Re: [Qemu-devel] [PATCH] main-loop: Fix build for w32 and w64

[Stefan Weil]

Am 28.04.2012 16:34, schrieb Paolo Bonzini:
> Il 27/04/2012 17:02, Stefan Weil ha scritto:
>
> > diff --git a/main-loop.c b/main-loop.c
> > index 0457bf2..24cf540 100644
> > --- a/main-loop.c
> > +++ b/main-loop.c
> > @@ -410,6 +410,7 @@ static int os_host_main_loop_wait(uint32_t timeout)
> > int ret, i;
> > PollingEntry *pe;
> > WaitObjects *w = &wait_objects;
> > + gint poll_timeout;
> > static struct timeval tv0;
> >
> > /* XXX: need to suppress polling by better using win32 events */
> > @@ -424,12 +425,12 @@ static int os_host_main_loop_wait(uint32_t timeout)
> > if (nfds >= 0) {
> > ret = select(nfds + 1, &rfds, &wfds, &xfds, &tv0);
> > if (ret != 0) {
> > - timeout = 0;
> > + /* TODO. */
>
> Please leave timeout = 0 here and...
>
> > }
> > }
> >
> > g_main_context_prepare(context, &max_priority);
> > - n_poll_fds = g_main_context_query(context, max_priority, &timeout,
> > + n_poll_fds = g_main_context_query(context, max_priority, &poll_timeout,
> > poll_fds, ARRAY_SIZE(poll_fds));
> > g_assert(n_poll_fds <= ARRAY_SIZE(poll_fds));
> >
> > @@ -439,7 +440,7 @@ static int os_host_main_loop_wait(uint32_t timeout)
> > }
> >
> > qemu_mutex_unlock_iothread();
> > - ret = g_poll(poll_fds, n_poll_fds + w->num, timeout);
>
> ... here do:
>
> if (poll_timeout == INFINITE || timeout < poll_timeout) {
> poll_timeout = timeout;
> }

With poll_timeout of type gint = int and INFINITE = 0xffffffff,
poll_timeout == INFINITE would never be true. What about

if (poll_timeout < 0 || timeout < poll_timeout) {

According to the glib documentation, "a negative value indicates
an infinite timeout".

> > + ret = g_poll(poll_fds, n_poll_fds + w->num, poll_timeout);
> > qemu_mutex_lock_iothread();
> > if (ret > 0) {
> > for (i = 0; i < w->num; i++) {
>
> Thanks for fixing the breakage.
>
> Paolo

Thanks for your review.

Stefan