PREDECESSORS: Please, c.f. detailed information in Bug 596280 app-emulation/vmware-workstation-12.5.2 version bump and Bug 614666 (CVE-2017-4902, CVE-2017-4903, CVE-2017-4904, CVE-2017-4905, VMSA-2017-0006) app-emulation/vmware-workstation: multiple vulnerabilities Would be nice to get the proposed 'copy & rename' ebuild into overlay at least. Thanks.
FYI 12.5.6 is out; "What's new Day 0 support of the Windows 10 Creators Update (version 1703). End users can create virtual machines with the Windows 10 Creators Update or by installing VMware Workstation 12 Pro on a Windows 10 Creators Update operating system. Bug fixes and security updates." FWIW i have stability issues with a Windows10 guest, not fixed by Tools 10.1.6 5214329 (which is for both 12.5.5 and 12.5.6), so i'm interested in the version bump.
(In reply to Adam Carter from comment #1) > FYI 12.5.6 is out Build 5528349 Last updated: 18 MAY 2017 Release-Notes: http://pubs.vmware.com/Release_Notes/en/workstation/12/workstation-1256-release-notes.html Repository: http://softwareupdate.vmware.com/cds/vmw-desktop/ws/12.5.6/5528349/linux/
Corresponding vmware-tools for 12.5.4 have been . . . *.10.1.5.5055693..* [ in vmware overlay ] http://softwareupdate.vmware.com/cds/vmw-desktop/ws/12.5.4/5192485/linux/packages/ Corresponding vmware-tools for 12.5.5 should already have been . . . *.10.1.6-5214329.* http://softwareupdate.vmware.com/cds/vmw-desktop/ws/12.5.5/5234757/linux/packages/ As of today, corresponding vmware-tools for 12.5.6 stay the same: . . . *.10.1.6-5214329.* http://softwareupdate.vmware.com/cds/vmw-desktop/ws/12.5.6/5528349/linux/packages/
Created attachment 473694 [details] vmware-workstation-12.5.6.5528349.ebuild
Created attachment 473696 [details] vmware-modules-308.5.6.ebuild
Created attachment 473698 [details] vmware-tools-10.1.6.5214329-r1256.ebuild -r1256: Read: "vmware-tools-10.1.6.5214329 ebuild, download from 12.5.6 packages"
$ grep -i "Unable" vmware-apploader-26792.log 2017-05-21T20:24:21.367+02:00| appLoader| I125: Unable to load libfontconfig.so.1. /opt/vmware/lib/vmware/lib/libz.so.1/libz.so.1: version `ZLIB_1.2.9' not found (required by /usr/lib64/libpng16.so.16) 2017-05-21T20:24:21.370+02:00| appLoader| W115: Unable to load libfontconfig.so.1 from /opt/vmware/lib/vmware/lib/libfontconfig.so.1/libfontconfig.so.1: libexpat.so.0: cannot open shared object file: No such file or directory 2017-05-21T20:24:21.370+02:00| appLoader| W115: Unable to load dependencies for /opt/vmware/lib/vmware/lib/libvmware.so/libvmware.so 2017-05-21T20:24:21.370+02:00| appLoader| W115: Unable to execute vmware. [IP-] [ ] media-libs/libpng-1.6.29:0/16 [IP-] [ ] media-libs/fontconfig-2.12.1-r1:1.0 [IP-] [ ] sys-libs/zlib-1.2.11:0 http://www.zlib.net/ : Current release: zlib 1.2.11 January 15, 2017 Version 1.2.11 has these key improvements over 1.2.10: Fix deflate stored bug when pulling last block from window Permit immediate deflateParams changes before any deflate input Due to the bug fixes, any installations of 1.2.9 or 1.2.10 should be immediately replaced with 1.2.11.
Created attachment 473700 [details] vmware-apploader-26792.log
? libexpat.so.0 ? # equery list -p dev-libs/expat [IP-] [ ] dev-libs/expat-2.2.0-r1:0 # equery files dev-libs/expat | grep libexpat.so /usr/lib64/libexpat.so /usr/lib64/libexpat.so.1 /usr/lib64/libexpat.so.1.6.2 +++ WORKAROUND +++ # cd /usr/lib64/ # ln -s libexpat.so.1 libexpat.so.0 +++ WORKSFORME +++ Late; so only a very quick test: - start vmware - start / stop Win 7
Manfred's ebuilds work for me using the libexpat symlink, and these hacks required for earlier versions; cp /lib64/libz.so.1 /opt/vmware/lib/vmware/lib/libz.so.1/ unset VMWARE_USE_SHIPPED_LIBS && vmware &
(In reply to Adam Carter from comment #10) > Manfred's ebuilds work for me using the libexpat symlink, @ Adam: Thanks for testing! > and these hacks > required for earlier versions > cp /lib64/libz.so.1 /opt/vmware/lib/vmware/lib/libz.so.1/ > unset VMWARE_USE_SHIPPED_LIBS && vmware & Just for case classification: Are you on (more or less) stable or running "~amd64" in /etc/portage/make.conf ? gcc-5 ? bundled-libs ? Because I don't need those hacks; so for comparison: - mostly stable, - sys-devel/gcc-5.4.0-r3 has been stabilized by now, - USE=+bundled-libs - app-emulation/vmware-workstation:0 - app-emulation/vmware-modules:0 - app-emulation/vmware-tools:0 Would be nice to identify why some need those "hacks" and others don't ...
Reference: Bug 619392 - app-emulation/vmware-modules-308.5.6 : version bump Bug 619394 - app-emulation/vmware-tools-10.1.6.5214329-r1256 : version bump
> Just for case classification: > > Are you on (more or less) stable > or running "~amd64" in /etc/portage/make.conf ? ~amd64 > gcc-5 ? 6, as ~amd64 recently moved to this. > bundled-libs ? Yes
(In reply to Manfred Knick from comment #9) > ? libexpat.so.0 ? > +++ WORKAROUND +++ > > # cd /usr/lib64/ > > # ln -s libexpat.so.1 libexpat.so.0 With efferre79's version already introduced into the overlay including this (better) "hack" + dosym /$(get_libdir)/libz.so.1 \ + "${VM_INSTALL_DIR}"/lib/vmware/lib/libz.so.1/libz.so.1 my workaround above is _not_ needed any more; cdyou can # cd /usr/lib64/ # unlink /libexpat.so.0 to clean. Thanks to efferre79 for all his recent updates into Overlay!
(In reply to Manfred Knick from comment #14) > Thanks to efferre79 for all his recent updates into Overlay! Does it work for you? I have not recompiled yet my system for >=gcc-5 so I hope it works also for those configurations.
Perhaps you might need the WORKAROUND from Bug 619530 - app-emulation/vmware-workstation-12.5.*: after minor version bump: network connections "get lost" in order to re-gain working (bridged) networks. Thanks.
(In reply to Fabio Rossi from comment #15) > Does it work for you? First basic tests with Win7/8.1/10.1607 suggest "yes". MY SETUP: _no_ global "~amd64" in /etc/portage/make.conf USE="..." - kernel: 4.9.29 (longterm) with linux-firmware-20170519:0 - sys-apps/openrc-0.26.2:0 comment #11: > - mostly stable, > - sys-devel/gcc-5.4.0-r3 has been stabilized by now, > - USE=+bundled-libs > > - app-emulation/vmware-workstation:0 > - app-emulation/vmware-modules:0 > - app-emulation/vmware-tools:0 - complete "emerge -e" after upgrade to gcc-5.4.0-r3 # equery list -F '[$location] [$mask] $cpv:$slot [$repo]' 'vmware-*' [I-O] [ ] app-emulation/vmware-modules-308.5.6:0 [vmware] [I-O] [ ] app-emulation/vmware-tools-10.1.6.5214329:0 [vmware] [I-O] [ ] app-emulation/vmware-workstation-12.5.6.5528349:0 [vmware] No further "hacks" like . . . unset VMWARE_USE_SHIPPED_LIBS && vmware & required. If anybody is interested in additional information for comparison, please let me know.
PROGRESS: It seems that Bug 510156 app-emulation/vmware-workstation-11 "/etc/init.d/vmware stop" fails if vsock module is built into kernel does _not_ apply any more: # /etc/init.d/vmware stop # /etc/init.d/vmware start just worked as expected for me, no need to # /etc/init.d/vmware zap
(In reply to Manfred Knick from comment #18) > PROGRESS: > > It seems that > > Bug 510156 > app-emulation/vmware-workstation-11 > "/etc/init.d/vmware stop" fails if vsock module is built into kernel > > does _not_ apply any more: > > # /etc/init.d/vmware stop > > # /etc/init.d/vmware start > > just worked as expected for me, > no need to > > # /etc/init.d/vmware zap Yes, I fixed that too
*** Bug 596280 has been marked as a duplicate of this bug. ***
For me (~amd64 with +bundled-libs) after removing the libz and libexpat hacks, i'm back to just requiring unset VMWARE_USE_SHIPPED_LIBS; $ vmware /usr/share/themes/Adwaita/gtk-2.0/main.rc:733: error: unexpected identifier `direction', expected character `}' /usr/share/themes/Adwaita/gtk-2.0/hacks.rc:28: error: invalid string constant "normal_entry", expected valid string constant Loop on signal 11. $ unset VMWARE_USE_SHIPPED_LIBS && vmware & [1] 6505 $ /usr/share/themes/Adwaita/gtk-2.0/main.rc:733: error: unexpected identifier `direction', expected character `}' /usr/share/themes/Adwaita/gtk-2.0/hacks.rc:28: error: invalid string constant "normal_entry", expected valid string constant (vmware-tray:6240): Gtk-WARNING **: Unable to locate theme engine in module_path: "adwaita", /usr/share/themes/Adwaita/gtk-2.0/main.rc:733: error: unexpected identifier `direction', expected character `}' (vmware-tray:6240): Gtk-WARNING **: Unable to locate theme engine in module_path: "adwaita", /usr/share/themes/Adwaita/gtk-2.0/hacks.rc:28: error: invalid string constant "normal_entry", expected valid string constant Gtk-Message: Failed to load module "canberra-gtk-module": libcanberra-gtk-module.so: cannot open shared object file: No such file or directory --> this works FYI, if I rebuild with -bundled-libs the GUI doesn't start and no messages are printed on the console. The last file load in the strace output is /opt/vmware/lib/vmware/lib/libvmwareui.so/libvmwareui.so, and; $ ldd /opt/vmware/lib/vmware/lib/libvmwareui.so/libvmwareui.so | grep -i found libcds.so => not found libgcr.so.0 => not found libview.so.3 => not found libbasichttp.so => not found libgvmomi.so => not found libvmwarebase.so => not found All of those files are in directories in /opt/vmware/lib/vmware/lib/, so I tried adding those directories to ld.so.conf and running ldconfig -v, but there was an error on one of them; /opt/vmware/lib/vmware/lib/libcds.so: ldconfig: file /opt/vmware/lib/vmware/lib/libcds.so/libcds.so is truncated /opt/vmware/lib/vmware/lib/libgcr.so.0: libgcr.so.0 -> libgcr.so.0 /opt/vmware/lib/vmware/lib/libview.so.3: libview.so.3 -> libview.so.3 /opt/vmware/lib/vmware/lib/libbasichttp.so: libbasichttp.so -> libbasichttp.so /opt/vmware/lib/vmware/lib/libgvmomi.so: libgvmomi.so -> libgvmomi.so /opt/vmware/lib/vmware/lib/libvmwarebase.so: libvmwarebase.so -> libvmwarebase.so /opt/vmware/lib/vmware/lib/libgck.so.0: libgck.so.0 -> libgck.so.0
(In reply to Adam Carter from comment #21) > For me (~amd64 with +bundled-libs) after removing the libz and libexpat > hacks, i'm back to just requiring unset VMWARE_USE_SHIPPED_LIBS; > > $ vmware > /usr/share/themes/Adwaita/gtk-2.0/main.rc:733: error: unexpected identifier > `direction', expected character `}' > /usr/share/themes/Adwaita/gtk-2.0/hacks.rc:28: error: invalid string > constant "normal_entry", expected valid string constant > Loop on signal 11. > > $ unset VMWARE_USE_SHIPPED_LIBS && vmware & > [1] 6505 > $ /usr/share/themes/Adwaita/gtk-2.0/main.rc:733: error: unexpected > identifier `direction', expected character `}' > /usr/share/themes/Adwaita/gtk-2.0/hacks.rc:28: error: invalid string > constant "normal_entry", expected valid string constant > > (vmware-tray:6240): Gtk-WARNING **: Unable to locate theme engine in > module_path: "adwaita", > /usr/share/themes/Adwaita/gtk-2.0/main.rc:733: error: unexpected identifier > `direction', expected character `}' > > (vmware-tray:6240): Gtk-WARNING **: Unable to locate theme engine in > module_path: "adwaita", > /usr/share/themes/Adwaita/gtk-2.0/hacks.rc:28: error: invalid string > constant "normal_entry", expected valid string constant > Gtk-Message: Failed to load module "canberra-gtk-module": > libcanberra-gtk-module.so: cannot open shared object file: No such file or > directory > > --> this works > > FYI, if I rebuild with -bundled-libs the GUI doesn't start and no messages > are printed on the console. The last file load in the strace output is > /opt/vmware/lib/vmware/lib/libvmwareui.so/libvmwareui.so, and; > $ ldd /opt/vmware/lib/vmware/lib/libvmwareui.so/libvmwareui.so | grep -i > found > libcds.so => not found > libgcr.so.0 => not found > libview.so.3 => not found > libbasichttp.so => not found > libgvmomi.so => not found > libvmwarebase.so => not found > > All of those files are in directories in /opt/vmware/lib/vmware/lib/, so I > tried adding those directories to ld.so.conf and running ldconfig -v, but > there was an error on one of them; > /opt/vmware/lib/vmware/lib/libcds.so: > ldconfig: file /opt/vmware/lib/vmware/lib/libcds.so/libcds.so is truncated > > /opt/vmware/lib/vmware/lib/libgcr.so.0: > libgcr.so.0 -> libgcr.so.0 > /opt/vmware/lib/vmware/lib/libview.so.3: > libview.so.3 -> libview.so.3 > /opt/vmware/lib/vmware/lib/libbasichttp.so: > libbasichttp.so -> libbasichttp.so > /opt/vmware/lib/vmware/lib/libgvmomi.so: > libgvmomi.so -> libgvmomi.so > /opt/vmware/lib/vmware/lib/libvmwarebase.so: > libvmwarebase.so -> libvmwarebase.so > /opt/vmware/lib/vmware/lib/libgck.so.0: > libgck.so.0 -> libgck.so.0 With bundled-libs unset, there's a work around in this thread: https://forums.gentoo.org/viewtopic-t-1063880.html
(In reply to Weihua Zhou from comment #22) > (In reply to Adam Carter from comment #21) > > For me (~amd64 with +bundled-libs) after removing the libz and libexpat > > hacks, i'm back to just requiring unset VMWARE_USE_SHIPPED_LIBS; > > > > $ vmware > > /usr/share/themes/Adwaita/gtk-2.0/main.rc:733: error: unexpected identifier > > `direction', expected character `}' > > /usr/share/themes/Adwaita/gtk-2.0/hacks.rc:28: error: invalid string > > constant "normal_entry", expected valid string constant > > Loop on signal 11. > > > > $ unset VMWARE_USE_SHIPPED_LIBS && vmware & > > [1] 6505 > > $ /usr/share/themes/Adwaita/gtk-2.0/main.rc:733: error: unexpected > > identifier `direction', expected character `}' > > /usr/share/themes/Adwaita/gtk-2.0/hacks.rc:28: error: invalid string > > constant "normal_entry", expected valid string constant > > > > (vmware-tray:6240): Gtk-WARNING **: Unable to locate theme engine in > > module_path: "adwaita", > > /usr/share/themes/Adwaita/gtk-2.0/main.rc:733: error: unexpected identifier > > `direction', expected character `}' > > > > (vmware-tray:6240): Gtk-WARNING **: Unable to locate theme engine in > > module_path: "adwaita", > > /usr/share/themes/Adwaita/gtk-2.0/hacks.rc:28: error: invalid string > > constant "normal_entry", expected valid string constant > > Gtk-Message: Failed to load module "canberra-gtk-module": > > libcanberra-gtk-module.so: cannot open shared object file: No such file or > > directory > > > > --> this works > > > > FYI, if I rebuild with -bundled-libs the GUI doesn't start and no messages > > are printed on the console. The last file load in the strace output is > > /opt/vmware/lib/vmware/lib/libvmwareui.so/libvmwareui.so, and; > > $ ldd /opt/vmware/lib/vmware/lib/libvmwareui.so/libvmwareui.so | grep -i > > found > > libcds.so => not found > > libgcr.so.0 => not found > > libview.so.3 => not found > > libbasichttp.so => not found > > libgvmomi.so => not found > > libvmwarebase.so => not found > > > > All of those files are in directories in /opt/vmware/lib/vmware/lib/, so I > > tried adding those directories to ld.so.conf and running ldconfig -v, but > > there was an error on one of them; > > /opt/vmware/lib/vmware/lib/libcds.so: > > ldconfig: file /opt/vmware/lib/vmware/lib/libcds.so/libcds.so is truncated > > > > /opt/vmware/lib/vmware/lib/libgcr.so.0: > > libgcr.so.0 -> libgcr.so.0 > > /opt/vmware/lib/vmware/lib/libview.so.3: > > libview.so.3 -> libview.so.3 > > /opt/vmware/lib/vmware/lib/libbasichttp.so: > > libbasichttp.so -> libbasichttp.so > > /opt/vmware/lib/vmware/lib/libgvmomi.so: > > libgvmomi.so -> libgvmomi.so > > /opt/vmware/lib/vmware/lib/libvmwarebase.so: > > libvmwarebase.so -> libvmwarebase.so > > /opt/vmware/lib/vmware/lib/libgck.so.0: > > libgck.so.0 -> libgck.so.0 > > With bundled-libs unset, there's a work around in this thread: > https://forums.gentoo.org/viewtopic-t-1063880.html I have just added latest 12.5.7 version into the overlay. Now vmware should work (at least here on my stable system) both with +bundled-libs and -bundled-libs. Please test it! If everything works it will be soon added to portage to remove the current package mask situation.
(In reply to Fabio Rossi from comment #23) Thanks, Fabio! Would you be so kind to copy-rename . . . app-emulation/vmware-tools-10.1.6.5214329:0::vmware to . . . app-emulation/vmware-tools-10.1.7.5813279:0::vmware please? Only difference to adapt: - BASE_URI="http://softwareupdate.vmware.com/cds/vmw-desktop/ws/ 12.5.6/5528349/linux/packages/" + BASE_URI="http://softwareupdate.vmware.com/cds/vmw-desktop/ws/ 12.5.7/5813279/linux/packages/" I just checked the page; it consistently contains: metadata.xml.gz vmware-tools-freebsd-10.1.6-5214329.x86_64.component.tar vmware-tools-linux-10.1.6-5214329.x86_64.component.tar vmware-tools-linuxPreGlibc25-10.1.6-5214329.x86_64.component.tar vmware-tools-netware-10.1.6-5214329.x86_64.component.tar vmware-tools-solaris-10.1.6-5214329.x86_64.component.tar vmware-tools-winPre2k-10.1.6-5214329.x86_64.component.tar vmware-tools-winPreVista-10.1.6-5214329.x86_64.component.tar vmware-tools-windows-10.1.6-5214329.x86_64.component.tar
Sorry - only just realized that it's still "-10.1.6.5214329." filenames in the "/12.5.7/5813279/" download directory !
(In reply to Manfred Knick from comment #25) > Sorry - only just realized that it's still > > "-10.1.6.5214329." filenames in the > "/12.5.7/5813279/" download directory ! right, for that reason I didn't bump vmware-tools
CONFIRMATION: First quick test, still with +bundled-libs: + VM Win7 starts + after reboot, both networks working proper directly (c.f. comment 16) + MSE updates + M$ updates + Tools updates + M$ Defrag + VMware Disk Defrag & Compact (to be continued with -bundled-libs / 8.1 / 10 )
Reference: VMware Workstation 12 Pro Version 12.5.7 | 22 June 2017 | Build 5813279 "This release includes some bug fixes and security updates." http://pubs.vmware.com/Release_Notes/en/workstation/12/workstation-1257-release-notes.html
(In reply to Manfred Knick from comment #27) > CONFIRMATION: > ... > (to be continued with -bundled-libs / 8.1 / 10 ) Re-built with "-bundled-libs", /etc/init.d/vmware { stop / start } , chgrp vmware /dev/vmnet* chmod g+rw /dev/vmnet* both ( 8.1 / 10 ) + start + with both networks kept + Defender updates + M$ updates / maintenance + VMware Tools (complete) upgrade - 10 shutdown takes a looong time ?
> I have just added latest 12.5.7 version into the overlay. Now vmware should > work (at least here on my stable system) both with +bundled-libs and > -bundled-libs. > > Please test it! If everything works it will be soon added to portage to > remove the current package mask situation. Thank you for updating the overlay!! :) I am having trouble using the 12.5.7 ebuild from the overlay. i emerged vmware-workstation with USE="-bundled-libs -cups", then ran 'env-update && source /etc/profile'. after reading the above comments, i wanted to make sure i did not have any "not found" libs.. i ran: ldd /opt/vmware/lib/vmware/lib/*/*.so* | grep -i "error\|found" which showed a few not found. i noticed ld.so.conf did not include either of the vmware paths, /opt/vmware/lib or /opt/vmware/lib/vmware/lib. /etc/env.d/90vmware did not contain any LDPATH, and only /opt/vmware/bin for ROOT and ROOTPATH. so i ran it with: LD_LIBRARY_PATH="${LD_LIBRARY_PATH}:/opt/vmware/lib:/opt/vmware/lib/vmware/lib" ldd /opt/vmware/lib/vmware/lib/*/*.so* | grep -i "error\|found" https://paste.pound-python.org/show/M4LUdlLp9teVDw91W5dE/ it would seem files that belong in /opt/vmware/lib and /opt/vmware/bin etc., are being installed into /opt/vmware/lib/vmware/lib and /opt/vmware/lib/vmware/bin. also the libs are placed inside a folder with the same name as the file, while ldd looks for the folder path, not the file inside. i.e. the paste above, and symlinks like "/opt/vmware/bin/vmware -> /opt/vmware/lib/vmware/bin/vmware"
(In reply to rjp421 from comment #30) > I am having trouble using the 12.5.7 ebuild from the overlay. ? ............^^^^^^^.^^^^^.^^^................................ ? Sorry, I could not understand which trouble you have in _using_ _which_ ... # ldd /opt/vmware/bin/* | grep -i "error\|found" does not detect anything at all. -- My USE="-bundled-libs cups doc ovftool server vix vmware-tools"
I apologize, my post was (In reply to Fabio Rossi from comment #23).
(In reply to Fabio Rossi from comment #23) It appears that my first comment (#30) is more of an upstream issue, as their installer bundle not only installs aimlessly into /usr/lib64/vmware/, but has a similarly confusing file structure like the one in /opt/vmware/ after installing the ebuild. this is a crude list of files after installing the bundle manually, please note the vmware file locations, compared to where the ebuild puts them. https://paste.pound-python.org/show/gwSEymX7xI7xSeCS8FrH/ also the /usr/bin/vmware is a bash script, not a symlink to the "other" vmware symlink to appLoader. the bash script runs the bootstrap and sets the bindir and libdir. installing from the bundle, vmware workstation opens. installing from the ebuild, it does not open. (In reply to Manfred Knick from comment #31) sorry, i meant installing, not USEing :)
(In reply to rjp421 from comment #33) > It appears that my first comment (#30) is more of an upstream issue, as > their installer bundle not only installs aimlessly into /usr/lib64/vmware/, Yes, indeed. That's part of the trouble Andreas (= dilfridge) and others have been fighting with for ages now ... > but has a similarly confusing file structure like the one in /opt/vmware/ > after installing the ebuild. Actually, the other way round ... > installing from the bundle, vmware workstation opens. > installing from the ebuild, it does not open. Please, disclose if you are working on "stable" or have "~amd64" enabled. @ Adam Carter , @ Weihua Zhou : ! Please NOTE: ! My BUG (this one = Bug 616958) is intended against "stable" _ONLY_ . ! ! Please, do _not_ hijack this Bug for "~amd64". Please, please: Open _separate_ bugs for "~amd64" issues. During the last years, VMware Bug reports have been extremely difficult to track because all different issues were lumped together under one and the same headline.
(In reply to Manfred Knick from comment #34) > (In reply to rjp421 from comment #33) > > > It appears that my first comment (#30) is more of an upstream issue, as > > their installer bundle not only installs aimlessly into /usr/lib64/vmware/, > Yes, indeed. That's part of the trouble Andreas (= dilfridge) and others > have been fighting with for ages now ... > > > but has a similarly confusing file structure like the one in /opt/vmware/ > > after installing the ebuild. > Actually, the other way round ... > > > installing from the bundle, vmware workstation opens. > > installing from the ebuild, it does not open. > Please, disclose if you are working on "stable" or have "~amd64" enabled. > > @ Adam Carter , > @ Weihua Zhou : > > ! Please NOTE: > ! My BUG (this one = Bug 616958) is intended against "stable" _ONLY_ . > ! > ! Please, do _not_ hijack this Bug for "~amd64". > > Please, please: Open _separate_ bugs for "~amd64" issues. > > During the last years, VMware Bug reports have been extremely difficult to > track > because all different issues were lumped together under one and the same > headline. Thank you for the reply and info! I am using "stable" as far as i know, with no ~amd64 for vmware-*.
(In reply to rjp421 from comment #35) > (In reply to Manfred Knick from comment #34) > > (In reply to rjp421 from comment #33) > > > > > It appears that my first comment (#30) is more of an upstream issue, as > > > their installer bundle not only installs aimlessly into /usr/lib64/vmware/, > > Yes, indeed. That's part of the trouble Andreas (= dilfridge) and others > > have been fighting with for ages now ... > > > > > but has a similarly confusing file structure like the one in /opt/vmware/ > > > after installing the ebuild. > > Actually, the other way round ... > > > > > installing from the bundle, vmware workstation opens. > > > installing from the ebuild, it does not open. > > Please, disclose if you are working on "stable" or have "~amd64" enabled. > > > > @ Adam Carter , > > @ Weihua Zhou : > > > > ! Please NOTE: > > ! My BUG (this one = Bug 616958) is intended against "stable" _ONLY_ . > > ! > > ! Please, do _not_ hijack this Bug for "~amd64". > > > > Please, please: Open _separate_ bugs for "~amd64" issues. > > > > During the last years, VMware Bug reports have been extremely difficult to > > track > > because all different issues were lumped together under one and the same > > headline. > > > Thank you for the reply and info! > > I am using "stable" as far as i know, with no ~amd64 for vmware-*. i was mrong, package.accept_keywords still has vmware workstation, modules, and tools keyworded with ~amd64. required by: https://github.com/gentoo/vmware/blob/master/app-emulation/vmware-workstation/vmware-workstation-12.5.7.5813279.ebuild i'm sorry again, i thought this was the bug for bumping to 12.5.7. is there a different thread to troubleshoot getting the version bumped? thanks!
(In reply to rjp421 from comment #36) > (In reply to rjp421 from comment #35) > i was mrong, Relax! Clarification: With "~amd64", we are talking about having this enabled _globally_ for _all_ packages in /etc/portage/make.conf (="bleeding edge"). You have no chance to emerge without "~" settings for the VMware packages, as you described correctly. Question remains: @ your place: "installing from the ebuild, it does not open" @ my place: "WORKSFORME" I'll attach a current "info" for you so you can compare:
Created attachment 480588 [details] emerge --info vmware-workstation > emerge--info__vmware-workstation.txt For rjp421 esp.
You are welcome to attach yours. BTW: Still don't know where you are having trouble with - ? Workstation ? ? Player ?
(In reply to Manfred Knick from comment #39) > You are welcome to attach yours. > BTW: Still don't know where you are having trouble with - > ? Workstation ? > ? Player ? Sorry for the delay, here is the emerge --info vmware-{workstation,modules,tools} https://paste.pound-python.org/show/AfcotKZfBx1YGSfeXbXj/ # equery list -F '[$location] [$mask] $cpv:$slot [$repo]' 'vmware-*' * Searching for vmware-* ... [I-O] [ ] app-emulation/vmware-modules-308.5.7:0 [vmware] [I-O] [ ] app-emulation/vmware-tools-10.1.6.5214329:0 [vmware] [I-O] [ ] app-emulation/vmware-workstation-12.5.7.5813279:0 [vmware] i am using workstation, not player, yet.. i could not find the apploader logs last time :\ but this time, still fails with libgdkmm or pangomm. running 'vmware' as user rjp instantly drops back to cmd with no output. /tmp/vmware-rjp/vmware-apploader-26728.log: https://paste.pound-python.org/show/x621TzmejCg1cxxz2por/ appLoader| I125: Loading shipped version of libgdkmm-2.4.so.1. appLoader| W115: Unable to load libgdkmm-2.4.so.1 from /opt/vmware/lib/vmware/lib/libgdkmm-2.4.so.1/libgdkmm-2.4.so.1: /usr/lib64/libpangomm-1.4.so.1: undefined symbol: _ZN4Glib6ObjectC2EOS0_ appLoader| W115: Unable to load dependencies for /opt/vmware/lib/vmware/lib/libvmware.so/libvmware.so appLoader| W115: Unable to execute vmware. this is with USE: vmware-workstation -bundled-libs vmware-modules -vmci -vsock is this part of the earlier issue, mentioned in the link from comment #22? also, can/should we create a separate bug for the existing issues? i am new here, and did not mean to turn your version bump into a support ticket :) i am rjp421 in #gentoo and #vmware on freenode if you prefer it thank you again for the assistance!
(In reply to rjp421 from comment #40) > (In reply to Manfred Knick from comment #39) > Sorry for the delay, here is the emerge --info > vmware-{workstation,modules,tools} > > https://paste.pound-python.org/show/AfcotKZfBx1YGSfeXbXj/ Already the very first lines: <cite> Portage 2.3.6 (python 2.7.12-final-0, hardened/linux/amd64, gcc-5.4.0, glibc-2.23-r4, 4.7.10-pentoo x86_64) ================================================================= System Settings ================================================================= System uname: Linux-4.7.10-pentoo-x86_64-Intel-R-_Core-TM-_CPU_870_@_2.93GHz-with-gentoo-2.3 </cite> state that you are _not_ working "Gentoo stable". > pentoo > location: /var/lib/layman/pentoo > masters: gentoo > priority: 50 You yourself must have installed Pentoo Layman Overlay, noting that "The Pentoo kernel includes grsecurity and PAX hardening and extra patches - with binaries compiled from a hardened toolchain with the latest nightly versions of some tools available." > appLoader| I125: Loading shipped version of libgdkmm-2.4.so.1. > appLoader| W115: Unable to load libgdkmm-2.4.so.1 from > /opt/vmware/lib/vmware/lib/libgdkmm-2.4.so.1/libgdkmm-2.4.so.1: > /usr/lib64/libpangomm-1.4.so.1: undefined symbol: _ZN4Glib6ObjectC2EOS0_ This reminds me @ former C++ issues. HINT: "libgdkmm-2.4.so.1" is probably supplied by "dev-cpp/gtkmm-2.24.5:2.4" - which is neither installed nor requested on my box. Further on, my --info does _not_ contain USE="gtk". > also, can/should we create a separate bug for the existing issues? ............^^^^^^..............................^^^ Managing VMware's code base (depending on libs some generations ago, compared to current gentoo stable) on "hardened", I dare say you perhaps might find yourself awakening on quite a rocky road. Sorry, but as far as I'm concerned, you are on your own. For your REFERENCE: Last year's Bug 592614 is about "hardened" (-> Hank Leininger). P.S.: It would have been helpful to publish your "--info" right from the beginning ;-) NOTABENE: This case does _not_ block bumping version 12.5.7 into Main Portage Tree.
(In reply to Manfred Knick from comment #41) > Sorry, but as far as I'm concerned, you are on your own. > Please do not go extreme and do not abandon our (Pentoo) users. Pentoo kernel is almost exactly the same as hardened-sources (with extra wifi patches as you already learned) We keep it in sync with the current stable hardened version so you should look at it as "stable" hardened Gentoo env. @Manfred feel free to install hardened-sources and post a new --info output to avoid such comments in future. P.S. sorry for the offtopic, I did not start it. Back to the topic: Manfred has spent several days on this issue and tested different setups and compared bundle install against ebuilds results. Please double check his outcome. Hopefully, it will help improve the ebuild (and make it work). Thank you.
(In reply to Anton Bolshakov from comment #42) @ Anton: !!! No offense intended at all !!! > > ... as far as I'm concerned, ... This is just _honest_ what I _personally_ can not supply. I would never dare to speak for others or for Gentoo VMware Project in general. (I am not the maintainer of this project.) > Please do not go extreme and do not abandon our (Pentoo) users. I do neither - nor. > Pentoo kernel is almost exactly the same as hardened-sources "hardened" is the point. (rjp421's "Linux-4.7.10" is not supported by Linux itself any more.) > @Manfred feel free to install hardened-sources and post a new --info output > to avoid such comments in future. Please, understand that - as probably most Gentoo supporters - I contribute in my limited spare free time. I don't have the ressources to "support" hardened. It would be wonderful if you (or anybody else?) would step in :-)
@ rjp421 : Suggestion: "Differential Diagnostics" : 1.) Make it work without Pentoo overlay: sys-kernel/gentoo-sources-4.9.36 (In case of errors, you might re-consider your special CFLAGS.) 2.) Install latest sys-kernel/hardened-sources-4.9.24 In https://bugs.gentoo.org/show_bug.cgi?id=592614#c4, Anton points to his former Bug #382793 (CANTFIX) containing Alex' valuable hint in https://bugs.gentoo.org/show_bug.cgi?id=382793#c25 : switch off these two PaX options: [ ] Enforce non-executable kernel pages [ ] Randomize kernel stack base Does this work with current VMware packages? If not: Please file a _separate_ Bug Report containing "[Hardened]" If yes: proceed to 3.) Re-Install Pentoo overlay Be aware: There are no pentoo-sources-4.9.xx yet. Latest: pentoo-sources-4.8.17-r2.ebuild Latest supported: pentoo-sources-4.4.8-r2.ebuild [https://www.kernel.org/] [https://github.com/pentoo/pentoo-overlay/tree/master/sys-kernel/pentoo-sources] Hope this helps! All the best Manfred
Created attachment 481518 [details] emerge--info__vmware__gentoo.txt
Created attachment 481520 [details] emerge--info__vmware__hardened.txt
Created attachment 481522 [details] emerge--info__vmware__gentoo__hardened__diff.txt
@ rjp421 @ Anton Bolshakov As a one-time-courtesy, I underwent installing latest ... sys-kernel/gentoo-sources-4.9.36 and ... sys-kernel/hardened-sources-4.9.24 with it's supplied 'default' options. Concerning the latter one: A) Compiling, it threw thousands of "mismatch" warnings. B) Re-emerging vmware-{workstation,modules,tools} succeeded. C) Installing vmware-drivers failed, even for multiple versions backwards. Thus I could not start up the specially crafted X11 setup (multiple driver-incompatible GPU / multiple separate-X11-screens in "Zaphod" mode), thus cannot report about VMware Workstation "starting" behaviour. Nevertheless, above, please find attached the requested respective "--info" outputs. ! Note : The "diff" is quasi empty ! ... pentoo-sources-4.7.10 and ... pentoo-sources-4.8.17-r2 are up to you. HTH. -- - I ask your understanding that I can not hijack stable production boxes - for {hardened|Pentoo} experiments again. -- For everything else concerning "hardened" or "Pentoo", _please_ file separate bugs, naming names in the title right from the beginning. I guess if this higly closed-source SW shall work within, somebody with 'hardened' experience has to step in and contribute: VMware Inc. will not deliver that f(l)avour. @ Anton Bolshakov: FYI: Please c.f. comment 2 as well as "Supported host operating systems for Workstation Pro 12.x" [ https://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2129859 ]. Kernel 4.9.0 was explicitly authorized in "12.5.3 Release Notes" under "Resolved Issues": [ http://pubs.vmware.com/Release_Notes/en/workstation/12pro/workstation-1253-release-notes.html ] Thanks.
Created attachment 481746 [details] emerge--info__vmware__hardened.txt
(In reply to Manfred Knick from comment #48) > > I underwent installing latest > ... sys-kernel/gentoo-sources-4.9.36 and > ... sys-kernel/hardened-sources-4.9.24 with it's supplied 'default' options. > it is not so simple to switch between default and hardened profiles. See https://wiki.gentoo.org/wiki/Hardened_Gentoo#Switching_to_a_Hardened_profile in short: you will need to recompile "world". P.S. I'm not interested in this bug and not subscribed. So I might not see replies and might not reply myself.
(In reply to Anton Bolshakov from comment #50) > it is not so simple to switch between default and hardened profiles. > See ... #Switching_to_a_Hardened_profile That's exactly why (comment #43) > I don't have the ressources to "support" hardened. My limited goal was, to support _your_ (!) request (comment #42) : > @Manfred feel free to install hardened-sources and post a new --info output with my attachments (comment #48) > please find attached the requested > respective "--info" outputs. > in short: you will need to recompile "world". NO. I myself neither "need" nor "will". > P.S. I'm not interested in this bug ...........^^^.^^^^^^^^^^
REFERENCE: Bug 623652 [vmware-overlay] app-emulation/vmware-modules-308.5.7 patch for kernel-4.12.0 CHECKED: As expected, no "Version 12.5.8" containing such a fix has been released by VMware yet.
INFORMATION // CONFIRMATION : # gcc-config --list-profiles [1] x86_64-pc-linux-gnu-5.4.0 [2] x86_64-pc-linux-gnu-6.3.0 * <--- # emerge -e ... system world . . . 12.5.7.5813279 WORKSFORME running multiple VMs concurrently, *stable* for some time already, no negative effects observed so far.
(In reply to Manfred Knick from comment #53) > INFORMATION // CONFIRMATION : Same applies to . . . sys-devel/gcc-6.4.0:6.4.0 now.
OBSOLETE now: 14.0.0 has been released REFERENCEs to new bugs: Bug 634770 - app-emulation/vmware-workstation-14.0.0 version bump [for "stable", excluding "~amd64"] Bug 634862 - app-emulation/vmware-modules ... version bump Bug 634854 - app-emulation/vmware-tools-10.1.15-6627299 : version bump
REFERENCE : successor: Bug 637922 - app-emulation/vmware-{workstation, player}-12.5.8 version bump [VMSA-2017-0018] multiple security vulnerabilities
