![](https://secure.gravatar.com/avatar/2ea1ea42072986bc3dd023464d0c2de6.jpg?s=120&d=mm&r=g)
28 Nov
2017
28 Nov
'17
7:53 a.m.
Hi Takashi
I wonder are these patches acceptable ? or not ? If acceptable, should I resend ?
Since the use-case to pass NULL is limited, I'd rather avoid the change. If there were more than handful users, maybe I would rethink, but currently not convincing enough.
And, the NULL check is needed only when the caller is outside the stream lock. Such a situation already smells like a buggy code. We don't need to make it easier :)
OK, I see Let's remove it so far Thanks
Best regards --- Kuninori Morimoto