[Bug 43009] New: TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

Bug ID: 43009
           Summary: TIM message manager window covered by an unresponsive
                    window
           Product: Wine-staging
           Version: 2.7
          Hardware: x86
               URL: http://dldir1.qq.com/qqfile/qq/TIM1.1.0/20837/TIM1.1.0
                    .exe
                OS: Linux
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: -unknown
          Assignee: [e-mail]
          Reporter: [e-mail]
                CC: [e-mail], [e-mail],
                    [e-mail], [e-mail]
      Distribution: ArchLinux

TIM is a cleaner version of QQ, successor of QQ Lite.

download and install it:
http://dldir1.qq.com/qqfile/qq/TIM1.1.0/20837/TIM1.1.0.exe

ecb6934f7cea33b1458308495fd965a274c6ba2a  TIM1.1.0.exe

winetricks riched20 so you can input your QQ number.

login in, then click your avatar on the left top corner, select "消息管理器".

with wine 2.7 you can use it normally. with wine-staging 2.7 movements and
clicks on the window cause nothing. Moving the window using the window manager,
you can move away an unresponsive window, underlying it there is a normal
window you can operate on. Moving the normal window the unresponsive one moves
back to cover it.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #1 from [e-mail] ---
Forgot to say, my window manager is awesome 3.5.9.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

Gijs Vermeulen  changed:

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

--- Comment #2 from Gijs Vermeulen  ---
If you can build wine from source, can you run a test to find the bad patch?
See bug 42988 for a way to do this.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #3 from [e-mail] ---
(In reply to Gijs Vermeulen from comment #2)
> If you can build wine from source, can you run a test to find the bad patch?
> See bug 42988 for a way to do this.

The command patchinstall.sh fails with:

Applying
/home/lilydjwg/src/wine-staging/patches/d3d11-Deferred_Context/0003-d3d11-Initial-implementation-for-deferred-contexts.patch
Applying: d3d11: Initial implementation for deferred contexts.
error: patch failed: dlls/d3d11/device.c:2260
error: dlls/d3d11/device.c: patch does not apply
Patch failed at 0001 d3d11: Initial implementation for deferred contexts.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #4 from Gijs Vermeulen  ---
(In reply to lilydjwg from comment #3)
> (In reply to Gijs Vermeulen from comment #2)
> > If you can build wine from source, can you run a test to find the bad patch?
> > See bug 42988 for a way to do this.
>
> The command patchinstall.sh fails with:
>
> Applying
> /home/lilydjwg/src/wine-staging/patches/d3d11-Deferred_Context/0003-d3d11-
> Initial-implementation-for-deferred-contexts.patch
> Applying: d3d11: Initial implementation for deferred contexts.
> error: patch failed: dlls/d3d11/device.c:2260
> error: dlls/d3d11/device.c: patch does not apply
> Patch failed at 0001 d3d11: Initial implementation for deferred contexts.

Did you try applying the patches to wine-git or wine 2.7?

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #5 from [e-mail] ---
(In reply to Gijs Vermeulen from comment #4)
> Did you try applying the patches to wine-git or wine 2.7?

I tried both git master and the wine-2.7 tag. They failed with different
patches.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #6 from Michael Müller  ---
You need to apply the corresponding patches for the Wine version you try to
build.

If you want to build Wine Staging 2.7, you need to checkout the 2.7 version of
Wine and Wine Staging:

cd wine; git checkout wine-2.7
cd wine-staging; git checkout v2.7

If you want to build the HEAD of Wine Staging, you need to get the related
upstream commit:

cd wine-staging; git checkout HEAD
cd wine; git checkout $(../wine-staging/patches/patchinstall.sh
--upstream-commit)

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #7 from [e-mail] ---
I've got it:

a473ddb697c4665c4a8a9e33b4e3e8400665996b is the first bad commit
commit a473ddb697c4665c4a8a9e33b4e3e8400665996b
Author: Dmitry Timoshkov 
Date:   Tue Nov 25 20:31:58 2014 +0100

user32: Try harder to find a target for mouse messages

:040000 040000 e35a9c01cdfdccb834a3dba603e9d2ba8228c497
bd2f52e0270953d08407d441850a8d11a10a6ef8 M      dlls

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #8 from Dmitry Timoshkov  ---
Created attachment 58168
  --> https://bugs.winehq.org/attachment.cgi?id=58168
patch

(In reply to lilydjwg from comment #0)
> TIM is a cleaner version of QQ, successor of QQ Lite.
>
> download and install it:
> http://dldir1.qq.com/qqfile/qq/TIM1.1.0/20837/TIM1.1.0.exe

The application doesn't run here, it crashes in one of its kernel drivers.
How do you workaround the crashes? Also, what is the application name that
one is supposed to run? is it Program Files/Tencent/TIM/Bin/TIM.exe? Maybe
TIMApp.exe? Something else?

Kernel drivers make me very reluctant to install and investigate this thing
under either a real Windows or in a VM.

Anyway, does the attached hack help?

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #9 from [e-mail] ---
It didn't crash here at the beginning. Later I found it disappeared after I
started it without showing any window. When I ran the second time it worked
though. I guess it's some timing issue between several executables it contains.

Make sure you kill all the processes after the installation is complete. Maybe
it's relevant.

The executable to run is TIM.exe. TIMApp.exe should also work.

Other changes I've made though it's not necessary (at least it doesn't crash):
* winetricks sandbox
* Windows version set to win10, but xp for TIM.exe (for the default browser)
* disable txplatform.exe and qqprotectupd.exe
* replace iphlpapi.dll with a native one (for addding buddies window to not
freeze)
* also I've disabled winemenubuilder.exe

I'll try the patch later.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #10 from [e-mail] ---
The patch didn't work. I applied the patch to the HEAD after applying all the
staging patches to wine master.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #11 from [e-mail] ---
BTW, run TIM.exe in the directory where it resides. It may have problems with
other working directory.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #12 from Dmitry Timoshkov  ---
Setting reported version to Windows 10 in winecfg is the key, this prevents
the application using its kernel drivers. Next required step is to install
riched20 with winetricks in order to be able to type anything.

Unfortunately since I can't read chinese it's impossible for me to create
an account and log in to reproduce the real problem. Is it possible to
create an account for testing purposes?

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #13 from [e-mail] ---
Oh really? I thought I successfully ran it with default Windows version once.
Maybe I mistook it.

There was a QQ account here but it doesn't work now. If you're willing you can
register one using this English form: https://ssl.zc.qq.com/en/index.html. It
may require a phone number, or requiring one when changing passwords.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #14 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #13)
> Oh really? I thought I successfully ran it with default Windows version
> once. Maybe I mistook it.
>
> There was a QQ account here but it doesn't work now. If you're willing you
> can register one using this English form:
> https://ssl.zc.qq.com/en/index.html. It may require a phone number, or
> requiring one when changing passwords.

It requires a phone number for the verification code, that's too much
I'm afraid for a testing account.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #15 from [e-mail] ---
Oh yeah, nowadays registration for nearly all Chinese websites requires a phone
number :-(

Anyway, can you tell me how to debug this?

PS: If I recall correctly this also happened for another program by Tencent
(and that one requires a phone number too).

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #16 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #15)
> Anyway, can you tell me how to debug this?

First thing to figure out is to get the properties of that unresponsive
window (class name, class style, window name, window style/exstyle).
Next step is to create the windows tree sorted by z-order the application
creates. This task could be done using tools like Spy++ and similar.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #17 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #16)
> (In reply to lilydjwg from comment #15)
> > Anyway, can you tell me how to debug this?
>
> First thing to figure out is to get the properties of that unresponsive
> window (class name, class style, window name, window style/exstyle).
> Next step is to create the windows tree sorted by z-order the application
> creates. This task could be done using tools like Spy++ and similar.

I'm using WinSpy++.

Class is TXGFLayerMask, no/empty caption, style
WS_{POPUP,VISIBLE,CLIPSIBLINGS,CLIPCHILDREN}, exstyle WS_EX_TOOLWINDOW, and
three are shown in gray: WS_EX_{LEFT,LTRREADING,RIGHTSCROLLBAR}. TIM.exe
crashes when I switch WinSpy++ to the "Class" tab.

It has two children and no siblings. Both have class TXGuiFoundation. The
parent window is the underlying window but the underlying window doesn't have
any children.

When I make it invisible by the dropdown menu provided by WinSpy++, the
unresponsive window disappears and the underlying window works normally.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

Dmitry Timoshkov  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW

--- Comment #18 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #17)
> Class is TXGFLayerMask, no/empty caption, style
> WS_{POPUP,VISIBLE,CLIPSIBLINGS,CLIPCHILDREN}, exstyle WS_EX_TOOLWINDOW, and
> three are shown in gray: WS_EX_{LEFT,LTRREADING,RIGHTSCROLLBAR}. TIM.exe
> crashes when I switch WinSpy++ to the "Class" tab.
>
> It has two children and no siblings. Both have class TXGuiFoundation. The
> parent window is the underlying window but the underlying window doesn't
> have any children.
>
> When I make it invisible by the dropdown menu provided by WinSpy++, the
> unresponsive window disappears and the underlying window works normally.

Thanks!

Please generate +class,+win,+msg,+event,+tid log (preferably running
'wine winver' in another xterm first to avoid start up noice in the log)
and attach it here. Attaching another log with the culprit patch reverted
would be also useful.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #19 from [e-mail] ---
Created attachment 58176
  --> https://bugs.winehq.org/attachment.cgi?id=58176
WINEDEBUG=+class,+win,+msg,+event,+tid with the issue

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #20 from [e-mail] ---
Created attachment 58177
  --> https://bugs.winehq.org/attachment.cgi?id=58177
WINEDEBUG=+class,+win,+msg,+event,+tid with the commit reverted

Note this one is generated on a different system (on which I bisected the
commit) than the previous log (on my work machine with wine-staging from the
official Arch repo).

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #21 from Dmitry Timoshkov  ---
Thanks for the logs. I'd need to write a test app to better see how it's
supposed to work.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #22 from Dmitry Timoshkov  ---
In short: the unresponsive window in question is a layered window with
an assigned color key. I guess that the window surface bitmap is used
as a layer mask, and is supposed to be "transparent" both visually and
for mouse clicks. winex11.drv calculates the mask shape and instructs
the WM about it, so WM sends the button click event to an underlying
window. But user32 side is not aware about that masked shape, and
currently there is no way to check for it from WINPOS_WindowFromPoint.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #23 from Dmitry Timoshkov  ---
It should be possible to verify my guess by disabling the shaping of
layered windows on winex11.drv side. In order to do that create a key
"ShapeLayeredWindows"="N" under "X11 Driver" section in the registry.
See https://wiki.winehq.org/Useful_Registry_Keys how to do it, it looks
like ShapeLayeredWindows setting is not documented there.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #24 from Dmitry Timoshkov  ---
(In reply to Dmitry Timoshkov from comment #23)
> It should be possible to verify my guess by disabling the shaping of
> layered windows on winex11.drv side. In order to do that create a key
> "ShapeLayeredWindows"="N" under "X11 Driver" section in the registry.
> See https://wiki.winehq.org/Useful_Registry_Keys how to do it, it looks
> like ShapeLayeredWindows setting is not documented there.

The test above should be done with the culprit patch reverted (or using
unpatched winehq source). Once the key above is added to the registry
the problem with unresponsive window should always present.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #25 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #23)
> It should be possible to verify my guess by disabling the shaping of
> layered windows on winex11.drv side. In order to do that create a key
> "ShapeLayeredWindows"="N" under "X11 Driver" section in the registry.
> See https://wiki.winehq.org/Useful_Registry_Keys how to do it, it looks
> like ShapeLayeredWindows setting is not documented there.

I get strange results.

With this setting, I get a white window that I can't move or close. But I can
minimize it, and after I restore the window it behaves normally with all the
functionality usable.

I also test the original issue after minimizing and restoring. The unresponsive
window is gone after that.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #26 from Dmitry Timoshkov  ---
Created attachment 58178
  --> https://bugs.winehq.org/attachment.cgi?id=58178
patch

Could you please test the attached patch? It was created against staging tree,
previous hack should be reverted, make sure that "ShapeLayeredWindows"="Y"
is set or that setting is removed from the registry.

The patch adds experimental support for a layered window region.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #27 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #25)
> > It should be possible to verify my guess by disabling the shaping of
> > layered windows on winex11.drv side. In order to do that create a key
> > "ShapeLayeredWindows"="N" under "X11 Driver" section in the registry.
> > See https://wiki.winehq.org/Useful_Registry_Keys how to do it, it looks
> > like ShapeLayeredWindows setting is not documented there.
>
> I get strange results.
>
> With this setting, I get a white window that I can't move or close. But I
> can minimize it, and after I restore the window it behaves normally with all
> the functionality usable.
>
> I also test the original issue after minimizing and restoring. The
> unresponsive window is gone after that.

Thanks for the testing, the result pretty much confirms my theory.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #28 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #26)
> Created attachment 58178 [details]
> patch
>
> Could you please test the attached patch? It was created against staging
> tree,
> previous hack should be reverted, make sure that "ShapeLayeredWindows"="Y"
> is set or that setting is removed from the registry.
>
> The patch adds experimental support for a layered window region.

Oops, I can't login anymore. The login window shows up with animation! (That
animation exists in Windows but I never saw it in Wine.) But then it seems to
freeze, and moving mouse over it will hide the mouse cursor. I can close it,
and it disappears with animation too.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #29 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #28)
> Oops, I can't login anymore. The login window shows up with animation! (That
> animation exists in Windows but I never saw it in Wine.) But then it seems
> to freeze, and moving mouse over it will hide the mouse cursor. I can close
> it, and it disappears with animation too.

I guess that's how the progress looks like sometimes :)

I'll attach a slightly trimmed down version of the patch that reduces
new functionality only to point_from_window() on the server side, and
leaves visible region calculations intact. Please give it a try.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #30 from Dmitry Timoshkov  ---
Created attachment 58179
  --> https://bugs.winehq.org/attachment.cgi?id=58179
patch2

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #31 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #30)
> Created attachment 58179 [details]
> patch2

The animation is gone, nothing else changes.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #32 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #31)
> The animation is gone, nothing else changes.

Please add +server,+relay to the debug flags and generate another log
with last patch applied. There is no need to start 'wine winver' in a
separate terminal this time beforehand.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #33 from Dmitry Timoshkov  ---
New log is going to be pretty large, please compress it before attaching.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #34 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #33)
> New log is going to be pretty large, please compress it before attaching.

The log is huge (1.6G uncompressed), I've uploaded it here:
https://transfer.sh/iy1J0/relay.log.xz

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #35 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #34)
> The log is huge (1.6G uncompressed), I've uploaded it here:
> https://transfer.sh/iy1J0/relay.log.xz

Thanks for the log. Unfortunately the log doesn't contain a window
with the class "TXGFLayerMask", and there is no any layered window
either. Did you try to reproduce the problem while creating the log?

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #36 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #35)
> Thanks for the log. Unfortunately the log doesn't contain a window
> with the class "TXGFLayerMask", and there is no any layered window
> either. Did you try to reproduce the problem while creating the log?

With last patch I can't login, so there is only one window that I can see: the
login window. I thought you were debugging why that window hides my mouse and
does not response.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #37 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #36)
> > Thanks for the log. Unfortunately the log doesn't contain a window
> > with the class "TXGFLayerMask", and there is no any layered window
> > either. Did you try to reproduce the problem while creating the log?
>
> With last patch I can't login, so there is only one window that I can see:
> the login window. I thought you were debugging why that window hides my
> mouse and does not response.

Sorry about not being clear enough, looks like I misunderstood your
description. I'd like to better see what is going on with the layered
window during its lifetime, so I'd need a log with "TXGFLayerMask" window
present. Then please generate the log using last debug flags and with
plain staging tree (i.e. without testing patches attached to this bug).

I'll keep the log you've created just in case I'll need to investigate
the login window bug caused by my experimental patch. Thanks!

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #38 from Dmitry Timoshkov  ---
Created attachment 58186
  --> https://bugs.winehq.org/attachment.cgi?id=58186
patch3

Thanks to your log i think that I've found the problem with my previous
version of the patch. Please try this patch.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #39 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #38)
> Created attachment 58186 [details]
> patch3
>
> Thanks to your log i think that I've found the problem with my previous
> version of the patch. Please try this patch.

With this patch, the original problem is still there. I see no changes compared
to the version without the patch.

I'm still uploading the log you asked for now. It's huge (4.4G uncompressed,
137M compressed) and makes wine very slow.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #40 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #39)
> (In reply to Dmitry Timoshkov from comment #38)
> > Created attachment 58186 [details]
> > patch3
> >
> > Thanks to your log i think that I've found the problem with my previous
> > version of the patch. Please try this patch.
>
> With this patch, the original problem is still there. I see no changes
> compared to the version without the patch.

Let me clarify things, so unresponsive window (original problem) still
exists, but the bug I introduced in patches 1,2 that didn't allow you
to login got fixed, right?
> I'm still uploading the log you asked for now. It's huge (4.4G uncompressed,
> 137M compressed) and makes wine very slow.

Thanks!

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #41 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #40)
> Let me clarify things, so unresponsive window (original problem) still
> exists, but the bug I introduced in patches 1,2 that didn't allow you
> to login got fixed, right?

Yes.
> > I'm still uploading the log you asked for now. It's huge (4.4G uncompressed,
> > 137M compressed) and makes wine very slow.
>
> Thanks!

Here's the log: https://transfer.sh/ZITef/relay.log.xz
(without any patches, only wine-staging)

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #42 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #41)
> > Let me clarify things, so unresponsive window (original problem) still
> > exists, but the bug I introduced in patches 1,2 that didn't allow you
> > to login got fixed, right?
>
> Yes.
> > > I'm still uploading the log you asked for now. It's huge (4.4G uncompressed,
> > > 137M compressed) and makes wine very slow.
> >
> > Thanks!
>
> Here's the log: https://transfer.sh/ZITef/relay.log.xz
> (without any patches, only wine-staging)

Hmm, did you test the patch3 then?

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #43 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #42)
> Hmm, did you test the patch3 then?

Yes I tested, but without logs. The log completed before I saw patch3.
Do you need it also? It'll take some time.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #44 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #43)
> > Hmm, did you test the patch3 then?
>
> Yes I tested, but without logs. The log completed before I saw patch3.
> Do you need it also? It'll take some time.

Yes, ideally I'd need the log with patch3 applied to see why it doesn't
work. I asked for the log without patches because the patches didn't
allow you to login and I needed something to investigate.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #45 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #44)
> Yes, ideally I'd need the log with patch3 applied to see why it doesn't
> work. I asked for the log without patches because the patches didn't
> allow you to login and I needed something to investigate.

Here's it: https://transfer.sh/ni3Up/relay2.log.xz

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #46 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #45)
> Here's it: https://transfer.sh/ni3Up/relay2.log.xz

Thanks for the log. It looks like that there is (an existing) bug in the code
that creates region from a bitmap.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #47 from Dmitry Timoshkov  ---
Created attachment 58202
  --> https://bugs.winehq.org/attachment.cgi?id=58202
patch4

Does the attached patch work better?

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #48 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #47)
> Created attachment 58202 [details]
> patch4
>
> Does the attached patch work better?

This one works, thanks!

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #49 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #48)
> This one works, thanks!

Great, thanks for testing. Does the animation (that you mentioned earlier)
work with this version of the patch?

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #50 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #49)
> (In reply to lilydjwg from comment #48)
> > This one works, thanks!
>
> Great, thanks for testing. Does the animation (that you mentioned earlier)
> work with this version of the patch?

I can't see the showing up animation, maybe because it's too slow. It shows
only part of the window, then the full window.

The disappearing animation and the switching (to settings) animation is fine.
And the problem that after the switch, some parts of the window didn't get
updated are fixed too :-)

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #51 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #50)
> I can't see the showing up animation, maybe because it's too slow. It shows
> only part of the window, then the full window.
>
> The disappearing animation and the switching (to settings) animation is
> fine. And the problem that after the switch, some parts of the window didn't
> get updated are fixed too :-)

Do any of these problems exist in not staging version of wine?

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #52 from [e-mail] ---
(In reply to Dmitry Timoshkov from comment #51)
> (In reply to lilydjwg from comment #50)
> > I can't see the showing up animation, maybe because it's too slow. It shows
> > only part of the window, then the full window.
> >
> > The disappearing animation and the switching (to settings) animation is
> > fine. And the problem that after the switch, some parts of the window didn't
> > get updated are fixed too :-)
>
> Do any of these problems exist in not staging version of wine?

There is no those animations in the normal version of wine. The UI updates
problem exists there too.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #53 from Dmitry Timoshkov  ---
(In reply to lilydjwg from comment #52)
> > > I can't see the showing up animation, maybe because it's too slow. It shows
> > > only part of the window, then the full window.
> > >
> > > The disappearing animation and the switching (to settings) animation is
> > > fine. And the problem that after the switch, some parts of the window didn't
> > > get updated are fixed too :-)
> >
> > Do any of these problems exist in not staging version of wine?
>
> There is no those animations in the normal version of wine. The UI updates
> problem exists there too.

Thanks, good to know.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #54 from [e-mail] ---
Hi, Dmitry Timoshkov, I've filed another bug 43085 with TIM. It's also about
unresponsible windows, would you like to take a look?

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

--- Comment #55 from Sebastian Lackner  ---
This issue should be fixed with Wine Staging 2.9. Please verify if everything
is working as expected and feel free to update the status of this bug report.

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

[e-mail] changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|NEW                         |RESOLVED

--- Comment #56 from [e-mail] ---
(In reply to Sebastian Lackner from comment #55)
> This issue should be fixed with Wine Staging 2.9. Please verify if
> everything is working as expected and feel free to update the status of this
> bug report.

I can confirm everything is working as expected. Thanks!

[Bug 43009] TIM message manager window covered by an unresponsive window

https://bugs.winehq.org/show_bug.cgi?id=43009

Alistair Leslie-Hughes  changed:

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

--- Comment #57 from Alistair Leslie-Hughes  ---
Closing Fixed Staging 3.14