On Thursday 29 June 2017 07:59 PM, Shreyansh Jain wrote: > Hello Ferruh, > [...] > >> >>> + >>> + if (rte_eal_process_type() == RTE_PROC_PRIMARY) >>> + rte_free(eth_dev->data->dev_private); >>> + >> >> no pmd uninit() ?
Just to clarify, were you asking about uninit of the driver? There is no such call from within normal flow - that is probably left to the application to perform. (rte_dpaa_driver_unregister) I just cross checked, this is not being done even for the PCI functions (rte_pci_unregister) - probably because we never see a case (for sample applications) where driver is unregistered. Am I missing something here? > > I will fix this. There is an internal commit that we had very > recently (a miss in previous series). There was some device/queue cleanup which was missing which I have added and will push in the next version. I mixed that up with 'PMD uninit' comment when I was replying to your previous email. > >> - Shreyansh