On 09/24/2012 08:19 AM, Brad King wrote: > On 09/23/2012 01:22 PM, Alexander Neundorf wrote: >> This is now in the export-sets-2 branch on stage, including a basic test. >> >> Comments ? Ok to merge into next ? > > Yes, thanks.
This commit is a problem: http://cmake.org/gitweb?p=cmake.git;a=commitdiff;h=21d1619b -------------------------------------------------------------------------- diff --git a/Tests/ExportImport/Import/A/CMakeLists.txt b/Tests/ExportImport/Import/A/CMakeLists.txt index 650af6a..ca7a67a 100644 --- a/Tests/ExportImport/Import/A/CMakeLists.txt +++ b/Tests/ExportImport/Import/A/CMakeLists.txt @@ -1,8 +1,9 @@ +# Import targets from the exported install tree. +include(${CMAKE_INSTALL_PREFIX}/lib/exp/exp.cmake) + # Import targets from the exported build tree. include(${Import_BINARY_DIR}/../Export/ExportBuildTree.cmake) -# Import targets from the exported install tree. -include(${CMAKE_INSTALL_PREFIX}/lib/exp/exp.cmake) # Try referencing an executable imported from the install tree. add_custom_command( -------------------------------------------------------------------------- The test has no dependence between exports from the build tree and exports from the install tree. The new cross-export dependency finding logic is artificially connecting them. Export sets created by export() and those created by install() should be *independent* groups. They cannot cross. It does not make sense to have a build tree refer to its own install tree or vice versa. -Brad -- Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Please keep messages on-topic and check the CMake FAQ at: http://www.cmake.org/Wiki/CMake_FAQ Follow this link to subscribe/unsubscribe: http://public.kitware.com/cgi-bin/mailman/listinfo/cmake-developers