summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Merge topic 'intel-llvm-compilers'Brad King2021-01-2966-74/+613
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | 9b5fd04905 Tests: Disable MFC test for IntelLLVM due to #18311 ccdf7c0005 Tests: Skip VSGNUFortran test for ifx. ddaf412dd0 Tests: Skip per file PDBs in PDBDirectoryAndName for IntelLLVM a619b32768 Tests: Update RunCMake.GenerateExportHeader for IntelLLVM 661b04d6a6 Tests: Skip old PrecompiledHeader test on IntelLLVM 01b1e9c40d Tests: Update CompileFeatures test for IntelLLVM f26f10184f Tests: Update Module.WriteCompilerDetectionHeader for IntelLLVM 3ceb364b1c Tests: Update RunCMake.FileAPI for IntelLLVM ... Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !5594
| * Tests: Disable MFC test for IntelLLVM due to #18311William R. Dieter2021-01-281-1/+8
| | | | | | | | | | | | | | | | | | clang-cl cannot deal with implicit dependencies in UTF16 files (see #18311). IntelLLVM inherits this behavior from Clang. The MFC test uses `rc` in the way described in #18311 and gets the same error described in #18311. Disable the test until #18311 is fixed. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * Tests: Skip VSGNUFortran test for ifx.William R. Dieter2021-01-281-0/+4
| | | | | | | | DLLEXPORT is not implemented in ifx yet, and required by this test.
| * Tests: Skip per file PDBs in PDBDirectoryAndName for IntelLLVMWilliam R. Dieter2021-01-281-2/+3
| | | | | | | | | | | | The IntelLLVM C compiler do not yet support PDBs per file with the /Fd option. Disable the parts of PDBDirectoryAndName that tries to generate them so that the test will pass.
| * Tests: Update RunCMake.GenerateExportHeader for IntelLLVMBrad King2021-01-281-1/+1
| | | | | | | | On Windows, the expected results are like Clang's.
| * Tests: Skip old PrecompiledHeader test on IntelLLVMBrad King2021-01-281-1/+1
| | | | | | | | | | This test uses hard-coded flags for a fixed set of compilers. We have first-class PCH support implemented for IntelLLVM.
| * Tests: Update CompileFeatures test for IntelLLVMBrad King2021-01-281-2/+2
| | | | | | | | We do not have granular features for this compiler.
| * Tests: Update Module.WriteCompilerDetectionHeader for IntelLLVMBrad King2021-01-281-2/+2
| |
| * Tests: Update RunCMake.FileAPI for IntelLLVMBrad King2021-01-281-1/+1
| |
| * Tests: Update MSVCRuntimeLibrary.Fortran test for IntelLLVMWilliam R. Dieter2021-01-281-2/+3
| | | | | | | | | | | | `ifx` needs the same command line arguments as `ifort`. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * Tests: Update Preprocess test for IntelLLVMWilliam R. Dieter2021-01-281-1/+6
| | | | | | | | | | | | | | | | | | | | Work around an icx bug in command line handling of preprocessor macros. A `\\\"` string in a macro passed to icx on the command line is not properly parsed in compiler versions 2021.1.0 and earlier. This problem is expected to be fixed, though a target release has no been set yet. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * Tests: Update CPack DEBUGINFO tests for IntelLLVMWilliam R. Dieter2021-01-281-1/+1
| | | | | | | | | | | | | | By default icx and icpx do not add build IDs when linking, so enable the build ID to make the test pass. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * Tests: Update Assembler test for IntelLLVMBrad King2021-01-281-1/+1
| | | | | | | | Enable parts of the test the same way we do for `Intel`.
| * Tests: Update Fortran tests for IntelLLVMWilliam R. Dieter2021-01-282-3/+3
| | | | | | | | | | | | Update checks for the `Intel` compiler id to match `IntelLLVM` too. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * FindBLAS: Add support for IntelLLVMWilliam R. Dieter2021-01-281-0/+3
| | | | | | | | Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * FindBoost: Add support for IntelLLVMWilliam R. Dieter2021-01-281-2/+4
| | | | | | | | | | | | Use the same code paths as the `Intel` compiler. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * FindOpenMP: Add support for IntelLLVMWilliam R. Dieter2021-01-281-0/+1
| | | | | | | | Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * FindMPI: Add support for IntelLLVMWilliam R. Dieter2021-01-281-1/+11
| | | | | | | | Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * IRSL: Add support for IntelLLVMWilliam R. Dieter2021-01-281-7/+7
| | | | | | | | | | | | Use the same code paths as the `Intel` compiler id. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * cmake: Reduce default recursion depth when compiling with IntelLLVMBrad King2021-01-281-0/+2
| | | | | | | | | | | | This compiler makes stack frames large enough (at least in Debug builds) that CMake language recursion to depth 1000 overflows the stack. Reduce the limit for this compiler.
| * Utilities: Suppress warnings in third-party code when using IntelLLVMBrad King2021-01-2812-12/+12
| |
| * IntelLLVM: Add support for Intel LLVM-based compilersWilliam R. Dieter2021-01-2829-12/+497
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Using a single ID 'IntelLLVM' for the suite of Intel compilers based on the LLVM backend. The 'IntelLLVM' ID are used for C, C++, and Fortran. Data Parallel C++ will be handled in a separate commit. The C and C++ definitions are based on the Clang definitions. The Intel LLVM-based C and C++ compilers are based on the Clang front end, so existing Clang options are more likely to be a good match than options for the older Intel compilers. Fortran is based on the older Fortran front end with the LLVM backend. It has a similar interface to the older versions, though many options are shared with the C and C++ compilers. Fixes: #21561 Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * GenerateExportHeader: Recognize Intel < 12 more strictlyWilliam R. Dieter2021-01-281-1/+1
| | | | | | | | | | | | Match exactly the `Intel` compiler id. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * Tests: Cleanup CPack message when files are expected and none are foundWilliam R. Dieter2021-01-281-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | Previously, when looking for exactly one file and the number of files was not one, the test printed out an error message saying that too many files were found, then printing the list of files. However, 0 is not 1 and also not too many files. To reduce confusion, this commit adds a different message when 0 files are found and 1 is expected. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * Tests: Update RunCMake.DependencyGraph to canonicalize paths when comparingWilliam R. Dieter2021-01-281-1/+2
| | | | | | | | | | | | | | A symbolic link in the path was causing false path mismatches. Resolving the symlinks makes both paths identical. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * Tests: Make GCOV test more robust to symbolic links in pathsWilliam R. Dieter2021-01-281-0/+19
| | | | | | | | | | | | | | | | Symbolic links in the path to the cmake build directory caused some paths to have '..' directories in the path, which are equivalent, but not a string match for the expected path. As a result, some tests fail. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * Tests: Fix Fortran test C function prototypesWilliam R. Dieter2021-01-281-2/+2
| | | | | | | | | | | | | | | | Several extern functions were declared without return type, which results in warnings. The functions are for calling Fortran subroutines, so there should not be a return value. Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * Tests: Use the CVS :local: access method for local directoriesWilliam R. Dieter2021-01-282-17/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | Windows absolute paths start with a drive letter, followed by the path. Without an explicit access method specification, CVS treats the drive letter as a host name. This changes explicitly adds `:local:` to local directory repository paths used in tests to avoid this confusion. See the CVS Manual [1] for more information. [1] https://www.gnu.org/software/trans-coord/manual/cvs/cvs.html#Repository Signed-off-by: William R. Dieter <william.r.dieter@intel.com>
| * Tests: Fix CMP0115 warnings in CompileFeatures testBrad King2021-01-281-1/+4
| | | | | | | | Pass source file extensions explicitly.
* | CMake Nightly Date StampKitware Robot2021-01-291-1/+1
| |
* | Merge branch 'release-3.19'Brad King2021-01-280-0/+0
|\ \ | |/ |/|
| * CMake 3.19.4v3.19.4Brad King2021-01-281-1/+1
| |
* | Merge topic 'nvhpc-compiler'Brad King2021-01-2821-11/+92
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 4de28ba4e6 Tests: Update tests for NVHPC compiler 90502d76b7 FindOpenMP: Add support for NVHPC compiler 76dee5831b FindOpenACC: Add support for NVHPC compiler ff7bab83c2 GenerateExportHeader: Add support for NVHPC compiler 83ea49818f NVHPC: Add support for NVIDIA HPC SDK compilers based on PGI Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !5475
| * | Tests: Update tests for NVHPC compilerTin Huynh2021-01-275-8/+8
| | | | | | | | | | | | Treat it just like PGI.
| * | FindOpenMP: Add support for NVHPC compilerTin Huynh2021-01-271-0/+1
| | |
| * | FindOpenACC: Add support for NVHPC compilerTin Huynh2021-01-271-1/+4
| | |
| * | GenerateExportHeader: Add support for NVHPC compilerBrad King2021-01-271-2/+2
| | | | | | | | | | | | Treat it just like PGI.
| * | NVHPC: Add support for NVIDIA HPC SDK compilers based on PGITin Huynh2021-01-2713-0/+77
| | | | | | | | | | | | | | | | | | | | | Identify the compilers as `NVHPC` to distinguish it from the older PGI compilers from which they evolved, and from other `NVIDIA` compilers. Fixes: #20887
* | | Merge topic 'ci-aarch64'Brad King2021-01-2814-89/+296
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 430224b6e8 gitlab-ci: add test:debian10-aarch64-ninja job 300a91d42d gitlab-ci: add build:centos7-aarch64 job d997637fa3 gitlab-ci: add rules infrastructure to allow non-MR jobs a54e43670d gitlab-ci: rename build:centos6-{release => x86_64} 616bf0d2f0 gitlab-ci: consolidate release and package job infrastructure 227ca689a7 ci: download cmake on aarch64 a601ed8e0e ci: download ninja on aarch64 54c2c5f889 ci: generalize ninja download script to support non-upstream builds ... Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !5745
| * | | gitlab-ci: add test:debian10-aarch64-ninja jobBrad King2021-01-275-0/+176
| | | |
| * | | gitlab-ci: add build:centos7-aarch64 jobBrad King2021-01-271-0/+10
| | | | | | | | | | | | | | | | | | | | Test building our release for the `aarch64` architecture. For now, leave this job out during merge requests.
| * | | gitlab-ci: add rules infrastructure to allow non-MR jobsBrad King2021-01-271-0/+6
| | | |
| * | | gitlab-ci: rename build:centos6-{release => x86_64}Brad King2021-01-271-7/+7
| | | | | | | | | | | | | | | | This makes room for a future equivalent for other architectures.
| * | | gitlab-ci: consolidate release and package job infrastructureBrad King2021-01-274-61/+27
| | | | | | | | | | | | | | | | | | | | | | | | Use `linux_release` in names associated with building or testing release-style binaries. Add a `.linux_package` helper to adjust the release binary build behavior for package-specific jobs.
| * | | ci: download cmake on aarch64Brad King2021-01-271-0/+5
| | | |
| * | | ci: download ninja on aarch64Brad King2021-01-271-0/+7
| | | |
| * | | ci: generalize ninja download script to support non-upstream buildsBrad King2021-01-271-9/+8
| | | |
| * | | ci: add docker spec to build ninja for linux-aarch64Brad King2021-01-273-0/+38
| | | |
| * | | ci: update to use CMake 3.19.3Brad King2021-01-272-12/+12
| |/ /
* | | CMake Nightly Date StampKitware Robot2021-01-281-1/+1
|/ /
generated by cgit v0.12 at 2025-09-13 18:38:00 (GMT)