On Mon, 3 Jun 2019, Richard Biener wrote:
>
> This mostly fixes PR90726, employing proper visited mechanisms to
> avoid exponential behavior when walking a GENERIC expression
> with shared trees. It also fixes the code-generation side where
> gimplification of such tree naturally explodes as wel
This mostly fixes PR90726, employing proper visited mechanisms to
avoid exponential behavior when walking a GENERIC expression
with shared trees. It also fixes the code-generation side where
gimplification of such tree naturally explodes as well - but
not by more optimal gimplifying but accounti
