On Thu, May 23, 2019 at 10:37 AM Jakub Janků <[email protected]> wrote: > > The pointer to loop must be set to NULL on unref. > > Quit signal handler can be called at any time, > using g_main_loop_quit() on a freed loop could lead to segfault.
I am not sure if this is enough to solve the race you describe (handler could still be called before loop is set to null, right?), but the change seems to go in the right direction. ack > > Signed-off-by: Jakub Janků <[email protected]> > --- > spice/spice-webdavd.c | 5 +++-- > 1 file changed, 3 insertions(+), 2 deletions(-) > > diff --git a/spice/spice-webdavd.c b/spice/spice-webdavd.c > index 6ad63c5..3fac28b 100644 > --- a/spice/spice-webdavd.c > +++ b/spice/spice-webdavd.c > @@ -243,7 +243,8 @@ quit (int sig) > if (sig == SIGINT || sig == SIGTERM) > quit_service = TRUE; > > - g_main_loop_quit (loop); > + if (loop) > + g_main_loop_quit (loop); > } > > static Client * > @@ -982,7 +983,7 @@ run_service (ServiceData *service_data) > > start_mux_read (mux_istream); > g_main_loop_run (loop); > - g_main_loop_unref (loop); > + g_clear_pointer (&loop, g_main_loop_unref); > > #ifdef G_OS_WIN32 > g_cancellable_cancel (map_drive_data.cancel_map); > -- > 2.21.0 > > _______________________________________________ > Spice-devel mailing list > [email protected] > https://lists.freedesktop.org/mailman/listinfo/spice-devel -- Marc-André Lureau _______________________________________________ Spice-devel mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/spice-devel
