RE: [PATCH v3 net-next 01/13] net: enetc: add initial netc-lib driver to support NTMP
From: Wei Fang
Date: Thu Mar 06 2025 - 20:47:19 EST
> On Tue, 4 Mar 2025 15:21:49 +0800 Wei Fang wrote:
> > +config NXP_NETC_LIB
> > + tristate "NETC Library"
>
> Remove the string after "tristate", the user should not be prompted
> to make a choice for this, since the consumers "select" this config
> directly.
>
Okay, I will remove it.
> > + help
> > + This module provides common functionalities for both ENETC and NETC
> > + Switch, such as NETC Table Management Protocol (NTMP) 2.0, common
> tc
> > + flower and debugfs interfaces and so on.
> > +
> > + If compiled as module (M), the module name is nxp-netc-lib.
>
> Not sure if the help makes sense for an invisible symbol either.
Yes, I think it can also be removed. Thanks.
>
> > config FSL_ENETC
> > tristate "ENETC PF driver"
> > depends on PCI_MSI
> > @@ -40,6 +50,7 @@ config NXP_ENETC4
> > select FSL_ENETC_CORE
> > select FSL_ENETC_MDIO
> > select NXP_ENETC_PF_COMMON
> > + select NXP_NETC_LIB
> > select PHYLINK
> > select DIMLIB
> > help
>
> > +#pragma pack(1)
>
> please don't blindly pack all structs, only if they are misaligned
> or will otherwise have holes.
Because these structures are in hardware buffer format and need
to be aligned, so for convenience, I simply used pack(1). You are right,
I should use pack() for structures with holes. Thanks.
>
> > +#if IS_ENABLED(CONFIG_NXP_NETC_LIB)
>
> why the ifdef, all callers select the config option
hm..., there are some interfaces of netc-lib are used in common .c files
in downstream, so I used "ifdef" in downstream. Now for the upstream,
I'm going to separate them from the common .c files. So yes, we can
remove it now.