Brad King authored 8 years ago and Kitware Robot committed 8 years ago

41199f8c CPackIFW: Add support for Promoting Updates
6ca6b0dd CPackIFW: Add more known QtIFW releases

Konstantin Podsvirov authored 8 years ago and Brad King committed 8 years ago

Add support for this feature added by QtIFW 2.0.3:

  http://doc.qt.io/qtinstallerframework/ifw-updates.html

Add a `cpack_ifw_update_repository` command as porcelain.

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

edcccde7 file: Sort GLOB results to make it deterministic (#14491)

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

d7233a04 GHS: Use shorter object file names on collision

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

5a21557b AIX,HP: Allow user to override the default runtime path (libpath)

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

030556b7 cmListFileLexer: Revise C++ coding style using clang-format
e4a92dab cmListFileLexer: Update to flex 2.6

Matthew Hanna authored 8 years ago and Brad King committed 8 years ago

The `CMAKE_PLATFORM_REQUIRED_RUNTIME_PATH` is used by CMake to always
append `-Wl,-blibpath:/usr/lib:/lib` to the link line.  This is needed
by default on these platforms but needs to be overridden in some use
cases (e.g. an environment in which one maintains versioned shared
libraries).  Change our logic to set this value only if it not already
set by the user, project, or toolchain file.

Manually extract the C++ portion of `cmListFileLexer.in.l` into a
temporary file, format it, and then move it back into the original file.
Manually format C++ code inside the lexer actions to match our style.
Then re-generate the lexer.

Revise the documented modifications we need to make to the
flex-generated source file according to the needs of the new version.
Update our own implementation to avoid warnings with flex types.

Reiner Herrmann authored 8 years ago and Brad King committed 8 years ago

Even though the `file(GLOB)` documentation specifically warns against
using it to collect a list of source files, projects often do it anyway.
Since it uses `readdir()`, the list of files will be unsorted.
This list is often passed directly to add_executable / add_library.
Linking binaries with an unsorted list will make it unreproducible,
which means that the produced binary will differ depending on the
unpredictable `readdir()` order.

To solve those reproducibility issues in a lot of programs (which don't
explicitly `list(SORT)` the list manually), sort the resulting list of
the `file(GLOB)` command.

A more detailed rationale about reproducible builds is available
[here](https://reproducible-builds.org/).

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

d9fd2f54 Revise C++ coding style using clang-format
82df6dea Empty commit at end of history preceding clang-format style transition
6a13f43f CONTRIBUTING: Add a section on coding style
bf451d9f Add a script to run clang-format on the entire source tree
1e90d78f Configure clang-format for CMake source tree
da60adc3 Tell Git to use a distinct conflict marker size in `.rst` files

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

c7006d2e cmGeneratorTarget: Protect macro code layout from clang-format

Kitware Robot authored 8 years ago and Brad King committed 8 years ago

Run the `Utilities/Scripts/clang-format.bash` script to update
all our C++ code to a new style defined by `.clang-format`.
Use `clang-format` version 3.8.

* If you reached this commit for a line in `git blame`, re-run the blame
  operation starting at the parent of this commit to see older history
  for the content.

* See the parent commit for instructions to rebase a change across this
  style transition commit.

This is an empty commit that precedes an automatic application of
clang-format to update the C++ style of our entire source tree.
This may be helpful to rebase a topic branch that was originally
based on a commit preceding the transition.  One may first rebase
the topic on this commit.  Then use one of the following approaches.

*   Rewrite the topic, including this commit, using `git filter-branch`
    `--tree-filter` with `Utilities/Scripts/clang-format.bash` to update
    the style in every commit.  Rebase the revised topic, excluding the
    rewrite of this commit, on the style transition commit.

OR

*   Add a `.git/info/grafts` entry to change the parent of the first
    commit in the topic from this commit to the style transition commit.
    Rewrite the topic using `git filter-branch --tree-filter` with
    `Utilities/Scripts/clang-format.bash` to update the style in every
    commit.  Then remove the graft, which was resolved by the filter.

See `git help filter-branch` and `git help repository-layout` for
details.

Document that we now define our coding style with `clang-format`.

List all sources in version control and filter out those that we should
not format for various reasons.  Then run the clang-format tool to do an
in-place update.

After discussion among developers we settled on the Mozilla style
with a few tweaks:

* Do not align operator arguments.

* Do not always break after a function return type.

* Limit to 79 columns instead of 80 to fit edge cases
  better in 80-column terminals as CMake has always done.

* Format for C++98 instead of C++11 because CMake is written
  in the former language.

Co-Author: Daniel Pfeifer <daniel@pfeifer-mail.de>

Use a long conflict marker size that is unlikely to be in conflict
with a `=======` section header underline.

The `IMPLEMENT_VISIT_IMPL` macro must preserve a space before the `>`
character in case the `DATATYPE` is a template type ending in `>`.
Manually format the macro layout as clang-format would except for this
space.  Then add markup to tell clang-format not to format this macro.

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

c2e8d515 FindBoost: Add checks for component-specific headers

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

328eff90 FindBoost: Update dependencies for Boost 1.61

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

df97b979 FindPkgConfig: optionally create imported target for the found libraries

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

f5089cfc CPack/RPM adding dist to release tag test
f7003a60 CPack/RPM release dist tag support

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

44ee2d71 CPack/RPM different package names

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

4461e8bb CPack/Deb cpack-deb-improvements release notes
7e940bf7 CPack/Deb test changes due to breaking changes
3b648894 CPack/Deb package release number in file name
316dd613 CPack/Deb proper package file naming
271e03f0 CPack/Deb per-component package architecture
23baaf8d CPack/Deb inter package dependencies
3a55a0e7 CPack/Deb proper component packages file naming
c7f388e7 CPack/Deb generation of postinst and postrm ldconfig files
2d589653 CPack/Deb generation of DEBIAN/shlibs control file

Brad King authored 8 years ago and Kitware Robot committed 8 years ago

9cb263d7 cmComputeLinkInformation: Do not mark interface library as a path

Roger Leigh authored 8 years ago and Brad King committed 8 years ago

This supplements the existing library checks, to
cater for the possibility that the libraries are
present but the headers are not.  This can happen
when the Boost collections is split up into
multiple packages and not all are installed,
and will avoid the checks silently passing when
the build would subsequently fail.

Packagers may now set their own rpm package
file names or request that rpmbuild tool
chooses one for them. It also supports handing
of situations where one spec file may produce
multiple rpm packages.

Tests and release notes

Harry Mallon authored 8 years ago and Domen Vrankar committed 8 years ago

Some Linux distros require Release tag
to be set to <ReleaseVersion><Dist>.

The empty string we add as a link item for an INTERFACE_LIBRARY target
is not a path, so do not mark it as such.  The generators currently
tolerate it either way, but only by accident.