adjust_pollset improvements

[icing]

There is the possibility in the existing implementation that for HTTP/2 the pollset adjustment could still lead to cpu busy loops. If HTTP/2 flow control blocks sending, the h2 filter would replace a POLLOUT in the pollset adjustment with a POLLIN. But the underlying cf-socket filter would still add a POLLOUT. A writeable socket would then lead to sends that are blocked in the HTTP/2 filters.

Fix:

• let multi_getsock() initialize the pollset in what the transfer state requires in regards to SEND/RECV
• change connection filters adjust_pollset() implementation to react on the presence of POLLIN/-OUT in the pollset and no longer check CURL_WANT_SEND/CURL_WANT_RECV
• cf-socket will no longer add POLLIN on its own
• http2 and http/3 filters will only do adjustments if the passed pollset wants to POLLIN/OUT for the transfer on the socket. This is similar to the HTTP/2 proxy filter and works in stacked filters.