Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Adam Manzanares <a.manzanares@samsung.com>
To:     Davidlohr Bueso <dave@stgolabs.net>
CC:     "alison.schofield@intel.com" <alison.schofield@intel.com>,
        "vishal.l.verma@intel.com" <vishal.l.verma@intel.com>,
        "ira.weiny@intel.com" <ira.weiny@intel.com>,
        "widawsk@kernel.org" <widawsk@kernel.org>,
        "dan.j.williams@intel.com" <dan.j.williams@intel.com>,
        "Jonathan.Cameron@huawei.com" <Jonathan.Cameron@huawei.com>,
        "linux-cxl@vger.kernel.org" <linux-cxl@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] cxl: Replace HDM decoder granularity magic numbers
Thread-Topic: [PATCH] cxl: Replace HDM decoder granularity magic numbers
Thread-Index: AQHYtklu6j3UH4AVtU69YQX5WSmWH627nsOAgAP2k4A=
Date:   Thu, 25 Aug 2022 05:48:22 +0000
Message-ID: <20220825054821.GC141173@bgt-140510-bm01>
In-Reply-To: <20220822171703.sfizbmn6retpxggi@offworld>
Accept-Language: en-US
Content-Language: en-US
Content-Type: text/plain; charset="us-ascii"
Content-ID: <3383A43F9A0F3A458DAAFF132F6005F5@ssi.samsung.com>
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
CMS-TYPE: 301P
References: <CGME20220822170552uscas1p1b1ee530bf38a14806010d65d1b593ab0@uscas1p1.samsung.com>
        <20220822170510.125082-1-a.manzanares@samsung.com>
        <20220822171703.sfizbmn6retpxggi@offworld>
Precedence: bulk

On Mon, Aug 22, 2022 at 10:17:03AM -0700, Davidlohr Bueso wrote:
> On Mon, 22 Aug 2022, Adam Manzanares wrote:
>=20
> > When reviewing the CFMWS parsing code that deals with the HDM decoders,
> > I noticed a couple of magic numbers. This commit replaces these magic n=
umbers
> > with constants defined by the CXL 2.0 specification.
>=20
> Please use 3.0 spec :)
>=20
> Actually the whole drivers/cxl/* could use updating the comments for 3.0.
>=20
> > Signed-off-by: Adam Manzanares <a.manzanares@samsung.com>
> > ---
> > drivers/cxl/cxl.h | 11 +++++++----
> > 1 file changed, 7 insertions(+), 4 deletions(-)
> >=20
> > diff --git a/drivers/cxl/cxl.h b/drivers/cxl/cxl.h
> > index f680450f0b16..ba3a66b5b9cd 100644
> > --- a/drivers/cxl/cxl.h
> > +++ b/drivers/cxl/cxl.h
> > @@ -61,6 +61,10 @@
> > #define CXL_HDM_DECODER0_SKIP_LOW(i) CXL_HDM_DECODER0_TL_LOW(i)
> > #define CXL_HDM_DECODER0_SKIP_HIGH(i) CXL_HDM_DECODER0_TL_HIGH(i)
> >=20
> > +/* HDM decoder control register constants CXL 2.0 8.2.5.12.7 */
>=20
> This now becomes 8.2.4.19.7.
>=20
> > +#define CXL_DECODER_MIN_GRANULARITY 256
> > +#define CXL_DECODER_MAX_GRANULARITY_ORDER 6
>=20
> Considering there is a single user, I don't think we need to add
> the CXL_DECODER_MAX_GRANULARITY_ORDER. And if a new one is added
> it would be better to keep symmetry (bytes) and just do
> CXL_DECODER_MAX_GRANULARITY, but that's probably the reason why
> it doesn't exist in the first place.

I wasn't so worried about the single user. I was hoping to give some additi=
onal
context for the constants used in the granularity calculations.=20

Given Dans suggestion for the name change are you more comfortable with the=
=20
proposed change. If not, feel free to suggest an alternative to eliminate t=
he
magic number.

>=20
> > +
> > static inline int cxl_hdm_decoder_count(u32 cap_hdr)
> > {
> > 	int val =3D FIELD_GET(CXL_HDM_DECODER_COUNT_MASK, cap_hdr);
> > @@ -71,9 +75,9 @@ static inline int cxl_hdm_decoder_count(u32 cap_hdr)
> > /* Encode defined in CXL 2.0 8.2.5.12.7 HDM Decoder Control Register */
> > static inline int cxl_to_granularity(u16 ig, unsigned int *val)
> > {
> > -	if (ig > 6)
> > +	if (ig > CXL_DECODER_MAX_GRANULARITY_ORDER)
> > 		return -EINVAL;
> > -	*val =3D 256 << ig;
> > +	*val =3D CXL_DECODER_MIN_GRANULARITY << ig;
> > 	return 0;
> > }
> >=20
> > @@ -96,7 +100,7 @@ static inline int cxl_to_ways(u8 eniw, unsigned int =
*val)
> >=20
> > static inline int granularity_to_cxl(int g, u16 *ig)
> > {
> > -	if (g > SZ_16K || g < 256 || !is_power_of_2(g))
> > +	if (g > SZ_16K || g < CXL_DECODER_MIN_GRANULARITY || !is_power_of_2(g=
))
>=20
> Looks good.
>=20
> Thanks,
> Davidlohr=