Settlers II: 10th Anniversary regression?

[Markus]

Hi,

when trying the demo of the above game ( 
http://appdb.winehq.org/appview.php?iVersionId=5346 ) with 0.9.30, it loads 
the main menu and allows you to enter a game.
In 0.9.32, it does not start anymore at all (it hangs after opening a window) 
with the following messages:
fixme:wininet:InternetSetOptionExA Flags 04000000 ignored
fixme:wininet:InternetSetOptionW Option INTERNET_OPTION_CONNECT_TIMEOUT 
(1000): STUB
fixme:wininet:InternetSetOptionExA Flags 04000000 ignored
fixme:wininet:InternetSetOptionW INTERNET_OPTION_SEND/RECEIVE_TIMEOUT not 
supported on protocol 1
err:ntdll:RtlpWaitForCriticalSection section 0x110020 "heap.c: main process 
heap section" wait timed out in thread 0009, blocked by 000d, retrying (60 
sec)

Could this be a regression in DirectPlay?

Regards,

[Stefan Dösinger]

Am Sonntag 04 März 2007 20:41 schrieb Markus:
> Hi,
>
> when trying the demo of the above game (
> http://appdb.winehq.org/appview.php?iVersionId=5346 ) with 0.9.30, it loads
> the main menu and allows you to enter a game.
> In 0.9.32, it does not start anymore at all (it hangs after opening a
> window) with the following messages:
> fixme:wininet:InternetSetOptionExA Flags 04000000 ignored
> fixme:wininet:InternetSetOptionW Option INTERNET_OPTION_CONNECT_TIMEOUT
> (1000): STUB
> fixme:wininet:InternetSetOptionExA Flags 04000000 ignored
> fixme:wininet:InternetSetOptionW INTERNET_OPTION_SEND/RECEIVE_TIMEOUT not
> supported on protocol 1
> err:ntdll:RtlpWaitForCriticalSection section 0x110020 "heap.c: main process
> heap section" wait timed out in thread 0009, blocked by 000d, retrying (60
> sec)

Unlikely, dplay is rather old. Can you do a regression test? Most likely one 
of my d3d patches killed it.

[Markus]

On Sunday 04 March 2007 21:12, you wrote:
> Unlikely, dplay is rather old. Can you do a regression test? Most likely
> one of my d3d patches killed it.

I would, but unfortunately I cannot compile Wine myself under (SUSE) x86_64 
even though I follow the instructions at http://wiki.winehq.org/WineOn64bit

{standard input}: Assembler messages:
{standard input}:38: Error: suffix or operands invalid for `push'
[...]

[Markus]

On Sunday 04 March 2007 23:24, Markus wrote:
> On Sunday 04 March 2007 21:12, you wrote:
> > Unlikely, dplay is rather old. Can you do a regression test? Most likely
> > one of my d3d patches killed it.
>
> I would, but unfortunately I cannot compile Wine myself under (SUSE) x86_64
> even though I follow the instructions at http://wiki.winehq.org/WineOn64bit
>
> {standard input}: Assembler messages:
> {standard input}:38: Error: suffix or operands invalid for `push'
> [...]

Ok, got compiling to work. The Wiki was quite far off as to which settings are 
needed for compiling under 64bit SUSE.

Anyway, the patch that stopped the demo from starting seems to be this one (I 
only checked dlls/wined3d patches):
---
401110538aa39214280367725f2d8cb80abbc3da is first bad commit
commit 401110538aa39214280367725f2d8cb80abbc3da
Author: Stefan Dösinger 
Date:   Mon Feb 19 15:18:45 2007 +0100

wined3d: Check volume texture creation errors.

Volumetexture::Release checks for NULL pointers, so releasing a partially
    created texture works.

:040000 040000 f7e20f7cadd15cfd36f8d997615d5c24f5a2e921 
2a25e3f59b0cbff4adcce0b372326da6e7389dcd M      dlls
---

Apparently there is another patch that went in between 0.9.30 and .31 that 
causes an "Invalid call" error message to show after the start screen. I'll 
do a test for that later.

Regards,

[Kai Blin]

On Monday 05 March 2007 15:23, Markus wrote:
> Ok, got compiling to work. The Wiki was quite far off as to which settings
> are needed for compiling under 64bit SUSE.

The X libraries moved elsewhere in openSUSE 10.2. Fixes to the Wiki welcome.

Kai

[Markus]

On Monday 05 March 2007 15:55, Kai Blin wrote:
> > Ok, got compiling to work. The Wiki was quite far off as to which
> > settings are needed for compiling under 64bit SUSE.
>
> The X libraries moved elsewhere in openSUSE 10.2. Fixes to the Wiki
> welcome.

I already updated the Wiki to reflect the settings I had to use under 10.2.

[Marcus Meissner]

On Mon, Mar 05, 2007 at 05:11:10PM +0100, Markus wrote:
> On Monday 05 March 2007 15:55, Kai Blin wrote:
> > > Ok, got compiling to work. The Wiki was quite far off as to which
> > > settings are needed for compiling under 64bit SUSE.
> >
> > The X libraries moved elsewhere in openSUSE 10.2. Fixes to the Wiki
> > welcome.
>
> I already updated the Wiki to reflect the settings I had to use under 10.2.

-m32 is fully automatic detected in configure, it is not necessary to mention.

Ciao, Marcus

[Markus]

On Monday 05 March 2007 17:30, you wrote:
> > I already updated the Wiki to reflect the settings I had to use under
> > 10.2.
>
> -m32 is fully automatic detected in configure, it is not necessary to
> mention.

I cannot confirm this. When not setting -m32 manually, I am getting the 
previously posted assembler message errors. In fact, when specifying it under 
CFLAGS only, the linker would try to create a 64bit library out of the 32bit 
object files. Thus LDFLAGS needed -m32 as well.

[Fabian Bieler]

On Tuesday 06 March 2007 05:02, Markus wrote:
> On Monday 05 March 2007 17:30, you wrote:
> > > I already updated the Wiki to reflect the settings I had to use under
> > > 10.2.
> >
> > -m32 is fully automatic detected in configure, it is not necessary to
> > mention.
>
> I cannot confirm this. When not setting -m32 manually, I am getting the
> previously posted assembler message errors. In fact, when specifying it
> under CFLAGS only, the linker would try to create a 64bit library out of
> the 32bit object files. Thus LDFLAGS needed -m32 as well.

Just a wild guess here, but have you got your CC environment variable set?
wine's configure uses the CC, LD and AS variables to set -m32 (or equivalent 
options). It does not, however overwrite the variables if they exist.

[Marcus Meissner]

On Mon, Mar 05, 2007 at 03:23:39PM +0100, Markus wrote:
> On Sunday 04 March 2007 23:24, Markus wrote:
> > On Sunday 04 March 2007 21:12, you wrote:
> > > Unlikely, dplay is rather old. Can you do a regression test? Most likely
> > > one of my d3d patches killed it.
> >
> > I would, but unfortunately I cannot compile Wine myself under (SUSE) x86_64
> > even though I follow the instructions at http://wiki.winehq.org/WineOn64bit
> >
> > {standard input}: Assembler messages:
> > {standard input}:38: Error: suffix or operands invalid for `push'
> > [...]
>
> Ok, got compiling to work. The Wiki was quite far off as to which settings are 
> needed for compiling under 64bit SUSE.

I reviewed and adjusted the page a bit.

Ciao, Marcus

[Dmitry Timoshkov]

"Marcus Meissner"  wrote:
> > Ok, got compiling to work. The Wiki was quite far off as to which settings are 
> > needed for compiling under 64bit SUSE.
>
> I reviewed and adjusted the page a bit.

Marcus, you seem to be a knowledgeable person how to set up X11 in a 64-bit
environment. Perhaps you have a clue what might be needed to properly setup
locale support in that case?

See for instance the bug http://bugs.winehq.org/show_bug.cgi?id=4704

[Marcus Meissner]

On Mon, Mar 05, 2007 at 11:53:07PM +0800, Dmitry Timoshkov wrote:
> "Marcus Meissner"  wrote:
> > > Ok, got compiling to work. The Wiki was quite far off as to which 
> > > settings are needed for compiling under 64bit SUSE.
> >
> > I reviewed and adjusted the page a bit.
>
> Marcus, you seem to be a knowledgeable person how to set up X11 in a 64-bit
> environment. Perhaps you have a clue what might be needed to properly setup
> locale support in that case?
>
> See for instance the bug http://bugs.winehq.org/show_bug.cgi?id=4704

I never had any problems setting it up, in fact it just worked all the time :(
So I am likely not of much help.

Ciao, Marcus

[Dmitry Timoshkov]

"Marcus Meissner"  wrote:
> > Marcus, you seem to be a knowledgeable person how to set up X11 in a 64-bit
> > environment. Perhaps you have a clue what might be needed to properly setup
> > locale support in that case?
> >
> > See for instance the bug http://bugs.winehq.org/show_bug.cgi?id=4704
>
> I never had any problems setting it up, in fact it just worked all the time :(
> So I am likely not of much help.

Well, at least you could just describe how it's set up in your system.
It appears that some systems don't provide locale support in an emulated
32-bit environment at all.

[Dmitry Timoshkov]

"Dmitry Timoshkov"  wrote:
> > > Marcus, you seem to be a knowledgeable person how to set up X11 in a 64-bit
> > > environment. Perhaps you have a clue what might be needed to properly setup
> > > locale support in that case?
> > >
> > > See for instance the bug http://bugs.winehq.org/show_bug.cgi?id=4704
> >
> > I never had any problems setting it up, in fact it just worked all the time :(
> > So I am likely not of much help.
>
> Well, at least you could just describe how it's set up in your system.
> It appears that some systems don't provide locale support in an emulated
> 32-bit environment at all.

A follow up: a working solution for a 64-bit system was posted in the bug above:

<quote>
Eventually I found the suitable hint here
http://ubuntuforums.org/showpost.php?p=2160001&postcount=11 : For whatever
reason the respective environment variable is no longer being called XLOCALEDIR
but instead XLOCALELIBDIR... :-(

When setting

export XLOCALELIBDIR=/usr/lib32/X11/locale

everything works.
</quote>

[Markus]

On Monday 05 March 2007 15:23, Markus wrote:
> Apparently there is another patch that went in between 0.9.30 and .31 that
> causes an "Invalid call" error message to show after the start screen. I'll
> do a test for that later.

I've tried to track down the above error by going back up to 0.9.29 but 
without success. Hopefully the debug output gives some hints. It is filed 
under http://bugs.winehq.org/show_bug.cgi?id=7656

Regards,

[Markus]

On Wednesday 07 March 2007 23:57, Markus wrote:
> On Monday 05 March 2007 15:23, Markus wrote:
> > Apparently there is another patch that went in between 0.9.30 and .31
> > that causes an "Invalid call" error message to show after the start
> > screen. I'll do a test for that later.
>
> I've tried to track down the above error by going back up to 0.9.29 but
> without success. Hopefully the debug output gives some hints. It is filed
> under http://bugs.winehq.org/show_bug.cgi?id=7656

Concerning this bug, can anyone tell me what info is needed exactly?

Thanks.

[Kai Blin]

On Sunday 04 March 2007 20:41, Markus wrote:
> when trying the demo of the above game [... it quits ...] with the following
> messages: 
> fixme:wininet:InternetSetOptionExA Flags 04000000 ignored
> fixme:wininet:InternetSetOptionW Option INTERNET_OPTION_CONNECT_TIMEOUT
> (1000): STUB
> fixme:wininet:InternetSetOptionExA Flags 04000000 ignored
> fixme:wininet:InternetSetOptionW INTERNET_OPTION_SEND/RECEIVE_TIMEOUT not
> supported on protocol 1
> err:ntdll:RtlpWaitForCriticalSection section 0x110020 "heap.c: main process
> heap section" wait timed out in thread 0009, blocked by 000d, retrying (60
> sec)
>
> Could this be a regression in DirectPlay?

No, I really doubt that. dplay will not call wininet at all. Is there a bug 
report in bugzilla for this?

Cheers,
Kai