From: Jan Stancek Date: Thu, 4 Jul 2019 04:08:39 -0400 (EDT) Subject: [LTP] [PATCH v2] lib: add tst_no_corefile to avoid corefile dumping In-Reply-To: <20190703072417.24091-1-liwang@redhat.com> References: <20190703072417.24091-1-liwang@redhat.com> Message-ID: <2030850327.31633394.1562227719657.JavaMail.zimbra@redhat.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ltp@lists.linux.it ----- Original Message ----- > If crash is expected in a testcase, we can avoid dumping core file > in calling this function. > > Signed-off-by: Li Wang > Cc: Jan Stancek > --- > > Notes: > v1 --> v2 > * add a paramenter to hide the message print > * add notes in test-writing-guidelines.txt > > doc/test-writing-guidelines.txt | 14 ++++++++++++-- > include/tst_safe_macros.h | 18 ++++++++++++++++++ > .../kernel/security/umip/umip_basic_test.c | 2 ++ > testcases/kernel/syscalls/ipc/shmat/shmat01.c | 16 +++------------- > 4 files changed, 35 insertions(+), 15 deletions(-) > > diff --git a/doc/test-writing-guidelines.txt > b/doc/test-writing-guidelines.txt > index c6d4e001d..1a77db6bf 100644 > --- a/doc/test-writing-guidelines.txt > +++ b/doc/test-writing-guidelines.txt > @@ -826,8 +826,8 @@ The 'TST_PROCESS_STATE_WAIT()' waits until process 'pid' > is in requested > It's mostly used with state 'S' which means that process is sleeping in > kernel > for example in 'pause()' or any other blocking syscall. > > -2.2.10 Signal handlers > -^^^^^^^^^^^^^^^^^^^^^^ > +2.2.10 Signals and signal handlers > +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > If you need to use signal handlers, keep the code short and simple. Don't > forget that the signal handler is called asynchronously and can interrupt > the > @@ -859,6 +859,16 @@ type defined in C99 but this one *DOES NOT* imply > 'volatile' (it's just a > 'typedef' to 'int'). So the correct type for a flag that is changed from a > signal handler is either 'volatile int' or 'volatile sig_atomic_t'. > > +If a crash (e.g. triggered by signal SIGSEGV) is expected in testing, you > can > +avoid dumping core file via calling this below tst_no_corefile() function. > +Note that this chanage will only effect on that process with this invoke. > And I'd rephrase it to: If a crash (e.g. triggered by signal SIGSEGV) is expected in testing, you can avoid creation of core files by calling tst_no_corefile() function. This takes effect for process (and its children) which invoked it, unless they subsequently modify RLIMIT_CORE. Note that LTP library will reap any processes that test didn't reap itself, and report any non-zero exit code as failure. One note below. > +the parameter 'verbose' is used as message print option. > + > +[source,c] > +------------------------------------------------------------------------------- > +void tst_no_corefile(int verbose); > +------------------------------------------------------------------------------- > + > 2.2.11 Kernel Modules > ^^^^^^^^^^^^^^^^^^^^^ > > diff --git a/include/tst_safe_macros.h b/include/tst_safe_macros.h > index 53a888c80..c4ddf84ef 100644 > --- a/include/tst_safe_macros.h > +++ b/include/tst_safe_macros.h > @@ -394,6 +394,24 @@ static inline int safe_setrlimit(const char *file, const > int lineno, > #define SAFE_SETRLIMIT(resource, rlim) \ > safe_setrlimit(__FILE__, __LINE__, (resource), (rlim)) > > +/* > + * Crash is expected, avoid dumping corefile. > + * 1 is a special value, that disables core-to-pipe. > + * At the same time it is small enough value for > + * core-to-file, so it skips creating cores as well. > + */ > +static inline void tst_no_corefile(int verbose) > +{ > + struct rlimit r; > + > + r.rlim_cur = 1; > + r.rlim_max = 1; > + SAFE_SETRLIMIT(RLIMIT_CORE, &r); SAFE_SETRLIMIT is fine if needs_root = 1. But if test runs as unprivileged user and RLIMIT_CORE is already 0, unprivileged user won't be able to increase it, so we get TBROK here.