From: Jan Stancek Date: Sun, 16 Jun 2019 03:34:02 -0400 (EDT) Subject: [LTP] [PATCH v3 3/3] syscalls/ioctl_ns0[156]: align stack and wait for child In-Reply-To: <20190613151419.GE29926@rei.lan> References: <1b6ddab9cd8e3620da9e37b1132e911280c22e32.1560410182.git.jstancek@redhat.com> <1320546599.27920597.1560420988025.JavaMail.zimbra@redhat.com> <20190613141714.GC29926@rei.lan> <1211798607.28016198.1560437855198.JavaMail.zimbra@redhat.com> <20190613151419.GE29926@rei.lan> Message-ID: <151498464.28356667.1560670442630.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 ----- > Hi! > > > If I remember correctly at some point we decided to clean up after tests > > > properly so that we don't upset various debugging tools, i.e. coverity, > > > valgrind, etc. and I think that you were part of that discussion. > > > > I recall I started with that position (free all), and I thought you > > turned me around after this many years :-). > > Well I didn't care that much, but I guess that I lean slightly to free > the memory :-). OK, so should I repost or is this good to go with free added to cleanup? > > > Do we have anything about this in style guide? I only found brief mention > > in "don't call cleanup from setup" section, which isn't even possible with > > newlib. > > I don't think so. I guess that we should write something down, once we > decide what is the prefered option. > > > ... > > You don't need to clean up the following: > > > > * +malloc(3)+'ed memory. > > * Read-only file descriptors in persistent paths (i.e. not > > temporary directories). > > Looks like this is terribly outdated, at least I tend to tell people to > close all filedescriptors to make things simpler. > > -- > Cyril Hrubis > chrubis@suse.cz >