|
From: Fredrik H. <fre...@ax...> - 2008-12-30 12:44:14
|
Hi, A new, long awaited release of check is now available. This release is mostly a bug-fix release, no big new features are available. Please test it out and report any problems you might have! http://prdownloads.sourceforge.net/check/check-0.9.6.tar.gz?download Thanks, /Fredrik Hugosson aka hugo303 RELEASE NOTES ============= Mon, Dec 29, 2009: Released Check 0.9.6 based on r453 (2008-12-01 22:14:15). * 'make distcheck' does not work out of the box. Disable the two top lines in doc/Makefile.am and it shall pass. Will have to be fixed later. * add call to AC_REPLACE_FUNCS([strsignal]) * add new rpl_strsignal following rpl_(re)malloc template, body of function due to Roland Illig -- hopefully closes 1629755 * add missing NULL argument to fail* varargs macro calls * define incorrect tests for __GNUC__ only -- both per Roland Illig in bug 1677391 * define CK_ATTRIBUTE_UNUSED for GCC >= 2.95, closing: [ 1674626 ] compile error with non defined __attribute__ compiler * Fixed error in documentation example Makefile.am, bug #1888237 * Fixed spelling (patch #1652630) * Handle NULL in srunner_add_suite(). Fixes #1624887 * add CK_FORK_GETENV to enum fork_status and delete CK_FORK_UNSPECIFIED * call strdup on result from strsignal to avoid clobbering it * add Torok Edwin to AUTHORS * add CK_TEST_RESULT_INVALID to enum test_result * add CK_CTX_INVALID to ck_result_ctx and don't use -1 anymore * count checks in setup() as well; patch due to Roland Stigge * use int __attribute__((unused)) _i instead of int _i __attribute__((unused)) per gcc-3.3.5 request from Sebastian Trahm * rename signal to _signal in _tcase_add_test to avoid a symbol clash with /usr/include/sys/signal.h on OS X. * define rpl_malloc and rpl_realloc for platforms where !malloc(0) and !realloc(0,0), such as AIX, because configure goes and redefines malloc/realloc in this case... * SVNChangeLog patch from Robert Collins * Incorporated patch from Debian for debian bug #395466. This fixes 'AM_PATH_CHECK causes "possibly undefined macro" errors'. * Added new Check fail API. Implemented on top of fail_unless. Future versions will reverse this so fail_unless is implemented on top of the ck_assert API. This API is not documented yet, will probably not be fully released until 0.9.8, when it will be possible to choose API. |