Hi everyone, I understand that when get_perthread_info() was introduced a decade ago, there were performance benefits of caching that result. But now that OIIO can cache this itself in a thread_local, is there any real benefit for the caller to still cache it? It's true that thread_local access can have overhead over non-thread_local, but compared to the much more expensive texture lookups being done I would think the overhead would go unnoticed.
I wonder if maybe this function should be removed/deprecated from the OIIO API and instead only be used internally?
_______________________________________________ Oiio-dev mailing list Oiio-dev@lists.openimageio.org http://lists.openimageio.org/listinfo.cgi/oiio-dev-openimageio.org
