| Commit message (Collapse) | Author | Age | Files | Lines |
|\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
91a7c0b51c CMake: fix violations of #pragma once check
aa0c99c55c clang-tidy module: add test for #pragma once check
c9af6f2ff6 clang-tidy module: add #pragma once check
68a4d97da7 clang-tidy module: allow header files in test cases
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: buildbot <buildbot@kitware.com>
Acked-by: Ben Boeckel <ben.boeckel@kitware.com>
Merge-request: !7933
|
| | |
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | | |
38c8807c5a Ninja: Avoid re-linking a Swift executable on every build
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !7901
|
| | |
| | |
| | |
| | |
| | |
| | | |
Swift doesn't emit swiftmodules for executables, so we shouldn't put it
in dependency graph. Ninja sees the "missing" dependency and always
tries to rebuild/re-link the target.
|
|\ \ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
6e3e8827fa Refactor: `cmGlobalGeneratorFactory::GetDocumentation` returns entry
e99a4acbcd cmGlobalGeneratorFactory.h: Simplify returning default values
7daadd304f cmake.cxx: Simplify `cmake::AppendExtraGeneratorsDocumentation()`
df2047c2ac cpack.cxx: Optimize calls to `std::osteam::operator<<`
f6180485bb cpack.cxx: Eliminate redundant `if`
d7c183f35c cpack.cxx: Deduplicate "Generators" section creation code
9c06f0dd71 cmake.cxx: Move `static` functions to an anonymous namespace
69918b07e1 cmDocumentationEntry: Drop all user provided ctors for C++ >= 14
...
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: buildbot <buildbot@kitware.com>
Merge-request: !7605
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Before, a documentation entry was in/out parameter.
Now it's a normal return value.
This also makes possible to eliminate defaulted default ctor
for `cmDocumentationEntry` for C++ 11.
Also, simplify `cmake::AppendGlobalGeneratorsDocumentation()`.
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
There is no need for them cuz:
- the last field has a default value
- all static instances use 2 arguments convertible to `std::string`
- "dynamic" instances used for _Generator_ doc entries access
fields diectly using default constructed instance
Moreover, compiler may generate move ctor/assign when needed.
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Use fixed size arrays of `cmDocumentationEntry` items instead of
open arrays of two `char` pointers when describe program options
help screens.
Also, drop `const char*[][2]` overloads of methods of
`cmDocumentation` and `cmDocumentationSection` classes in the sake
of generic (template) appenders introduced earlier.
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | | |
Instead of one char at time copy (append), use lines.
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | | |
Cuz normally required text width can't be negative. It was `int`…
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Fix `cmDocumentationFormatter::PrintColumn()` method to eliminate
an indentation reset side effect.
|
| | | |
| | | |
| | | |
| | | |
| | | | |
The `cmDocumentationFormatter::PrintSection()` method ignores
the currently set indentation level and use it's own.
|
| | | |
| | | |
| | | |
| | | | |
… instead of `std::string`.
|
| | | |
| | | |
| | | |
| | | | |
… instead of manually print spaces to align fields.
|
| | | |
| | | |
| | | |
| | | | |
Was `const char*`.
|
| | | |
| | | |
| | | |
| | | | |
Also, make it `const` method cuz it's not modify the state.
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Can be added later it really need.
Also, drop ctor/dtor to make the class trivially destructable.
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
38cbf5e15b Genex: Add $<BUILD_LOCAL_INTERFACE:...> genex
37b5c78688 cmGeneratorExpression: Refactor stripExportInterface() to use enum class
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: buildbot <buildbot@kitware.com>
Merge-request: !7919
|
| | | | |
| | | | |
| | | | |
| | | | | |
Fixes: #23209
|
| | |_|/
| |/| | |
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
8437141b53 Genex: Fix TARGET_PROPERTY lookup scope in transitive usage requirements
Acked-by: Kitware Robot <kwrobot@kitware.com>
Tested-by: buildbot <buildbot@kitware.com>
Acked-by: Robert Maynard <robertjmaynard@gmail.com>
Merge-request: !7928
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
When `$<TARGET_PROPERTY:tgt,prop>` is used in an `INTERFACE_*` target
property for usage requirements, it may be evaluated in the context of a
dependent target in another directory. Look up the `tgt` name in the
directory of the target whose property holds the expression so that
imported targets isolated to that directory are visible.
Fixes: #24163
|
| |/ / /
|/| | | |
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
48db261e69 cmake --build: Use both inter- and intra-project parallelism with msbuild
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !7926
|
| |/ / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Since commit 1ab3881ec9 (cmake: Add options for parallel builds to
--build mode, 2018-04-14, v3.12.0-rc1~42^2), `cmake --build --parallel`
has added two options to the msbuild command line:
* `/m` enables inter-project parallelism in msbuild.
* `/p:CL_MPCount=1` suppresses intra-project parallelism in cl.
The latter was used to avoid `O(N^2)` compilation threads on `N`
processors. In practice however, projects often have many source files
in a few targets, so `--parallel` actually reduces parallelism.
Drop the latter option to restore intra-project parallelism.
Users can always add `-- /p:CL_MPCount=1` themselves.
Fixes: #20564
|