Re: [PATCH v3 4/6] x86/hygon: Add Family 0x18 DF node enumeration

[Lin Wang]

Hi Yazen,

Thanks for checking！

Yes, "CDD" is intentional and is the Hygon term for "Core and DDR Die". 
It is not a mix-up with the AMD "CCD" terminology.

On the supported Hygon Family 0x18 MCM systems, a CDD contains CPU core
complexes together with UMCs. The IOD provides the I/O and fabric
functions and does not contain UMCs. Therefore, this organization is
different from AMD systems where the memory controllers reside on the
I/O die.

The DFID-based classification reflects this Hygon-specific hardware
organization: DFIDs below 4 identify IODs, while CDDs start at DFID 4.
The latter are consequently the nodes used for CPU topology mapping and
per-node UMC access.

I will spell out CDD as "Core and DDR Die" at its first use and note
that it is Hygon-specific terminology to avoid confusion with AMD CCDs.

Thx & Br,

Lin

On 7/20/2026 4:36 AM, Yazen Ghannam wrote:
> > +/* DF function indices supported by hygon_node_get_func(). */
> > +#define HYGON_DF_F3	3	/* misc */
> > +#define HYGON_DF_F4	4	/* link */
> > +
> > +/* Hygon compute dies (CDD) start at DFID 4; IO dies occupy DFIDs 0-3. */
> I noticed the regular to of "CDD" to mean compute dies. Is this the
> official use for Hygon systems?
>
> It is very similar to "CCD" that is used for AMD systems. So I just
> wanted to make sure it's not a mixup.
>
> > +#define HYGON_CDD_DFID_BASE	4
> > +
> > +/**
> > + * enum hygon_node_type - DF node type, derived from DFID
> > + * @HYGON_NODE_IOD: I/O die (DFID < HYGON_CDD_DFID_BASE); no UMC.
> > + * @HYGON_NODE_CDD: compute die (DFID >= HYGON_CDD_DFID_BASE); hosts
> > + *                  CPU cores and UMC controllers.
> This is the opposite of AMD systems, so is this a Hygon-specific
> implementatoin detail?
>
> On AMD systems, the compute die holds the CPU cores and cache while the
> I/O die holds the 'uncore' logical which includes the memory
> controllers.