e-mail] changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #34 from [e-mail] ---
This commmit
http://source.winehq.org/git/wine.git/commit/7bc003287806ad9509162c444493e27...

Added

+            {"TURKS",   CARD_AMD_RADEON_HD6600},

But only for the mesa.

I think it's safe to send this one line patch to upstream.

This is similar issue

https://source.winehq.org/patches/data/128628

Only mesa part was added for rx480 and it uses fallback on closed source
drivers.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

Józef Kucia  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #35 from Józef Kucia  ---
It's likely that this bug is fixed by 1df961bd3d1a5d4d4badfe364ca51e9f6cc88fd3.
Please retest.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #36 from Sergey Isakov  ---
(In reply to Józef Kucia from comment #35)
> It's likely that this bug is fixed by
> 1df961bd3d1a5d4d4badfe364ca51e9f6cc88fd3. Please retest.

It's a pity I have no more this game. But I may say that the similar bug 40180
is not resolved.
wine-2.0-rc4-38-ga528847

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #37 from Sergey Isakov  ---
(In reply to Józef Kucia from comment #35)
> It's likely that this bug is fixed by
> 1df961bd3d1a5d4d4badfe364ca51e9f6cc88fd3. Please retest.

It's a pity I have no more this game. But I may say that the similar bug 40180
is not resolved.
wine-2.0-rc4-38-ga528847

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #38 from [e-mail] ---
(In reply to Sergey Isakov from comment #37)
> (In reply to Józef Kucia from comment #35)
> > It's likely that this bug is fixed by
> > 1df961bd3d1a5d4d4badfe364ca51e9f6cc88fd3. Please retest.
>
> It's a pity I have no more this game. But I may say that the similar bug
> 40180 is not resolved.
> wine-2.0-rc4-38-ga528847

Either this bug or bug 38216 should be closed you yourself said that the issue
is the same and it can be prevented by the patch that can be seen on both bugs.
It's still a regression.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #39 from Sergey Isakov  ---
*** Bug 40180 has been marked as a duplicate of this bug. ***

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #40 from Józef Kucia  ---
(In reply to Sergey Isakov from comment #37)
> It's a pity I have no more this game. But I may say that the similar bug
> 40180 is not resolved.
> wine-2.0-rc4-38-ga528847

Please attach a +d3d,+wgl trace.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #41 from Sergey Isakov  ---
(In reply to Józef Kucia from comment #40)
> (In reply to Sergey Isakov from comment #37)
> > It's a pity I have no more this game. But I may say that the similar bug
> > 40180 is not resolved.
> > wine-2.0-rc4-38-ga528847
>
> Please attach a +d3d,+wgl trace.

I tested again bug 40180 and see good picture. Something changed in my
environment?
Looks like the bug is resolved!

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #42 from Józef Kucia  ---
(In reply to Sergey Isakov from comment #41)
> I tested again bug 40180 and see good picture. Something changed in my
> environment?
> Looks like the bug is resolved!

A +d3d,+wgl trace still could be useful in order to confirm that this bug is
solved thanks to WGL_WINE_query_renderer.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

--- Comment #43 from Sergey Isakov  ---
Created attachment 56831
  --> https://bugs.winehq.org/attachment.cgi?id=56831
log from HeroesVI +d3d,+wgl

OK.
Original size=1,44Gb can't be attached
So I made
cat LastRunWine.log | head -n 500000 >HeroesVI-1.log
Is it enough?

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

Matteo Bruni  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
      Fixed by SHA1|                            |1df961bd3d1a5d4d4badfe364ca
                   |                            |51e9f6cc88fd3
         Resolution|---                         |FIXED

--- Comment #44 from Matteo Bruni  ---
(In reply to Sergey Isakov from comment #43)
> Created attachment 56831 [details]
> log from HeroesVI +d3d,+wgl
>
> Is it enough?

Yes.

[Bug 38216] Wrong screen in Civilization V

https://bugs.winehq.org/show_bug.cgi?id=38216

Alexandre Julliard  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |CLOSED

--- Comment #45 from Alexandre Julliard  ---
Closing bugs fixed in 2.0-rc5.