> + Ferruh > > <snip> > > Hi Joyce/Ferruh, > I do not think the port_status changes need to be handled atomically. The > changes to port_status do not seem to be happening from multiple threads. > It seems to be getting modified during initialization or through the test pmd > prompt. > > Do we really need atomic operations on port_status? >
Hi Honnappa, Seem you are right, I shall remove the atomic operations for port_status. Joyce > > > > Replace rte_atomic16_cmpset operation with compiler atomic CAS > > operation for port status sync in testpmd case. > > > > Signed-off-by: Joyce Kong <joyce.k...@arm.com> > > Reviewed-by: Ruifeng Wang <ruifeng.w...@arm.com> > > --- > > app/test-pmd/testpmd.c | 75 > > +++++++++++++++++++++++++++--------------- > > 1 file changed, 48 insertions(+), 27 deletions(-) > >
