Re: [PATCH v4 5/6] mm: cma: add cma_alloc_frozen{_compound}()

[Kefeng Wang <wangkefeng.wang@huawei.com>]

On 2025/10/24 9:12, Andrew Morton wrote:
> On Thu, 23 Oct 2025 19:59:39 +0800 Kefeng Wang <wangkefeng.wang@huawei.com> wrote:
> 
>> --- a/mm/internal.h
>> +++ b/mm/internal.h
>> @@ -515,13 +515,6 @@ static inline void set_page_refcounted(struct page *page)
>>   
>>   static inline void set_pages_refcounted(struct page *page, unsigned long nr_pages)
>>   {
>> -	unsigned long pfn = page_to_pfn(page);
>> -
>> -	if (PageHead(page)) {
>> -		set_page_refcounted(page);
>> -		return;
>> -	}
>> -
>>   	for (; nr_pages--; pfn++)
>>   		set_page_refcounted(pfn_to_page(pfn));
>>   }
> 
> pfn is undefined here.
> 
> I could fix, but what you sent clearly wasn't what you tested, so I'll
> drop this version of this series.
> 
> 

I do make some test with or without "hugetlb_cma= cmdline", which will
allocate gigantic from hugetlb_cma_reserve or by alloc_contig_range but
make a mistake when rebase this, sorry about that.