Vm Workstation 10.0 does not work on Ubuntu 13.10, kernel 3.11.0-13-generic x86_64

[fw114]

after WS 10 install and starting VM WS 10 i get

Failed to open device "/dev/vmci": Datei oder Verzeichnis nicht gefunden

Please make sure that the kernel module 'vmci' is loaded.

Module DevicePowerOn power on failed.

Failed to start the virtual machine

ok so ->

Starting VMware services:
   Virtual machine monitor                                             done
   Virtual machine communication interface                            failed
   VM communication interface socket family                           failed
   Blocking file system                                                done
   Virtual ethernet                                                    done
   VMware Authentication Daemon                                        done

ok next step ->

sudo vmware-modconfig --console --install-all

Stopping VMware services:
   VMware Authentication Daemon                                        done
   VM communication interface socket family                            done
   Virtual machine communication interface                             done
   Virtual machine monitor                                             done
   Blocking file system                                                done
Using 2.6.x kernel build system.
make: Gehe in Verzeichnis '/tmp/modconfig-Xh2oEK/vmmon-only'
/usr/bin/make -C /lib/modules/3.11.0-13-generic/build/include/.. SUBDIRS=$PWD SRCROOT=$PWD/. \
          MODULEBUILDDIR= modules
make[1]: Betrete Verzeichnis '/usr/src/linux-headers-3.11.0-13-generic'
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/linux/driver.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/linux/driverLog.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/linux/hostif.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/common/apic.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/common/comport.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/common/cpuid.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/common/hashFunc.o
/tmp/modconfig-Xh2oEK/vmmon-only/linux/driver.c:1342:1: warning: always_inline function might not be inlinable [-Wattributes]
LinuxDriverSyncReadTSCs(uint64 *delta) // OUT: TSC max - TSC min
^
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/common/memtrack.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/common/phystrack.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/common/task.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/common/vmx86.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmmon-only/vmcore/moduleloop.o
  LD [M]  /tmp/modconfig-Xh2oEK/vmmon-only/vmmon.o
  Building modules, stage 2.
  MODPOST 1 modules
  CC      /tmp/modconfig-Xh2oEK/vmmon-only/vmmon.mod.o
  LD [M]  /tmp/modconfig-Xh2oEK/vmmon-only/vmmon.ko
make[1]: Verlasse Verzeichnis '/usr/src/linux-headers-3.11.0-13-generic'
/usr/bin/make -C $PWD SRCROOT=$PWD/. \
          MODULEBUILDDIR= postbuild
make[1]: Betrete Verzeichnis '/tmp/modconfig-Xh2oEK/vmmon-only'
make[1]: »postbuild« ist bereits aktualisiert.
make[1]: Verlasse Verzeichnis '/tmp/modconfig-Xh2oEK/vmmon-only'
cp -f vmmon.ko ./../vmmon.o
make: Verlasse Verzeichnis '/tmp/modconfig-Xh2oEK/vmmon-only'
Using 2.6.x kernel build system.
make: Gehe in Verzeichnis '/tmp/modconfig-Xh2oEK/vmnet-only'
/usr/bin/make -C /lib/modules/3.11.0-13-generic/build/include/.. SUBDIRS=$PWD SRCROOT=$PWD/. \
          MODULEBUILDDIR= modules
make[1]: Betrete Verzeichnis '/usr/src/linux-headers-3.11.0-13-generic'
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/driver.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/hub.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/userif.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/netif.o
/tmp/modconfig-Xh2oEK/vmnet-only/hub.c: In function ‘VNetHubFindHubByID’:
/tmp/modconfig-Xh2oEK/vmnet-only/hub.c:132:49: warning: argument to ‘sizeof’ in ‘memcmp’ call is the same expression as the first source; did you mean to provide an explicit length? [-Wsizeof-pointer-memaccess]
         memcmp(idNum, currHub->id.pvnID, sizeof idNum))) {
                                                 ^
/tmp/modconfig-Xh2oEK/vmnet-only/hub.c: In function ‘VNetHubAlloc’:
/tmp/modconfig-Xh2oEK/vmnet-only/hub.c:315:36: warning: argument to ‘sizeof’ in ‘memcpy’ call is the same pointer type ‘uint8 *’ as the destination; expected ‘uint8’ or an explicit length [-Wsizeof-pointer-memaccess]
   memcpy(hub->id.pvnID, id, sizeof id);
                                    ^
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/bridge.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/filter.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/procfs.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/smac_compat.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/smac.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/vnetEvent.o
  CC [M]  /tmp/modconfig-Xh2oEK/vmnet-only/vnetUserListener.o
  LD [M]  /tmp/modconfig-Xh2oEK/vmnet-only/vmnet.o
  Building modules, stage 2.
  MODPOST 1 modules
  CC      /tmp/modconfig-Xh2oEK/vmnet-only/vmnet.mod.o
  LD [M]  /tmp/modconfig-Xh2oEK/vmnet-only/vmnet.ko
make[1]: Verlasse Verzeichnis '/usr/src/linux-headers-3.11.0-13-generic'
/usr/bin/make -C $PWD SRCROOT=$PWD/. \
          MODULEBUILDDIR= postbuild
make[1]: Betrete Verzeichnis '/tmp/modconfig-Xh2oEK/vmnet-only'
make[1]: »postbuild« ist bereits aktualisiert.
make[1]: Verlasse Verzeichnis '/tmp/modconfig-Xh2oEK/vmnet-only'
cp -f vmnet.ko ./../vmnet.o
make: Verlasse Verzeichnis '/tmp/modconfig-Xh2oEK/vmnet-only'
Starting VMware services:
   Virtual machine monitor                                             done
   Virtual machine communication interface                            failed
   VM communication interface socket family                           failed
   Blocking file system                                                done
   Virtual ethernet                                                    done
   VMware Authentication Daemon                                        done
Unable to start services

so what now ?

[fw114]

do i ask the wrong questions that i dont get any answers ? or is it so easy that i have to see it ?

[continuum]

your question has been asked in the right place - and no - it is not too easy ...

The WS 10 installer probably does not know how to handle such a new kernel - as always in such cases someone will sooner or later create a patch to make it work but you have to search for 3rd party patches like that.

As ugly workaround you could try to run WS without vmci - to do that you have to disable vmci in the configfile of all your existing VMs

[fw114]

Hi,

thank your for your answer. i tested virtualbox right now and have to say there are no kernel issues at all, also with the newest kernel.

now i have to see what issues will rise on daily basis with virtualbox.

what i dont understand is why vmware allways have to deal with kernel issues. the kernel issues suck that much, that i now installed virtualbox....

[kjansenz]

I have the same error with WS 9.0.2, WS9.0.3 and WS 10.0 on this kernel in ubuntu 13.10 with kernel 3.11.0-13-generic.

When I switch to the 3.8.0-12-generic kernel from ubuntu 13.04 everything compiles fine ans workstation-server works.

I tried the patches in other threads. The modules seem to compile fine, but in /var/log/messages, vmci is not happy.

But another problem arises, WS 9.0 and WS 10.0 run fine on the older kernel, but when you connect to workstation-server the ui crashes.

The log for the vm crash is available on vmware-ui error saucy salamander - Pastebin.com

[esotericbyte]

I've been running Fedora 19 kernel 3.11.1-200.fc19.x86_64 and it's got the same/similar problems running VMware workstation 9.

I really need my vm now so I'm just downgrading my kernel and also upgrading VMWare WS to 9.0.3.  I'm downgrading my kernel to 3.9 and staying there for awhile. My system has been unstable ever since kernel 3.10+ anyway. That might just be Fedora or graphics card issues or something that would not effect you guys, but I've wondered also if the patches are less that perfect fixes. No log messages would point to that but I've had system freezes that might not make it into the logs.

I'm really hoping 9.0.3 and downgrading kernel to 3.9 will solve the issue for me for now. If i need patches for vmw ws 9.0.3 to work on 3.9 I'm going to be very sad.

[update: no patches were needed to run 9.0.3 on kernel 3.9, and stable so far in host and guest, yey!  It was not as hard or time-consuming as i expected to downgrade the kernel make it the default boot and prevent the package manager from upgrading / changing this.]

I don't think I'll upgrade to 10 anytime soon.  I've toyed with the idea of running a current kernel in a VM and using a so-stable-it's-boring OS with conservative kernel updates as my host but that does not seem very practical because it affects all the libs and applications as well and just and seems so retro.  I wish vmware would deal with these issues more pro-actively.

It would be nice also if there was some decent distro that did not upgrade the kernel unless vmware modules at least had a patch available but both issues deal with the open-source / closed-source divide.

VMWare Kernel module issues has been a problem off and on for too long. I'm frustrated and thought of trying virtualbox or even just kvm as well.

[kjansenz]

For your information,

I have tracked the GUI problem down to a mismatched version of libgtkmm and libcurl, which were not updated when i moved from 12.04 to 13.04 and then to 13.10. Sorry for the confusion.

Furthermore ,downgrading to an older kernel made VMWare 9 and 10 run again.

Also, I don't think this is is a VMWare issue, but one with the faulty 3.11 kernel. Until now a have never experienced these problems with Ubuntu LTS editions.

I have followed the following steps to solve this issue on 13.10 for now;

1. Downgrade to kernel version 3.8.0-29-generic by downloading this one and the header files from the ubuntu 13.04 repo.

2. reboot to the older kernel and run make oldconfig in /usr/src/Linux-headers-3.8.0-29-generic to make sure that the config for the header files match the running kernel.

3. install VMWare Workstation 9.0.3

[bonnie201110141]

Have you tried Workstation10.0.1? It is free update for Workstation10.0.0. Workstation 10.0.1 supports Ubuntu 13.10 GA version.

[jbankstonga]

I bought WS 10.0.x when I saw that Ubuntu was releasing 14.04-LTS being pretty sure that WS would have to be upgraded. Much to my dismay, WS 10 wouldn't compile the network modules and I was sure I had been snarkled. Chasing this back, I settled on WS 9.0.4 and Ubuntu desktop 12.0.4.4-LTS running kernel 3.11.0-26 and runs fine. Any attempts to upgrade the kernel beyond this level renders the error in the VM network modules. I searched it back in Google and finally found where there's a change in 3.13 and above kernel in how the net modules are identified and compiled, but couldn't figure out how to apply the patch that supposedly bypasses the error.

I'm sticking at Ubuntu desktop 12.04.4 and kernel 3.11.0-26 with WS 9.0.4, everything is stable as a rock for me. This is my main work laptop, and I can't afford to have this thing unstable.

-Jeff

[bonnie201110141]

Let me double confirm, is it 3.11 kernel not working for WS 10.0.x or 3.13?