From: Cyril H. <ch...@su...> - 2010-11-01 18:55:38
|
Hi! > The branch, master, has been updated > via 12bb83d825c7b58bb4fb3322620591c9fdf4ff5c (commit) > from 1aba4c6068d5164b38e2ea9248f4be1a2721933c (commit) > > > - Log ----------------------------------------------------------------- > commit 12bb83d825c7b58bb4fb3322620591c9fdf4ff5c > Author: Subrata Modak <sub...@su...> > Date: Wed Oct 13 12:45:43 2010 +0530 > > securebits: add secure_keepcaps testcases. > This adds basic tests of the keepcaps securebits settings. Lots more securebits tests to come (see my email from one or 1.5 years ago, and, heck, write them if you have time :). Signed-off-by: Serge E. Hallyn <ser...@ca...>. > > ----------------------------------------------------------------------- > > Summary of changes: > m4/ltp-securebits.m4 | 24 +++ > runtest/securebits | 2 + > testcases/kernel/security/Makefile | 5 +- > testcases/kernel/security/securebits/Makefile | 28 ++++ > testcases/kernel/security/securebits/README | 10 ++ > .../kernel/security/securebits/check_keepcaps.c | 161 ++++++++++++++++++++ > .../kernel/security/securebits/run_securebits.sh | 20 +++ Looks like there is m4 function for securebits but it's not called from configure.ac, so there is no way this could be detected. And even the m4/ltp-securebits.m4 is broken. This effectively breaks build on system without <linux/securebits.h>. Anyway it looks like HAVE_SETCAP from security/Makefile doesn't work either as I'm getting "checking for setcap... false" but these tests compiles anyway. (There is HAVE_SETCAP=false in m4/Makefile but it's not included from Makefile system anywhere (at least I couldn't find it)). Garrett is there any correct way how we could propagate information about missing headers into the Makefile so we could skip some subtrees accordingly? -- Cyril Hrubis ch...@su... |