| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Blender builds against version 7.0.0, but does not build against
version 7.1.0 when linking against the openvdb headers,
complaining that make_unique is not in std, even though the file
includes memory.
Given that 7.0.0 is the last version of openvdb that still includes
ABI 5 compatiblity, it might be worthwhile to keep it around.
Signed-off-by: Adrian Grigo <agrigo2001@yahoo.com.au>
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This bumps the version of openvdb to 7.1.0
Upstream have deprecated ABI prior to 6 with this version.
It adds an abi7-compat USE flag to build against ABI 7, and makes use
of avx and sse4.2 cpu flags if set. There are additional use flags to
control building the utilities, and linking against numpy is now
optional. The library can also be built as a static library if desired.
Thanks to Bernd Waibel for some of the initial code and digging into
the available cmake build options.
Other changes:
* Similar to version 6, the gnuinstalldirs patch no longer needs
BLOSC_LOCATION, GLFW3_LOCATION or TBB_LOCATION so these have been
removed.
* The new multilib header patch needs to know the CHOST so it is set
in the arguments passed to cmake.
* USE_GLFW3 is no longer used by the build system so it has been
removed also.
Signed-off-by: Adrian Grigo <agrigo2001@yahoo.com.au>
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Version bump to openvdb 6.2.1, requiring an additional use flag for
abi6-compat.
This version requires cmake 3.16 or higher.
Other changes:
* Compared to version 5, the gnuinstalldirs patch no longer needs
BLOSC_LOCATION, GLFW3_LOCATION or TBB_LOCATION so these have been
removed.
* The new multilib header patch needs to know the CHOST so it is set
in the arguments passed to cmake.
* USE_GLFW3 is no longer used by the build system so it has been
removed also.
* Update the github upstream source as dreamworksanimation now redirects to
AcademySoftwareFoundation.
Signed-off-by: Adrian Grigo <agrigo2001@yahoo.com.au>
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
| |
If openvdb is built with the python USE flag, then it is necessary to
build boost with python. Failure to do so produces compile failures.
Use the ? operator to ensure that this occurs
Signed-off-by: Adrian Grigo <agrigo2001@yahoo.com.au>
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The initial ebuilds for openvdb assumed that they would build the same
abi as the major version, unless abiX-compat was given to force
building against a legacy version.
This makes it difficult for other ebuilds to determine which ABI
openvdb supports as they can not rely on a USE flag being set to
ensure that openvdb uses the same version they are building against.
Starting with openvdb 5, openvdb also requires
OPENVDB_ABI_VERSION_NUMBER to be set with the version to build against.
This is also hard to determine if the user does not have a USE flag
to determine which version is being built against.
To fix these problems, I am using abiX-compat where x is 3,4,5...
to determine the appropriate number for OPENVDB_ABI_VERSION_NUMBER
and other ebuilds can use these flags to determine which openvdb
abi to build against. It is required that openvdb and all programs
linking with it build against the same openvdb version.
These use flags are no longer set by default in the ebuild, so
the user must ensure they set abiX-compat in package.use for openvdb
and any other package using openvdb (which will be the new versions
of blender and openimageio).
These use flags must be set even if the user wants to build againt the
latest supported abi version (eg abi5-compat for openvdb-5.2.0), which
is a change from the old behaviour.
I personally think this would be better hidden away inside an eclass
and a USE_EXPAND variable. I have submitted a proposal and would
be happy to prepare a PR if desired. This can be found at https://
archives.gentoo.org/gentoo-dev/message/1df75c608c83530b43c6ea67e1db8930
Signed-off-by: Adrian Grigo <agrigo2001@yahoo.com.au>
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
| |
Openvdb builds with python 3.7 and 3.8.
Support for python 3.6 has been removed.
Signed-off-by: Adrian Grigo <agrigo2001@yahoo.com.au>
Closes: https://bugs.gentoo.org/737390
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
Apply the patches from 4.0.2 for const correctness with newer
c-blosc and building the documentation to openvdb 5.2.0
Signed-off-by: Adrian Grigo <agrigo2001@yahoo.com.au>
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
Old versions of boost and c-blosc are no longer in the tree, so remove
unnecessary minimum version specifications
Signed-off-by: Adrian Grigo <agrigo2001@yahoo.com.au>
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The c-blosc 1.19 function blosc_compcode_to_compname now expects a
const char** rather than char**. This patch uses const_cast to fix the
compiler error.
When building with doc USE flag enabled, the doc target is not built by
default and the missing docs cause an install failure. This patch
ensures that when the doc target is defined, it will be built.
With these two patches, openvdb builds with all USE flags enabled.
Signed-off-by: Adrian Grigo <agrigo2001@yahoo.com.au>
Closes: https://bugs.gentoo.org/734102
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
The minimum set of changes to pass repoman checks is to update
to EAPI 7 and switch the cmake-utils eclass to cmake
Signed-off-by: Adrian Grigo <agrigo2001@yahoo.com.au>
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Sam James <sam@gentoo.org>
|
|
|
|
|
| |
Bug: https://bugs.gentoo.org/718450
Signed-off-by: Richard Freeman <rich0@gentoo.org>
|
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/718450
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Richard Freeman <rich0@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.3, Repoman-2.3.23
Signed-off-by: Louis Sautier <sbraz@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.3, Repoman-2.3.23
Signed-off-by: Louis Sautier <sbraz@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Thomas Deutschmann <whissi@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Thomas Deutschmann <whissi@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.3, Repoman-2.3.23
Signed-off-by: Louis Sautier <sbraz@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Thomas Deutschmann <whissi@gentoo.org>
|
|
|
|
|
|
| |
Bug: https://bugs.gentoo.org/737270
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Thomas Deutschmann <whissi@gentoo.org>
|
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/737270
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Thomas Deutschmann <whissi@gentoo.org>
|
|
|
|
|
| |
Bug: https://bugs.gentoo.org/738798
Signed-off-by: Michał Górny <mgorny@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.3, Repoman-2.3.23
Signed-off-by: Louis Sautier <sbraz@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.3, Repoman-2.3.23
Signed-off-by: Louis Sautier <sbraz@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.3, Repoman-2.3.23
Signed-off-by: Louis Sautier <sbraz@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
| |
* Migrate to cmake.eclass
* Add a systemd unit
* Add Python 3.8 and 3.9 support
* Add a debug USE flag
* Move some dependencies to BDEPEND
Package-Manager: Portage-3.0.3, Repoman-2.3.23
Signed-off-by: Louis Sautier <sbraz@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.3, Repoman-2.3.23
Signed-off-by: Louis Sautier <sbraz@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.3, Repoman-2.3.23
Signed-off-by: Louis Sautier <sbraz@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.4, Repoman-2.3.23
Signed-off-by: Mike Pagano <mpagano@gentoo.org>
|
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/711478
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Miroslav Šulc <fordfrog@gentoo.org>
|
|
|
|
|
|
|
|
|
| |
based on work of RarogCmex <denis7774@gmail.com> - thanks
Bug: https://github.com/gentoo/gentoo/pull/14834
Bug: https://bugs.gentoo.org/709282
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Miroslav Šulc <fordfrog@gentoo.org>
|
|
|
|
|
|
|
| |
thanks sam for pointing this out
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Miroslav Šulc <fordfrog@gentoo.org>
|
|
|
|
| |
Signed-off-by: Michał Górny <mgorny@gentoo.org>
|
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/690676
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Miroslav Šulc <fordfrog@gentoo.org>
|
|
|
|
|
|
| |
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Closes: https://bugs.gentoo.org/295168
Signed-off-by: Jeroen Roovers <jer@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Before the change libgccjit.so was installed to /usr/lib.
Relocation code in toolchain.eclass:gcc_movelibs() only
handles /usr/$(get_libdir) -> $LIBDIR location.
As a result gcc did not relocate libraries to version-speciifc
firectory: caused collisions and installed library into wrong
LIBDIR.
The change handles libgccjit special case by relocating it
explicitly.
Reported-by: Michał Górny
Bug: https://bugs.gentoo.org/583010
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
|
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/737462
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/735702
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
|
| |
Closes: https://bugs.gentoo.org/735700
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As with sys-devel/flex, virtual/yacc would only be needed if the parser
input is older than its output, which in the current versions is not the
case (perhaps depending on the type of filesystem the sources are
temporarily and temperamentally stored on, but touching the output file
would then have the same effect).
Not an actual revert as the incidental 0.5.0 EAPI=7 changes remain.
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Jeroen Roovers <jer@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-3.0.4, Repoman-3.0.1
Signed-off-by: Jeroen Roovers <jer@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|
|
|
|
|
| |
Package-Manager: Portage-2.3.103, Repoman-2.3.23
Signed-off-by: Mikle Kolyada <zlogene@gentoo.org>
|