RE: [PATCH v3 1/3] doc: announce bonding macro change

["Morten Brørup" <mb@smartsharesystems.com>]

> From: Chaoyong He [mailto:chaoyong.he@corigine.com]
> Sent: Tuesday, 18 July 2023 10.29
> 
> From: Long Wu <long.wu@corigine.com>
> 
> In order to support inclusive naming, some of the macro in DPDK will
> need to be renamed. Do this through deprecation process now for 23.07.
> 
> Signed-off-by: Long Wu <long.wu@corigine.com>
> Reviewed-by: Chaoyong He <chaoyong.he@corigine.com>
> ---
>  app/test-pmd/testpmd.c                 | 2 +-
>  doc/guides/rel_notes/deprecation.rst   | 4 ++++
>  drivers/net/bonding/rte_eth_bond_api.c | 6 +++---
>  lib/ethdev/rte_ethdev.h                | 5 +++--
>  4 files changed, 11 insertions(+), 6 deletions(-)
> 
> diff --git a/app/test-pmd/testpmd.c b/app/test-pmd/testpmd.c
> index c6ad9b18bf..938ca035d4 100644
> --- a/app/test-pmd/testpmd.c
> +++ b/app/test-pmd/testpmd.c
> @@ -4248,7 +4248,7 @@ uint8_t port_is_bonding_slave(portid_t slave_pid)
>  			slave_pid);

We might want to get rid of the term "slave" in function, parameter and variable names in testpmd too. This is not important now, and can be postponed to a later patch.

>  		return 0;
>  	}
> -	if ((*dev_info.dev_flags & RTE_ETH_DEV_BONDED_SLAVE) || (port-
> >slave_flag == 1))
> +	if ((*dev_info.dev_flags & RTE_ETH_DEV_BONDING_MEMBER) || (port-
> >slave_flag == 1))

Can we please standardize on using only "bond" and "BOND" everywhere in this driver, instead of both "bond" (in function names) and "BONDING" (in enum values)?

The source code file are also named "rte_eth_bond...", while the directory name is /drivers/net/bonding/.

We are about to rename anyway, so let's do it as good as we can.


Also, are the bonding and balancing modes defined in /drivers/net/bonding/rte_eth_bond.h missing the RTE_ prefix, or are they private?


PS: Sorry about joining the discussion late (and possibly ignoring previous discussions on this). I reacted to the poll for ACKs to this series.