[libvirt PATCH] tests: Use freecon in testSELinuxCheckLabels

[Daniel P. Berrangé]

On Fri, Feb 11, 2022 at 03:16:35PM +0100, Ján Tomko wrote:
> On a Friday in 2022, Jiri Denemark wrote:
> > On Fri, Feb 11, 2022 at 14:34:04 +0100, Jiri Denemark wrote:
> > > This partially reverts commit 0fc4a43d248b86fd54ad7323beb66faec8c1043c.
> > > 
> > > Signed-off-by: Jiri Denemark <jdenemar at redhat.com>
> > > ---
> > >  tests/securityselinuxlabeltest.c | 4 +++-
> > >  1 file changed, 3 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/tests/securityselinuxlabeltest.c b/tests/securityselinuxlabeltest.c
> > > index dfe9d38d34..83532d8090 100644
> > > --- a/tests/securityselinuxlabeltest.c
> > > +++ b/tests/securityselinuxlabeltest.c
> > > @@ -247,8 +247,10 @@ testSELinuxCheckLabels(testSELinuxFile *files, size_t nfiles)
> > >              virReportError(VIR_ERR_INTERNAL_ERROR,
> > >                             "File %s context '%s' did not match expected '%s'",
> > >                             files[i].file, ctx, files[i].context);
> > > +            freecon(ctx);
> > >              return -1;
> > >          }
> > > +        freecon(ctx);
> > >      }
> > >      return 0;
> > >  }
> > 
> > Self-NACK
> > 
> > getfilecon is mocked in this test so not using freecon() was actually
> > correct.
> > 
> 
> Even if it was from libselinux [0], is there something wrong with using
> plain free for a char * variable?

SELinux uses to use 'security_context_t' data type for all its
APIs, whicih was a typedef for 'char *'.  They stopped doing
that and now just use 'char *'. The docs for freecon still say
it is to be used to free memory from anmy other get* function,
but impl is just delegating to free()

Regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|