15/07/2021 14:55, David Marchand:
> On Wed, Jul 7, 2021 at 12:56 PM Ivan Malov <ivan.ma...@oktetlabs.ru> wrote:
> > diff --git a/drivers/net/sfc/sfc_flow.h b/drivers/net/sfc/sfc_flow.h
> > index bd3b374d68..5b1f34aa7a 100644
> > --- a/drivers/net/sfc/sfc_flow.h
> > +++ b/drivers/net/sfc/sfc_flow.h
> > @@ -136,13 +136,21 @@ typedef int (sfc_flow_item_parse)(const struct
> > rte_flow_item *item,
> >
> > struct sfc_flow_item {
> > enum rte_flow_item_type type; /* Type of item */
> > + const char *name; /* Item name */
> > enum sfc_flow_item_layers layer; /* Layer of item */
> > enum sfc_flow_item_layers prev_layer; /* Previous layer of item */
> > enum sfc_flow_parse_ctx_type ctx_type; /* Parse context type */
> > sfc_flow_item_parse *parse; /* Parsing function */
> > };
> >
> > -int sfc_flow_parse_pattern(const struct sfc_flow_item *flow_items,
> > +#define SFC_FLOW_ITEM(_name) \
> > + .type = RTE_FLOW_ITEM_TYPE_##_name, \
> > + .name = #_name
>
> Using this concatenation (same for SFC_FLOW_ACTION) confuses
> devtools/parse-flow-support.sh script.
>
> $ ./devtools/check-doc-vs-code.sh
> rte_flow doc out of sync for sfc
> item
> item pf
> item phy_port
> item port_id
> item pppoes
> item tcp
> item vf
> action
>
> Either net/sfc keeps full RTE_FLOW_x_TYPE_y form like other drivers,
> or the check script must be updated (see exclude/include helpers).
Generating code with macros is also counter-intuitive for grep.
I think it should be avoided and keep full form RTE_FLOW_x_TYPE_y.
