Re: [PATCH v3 07/10] cxl/pci: Split cxl_pci_setup_regs()

[Dan Williams <dan.j.williams@intel.com>]

On Wed, Oct 13, 2021 at 5:12 PM Ben Widawsky <ben.widawsky@intel.com> wrote:
>
> On 21-10-13 15:49:30, Dan Williams wrote:
> > On Wed, Oct 13, 2021 at 3:45 PM Ben Widawsky <ben.widawsky@intel.com> wrote:
> > >
> > > On 21-10-09 09:44:34, Dan Williams wrote:
> > > > From: Ben Widawsky <ben.widawsky@intel.com>
> > > >
> > > > In preparation for moving parts of register mapping to cxl_core, split
> > > > cxl_pci_setup_regs() into a helper that finds register blocks,
> > > > (cxl_find_regblock()), and a generic wrapper that probes the precise
> > > > register sets within a block (cxl_setup_regs()).
> > > >
> > > > Move the actual mapping (cxl_map_regs()) of the only register-set that
> > > > cxl_pci cares about (memory device registers) up a level from the former
> > > > cxl_pci_setup_regs() into cxl_pci_probe().
> > > >
> > > > With this change the unused component registers are no longer mapped,
> > > > but the helpers are primed to move into the core.
> > > >
> > > > Signed-off-by: Ben Widawsky <ben.widawsky@intel.com>
> > > > [djbw: rebase on the cxl_register_map refactor]
> > > > [djbw: drop cxl_map_regs() for component registers]
> > > > Signed-off-by: Dan Williams <dan.j.williams@intel.com>
> > >
> > > [snip]
> > >
> > > Did you mean to also drop the component register handling in cxl_probe_regs()
> > > and cxl_map_regs()?
> >
> > No, because that has a soon to be added user, right?
>
> In the current codebase, the port driver gets the offset from cxl_core, not
> through the pci driver. I know you wanted this to be passed from cxl_pci (and
> indeed it was before). Currently however, the functionality is subsumed by
> cxl_find_regblock and is used by cxl_pci (for device registers), cxl_acpi (to
> get the CHBCR) and cxl_core (to get the component register block for switches).
>
> I have no user in cxl_pci for the component registers, and as we discussed, we
> have no good way to share them across modules.

Are you saying that cxl_probe_regs() will not move to the core in your
upcoming series? I was expecting that cxl_find_regblock() and
cxl_probe_regs() go hand in hand.

>
> We can ignore this for now though and discuss it on the list when I post. If
> there is a better way to handle this, I'm open to it.

It's hard to have discussions about API uses without the patches, but
I'm ok to leave further cxl_probe_regs() refactoring to your series.