Re: [PATCH v3 06/16] x86/virt/tdx: Improve PAMT refcounters allocation for sparse memory

[Huang, Kai]

On Wed, 2025-10-01 at 19:00 +0000, Edgecombe, Rick P wrote:
> I guess we are really doing two separate calculations. First calculate the range
> of refcounts we need, and then calculate the range of vmalloc space we need. How
> about this, is it clearer to you? It is very specific about what/why we actually
> are doing, but at the expense of minimizing operations. In this slow path, I
> think clarity is the priority.
> 
> static int alloc_pamt_refcount(unsigned long start_pfn, unsigned long end_pfn)
> {
> 	unsigned long refcount_first, refcount_last;
> 	unsigned long mapping_start, mapping_end;
> 
> 	/*
> 	 * 'start_pfn' is inclusive and 'end_pfn' is exclusive. Find the
> 	 * range of refcounts the pfn range will need.
> 	 */
> 	refcount_first = (unsigned long)tdx_find_pamt_refcount(start_pfn);
> 	refcount_last   = (unsigned long)tdx_find_pamt_refcount(end_pfn - 1);
> 
> 	/*
> 	 * Calculate the page aligned range that includes the refcounts. The
> 	 * teardown logic needs to handle potentially overlapping refcount
> 	 * mappings resulting from the alignments.
> 	 */
> 	mapping_start = round_down(refcount_first, PAGE_SIZE);
> 	mapping_end   = round_up(refcount_last + sizeof(*pamt_refcounts),
> PAGE_SIZE);
> 
> 
> 	return apply_to_page_range(&init_mm, mapping_start, mapping_end -
> mapping_start,
> 				   pamt_refcount_populate, NULL);
> }

Yeah it's better and clearer.  LGTM.  Thanks.