On Wed, Feb 8, 2023 at 10:26 PM Tyler Retzlaff <roret...@linux.microsoft.com> wrote: > > Announce deprecation of rte_ctrl_thread_create API which is the final > remaining stable API exposing pthread_t. > > Provide an equivalent replacement API rte_thread_create_control that > uses the new rte_thread_t type. > > Provide a unit test for the new rte_thread_create_control API > (test code provided by David Marchand.) > > Add missing doxygen index for thread / rte_thread.h. > > Notice! > > To limit compatibility regression risk and ease removal of the > existing rte_ctrl_thread_create in the future duplicate most of the > existing implementation rather than try to have it accommodate both > public API contracts. > > The duplication, the union introduced to support it along with remaining > internal pthread_xxx calls will be removed when rte_ctrl_thread_create is > finally removed. > > The old unit test for rte_ctrl_thread_create is kept in place to guarantee > correct behavior while deprecated and will be removed when > rte_ctrl_thread_create is finally removed. > > Series-acked-by: Morten Brørup <m...@smartsharesystems.com> > Reviewed-by: Mattias Rönnblom <mattias.ronnb...@ericsson.com> > > v6: > * fix missing comma in doxygen index > * combine deprecation notice for rte_ctrl_thread_create into notice > of removal of rte_thread_setname. > * remove test_lcore.c test and add test provided by David Marchand > to test_threads.c. > * rename the function to start with rte_thread to be consistent > with the rest of functions in the rte_thread.h APIs.
Thank you Tyler. The series lgtm. The deprecation notice is missing one ack, I'll still wait a bit. If we don't get it soon, I'll apply the first two patches for rc1. -- David Marchand