Re: [PATCH] PCI: brcmstb: Sort enums, pcie_offsets[], pcie_cfg_data, .compatible strings

[florian.fainelli@broadcom.com]

On 9/2/24 13:54, 'Bjorn Helgaas' via BCM-KERNEL-FEEDBACK-LIST,PDL wrote:
> From: Bjorn Helgaas <bhelgaas@google.com>
> 
> Sort enum pcie_soc_base values.
> 
> Rename pcie_offsets_bmips_7425[] to pcie_offsets_bcm7425[] to match BCM7425
> pcie_soc_base enum, bcm7425_cfg, and "brcm,bcm7425-pcie" .compatible
> string.
> 
> Rename pcie_offset_bcm7278[] to pcie_offsets_bcm7278[] to match other
> "pcie_offsets" names.
> 
> Rename pcie_offset_bcm7712[] to pcie_offsets_bcm7712[] to match other
> "pcie_offsets" names.
> 
> Sort pcie_offsets_*[] by SoC name, move them all together, indent values
> for easy reading.
> 
> Sort pcie_cfg_data structs by SoC name.
> 
> Sort .compatible strings by SoC name.
> 
> No functional change intended.
> 
> Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>

Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>

There was some appeal in preserving the enumeration in the order in 
which the chips have been produced, like 7425, 7435 etc.. but I see we 
probably got some parts of the timeline wrong at some point anyway.

Should we push a bit further an accept a little bit of duplication and 
simply embed the array of register offsets directly into the 
pcie_cfg_data structure, rather than a pointer to to an array?
-- 
Florian