summaryrefslogtreecommitdiffstats
path: root/Source
Commit message (Collapse)AuthorAgeFilesLines
* CMake 3.22.0v3.22.0Brad King2021-11-181-1/+1
|
* MINGW-w64: Fix string(TIMESTAMP) build on 32bits.Antons Jeļkins2021-11-181-3/+9
| | | | | | | Rephrase the string(TIMESTAMP) implementation not to cause gcc-11 ICE on MSYS2/mingw32. Fixes: #22916
* cmTimestamp: Declare component buffer before MinGW-specific codeAntons Jeļkins2021-11-181-2/+2
|
* Merge topic 'IntelLLVM-Fortran-copy-mod' into release-3.22Brad King2021-11-181-1/+1
|\ | | | | | | | | | | | | 209b896f2c IntelLLVM: Enable Fortran module rebuild avoidance in Makefile generators Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !6740
| * IntelLLVM: Enable Fortran module rebuild avoidance in Makefile generatorsBrad King2021-11-171-1/+1
| | | | | | | | | | | | | | | | | | | | The Makefile generators use an internal `cmake -E cmake_copy_f90_mod` tool to avoid rebuilding module consumers when the `.mod` content changes only in a trivial way (e.g. the time it was built). This is done with logic specific to each vendor's module file format. Enable the "Intel" format support when using the IntelLLVM compiler (ifx) too. Issue: #22922
| * Merge branch 'fix-ifdef-windows' into release-3.20Brad King2021-10-154-4/+4
| |\ | | | | | | | | | Merge-request: !6631
| * | CMake 3.20.6v3.20.6Brad King2021-09-201-1/+1
| | |
* | | CMake 3.22.0-rc3v3.22.0-rc3Brad King2021-11-121-1/+1
| | |
* | | CMP0128: Enable/disable extensions if standard same as defaultRaul Tambre2021-11-111-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This was intended to be part of the initial MR (!6177), but accidentally went missing when debugging nightly failures on less common systems. Noticed during !6711 review as the comment about this behaviour didn't match the code. Documentation for CMP0128 is updated to remove a false case and note the two cases related to this. Fixes #22224.
* | | Merge topic 'nmc-reuse-from' into release-3.22Brad King2021-11-091-3/+3
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | 244550997f PCH: Fixed Ninja Multi-Config and REUSE_FROM for MSVC Acked-by: Kitware Robot <kwrobot@kitware.com> Tested-by: buildbot <buildbot@kitware.com> Merge-request: !6713
| * | | PCH: Fixed Ninja Multi-Config and REUSE_FROM for MSVCCristian Adam2021-11-081-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The copy_idb_pdb.cmake script would be executed for every configuration for all configurations. Debug would still want to get the RelWithDebInfo files, and the other way around.
* | | | Merge topic 'msvc-pch-reuse-config' into release-3.22Brad King2021-11-091-2/+2
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 95fa27d94d PCH: Fix REUSE_FROM when Debug and RelWithDebInfo configs differ Acked-by: Kitware Robot <kwrobot@kitware.com> Acked-by: buildbot <buildbot@kitware.com> Merge-request: !6712
| * | | | PCH: Fix REUSE_FROM when Debug and RelWithDebInfo configs differCristian Adam2021-11-081-2/+2
| |/ / / | | | | | | | | | | | | | | | | | | | | | | | | If the Debug configuration contains `/ZI` but the RelWithDebInfo doesn't then the copy_pdb_idb.cmake script will cause problems due to the fact that it was common for both configurations but they are incompatible with each other.
* | | | Merge topic 'vs-framework-version' into release-3.22Brad King2021-11-085-39/+124
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | d51246c662 VS: Default TargetFrameworkVersion to v4.7.2 for VS 2022 f97f8537f3 VS: Model a default target framework e40cedddc0 cmVisualStudio10TargetGenerator: Refactor target framework selection 78782cc7dc cmGlobalVisualStudio8Generator: Refactor SetGeneratorPlatform Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !6699
| * | | | VS: Default TargetFrameworkVersion to v4.7.2 for VS 2022Brad King2021-11-061-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | MSBuild defaults to v4.0 but VS 2022 does not install it anymore. Explicitly specify a newer framework version by default. Use a version that VS 2022 installs without selecting a separate component. Fixes: #22835
| * | | | VS: Model a default target frameworkBrad King2021-11-064-10/+64
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add fields to the VS generator to select a target framework. Migrate the existing default for VS 12 .NET CF for Windows CE. Report the values in `CMAKE_VS_*` variables and use them for the CSharp compiler id project too. Issue: #22849
| * | | | cmVisualStudio10TargetGenerator: Refactor target framework selectionBrad King2021-11-061-35/+53
| | | | | | | | | | | | | | | | | | | | Split the selection from the generation of the resulting element(s).
| * | | | cmGlobalVisualStudio8Generator: Refactor SetGeneratorPlatformBrad King2021-11-031-4/+11
| | | | | | | | | | | | | | | | | | | | | | | | | Re-organize the method to make room for additional `CMAKE_GENERATOR_PLATFORM` processing.
* | | | | Ninja Multi-Config: Fix custom command target dependencies in cross-configsBrad King2021-11-041-4/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Generator expressions in a non-cross custom command's `COMMAND` arguments are evaluated in the command config. Target-level dependencies implied by `TARGET_FILE` must therefore be cross dependencies. This is important to generate proper target-level dependencies on the cross-config build statements for the target to which the custom command is attached. Fixes: #22855
* | | | | Ninja Multi-Config: Fix internal cross-config target dependency orderingBrad King2021-11-041-4/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In commit 7abc3d61ac (Ninja Multi-Config: Fix issue with framework dependencies and Autogen, 2020-02-13, v3.17.0-rc2~18^2) the `cmLinkItem` comparison operator was updated to order identical strings by the cross-config boolean. We need to order identical targets that way too in order to represent both a cross and non-cross dependency on the same target. Issue: #22855
* | | | | Ninja Multi-Config: Fix cross-config custom command dependency tracingBrad King2021-11-044-5/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Process `CMAKE_CROSS_CONFIGS` and friends to properly configure the generator for cross-config behavior before custom command dependency tracing.
* | | | | Merge topic 'msvc-cxx-modules-scanDependencies' into release-3.22Brad King2021-11-041-1/+1
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 9ed1d7bee6 cmScanDepFormat: Accept P1689r4 files with version 1 Acked-by: Kitware Robot <kwrobot@kitware.com> Acked-by: Ben Boeckel <ben.boeckel@kitware.com> Merge-request: !6696
| * | | | | cmScanDepFormat: Accept P1689r4 files with version 1Brad King2021-11-031-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | VS 2022's `cl` 19.30 has a `-scanDependencies` option that produces the P1689r4 format. It reports the "version" field with value "1".
| * | | | | CMake 3.21.4v3.21.4Brad King2021-10-271-1/+1
| | | | | |
| * | | | | Merge branch 'fix-ifdef-windows' into release-3.21Brad King2021-10-154-4/+4
| |\ \ \ \ \ | | | |_|_|/ | | |/| | | | | | | | | Merge-request: !6631
* | | | | | Ninja: Fix creation of Windows import library directoryBrad King2021-11-031-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Call `EnsureParentDirectoryExists` with our internal representation of the path, not the representation quoted/encoded for Ninja. Fixes: #22841
* | | | | | cmNinjaTargetGenerator: Replace "their" with "there" in commentBrad King2021-11-031-1/+1
| |_|/ / / |/| | | |
* | | | | cmCTestRunTest: get the default value from the environmentBen Boeckel2021-10-291-1/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This only works due to some assumptions about how the `ENVIRONMENT` property is processed. Comments have been added to notify anyone modifying the behavior about where to look. Fixes: #22819
* | | | | CMake 3.22.0-rc2v3.22.0-rc2Brad King2021-10-281-1/+1
| | | | |
* | | | | Merge topic 'builtin-help-versionadded' into release-3.22Brad King2021-10-282-0/+6
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 054754359a cmRST: support `versionadded` and `versionchanged` directives Acked-by: Kitware Robot <kwrobot@kitware.com> Acked-by: buildbot <buildbot@kitware.com> Merge-request: !6667
| * | | | | cmRST: support `versionadded` and `versionchanged` directivesBen Boeckel2021-10-272-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This makes `versionadded` and `versionchanged` directives show up in `cmake --help-*` output instead of disappearing (and potentially making empty sections). Fixes: #22808
* | | | | | Merge topic 'envmod-support-mod-to-new-variables' into release-3.22Brad King2021-10-271-6/+2
|\ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 69f95cf1d9 cmCTestRunTest: fix modifying non-existent envvars Acked-by: Kitware Robot <kwrobot@kitware.com> Acked-by: buildbot <buildbot@kitware.com> Merge-request: !6663
| * | | | | | cmCTestRunTest: fix modifying non-existent envvarsBen Boeckel2021-10-261-6/+2
| |/ / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When appending or modifying to a variable that has not been modified before, the iterator was not valid, but it was used to insert into the map again. Instead, just use indexing into the map since we know it will exist by the end of the function anyways. Fixes: #22796
* | | | | | cmGlobalVisualStudioVersionedGenerator: Fix repeating SetGeneratorInstanceBrad King2021-10-262-2/+6
|/ / / / / | | | | | | | | | | | | | | | | | | | | | | | | | Fix logic added by commit 8917b8512f (cmGlobalVisualStudioVersionedGenerator: Allow repeating SetGeneratorInstance, 2021-10-20) to avoid repeating work.
* | | | | Restore honoring CMAKE_GENERATOR_INSTANCE in toolchain fileBrad King2021-10-204-33/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Revert the changes from commit d5b5c19278 (cmGlobalGenerator: FindMakeProgram() before CMakeDetermineSystem, 2020-06-15, v3.19.0-rc1~619^2~3) and commit ef91fb02f3 (cmGlobalGenerator: FindMakeProgram() at a generator-specific time, 2020-11-23, v3.19.1~2^2). We must delay selecting the location of MSBuild until after an instance of Visual Studio has been selected. It is now safe to revert the ordering because the motivating use case (sysroot detection in Platform/Android-Determine) has been implemented another way. Fixes: #22782
* | | | | cmake_host_system_information: Add undocumented VS_MSBUILD_COMMAND keyBrad King2021-10-201-0/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When using the Visual Studio generator for VS 10 or above, offer this key to get the location of the MSBuild command before the first `project()` or `enable_language()` command has finished running. This will be needed only by one of our own modules, so leave it undocumented for now.
* | | | | cmGlobalVisualStudio10Generator: Add method to find MSBuild earlyBrad King2021-10-204-0/+21
| | | | | | | | | | | | | | | | | | | | | | | | | Add a way to find MSBuild before the main `FindMakeProgram` code path has executed.
* | | | | cmGlobalGenerator: Add method to check if generator is at least VS 10Brad King2021-10-202-0/+4
| | | | |
* | | | | cmGlobalVisualStudioVersionedGenerator: Allow repeating SetGeneratorInstanceBrad King2021-10-202-0/+9
| | | | |
* | | | | cmGlobalVisualStudio10Generator: Allow subclasses to reset MSBuild searchBrad King2021-10-202-2/+1
| | | | | | | | | | | | | | | | | | | | While at it, convert to inline initialization.
* | | | | Merge topic 'fix-ifdef-windows' into release-3.22Brad King2021-10-184-4/+4
|\ \ \ \ \ | | |/ / / | |/| | | | | | | | | | | | | | | | | | | | | | | 40e73c5ac4 Source: Fix typo in _WIN32 preprocessor checks Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !6631
| * | | | Source: Fix typo in _WIN32 preprocessor checksBrad King2021-10-154-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Since CMake's default compiler flags with MSVC include `-DWIN32` for historical reasons, a few preprocessor conditions were accidentally checking for `WIN32` instead of `_WIN32`. The corresponding blocks were left out when compiling official binaries for `cmake.org` because we hard-code compiler flags without `-DWIN32`. Fixes: #22764
| * | | | CMake 3.19.8v3.19.8Brad King2021-04-061-1/+1
| | | | |
* | | | | CMake 3.22.0-rc1v3.22.0-rc1Brad King2021-10-141-1/+1
| | | | |
* | | | | cmBinUtilsMacOSMachOLinker: improve performance by memoizing otool callsSeth R Johnson2021-10-122-9/+43
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Libraries with many repeated transitive dependencies (e.g. Trilinos) can result in very long runtime dependency call times, especially if system calls are made more expensive by antivirus software. This change caches the results of the calls to otool for efficiency.
* | | | | cmFileCommand: improve error messageSeth R Johnson2021-10-121-3/+6
| | | | | | | | | | | | | | | | | | | | | | | | | Print full list of unresolved dependencies and a phrase that indicates what about the file is unresolved.
* | | | | Begin 3.22 release versioningBrad King2021-10-081-3/+3
| |_|/ / |/| | |
* | | | CMake Nightly Date StampKitware Robot2021-10-081-1/+1
| | | |
* | | | CMake Nightly Date StampKitware Robot2021-10-071-1/+1
| | | |
* | | | Merge topic 'bootstrap-qt6'Brad King2021-10-061-2/+2
|\ \ \ \ | | |/ / | |/| | | | | | | | | | | | | | | | | | | | | | 6458a32242 bootstrap: Enable cmake_language command to support cmake-gui with Qt 6.2 Acked-by: Kitware Robot <kwrobot@kitware.com> Reviewed-by: Andrew Maclean <andrew.amaclean@gmail.com> Merge-request: !6593
generated by cgit v0.12 at 2025-01-18 08:50:23 (GMT)