Re: [PATCH v5 1/1] cxl/core/regs: Refactor out functions to count regblocks of given type

[Dave Jiang <dave.jiang@intel.com>]

On 1/22/25 2:14 AM, Dan Williams wrote:
> Some minor comments that Dave can fixup:
> 
> Huaisheng Ye wrote:
>> In commit d717d7f3df18494baafd9595fb4bcb9c380d7389, cxl_count_regblock was
> 
> When referencing a commit the typical form is 
> 
> d717d7f3df18 ("cxl: Add functions to get an instance of / count regblocks of a given type")
> 
> ...however in this case the commit history does not matter. Typically
> commit history is useful for fixes to identify what kernels are
> suspectible to the problem. In this case it is just a cleanup.
> 
> Also, for what this patch is I think the commit message can afford to be
> more concise and just say:
> 
> ---
> cxl_find_regblock_instance() counts the number of instances of a register
> block as a side effect of searching through all available register blocks.
> cxl_count_regblock() throws away that work and recounts all the register
> blocks by asking cxl_find_regblock_instance() to redo work it has
> already done until it finally returns an error, that is needlessly wasteful.
> 
> Let cxl_count_regblock() leverage the counting that
> cxl_find_regblock_instance() already does by passing in a sentinel value
> (CXL_INSTANCES_COUNT) that triggers the count to be returned.
> ---
> 
> Otherwise, patch looks good.
> 
> Reviewed-by: Dan Williams <dan.j.williams@intel.com>

Applied with the suggested commit log replacement