| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
| |
All $Header$ lines are not anymore needed as indicated in
https://bugs.gentoo.org/show_bug.cgi?id=611234
|
| |
|
|
|
|
|
| |
All $Id$ lines are not anymore needed as indicated in
https://bugs.gentoo.org/show_bug.cgi?id=611234
|
|
|
|
| |
This should fix https://bugs.gentoo.org/show_bug.cgi?id=619746
|
|
|
|
|
|
|
| |
Let's clean the overlay keeping only last version of each major release,
this simplifies the management of the files. At a certain point it might
be even possible to import older versions that are now package-masked in
the main Gentoo repository.
|
| |
|
| |
|
|
|
|
|
| |
Excluded right now the associated version of vmware-tools because it has
the same name of previous 12.5.6. Is it changed anything?
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
built into kernel
The old init scripts fail when the vsock/vmci are built into kernel
not as modules. This should close the bug #510156. The systemd services
fail but don't block the initialization so I haven't touched them.
|
| |
|
| |
|
|
|
|
| |
This fixes https://github.com/gentoo/vmware/issues/30
|
|
|
|
| |
condition.
|
| |
|
|
|
|
|
|
|
|
| |
* Split the patch for kernel 4.6 in two parts, the first one has been
modified to be compatible with the changes in the next module 308.5.2
(so it's possible to share a single patch between 308.1 and 308.5)
* Fixed the patch for kernel 4.9 to be shared between 308.1 and 308.5
* Added patch for kernel 4.10 (https://github.com/gentoo/vmware/pull/29)
|
|\ |
|
|/
|
|
|
| |
4.9 introduced a change to the get_user_pages() API:
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=768ae309a96103ed02eb1e111e838c87854d8b51
|
| |
|
| |
|
|
|
|
| |
This reverts commit 84ce9f6be8f1b115b9a3c454f3d9d8b2ff6db253.
|
|
|
|
|
| |
upstream kernel commit 4b9d0fab7166c9323f06d708518a35cf3a90426c renamed
a symbol, add a patch to fix the problem. This solves the bug #591284.
|
| |
|
|\ |
|
| |
| |
| |
| |
| |
| |
| | |
Latest 4.7 kernel doesn't export anymore the readlink_copy() function:
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=bfe8804d908a791b16e3686c101f0d7eca9fb5b9
This fixes bug #590220
|
| | |
|
| |
| |
| |
| |
| | |
Fix compilation with latest 4.7-rc1 taking example from
kernel commit 860e9538a9482bb84589f7d0718a7e6d0a944d58
|
| |
| |
| |
| |
| |
| |
| |
| | |
The old patch 308-4.05-00-vmblock-follow_link.patch was not complete
(probably produced before final 4.5-rc1 release). The problem was
triggering a build failure on some systems.
The patch for version 304 was already correct.
|
| |
| |
| |
| | |
This closes https://github.com/gentoo/vmware/pull/25
|
| |
| |
| |
| | |
This closes https://github.com/gentoo/vmware/pull/23 and bug #583244
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| | |
Revert partially part of commit 050003bb7077b9121a81452e28779c343d1f0004.
Forcing embedded glib causes other problems when VMWARE_USE_SHIPPED_LIBS
is not set (bug #578956).
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Commit 052215b3c1e4b532b9d7a7872a0f46ecdb10f4cb has drastically reduced the
amount of not needed deps... introducing other problems.
1) The removal of gnome-base/librsvg exposes a problem with
libconf/lib/gtk-2.0/2.10.0/loaders/svg_loader.so (only with
bundled-libs), which is not able to find the bundled librsvg-2.so.2.
2) Some binaries (bin/vmware-vmx*) depends on libXinerama.so.1 and libXcursor.so.1
but with bundled-libs they are not able to find the embedded version and
fail if those libs are not installed in the system. patchelf --set-rpath doesn't work
with bin/vmware-vmw-debug so I moved the deps on libXinerama and libXcursor
to be always set.
Moreover, with gcc-5, bundled-libs is mandatory otherwise there is a problem
with symbols (https://bugs.gentoo.org/show_bug.cgi?id=578070#c8). Even
with VMWARE_USE_SHIPPED_LIBS set, appLoader tries to dynamically load at
runtime the best version of glib and fontconfig leading to a mix of
system/bundled libs. Let's fix some rpath to be sure that with
bundled-libs only the libs embedded with vmware are used!
Tested both on amd64 and ~amd64 systems.
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| | |
The dependencies were initially obtained from ldd which works
recursively so it shows also all the dynamically linked libs of
the primary DT_NEEDED libs... Basically ldd shows much more
dependencies, the new configuration has been tested on a freshly
built ~amd64 system.
|
|/ |
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| | |
This restrict means that libraries from *this* package are not preserved.
We have the exact opposite problem, we want that this package does not *cause*
other libraries to be preserved...
Package-Manager: portage-2.2.27
|
| |
| |
| |
| |
| |
| |
| | |
This isn't really clean; probably parts of the package still need
0.9.8... better than installing broken links though. More later.
Package-Manager: portage-2.2.27
|
| |
| |
| |
| |
| |
| | |
bug 559798
Package-Manager: portage-2.2.27
|
| |
| |
| |
| |
| |
| | |
on bug 559798
Package-Manager: portage-2.2.27
|
| |
| |
| |
| | |
Package-Manager: portage-2.2.28
|
| |
| |
| |
| | |
Package-Manager: portage-2.2.28
|
| |
| |
| |
| | |
Package-Manager: portage-2.2.28
|