Message-ID: <1c317ed5e3f70b403206e6fdb181c1d0573e8b17.camel@microchip.com>
Subject: Re: [PATCH net-next 04/10] net: microchip: sparx5: Use chain ids
 without offsets when enabling rules
From:   Steen Hegelund <steen.hegelund@microchip.com>
To:     Dan Carpenter <error27@gmail.com>
CC:     "David S . Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>,
        <UNGLinuxDriver@microchip.com>,
        Randy Dunlap <rdunlap@infradead.org>,
        Casper Andersson <casper.casan@gmail.com>,
        "Russell King" <rmk+kernel@armlinux.org.uk>,
        Wan Jiabing <wanjiabing@vivo.com>,
        "Nathan Huckleberry" <nhuck@google.com>,
        <linux-kernel@vger.kernel.org>, <netdev@vger.kernel.org>,
        <linux-arm-kernel@lists.infradead.org>,
        "Daniel Machon" <daniel.machon@microchip.com>,
        Horatiu Vultur <horatiu.vultur@microchip.com>,
        Lars Povlsen <lars.povlsen@microchip.com>,
        Michael Walle <michael@walle.cc>
Date:   Mon, 13 Feb 2023 16:46:59 +0100
In-Reply-To: <Y+pTlf+2o0mVEErX@kadam>
References: <20230213092426.1331379-1-steen.hegelund@microchip.com>
         <20230213092426.1331379-5-steen.hegelund@microchip.com>
         <Y+oZjg8EkKp46V9Z@kadam>
         <b755fa1c818639a1e7c11ab3b2ac56443757ac3c.camel@microchip.com>
         <Y+pSKQdcpMw3YGvh@kadam> <Y+pTlf+2o0mVEErX@kadam>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.46.3 
MIME-Version: 1.0
Precedence: bulk

Hi Dan,

On Mon, 2023-02-13 at 18:13 +0300, Dan Carpenter wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know th=
e
> content is safe
>=20
> On Mon, Feb 13, 2023 at 06:07:21PM +0300, Dan Carpenter wrote:
> > On Mon, Feb 13, 2023 at 01:48:50PM +0100, Steen Hegelund wrote:
> > > Hi Dan,
> > >=20
> > > On Mon, 2023-02-13 at 14:05 +0300, Dan Carpenter wrote:
> > > > EXTERNAL EMAIL: Do not click links or open attachments unless you k=
now
> > > > the
> > > > content is safe
> > > >=20
> > > > On Mon, Feb 13, 2023 at 10:24:20AM +0100, Steen Hegelund wrote:
> > > > > diff --git a/drivers/net/ethernet/microchip/vcap/vcap_api.c
> > > > > b/drivers/net/ethernet/microchip/vcap/vcap_api.c
> > > > > index 68e04d47f6fd..9ca0cb855c3c 100644
> > > > > --- a/drivers/net/ethernet/microchip/vcap/vcap_api.c
> > > > > +++ b/drivers/net/ethernet/microchip/vcap/vcap_api.c
> > > > > @@ -1568,6 +1568,18 @@ static int vcap_write_counter(struct
> > > > > vcap_rule_internal *ri,
> > > > > =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return 0;
> > > > > =C2=A0}
> > > > >=20
> > > > > +/* Return the chain id rounded down to nearest lookup */
> > > > > +static int vcap_round_down_chain(int cid)
> > > > > +{
> > > > > +=C2=A0=C2=A0=C2=A0=C2=A0 return cid - (cid % VCAP_CID_LOOKUP_SIZ=
E);
> > > > > +}
> > > > > +
> > > > > +/* Return the chain id rounded up to nearest lookup */
> > > > > +static int vcap_round_up_chain(int cid)
> > > > > +{
> > > > > +=C2=A0=C2=A0=C2=A0=C2=A0 return vcap_round_down_chain(cid + VCAP=
_CID_LOOKUP_SIZE);
> > > >=20
> > > > Just use the round_up/down() macros.
> > >=20
> > > The only round up/down macros that I am aware of are:
> > >=20
> > > =C2=A0* round_up - round up to next specified power of 2
> > > =C2=A0* round_down - round down to next specified power of 2
> > >=20
> > > And I cannot use these as the VCAP_CID_LOOKUP_SIZE is not a power of =
2.
> > >=20
> > > Did I miss something here?
> > >=20
> >=20
> > Oh wow.=C2=A0 I didn't realize they needed to be a power of 2.=C2=A0 So=
rry!
>=20
> The correct macros are roundup/down().=C2=A0 Those don't have the power o=
f
> two requirement.

Yep - You are right, these are a bit further down in the math.h file, and I
never noticed them before :-)

I will update my code to use these instead.

>=20
> regards,
> dan carpenter
>=20

Thanks for the investigation!

BR
Steen