summaryrefslogtreecommitdiffstats
path: root/Source
Commit message (Collapse)AuthorAgeFilesLines
* CMake Nightly Date StampKitware Robot2018-10-171-1/+1
|
* Merge topic 'ctest_done_bug_fix'Brad King2018-10-161-1/+11
|\ | | | | | | | | | | | | 60ca88d726 ctest_submit: Generate Done.xml at submit time Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2477
| * ctest_submit: Generate Done.xml at submit timeZack Galbreath2018-10-151-1/+11
| | | | | | | | | | | | | | | | | | | | | | This fixes a bug where calling `ctest_submit()` would result in a Done.xml file with an empty <buildId> element. CDash responds with a buildId when CTest submits a file. For this reason, Done.xml needs to generated after CTest has submitted some other file to CDash. This change also has the benefit of making Done.xml's timestamp as late as possible, giving us a more accurate record of how long the entire build took to complete.
* | CMake Nightly Date StampKitware Robot2018-10-161-1/+1
| |
* | CMake Nightly Date StampKitware Robot2018-10-151-1/+1
| |
* | CMake Nightly Date StampKitware Robot2018-10-141-1/+1
| |
* | CMake Nightly Date StampKitware Robot2018-10-131-1/+1
| |
* | Merge topic 'cppcheck-exit-code'Brad King2018-10-121-3/+7
|\ \ | | | | | | | | | | | | | | | | | | 3b80cd77fb Fail the build if cppcheck returns a non-zero exit code Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2459
| * | Fail the build if cppcheck returns a non-zero exit codeHarald Brinkmann2018-10-111-3/+7
| | | | | | | | | | | | | | | This allows the build failure to be tuned with cppcheck's options --error-exitcode=<n> and --exitcode-suppressions=<file>.
* | | Merge topic 'revert-install-code-script-genex'Brad King2018-10-122-44/+9
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | fd0e40f166 Merge branch 'backport-revert-install-code-script-genex' into revert-install-code-script-genex 0f48fbaa72 install: Revert CODE,SCRIPT support for generator expressions Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2474
| * \ \ Merge branch 'backport-revert-install-code-script-genex' into ↵Brad King2018-10-112-44/+9
| |\ \ \ | | |_|/ | |/| | | | | | revert-install-code-script-genex
| | * | install: Revert CODE,SCRIPT support for generator expressionsBrad King2018-10-112-44/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Revert commit v3.13.0-rc1~441^2 (install: Teach CODE,SCRIPT modes to evaluate generator expressions, 2018-05-29). Unfortunately it has been found to break existing code in a real project, e.g. install(CODE [[ message("$<FOOBAR>") ]]) Address this regression by reverting support for the 3.13 release series. Support can be restored later with a policy for compatibility. Issue: #15785 Fixes: #18435
* | | | CMake Nightly Date StampKitware Robot2018-10-121-1/+1
|/ / /
* | | Merge topic 'remove-AddCompileDefinitions'Brad King2018-10-116-31/+9
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | 8f076acdb0 cmLocalGenerator: Remove AddCompileDefinitions method Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2470
| * | | cmLocalGenerator: Remove AddCompileDefinitions methodBrad King2018-10-106-31/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | This method offers the same definitions as `GetTargetDefines` except that it excludes the "export" macro on shared libraries. Update call sites to use `GetTargetDefines` instead. Some of them were incorrectly excluding the export macro.
* | | | Merge topic 'vs-cuda-pdb'Brad King2018-10-111-7/+18
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | faf3d7d224 VS: Add workaround for CUDA compiler PDB location with space 592064e026 VS: Drop workaround for CUDA compiler PDB location on CUDA 9.2+ fb378fc4d7 Tests: Fix Cuda test project names Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2473
| * | | | VS: Add workaround for CUDA compiler PDB location with spaceBrad King2018-10-101-0/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | CUDA Toolkit Visual Studio Integration for version 9.2 and above does honor the `ClCompile.ProgramDataBaseFileName` field when telling `nvcc` how to invoke `cl`. Unfortunately it does not quote paths with spaces correctly: -Xcompiler "... /Fd"C:\path\with space\foo.pdb" ..." Work around this by converting the PDB location to a relative path. Likely we could always do this, but for now make a minimal change just for CUDA support. Fixes: #18440
| * | | | VS: Drop workaround for CUDA compiler PDB location on CUDA 9.2+Brad King2018-10-101-7/+11
| | |/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The workaround added by commit v3.12.0-rc1~227^2 (VS: Add workaround for CUDA compiler PDB location, 2018-04-13) is not necessary on CUDA 9.2+ because the CUDA Toolkit Visual Studio Integration has fixed the original bug and forwards the `ProgramDataBaseFileName` to the host compiler itself. Make the workaround conditional on the CUDA version. Issue: #18440
* | | | Merge topic 'cmake_policy-get_warning'Brad King2018-10-113-1/+35
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 0d988f98e5 cmake_policy: Add undocumented GET_WARNING command f9f96598df Help: Convert FindOpenGL documentation to block comment Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2472
| * | | | cmake_policy: Add undocumented GET_WARNING commandKyle Edwards2018-10-103-1/+35
| | |/ / | |/| | | | | | | | | | | | | | | | | | | | | | This command is intended for modules that issue policy warnings so they can get the warning string from CMake in a uniform manner, rather than duplicating the string. Several modules been updated to include an example of the usage of this new command.
* | | | Merge topic 'install-subdirectory-order'Brad King2018-10-119-19/+195
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | b56f2db87a Testing: Add test for CMP0082 fc8955e889 add_subdirectory: Run subdirectory install rules in correct order 514f0b572e Testing: Update hard-coded line numbers to [0-9]+ in some tests Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2434
| * | | | add_subdirectory: Run subdirectory install rules in correct orderKyle Edwards2018-10-109-19/+195
| |/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Before this change, install rules created by add_subdirectory() would be executed after all of the top-level install rules, even if they were declared before the top-level rules. This change adds a new policy, CMP0082, which interleaves the add_subdirectory() install rules with the other install rules so they are run in the correct order.
* | | | Merge topic 'ctest-done'Brad King2018-10-114-28/+76
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | a6e0158712 ctest_submit: Add support for a "Done" part f460bbd4c8 ctest_submit: Refactor file list to use a vector instead of a set Acked-by: Kitware Robot <kwrobot@kitware.com> Acked-by: Zack Galbreath <zack.galbreath@kitware.com> Merge-request: !2405
| * | | | ctest_submit: Add support for a "Done" partBetsy McPhail2018-10-103-0/+41
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Teach CTest to submit Done.xml. Submission of this file indicates to CDash that a build is complete and no more files will be uploaded. It contains the build id returned by CDash and the current time. This file is submitted last for a given build when using the `ctest_submit()` command. If submitting by PARTS, use `ctest_submit(PARTS Done)`.
| * | | | ctest_submit: Refactor file list to use a vector instead of a setBetsy McPhail2018-10-102-28/+35
| | | | | | | | | | | | | | | | | | | | Remove duplicates in a way that preserves order.
* | | | | CMake Nightly Date StampKitware Robot2018-10-111-1/+1
| |/ / / |/| | |
* | | | CMake Nightly Date StampKitware Robot2018-10-101-1/+1
| | | |
* | | | Merge topic 'string-TIMESTAMP-unset-TZ'Brad King2018-10-091-1/+11
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | f23760ffac string(TIMESTAMP): Fix unset TZ Acked-by: Kitware Robot <kwrobot@kitware.com> Acked-by: Attila Krasznahorkay <attila.krasznahorkay@gmail.com> Merge-request: !2464
| * | | | string(TIMESTAMP): Fix unset TZJose Luis Blanco-Claraco2018-10-091-1/+11
| | |/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The logic added by commit v3.6.0-rc1~338^2 (CMake: Extend TIMESTAMP sub-commands with new unix time format specifier, 2016-02-16) to restore the `TZ` environment variable does not properly handle the case in which the variable was originally not set. Unset the variable in this case. Fixes: #18431
* | | | Merge topic 'vs-flag-map-Qspectre'Brad King2018-10-093-0/+11
|\ \ \ \ | |_|_|/ |/| | | | | | | | | | | | | | | | | | | bb60ed6e72 VS: Add flag table entry for -Qspectre Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2462
| * | | VS: Add flag table entry for -QspectreBrad King2018-10-083-0/+11
| |/ / | | | | | | | | | | | | | | | | | | Add special logic to map this flag to a top-level build setting instead of being in ClCompile. Fixes: #18426
* | | CMake Nightly Date StampKitware Robot2018-10-091-1/+1
| | |
* | | Merge topic 'project-always-set-desc-url'Brad King2018-10-081-12/+8
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | f8a086a86b project(): Ensure DESCRIPTION and HOMEPAGE_URL variables are set b27247c8df project(): Add tests verifying variables set by second project() call Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2457
| * | | project(): Ensure DESCRIPTION and HOMEPAGE_URL variables are setCraig Scott2018-10-071-12/+8
| |/ / | | | | | | | | | | | | | | | | | | | | | If a project() call does not have DESCRIPTION or HOMEPAGE_URL options, it must still set the relevant variables or else those variables will inherit values from an earlier project() call. That is inconsistent with how VERSION is handled and is likely to be unexpected. The docs were also ambiguous about what should happen in such cases.
* | | CMake Nightly Date StampKitware Robot2018-10-081-1/+1
| | |
* | | CMake Nightly Date StampKitware Robot2018-10-071-1/+1
| | |
* | | CMake Nightly Date StampKitware Robot2018-10-061-1/+1
| | |
* | | CMake Nightly Date StampKitware Robot2018-10-051-1/+1
| |/ |/|
* | Merge topic 'misc-typos'Brad King2018-10-043-4/+4
|\ \ | | | | | | | | | | | | | | | | | | 6b4c5a0365 Fix misc. typos Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2440
| * | Fix misc. typosluz.paz2018-10-033-4/+4
| |/ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Found via `codespell -q 3 -I ../cmake-whitelist.txt --skip="./Utilities"` where the whitelist consists of ``` aci ans behaviour buil convertor dum earch ect emmited emmitted helpfull iff isnt ith lowercased mose nd nknown nto objext ot pathes pevents splitted substract superceded supercedes te tim todays uint upto whitespaces ```
* | CMake Nightly Date StampKitware Robot2018-10-041-1/+1
| |
* | Begin post-3.13 developmentBrad King2018-10-031-2/+2
| |
* | CMake 3.13.0-rc1 version updateBrad King2018-10-031-3/+3
|/
* Merge topic 'ctest-num-width'Brad King2018-10-031-2/+6
|\ | | | | | | | | | | | | 6d28884617 cmCTestRunTest: Avoid float/int conversions in number width logic Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2433
| * cmCTestRunTest: Avoid float/int conversions in number width logicBrad King2018-10-021-2/+6
| | | | | | | | | | | | | | | | | | | | | | Use of `std::log10` added by commit 02c5091c90 (cmCTestRunTest: Simplify number width computation, 2018-09-08) broke our number width computation on some platforms where static_cast<int>(std::log10(static_cast<size_t>(10))) somehow produces `0` instead of `1`. Re-implement the logic to avoid floating-point computations.
* | Merge topic 'update-kwsys'Brad King2018-10-031-0/+2
|\ \ | | | | | | | | | | | | | | | | | | | | | 0da3340b9c Merge branch 'upstream-KWSys' into update-kwsys b6a689c6f8 KWSys 2018-10-03 (9d6873b1) Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2437
| * | Merge branch 'upstream-KWSys' into update-kwsysBrad King2018-10-031-0/+2
| | | | | | | | | | | | | | | * upstream-KWSys: KWSys 2018-10-03 (9d6873b1)
* | | Merge topic 'fix-csharp-target-type'Brad King2018-10-033-6/+12
|\ \ \ | |/ / |/| | | | | | | | | | | | | | | | | 375b420fdf CSharp: Fix regression in VS project type selection 8b21aa0af0 VS: Fix CSharp flag selection when linking to a static C++ library Acked-by: Kitware Robot <kwrobot@kitware.com> Merge-request: !2427
| * | CSharp: Fix regression in VS project type selectionBrad King2018-10-022-3/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | A that target contains only `.cs` sources should be generated as a `.csproj` project even if it links to non-CSharp static libraries. The latter case was broken by refactoring in commit v3.12.0-rc1~160^2~7 (remove TargetIsCSharpOnly() and use methods from cmGeneratorTarget, 2018-03-19). The reason is that the `HasLanguage` method added by commit v3.12.0-rc1~239^2~6 (cmGeneratorTarget: add HasLanguage() as wrapper for GetLanguages(), 2018-03-19) enforces its "exclusive" check on the combined set of source file languages and the link language. To restore the original `TargetIsCSharpOnly` semantics, update `HasLanguage` to enforce exclusiveness only on the list of sources. Fixes: #18239
| * | VS: Fix CSharp flag selection when linking to a static C++ libraryBrad King2018-10-021-3/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When a CSharp target links to a static C++ library, CMake will compute the link language as C++ instead of CSharp. That may be incorrect and needs further investigation, but it does not affect how VS drives C# linking. However, it does break our flag language selection logic and causes C++ flags to be used for CSharp. In particular, this drops the `-platform:x86` flag on 32-bit builds. Fix this by always selecting the CSharp flags when generating a `.csproj` project type. Issue: #18239