summaryrefslogtreecommitdiff
path: root/regress
AgeCommit message (Collapse)AuthorFilesLines
2005-06-09Fix test to that it's possible to check correctness whenjlam1-2/+2
COMPILER_RPATH_FLAG is "-Wl,-rpath,".
2005-06-08Add two new tests for rpath argument merging and splitting.jlam3-1/+33
2005-06-02Fix typo.wiz1-2/+2
2005-06-02add tests for groups of mutually exclusive optionsdillo5-2/+47
2005-06-02adapt to quoting fixesdillo1-1/+1
2005-06-01test legacy var with `-' in option namedillo2-6/+6
2005-05-31fix comments describing test casesdillo6-12/+14
2005-05-31To ease cleanup of the options namespace, add code to support legacydillo3-2/+20
option names: PKG_OPTIONS_LEGACY_OPTS+= old:new If PKG_DEFAULT_OPTIONS or PKG_OPTIONS.foo contains option old (or -old) it is rewritten to new (or -new) and a warning is issued by the supported-options-message target.
2005-05-28two new test cases:dillo5-2/+47
- all: test overrides and PKG_LEGACY_VARS - unsupported: test that unsupported option in PKG_OPTIONS.xxx causes failure
2005-05-22Remove USE_GNU_TOOLS and replace with the correct USE_TOOLS definitions:jlam1-2/+2
USE_GNU_TOOLS -> USE_TOOLS awk -> gawk m4 -> gm4 make -> gmake sed -> gsed yacc -> bison
2005-05-19Simplified how the testcases are invoked.rillig2-8/+4
2005-05-19Added the pkg-options regression test.rillig1-1/+2
2005-05-16Added the ``order'' testcase, which works since bsd.options.mk, revisionrillig1-2/+2
1.22.
2005-05-16Imported testsuite for the bsd.options.mk framework.rillig6-0/+65
2005-05-15Added an example workaround to bug1.mk.rillig2-10/+30
2005-05-15Found a bug in the pkgsrc bmake and NetBSD-1.6.2 /usr/bin/make, which hasrillig3-2/+39
been fixed at least in NetBSD-2.99.15. Wrote a testcase for it.
2005-05-15Added make-quoting.rillig1-1/+2
2005-05-15Imported make-quoting, a regression test to make sure various weird quotingrillig18-0/+181
styles stay valid when updating the pkgsrc make(1) program. The test cases also serve as kind of demonstration how things work inside make(1).
2005-05-08Add and enable bootstrap-install-sh.jlam1-1/+2
2005-05-08Regression test for the bootstrap install-sh script which is a BSD installjlam3-0/+430
replacement.
2005-05-02Remove the need for ${FIND} in the top-level make.jlam2-18/+42
2005-05-02Remove NO_BUILD, which was causing the wrappers to not be created afterjlam1-2/+1
revision 1.1604 of bsd.pkg.mk.
2005-04-11Remove USE_BUILDLINK3 and NO_BUILDLINK; these are no longer used.tv3-6/+3
2005-02-12New test to verify that the USE_GNU_TOOLS mechanism works while buildinggavan3-1/+58
a different GNU tool. For now, this test fails if flex or bison are already installed. This is suboptimal, but the alternatives (choosing a different PKG_DBDIR and LOCALBASE) are difficult to implement at the individual test level.
2005-01-26Add new regression tests to test that we correctly handle full paths tojlam2-0/+16
shared libraries and shared modules.
2005-01-10Update to work with current bsd.wrapper.mk.jlam1-13/+5
2004-11-09Add regression tests to check that -DFOO="/usr/pkg/lib/libfoo.la" isn'tjlam2-0/+16
transformed by the wrapper scripts. This is the test case for the breakage in compiling audio/arts reported several weeks ago on tech-pkg@.
2004-09-25Add a test for the "rmdir" transformation.jlam2-1/+14
2004-09-08Add and enable print-plist.jmmv1-1/+2
2004-09-08Initial import of the print-plist regression test:jmmv4-0/+55
This test verifies that print-PLIST works, as well as related functionality such as PRINT_PLIST_AWK. Reviewed by gavan@.
2004-09-07Add a test for correctness of oddly-quoted arguments.jlam1-0/+8
2004-09-07Add missing spec file which was causing pkg_regress to not run this test.jlam1-0/+12
2004-08-30Check for the right output on a platform that has _USE_RPATH == no.jlam4-8/+36
2004-08-30Add and enable buildlink-libtool.jlam1-1/+2
2004-08-30Regression test for the buildlink libtool manipulations. Start withjlam3-0/+110
testing whether munging dependency_libs in installed libtool archives works properly.
2004-08-29Correct typo in comment.jlam2-4/+4
2004-08-29Make the lib-expand test do a little more: see that repeated library argsjlam1-2/+2
are nuked.
2004-08-28Provide a way to report the number of subtests passed and failed.jlam2-4/+36
2004-08-28Don't die as soon as the first test fails. Instead, run them all,jlam2-9/+27
then report how many failed and which ones. Return with an exit code indicating how many tests failed.
2004-08-28Allow this test to be used with both the current buildlink3 frameworkjlam1-6/+17
and the forthcoming wrapper script framework.
2004-08-28Add some more tests for "l" and "rm" BUILDLINK_TRANSFORM commands.jlam4-3/+38
2004-08-28Add and enable buildlink-unwrap.jlam1-1/+2
2004-08-28Regression test for whether *-config scripts and other similar files arejlam4-0/+139
correctly unwrapped (unbuildinkified).
2004-08-28Add some comments.jlam1-2/+9
2004-08-28Add spec file so that this can be driven by pkg_regress.jlam1-0/+12
2004-08-28Add NetBSD tag.jlam1-0/+2
2004-08-28Add pkgfail, a simple regression test which ensures that PKG_FAIL_REASON works.gavan3-1/+66
This test serves as a simple example of how to write a regression test.
2004-08-27Revert previous. The fact that they don't pass is something to fix, notjlam1-3/+1
to avoid (although this is fixed in the soon-to-be-committed wrapper framework).
2004-08-27Don't run the *-slashdot tests using the current buildlink3 framework,jlam1-1/+3
since they're known to be broken (see PR pkg/24760 and PR pkg/25500).
2004-08-27Make this work with the current incarnation of the buildlink3 framework.jlam1-8/+8