Hi Ferruh, On Monday 29 August 2016 10:50 PM, Ferruh Yigit wrote: > On 8/26/2016 2:56 PM, Shreyansh Jain wrote: >> Now that all pdev are pci drivers, we don't need to register crypto and >> ethdev >> drivers through a dedicated channel. >> >> Signed-off-by: David Marchand <david.marchand at 6wind.com> >> Signed-off-by: Shreyansh Jain <shreyansh.jain at nxp.com> > > .... > >> -void rte_eth_driver_register(struct eth_driver *eth_drv); > > Function needs to be removed from .map file
Actually here I was a little confused. This is part of DPDK_2.2 - an old release. Would removing from an old release break compatibility for any system using the new release but still adhering to older APIs? Also, there is the case of this comment [1] from Neil Horman (and counter argument [2]) about not removing 'rte_cryptodev_pmd_driver_register' from map file because there was no deprecation notice. Overall situation is that I have maintained the previous patch of 'rte_cryptodev_pmd_driver_register' removal but refrained from removing new symbol - waiting for a little more clarification. Is it OK to remove a symbol from an older release without a deprecation notice? Or, should a deprecation notice follow this patch? [1] http://dpdk.org/ml/archives/dev/2016-June/042439.html [2] http://dpdk.org/ml/archives/dev/2016-June/042444.html > >> - >> -/** >> * Convert a numerical speed in Mbps to a bitmap flag that can be used in >> * the bitmap link_speeds of the struct rte_eth_conf >> * >> > > - Shreyansh