Package: cl-rt Version: 20040621-3 Severity: normal
The cl-rt regression test package does not pass its own self-tests. To reproduce: Load the system into the Lisp implementation (using ASDF). When prompted enter a name for a scratch file (e.g. "foo") Load rt-test.lisp execute (rt:do-tests) Result: CL-USER> (rt:do-tests) Doing 25 pending tests of 25 tests total. Test DEFTEST-1 failed ... (additional test failures reported) ...15 out of 30 total tests failed: DEFTEST-1, DEFTEST-3, DO-TEST-1, DO- TEST-2, REM-TEST-1, REM-TEST-2, REM-TEST-3, REM-TEST-4, DO-TESTS-1, DO-TESTS-2, DO-TESTS-4, DO- TESTS-5,
CONTINUE-TESTING-1, (TEMP), (T 2). NIL Expected result: CL-USER> (rt:do-tests) Doing 25 pending tests of 25 tests total. DEFTEST-1 DEFTEST-2 DEFTEST-3 DEFTEST-4 DO-TEST-1 DO-TEST-2 DO-TEST-3 GET-TEST-1 GET-TEST-2 GET-TEST-3 GET-TEST-4 GET-TEST-5 REM-TEST-1 REM-TEST-2 REM-TEST-3 REM-TEST-4 REM-TEST-5 REM-ALL-TESTS-1 REM-ALL-TESTS-2 DO-TESTS-1 DO-TESTS-2 DO-TESTS-3 DO-TESTS-4 DO-TESTS-5 CONTINUE-TESTING-1 No tests failed. TRoot cause: rt-test.lisp was not updated to account for the new format of the test database introduced with "pfdietz's modifications". The failing tests are meant to work with an empty test database, but
the code to empty the database is for the old, simpler format.Fix: the attached Git-formatted patches applied to rt-test.lisp fix the problem.
0001-Add-with-empty-test-database-macro.patch
Description: Binary data
0002-Fix-self-tests.patch
Description: Binary data
