[Spice-devel] [PATCH] gtask-helper: include functions to check coroutine context

[Fabiano Fidêncio]

Toso,

On Sat, May 14, 2016 at 10:05 PM, Victor Toso <lists at victortoso.com> wrote:
> Hi,
>
> On Sat, May 14, 2016 at 06:34:41PM +0200, Fabiano Fidêncio wrote:
>> On Sat, May 14, 2016 at 5:03 PM, Victor Toso <victortoso at redhat.com> wrote:
>> > If we are in coroutine context, fallback to return in idle; Otherwise,
>> > let GTask decide if callback should be called or not.
>> > ---
>> >  src/gtask-helper.h | 45 +++++++++++++++++++++++++++++++++++++++++++++
>> >  1 file changed, 45 insertions(+)
>> >
>> > diff --git a/src/gtask-helper.h b/src/gtask-helper.h
>> > index 81c041f..8771021 100644
>> > --- a/src/gtask-helper.h
>> > +++ b/src/gtask-helper.h
>> > @@ -29,6 +29,51 @@ void g_task_helper_return_error_in_idle(GTask *task, GError *error);
>> >  void g_task_helper_return_new_error_in_idle(GTask *task, GQuark domain, gint code, const char *format, ...);
>> >  void g_task_helper_return_pointer_in_idle(GTask *task, gpointer result, GDestroyNotify result_destroy);
>> >
>> > +#define g_task_helper_return_boolean(...) \
>> > +    do { \
>> > +        if (coroutine_self_is_main()) { \
>> > +            g_task_return_boolean(__VA_ARGS__); \
>> > +        } else { \
>> > +            g_task_helper_return_boolean_in_idle(__VA_ARGS__); \
>> > +        } \
>> > +    } while (0)
>> > +
>> > +#define g_task_helper_return_int(...) \
>> > +    do { \
>> > +        if (coroutine_self_is_main()) { \
>> > +            g_task_return_int(__VA_ARGS__); \
>> > +        } else { \
>> > +            g_task_helper_return_int_in_idle(__VA_ARGS__); \
>> > +        } \
>> > +    } while (0)
>> > +
>> > +#define g_task_helper_return_error(...) \
>> > +    do { \
>> > +        if (coroutine_self_is_main()) { \
>> > +            g_task_return_error(__VA_ARGS__); \
>> > +        } else { \
>> > +            g_task_helper_return_error_in_idle(__VA_ARGS__); \
>> > +        } \
>> > +    } while (0)
>> > +
>> > +#define g_task_helper_return_new_error(...) \
>> > +    do { \
>> > +        if (coroutine_self_is_main()) { \
>> > +            g_task_return_new_error(__VA_ARGS__); \
>> > +        } else { \
>> > +            g_task_helper_return_new_error_in_idle(__VA_ARGS__); \
>> > +        } \
>> > +    } while (0)
>> > +
>> > +#define g_task_helper_return_pointer(...) \
>> > +    do { \
>> > +        if (coroutine_self_is_main()) { \
>> > +            g_task_return_pointer(__VA_ARGS__); \
>> > +        } else { \
>> > +            g_task_helper_return_pointer_in_idle(__VA_ARGS__); \
>> > +        } \
>> > +    } while (0)
>> > +
>> >  G_END_DECLS
>> >
>> >  #endif /* __G_TASK_HELPER_H__ */
>> > --
>> > 2.5.5
>> >
>> > _______________________________________________
>> > Spice-devel mailing list
>> > Spice-devel at lists.freedesktop.org
>> > https://lists.freedesktop.org/mailman/listinfo/spice-devel
>>
>> NACK!
>> This is not what we agreed on IRC yesterday :-\
>> Anyways, with https://lists.freedesktop.org/archives/spice-devel/2016-May/029040.html
>> this patch doesn't seem to be needed anymore.
>
> Yes, this is not needed. No, it isn't what we agreed on IRC, just the
> idea that is. I realized while testing that this would not needed but I
> thought it would be good to share the patch anyway. I should have
> mentioned this in the email.
>
> Sorry for taking your time and thanks for reviewing.

There is nothing to be sorry about!
I'd like to have the g_task_helper_* in anyways, but without the
smartness (I mean, just returning in idle).
I do believe it's cleaner/better than just adding an idle. But it's
not up to me.

>   toso
>
>> _______________________________________________
>> Spice-devel mailing list
>> Spice-devel at lists.freedesktop.org
>> https://lists.freedesktop.org/mailman/listinfo/spice-devel
> _______________________________________________
> Spice-devel mailing list
> Spice-devel at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/spice-devel


Best Regards,
-- 
Fabiano Fidêncio