2015-10-30 22:16, Zhe Tao: > The vPMD RX function uses the multi-buffer and SSE instructions to > accelerate the RX speed, but now the pktype cannot be supported by the vPMD > RX, > because it will decrease the performance heavily.
So we cannot rely on the pktype information? Depending on which Rx path is used, we will have the information or not. > Signed-off-by: Zhe Tao <zhe.tao at intel.com> > +CONFIG_RTE_LIBRTE_I40E_RX_OLFLAGS_ENABLE=y When we'll have some devargs options for PCI dev, we could move this option to runtime.
