[Bug 12363] New: Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Summary: Exact Audio Copy hangs early when in "native win32"
                    mode.
           Product: Wine
           Version: CVS/GIT
          Platform: Other
               URL: http://exactaudiocopy.de
        OS/Version: other
            Status: NEW
          Keywords: download
          Severity: normal
          Priority: P2
         Component: -unknown
        AssignedTo: [e-mail]
        ReportedBy: [e-mail]

This is the next problem after bug 12361 and bug 12359.

Setting "native win32" mode in EAC options
as described at http://forums.gentoo.org/viewtopic-t-386370.html
causes Exact Audio Copy to lock up on the next startup.
It seems to wait forever for an event inside a windows
message handler for an app-defined message.

The next time you start EAC, it defaults back to the non-crashy
ASPI method, but that doesn't find discs for me (bug 12359).

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #1 from Dan Kegel   2008-04-04 17:14:54 ---
Created an attachment (id=11854)
 --> (http://bugs.winehq.org/attachment.cgi?id=11854)
+relay,+message log of hang

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #2 from Dan Kegel   2008-04-04 17:48:43 ---
Lots of people running into this kind of thing.
http://ubuntuforums.org/showthread.php?t=142784&page=8

http://www.winehq.org/pipermail/wine-users/2007-June/027348.html
has a plausible recipe:

--- snip ---
1. Insert a data CD in the cd-drive. The OS should mount the CD.
2. Run winecfg and set up the CD drive in drives tab. Eg. F:, 
drive mapping=path = /media/cdrom, in advanced mode set 
type to CD-ROM. Apply, OK.
3. Run EAC while the DATA CD is inserted in the drive. Set the 
interface to 'native Win32 interface for WinNT/2000/XP'.  Exit 
EAC.
4. Run EAC again (the DATA CD is still inserted and mounted).
EAC should start and show the data CD. Replace the data CD 
with audio CD without exiting EAC.

A data CD has to be mounted when EAC is started and when 
winecfg is run.
--- snip ---
Unfortunately it still hangs for me when starting EAC after
setting native win32 interface in eac, either with today's
wine, wine-0.9.53, or wine-0.9.42 (though I didn't try reinstalling
with the older ones).

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Alexander Smith  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |alex.smith.ixium@googlemail.
                   |                            |com

--- Comment #3 from Alexander Smith   2008-04-06 13:49:24 ---
This is the bug I'm getting to with all recent versions of Wine (as described
here: http://xithium.wordpress.com/2008/04/04/41/). It happens no matter which
interface I use in Exact Audio Copy.
This does not happen with wine 0.9.46 for me. EAC works fine on that wine
version
Here's the output I get with wine 0.9.58:

--snip--
worf@ubuntu:~/.wine/drive_c/Program Files/Exact Audio Copy$ wine EAC.exe 
fixme:process:IsWow64Process (0xffffffff 0x6efba8) stub!
fixme:ntdll:NtSetInformationToken 0x3c 4 0x61166988 44
fixme:ntdll:NtSetInformationToken 0x3c 6 0x6ef920 4
fixme:advapi:ImpersonateLoggedOnUser ((nil))
fixme:advapi:ImpersonateLoggedOnUser ((nil))
fixme:advapi:ImpersonateLoggedOnUser (0xa4)
fixme:advapi:ImpersonateLoggedOnUser (0xc8)
fixme:advapi:ImpersonateLoggedOnUser (0xa4)
fixme:advapi:ImpersonateLoggedOnUser (0xa4)
fixme:aspi:SendASPI32Command ASPI: Partially implemented SC_HA_INQUIRY for
adapter 0.
err:aspi:SCSI_OpenDevice Failed to open device /dev/sg0: Permission denied
fixme:aspi:SendASPI32Command ASPI: Partially implemented SC_HA_INQUIRY for
adapter 1.
fixme:process:IsWow64Process (0xffffffff 0x6efba8) stub!
fixme:ntdll:NtSetInformationToken 0x3c 4 0x61166988 44
fixme:ntdll:NtSetInformationToken 0x3c 6 0x6ef920 4
fixme:advapi:ImpersonateLoggedOnUser ((nil))
fixme:advapi:ImpersonateLoggedOnUser ((nil))
fixme:advapi:ImpersonateLoggedOnUser (0xa4)
fixme:advapi:ImpersonateLoggedOnUser (0xc8)
fixme:advapi:ImpersonateLoggedOnUser (0xa4)
fixme:advapi:ImpersonateLoggedOnUser (0xa4)
fixme:mountmgr:harddisk_ioctl unsupported ioctl 4d014

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #4 from Austin English   2008-04-07 21:33:08 ---
(In reply to comment #3)
> This is the bug I'm getting to with all recent versions of Wine (as described
> here: http://xithium.wordpress.com/2008/04/04/41/). It happens no matter which
> interface I use in Exact Audio Copy.
> This does not happen with wine 0.9.46 for me. EAC works fine on that wine
> version
> Here's the output I get with wine 0.9.58:

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

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #5 from Alexander Smith   2008-04-09 09:47:57 ---
(In reply to comment #4)
> (In reply to comment #3)
> > This is the bug I'm getting to with all recent versions of Wine (as described
> > here: http://xithium.wordpress.com/2008/04/04/41/). It happens no matter which
> > interface I use in Exact Audio Copy.
> > This does not happen with wine 0.9.46 for me. EAC works fine on that wine
> > version
> > Here's the output I get with wine 0.9.58:
>
> Can you run a regression test:
> http://wiki.winehq.org/RegressionTesting

I followed those instructions to run a regression test (between versions 0.9.46
and 0.9.56, but nothing showed up. Each time I recompiled and tested EAC, the
bug didn't appear.
The only difference between the versions that do work, and the versions that
don't, is the following line in the output:
fixme:mountmgr:harddisk_ioctl unsupported ioctl 4d014

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #6 from Alexander Smith   2008-04-24 14:28:46 ---
Here's something interesting:
I tried the regression test again on Ubuntu 8.04 Hardy, but even when I
compiled 0.9.46 (the version that worked on Ubuntu Gutsy), the problem still
occurred. And the problem *did not* appear, even when I compiled 0.9.56 on
Gutsy (0.9.56 did not work on Fedora *).
So that got me thinking, the only major change I could see was the kernel
version (Gutsy used 2.6.22, Fedora & Hardy used 2.6.24) - so I booted a 2.4.22
kernel in Ubuntu Hardy with wine 0.9.59 and sure enough - it works!
In short, the version of wine does not seem to affect the bug, the kernel
version does. The bug appears on 2.6.24, but not 2.6.22.
I don't know if any of that helps (or even makes sense :D), but I thought I'd
report it anyway.

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #7 from Dan Kegel   2008-04-24 15:00:05 ---
Try http://wiki.winehq.org/PreloaderPageZeroProblem

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #8 from Alexander Smith   2008-04-24 15:31:00 ---
(In reply to comment #7)
> Try http://wiki.winehq.org/PreloaderPageZeroProblem

Just tried it, but it made no difference.

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Andrew Lau  changed:

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

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #9 from Austin English   2008-12-25 16:49:49 ---
Still present in git.

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
            Version|CVS/GIT                     |unspecified

--- Comment #10 from Austin English   2009-01-19 15:14:50 ---
Removing deprecated CVS/GIT version tag. Please retest in current git. If the
bug is still present in today's wine, but was not present in some earlier
version of wine, please update version field to earliest known version of wine
that had the bug. Thanks!

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Andrej  changed:

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

--- Comment #11 from Andrej   2010-06-18 07:35:20 ---
I can confirm this behaviour with Wine 1.2-rc3 and EAC v99pb5.

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Saulius K.  changed:

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

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Nick Bowler  changed:

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

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #12 from Nick Bowler  2010-09-12 13:15:49 CDT ---
I think I see what's wrong here.

I have drive "D:" configured to /mnt/cdrom, and the 'd::' dosdevices symlink
points to /dev/sr0.

With nothing mounted, the permissions on /mnt/cdrom are:

drwx------ 2 root root 4.0K Dec  9  2009 /mnt/cdrom

This normally isn't a problem, because when a filesystem is mounted, the
permissions on that filesystem supersede the permissions on the mount point.

However, audio CDs do not have filesystems which get mounted, so upon starting
EAC these permissions are still there, and wine hangs!  chmod go+rx /mnt/cdrom
allows EAC to start normally for me.

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Jan Buecken  changed:

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

--- Comment #13 from Jan Buecken  2010-09-13 10:04:49 CDT ---
(In reply to comment #12)
> I think I see what's wrong here.
>
> I have drive "D:" configured to /mnt/cdrom, and the 'd::' dosdevices symlink
> points to /dev/sr0.
>
> With nothing mounted, the permissions on /mnt/cdrom are:
>
> drwx------ 2 root root 4.0K Dec  9  2009 /mnt/cdrom
>
> This normally isn't a problem, because when a filesystem is mounted, the
> permissions on that filesystem supersede the permissions on the mount point.
>
> However, audio CDs do not have filesystems which get mounted, so upon starting
> EAC these permissions are still there, and wine hangs!  chmod go+rx /mnt/cdrom
> allows EAC to start normally for me.

Yeah, and this is the solution for bug 12359 !!! Thanks. Never thought the
mountpoint is relevant...

Hint: I changed the group of /mnt/cdrom to "cdrom" with chown root:cdrom
/mnt/cdrom additionally.

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Andrew Lau  changed:

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

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Martin Thierer  changed:

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

--- Comment #14 from Martin Thierer  2010-12-28 17:02:52 CST ---
(In reply to comment #12)
> However, audio CDs do not have filesystems which get mounted, so upon starting
> EAC these permissions are still there, and wine hangs!  chmod go+rx /mnt/cdrom
> allows EAC to start normally for me.

I couldn't believe this, but was desperate enough to try :) And it really
works...

Thanks a lot for the suggestion!

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

Rafal Stanilewicz  changed:

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

--- Comment #15 from Rafal Stanilewicz  2012-03-06 05:07:42 CST ---
Shouldn't wineserver user be added to some cduser group? Perhaps it would solve
such problems with many applications using audio cds...

Just thinking...

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #16 from Nick Bowler  2012-03-06 08:28:47 CST ---
(In reply to comment #15)
> Shouldn't wineserver user be added to some cduser group? Perhaps it would solve
> such problems with many applications using audio cds...
>
> Just thinking...

In this instance, the wineserver user *already* has permissions to access the
cdrom device, which should be sufficient.

There is no reason for read/execute permissions to be required on the mount
point when dealing with audio CDs.  There is *nothing there* to access.

OTOH, there's also no reason for the mount point to not be world
readable/executable, but it is nevertheless the default in several GNU/Linux
distributions.

At the very least, Wine should display an informative error message rather
than just hanging forever.

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

François  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |francois.melchior@belgacom.
                   |                            |net

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #17 from François  2012-03-21 15:04:22 CDT ---
The versions 1.0beta1 and 1.0beta2 of EAC were problematic since they had the
"native" mode set by default (so to use them one had to put a data disc into
the drive, mount it, launch EAC and change the option!)

The 1.0beta3 also has the "native" mode as default, but does not hang (it
starts but doesn't recognize the cd drive!)

I also wanted to know: is there any means to move a message from one bug to
another? because there is a lot of comments here that are in fact related to
bug 12359 (ASPI mode - that works - but problems with filesystem rights on the
CD mountpoint)

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #18 from Nick Bowler  2012-03-21 16:20:25 CDT ---
(In reply to comment #17)
> The versions 1.0beta1 and 1.0beta2 of EAC were problematic since they had the
> "native" mode set by default (so to use them one had to put a data disc into
> the drive, mount it, launch EAC and change the option!)

You should just have to change the permsisions of the mount point as described
in comment #12.  No need for this dance with a data cd.
> The 1.0beta3 also has the "native" mode as default, but does not hang (it
> starts but doesn't recognize the cd drive!)

Does it work with changing the permissions on the mount point as described
above?  I've not tested this since 0.99 prebeta 5, so it's possible that newer
versions have new issues.  However, those new issues are probably different
bugs from this one.
> I also wanted to know: is there any means to move a message from one bug to
> another? because there is a lot of comments here that are in fact related to
> bug 12359 (ASPI mode - that works - but problems with filesystem rights on the
> CD mountpoint)

To which comments are you referring?  This bug is about "native win32".  The
mount point permissions issue affects EAC in "native win32" mode.  Fixing the
mount point permissions allows EAC to work (at least with 0.99 prebeta 5) in
"native win32" mode.

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

http://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #19 from François  2012-03-21 18:28:12 CDT ---
1.Correction: EAC version 1.0beta1 has "ASPI" mode by default and so doesn't
show
this bug if we do not modify the configuration...

2.(In reply to comment #18) Thank you! I finaly managed to use the "mount point
method"! I had already tried it... but not enough, and so was convinced that
all that "mount point" stuff was related to the other bug!

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

https://bugs.winehq.org/show_bug.cgi?id=12363

Olibuntu  changed:

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

--- Comment #20 from Olibuntu  ---
The workaround from comment #12 does not to work for me in recent versions of
*Ubuntu, since there is no fixed mount point for optical drives any more, but
the mount point is created dynamically in /media when a medium is inserted.

Any ideas about this?

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

https://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #21 from Jan Buecken  ---
(In reply to Olibuntu from comment #20)
> The workaround from comment #12 does not to work for me in recent versions
> of *Ubuntu, since there is no fixed mount point for optical drives any more,
> but the mount point is created dynamically in /media when a medium is
> inserted.
>
> Any ideas about this?

Since audio cds never get mounted, I believe it just depends hard on
/mnt/cdrom.
You may try

Create the directory manually

sudo mkdir -p /mnt/cdrom

then apply the workaround

sudo chmod go+rx /mnt/cdrom

Now try to start EAC.
Not tested.

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

https://bugs.winehq.org/show_bug.cgi?id=12363

--- Comment #22 from Olibuntu  ---
(In reply to Jan Buecken from comment #21)
> Since audio cds never get mounted, I believe it just depends hard on
> /mnt/cdrom.
> You may try
>
> Create the directory manually
>
> sudo mkdir -p /mnt/cdrom
>
> then apply the workaround
>
> sudo chmod go+rx /mnt/cdrom
>
> Now try to start EAC.
> Not tested.

It does not work for me.

I couldn't find anything helpful in an strace output either.

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

[WineHQ Bugzilla]

https://bugs.winehq.org/show_bug.cgi?id=12363

Anastasius Focht  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
      Fixed by SHA1|                            |f556d0e93679a10281430835a6e
                   |                            |01ce7f12269bd
                URL|http://exactaudiocopy.de    |https://web.archive.org/web
                   |                            |/20091223200315/http://www.
                   |                            |exactaudiocopy.de/eac-0.99p
                   |                            |b4.exe
                 CC|                            |[e-mail]
             Status|NEW                         |RESOLVED
          Component|-unknown                    |kernel32
                 OS|other                       |Linux
            Version|unspecified                 |0.9.42.
           Hardware|Other                       |x86

--- Comment #23 from Anastasius Focht  ---
Hello folks,

EAC works fine in all modes here, including "native win32" mode.
Successfully ripped some tracks from audio CD.

I could reproduce the hang problem with old Wine versions as well.

--- snip ---
$ pwd
/home/focht/.wine/drive_c/Program Files (x86)/Exact Audio Copy

$ ll
total 40716
-rw-rw-r--. 1 focht focht   208896 Dec  5  2007 AccurateRip.dll
-rw-rw-r--. 1 focht focht      293 Jan  2 00:43 CDDB.DAT
-rw-rw-r--. 1 focht focht        0 Jan  2 00:38 CDFILES.DAT
-rw-rw-r--. 1 focht focht        0 Jan  2 00:43 CDQUEUE.DAT
drwxrwxr-x. 2 focht focht     4096 Jan  2 00:38 CDRDAO
drwxrwxr-x. 2 focht focht     4096 Jan  2 00:38 Documentation
-rwxrwxr-x. 1 focht focht  3245056 Jan 23  2008 EAC.exe
-rw-rw-r--. 1 focht focht        4 Jan  2 00:44 ENCQUEUE.DAT
drwxrwxr-x. 2 focht focht     4096 Jan  2 00:38 Flac
drwxrwxr-x. 2 focht focht     4096 Jan  2 00:38 Languages
-rw-rw-r--. 1 focht focht    38888 Jun 28  2007 Legal.rtf
-rw-rw-r--. 1 focht focht    55822 Jan 23  2008 News.rtf
drwxrwxr-x. 2 focht focht     4096 Jan  2 00:38 Profiles
-rw-rw-r--. 1 focht focht 37902524 Jan  2 00:43 Track01.wav
-rwxrwxr-x. 1 focht focht    39436 Jan  2 00:38 uninst.exe
-rw-rw-r--. 1 focht focht   161580 Oct 11  2007 WMCmd.vbs

$ wine ./EAC.exe
...
--- snip ---

I did a quick reverse regression test using prebuilt Wine versions, and it
seems this was fixed in between Wine 1.7.43 and Wine 1.7.44

Armed with that info it was an easy guess:
https://source.winehq.org/git/wine.git/commitdiff/f556d0e93679a10281430835a6...
("kernel32: Return correct device type for cd devices without medium."). Yes,
validated.

Thanks Michael

Part of Wine 1.7.44

$ sha1sum eac-0.99pb4.exe 
28d765cace5efa9c9fcc640887ac5ea66d471a78  eac-0.99pb4.exe

$ du -sh eac-0.99pb4.exe 
2.5M    eac-0.99pb4.exe

$ wine --version
wine-5.0-rc3-1-g742f317806

Regards

[Bug 12363] Exact Audio Copy hangs early when in "native win32" mode.

[WineHQ Bugzilla]

https://bugs.winehq.org/show_bug.cgi?id=12363

Alexandre Julliard  changed:

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

--- Comment #24 from Alexandre Julliard  ---
Closing bugs fixed in 5.0-rc4.