I thought I added this fix before but apparently not. Looks good to me. Thank you for bringing this up. Hajime > On Jul 27, 2017, at 8:07 PM, Tanu Kaskinen <tanuk at iki.fi> wrote: > > CID: 1398155 > --- > src/modules/raop/raop-sink.c | 7 +++++++ > 1 file changed, 7 insertions(+) > > diff --git a/src/modules/raop/raop-sink.c b/src/modules/raop/raop-sink.c > index e5d219e87..4d13927fc 100644 > --- a/src/modules/raop/raop-sink.c > +++ b/src/modules/raop/raop-sink.c > @@ -391,6 +391,13 @@ static void thread_func(void *userdata) { > if (!pa_raop_client_can_stream(u->raop)) > continue; > > + /* This assertion is meant to silence a complaint from Coverity about > + * pollfd being possibly NULL when we access it later. That's a false > + * positive, because we check pa_raop_client_can_stream() above, and if > + * that returns true, it means that the connection is up, and when the > + * connection is up, pollfd will be non-NULL. */ > + pa_assert(pollfd); > + > if (u->memchunk.length <= 0) { > if (u->memchunk.memblock) > pa_memblock_unref(u->memchunk.memblock); > -- > 2.13.2 >