Re: [dpdk-dev] [PATCH v3 5/6] net/ena: rework RSS configuration

[Thomas Monjalon <thomas@monjalon.net>]

23/07/2021 11:49, Michał Krawczyk:
> pt., 23 lip 2021 o 11:43 Thomas Monjalon <thomas@monjalon.net> napisał(a):
> >
> > 23/07/2021 11:39, Michał Krawczyk:
> > > pt., 23 lip 2021 o 11:22 Thomas Monjalon <thomas@monjalon.net> napisał(a):
> > > >
> > > > 14/07/2021 12:43, Michal Krawczyk:
> > > > > +               PMD_DRV_LOG(WARNING,
> > > > > +                       "Setting RSS hash fields is not supported. Using default values: 0x%llx\n",
> > > > > +                       ENA_ALL_RSS_HF);
> > > > >
> > > >
> > > > checkpatch is returning this warning:
> > > >         Using %l format, prefer %PRI*64 if type is [u]int64_t
> > > >
> > > > Indeed rss_hf is uint64_t.
> > > >
> > > >
> > >
> > > Hi Thomas,
> > >
> > > I'm aware of that checkpatch warning, but changing format to %PRIx64
> > > results in another (compilation) warning:
> > >
> > > [1205/2572] Compiling C object
> > > drivers/libtmp_rte_net_ena.a.p/net_ena_ena_rss.c.o
> > > In file included from ../drivers/net/ena/ena_rss.c:7:
> > > ../drivers/net/ena/ena_rss.c: In function ‘ena_rss_hash_set’:
> > > ../drivers/net/ena/ena_logs.h:35:3: warning: format ‘%lx’ expects
> > > argument of type ‘long unsigned int’, but argument 5 has type ‘long
> > > long unsigned int’ [-Wformat=]
> > >    35 |   "%s(): " fmt, __func__, ## args)
> > >       |   ^~~~~~~~
> > > ../drivers/net/ena/ena_rss.c:427:3: note: in expansion of macro ‘PMD_DRV_LOG’
> > >   427 |   PMD_DRV_LOG(WARNING,
> > >       |   ^~~~~~~~~~~
> > > In file included from ../lib/mempool/rte_mempool.h:40,
> > >                  from ../lib/mbuf/rte_mbuf.h:38,
> > >                  from ../lib/net/rte_ether.h:23,
> > >                  from ../drivers/net/ena/ena_ethdev.h:10,
> > >                  from ../drivers/net/ena/ena_rss.c:6:
> > > /usr/include/inttypes.h:121:34: note: format string is defined here
> > >   121 | # define PRIx64  __PRI64_PREFIX "x"
> >
> >
> > Sorry I don't have time to dig, but I'm sure you can fix it.
> >
> >
> 
> Yeah, sure, casting ENA_RSS_HF to uint64_t in this log should do the
> trick - I'll fix that in the v4.

I think you should declare ENA_RSS_HF with UINT64_C()