Bugzilla Bug 177876

Version 6 now supports Vista

Reproducible: Always

Steps to Reproduce:
1. emerge -v vmware-workstation
2.
3.

Actual Results:  
version 5.5.3.34685 is installed

Expected Results:  
version 6.0 installed

vmware.com

Just picked up the license for this.  If you want an ebuild tester, I'll help
test.

I just upgraded my license, and would be happy to be a tester as well for the
amd64 platform.

Created an attachment (id=119722) [details]
vmware workstation 6 amd64 ebuild

This ebuild works for me.  It ought to set MY_P based on the platform though
...

Note that with workstation 6 you don't want vmware-modules installed (at least
not until the installed modules match, which currently they don't).

Got a license also, willing to test ebuild on x86.

Created an attachment (id=120597) [details]
Output of emerge --info

Update: tried the amd64 ebuild posted above, changing the last part of MY_P to
.i386, and seems to work on x86.

A few errors were reported during installation due to difference in file names
from x86 and and64, doesn't appear to have caused a problem.

The EULA.txt file is not installed correctly. 'touch
/opt/vmware/workstation/lib/share/EULA.txt' solves this problem.

To get workstation to run, I have to run: $ VMWARE_USE_SHIPPED_GTK=yes vmware
otherwise this error is reported: /opt/vmware/workstation/lib/bin/vmware:
symbol lookup error:
/opt/vmware/workstation/lib/lib/libvmwareui.so.0/libvmwareui.so.0: undefined
symbol: _ZN3Gtk13RecentManager11get_defaultEv

emerge --info attached in case it'd be useful.

No /etc/init.d/vmware created on install.

Vmware Workstation icon in launcher did not work either, had to change it to
point to /opt/vmware/workstation/bin/vmware

I have run the x86 version with changing ebuild file; then linking doc/EULA.
The initscripts are installed in /etc/vmware/init.d

A fresh VM does not run properly. It gave errors about "Cannot create
ethernet0. too many pci devices" I removed the ethernet card from the vm.
Then, after the boot screen bar fills, the vm exits with an "Unexpected signal:
11" error. I do not know if it is a problem about my system or the ebuild
though.

emerge error with mm-sources kernel 2.6.22-rc1-mm1.  See attachments.

Bug described at the following url:
http://www.vmware.com/community/thread.jspa?messageID=656350

error synopsis
/opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:
In function 'VNetCopyDatagramToUser':
/opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:633:
error: 'const struct sk_buff' has no member named 'h'
/opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:633:
error: 'const struct sk_buff' has no member named 'nh'
/opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:639:
error: 'const struct sk_buff' has no member named 'h'

Created an attachment (id=121722) [details]
emerge --info for 2.6.22-rc1-mm1 problem

Created an attachment (id=121724) [details]
emerge build log for 2.6.22-rc1-mm1 problem

I am recreating this with 16 - I noticed that my overlay did not work.  Will
port update

(In reply to comment #8)
> emerge error with mm-sources kernel 2.6.22-rc1-mm1.  See attachments.
> 
> Bug described at the following url:
> http://www.vmware.com/community/thread.jspa?messageID=656350
> 
> error synopsis
> /opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:
> In function 'VNetCopyDatagramToUser':
> /opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:633:
> error: 'const struct sk_buff' has no member named 'h'
> /opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:633:
> error: 'const struct sk_buff' has no member named 'nh'
> /opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:639:
> error: 'const struct sk_buff' has no member named 'h'
> 

I have confirmed that this issue occurs with vmware-modules-1.0.0.16 as well.

(In reply to comment #11)
> I am recreating this with 16 - I noticed that my overlay did not work.  Will
> port update
> 
> (In reply to comment #8)
> > emerge error with mm-sources kernel 2.6.22-rc1-mm1.  See attachments.
> > 
> > Bug described at the following url:
> > http://www.vmware.com/community/thread.jspa?messageID=656350
> > 
> > error synopsis
> > /opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:
> > In function 'VNetCopyDatagramToUser':
> > /opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:633:
> > error: 'const struct sk_buff' has no member named 'h'
> > /opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:633:
> > error: 'const struct sk_buff' has no member named 'nh'
> > /opt/tmp/portage/portage/app-emulation/vmware-modules-1.0.0.15-r1/work/vmnet-only/userif.c:639:
> > error: 'const struct sk_buff' has no member named 'h'
> > 
> 

An issue with GTK:

http://www.vmware.com/community/message.jspa?messageID=634653

Stephen, that particular missing symbol has been fixed by requiring that gtkmm
be built with USE="accessibility", which is now enforced by the ebuild which is
currently masked in the main tree. Dev-zero just pointed out I need to add it
to the vmware-player-2 ebuild, which I'll be doing this weekend.

I have a few other things to fix up this weekend concerning the download
locations, after which it will be unmasked.  Save for downloading issues, the
ebuild is done and usable.  I hope this helps...  5:)

The 2.6.22 kernel issues are a different problem, and anyone interested should
keep an eye on bug 182595.  Thanks...  5:)

It would be nice if the ebuild told you that you need
vmware-any-any-update109.tar.gz   as well as the VMWare workstation tarball. 
Fetch restrictions are on and there is no indicator that this file is also
required.

I understand, there's another bug about that (bug 183221).  Sorry for the
delay, I'm still working on ensuring everything's ok pending unmasking.  Sadly
with the new kernel release, there's still a bit of work to do...

Trying to go full screen using VMWare Workstation 6 installed via the layman
overlay (from bug 182595) I have noticed that I can't switch to full screen
mode.  The program tries to switch and then just disappears.

I'm getting the following as the error:
/opt/vmware/workstation/lib/bin/vmware: symbol lookup error:
/opt/vmware/workstation/lib/bin/vmware: undefined symbol:
ViewAutoDrawer_SetOffset


A quick google has shown requests for libview-0.6.0 to be backported for Ubuntu
Fiesty to fix the error, so I'm guessing its to do with this.

Found a ebuild for libview-0.6.0 in bug 170683, which I can't seem to make
work, so the above is untested yet.

OK, I managed to get libview-0.6.0 installed and this fixed the full screening
issue.

All works brilliantly now.

Vinnie, could you please test that going back to libview-0.5.6-r1 brings back
the problem?  Often a recompile of the libview will solve several of the
missing symbol issues, and I've been using workstation-6 here fine with the old
version of libview.  Currently the only spontaneous crasher issue is the
snapshot manager...

Just downgraded libview and tried it out and it still crashes when changing
into full screen mode.

Went back to version 0.6.0 and fullscreen worked again

I installed vmware-workstation-6.0.0.45731 (from the vmware overlay) today, on
amd64. It installed fine, but it wouldn't even run with libview-0.5.6,
complaining about unresolved symbols. I installed libview-0.6.0 (bug #170683),
and now it runs fine.

I'm just about to join the testing brigade. I see that vmware-workstation-6 and
vmware-modules are still masked. I've unmasked them and I'm in the process of
testing them. I'll uninstall my vanilla copy from VMware first, which I've been
running since it first hit the market with no issues. :) Anything else I need
to worry about? Host is amd64 -- I have no plans to ever run this on an x86,
since my x86s don't have enough RAM to make it usable.:)

(In reply to comment #22)
> I'm just about to join the testing brigade. I see that vmware-workstation-6 and
> vmware-modules are still masked. I've unmasked them and I'm in the process of
> testing them. I'll uninstall my vanilla copy from VMware first, which I've been
> running since it first hit the market with no issues. :) Anything else I need
> to worry about? Host is amd64 -- I have no plans to ever run this on an x86,
> since my x86s don't have enough RAM to make it usable.:)
> 

Oh, yeah ... kernel is 2.6.22-r2

--------------------------- ACCESS VIOLATION SUMMARY
---------------------------
LOG FILE =
"/var/log/sandbox/sandbox-app-emulation_-_vmware-modules-1.0.0.16-1130.log"

open_wr:   /usr/src/linux-2.6.22-gentoo-r2/-.gcda
open_wr:   /usr/src/linux-2.6.22-gentoo-r2/-.gcda
open_wr:   /usr/src/linux-2.6.22-gentoo-r2/-.gcda
open_wr:   /usr/src/linux-2.6.22-gentoo-r2/-.gcda
open_wr:   /usr/src/linux-2.6.22-gentoo-r2/-.gcda
open_wr:   /usr/src/linux-2.6.22-gentoo-r2/-.gcda
--------------------------------------------------------------------------------
DreamScape ~ # 

Should I post the log file, or is this a known issue?

FEATURES='-sandbox' got it installed. It's coming up, but issues some error
messages on the console:

znmeb@DreamScape ~ $ /opt/vmware/workstation/bin/vmware
/usr/share/themes/Clearlooks/gtk-2.0/gtkrc:62: error: unexpected identifier
`animation', expected character `}'
Fontconfig warning: no <cachedir> elements found. Check configuration.
Fontconfig warning: adding <cachedir>/var/cache/fontconfig</cachedir>
Fontconfig warning: adding <cachedir>~/.fontconfig</cachedir>
/usr/share/themes/Clearlooks/gtk-2.0/gtkrc:62: error: unexpected identifier
`animation', expected character `}'

Ok, vmware-workstation 6 is unmasked.  vmware-player-2 is unmasked.  The
vmware-modules-1.0.0.16-r1 are using the vmware-any-any-update sources, and are
unmasked, so this bug is done.  There is a known issue with an old libview, but
there's a separate bug for that now (bug 192968) and there's now a libview in
the vmware overlay, which will hopefully be hitting the main tree in the next
couple of weeks.

Things to look foward to in the gentoo/vmware world include:

* open-vm-tools - one single set of tools for workstation/player/server that
are completely open source
* drag-and-drop - support is being developed for drag and drop in
workstation-6, player-2 and open-vm-tools

Thanks for your patience.  This bug is now finally FIXED.  5:)