[PATCH 3/3] arm: extend pfn_valid to take into accound freed memory map alignment

[Kefeng Wang]

On 2021/5/18 23:52, Mike Rapoport wrote:
> On Tue, May 18, 2021 at 08:49:43PM +0800, Kefeng Wang wrote:
>>
>>
>> On 2021/5/18 17:06, Mike Rapoport wrote:
>>> From: Mike Rapoport <rppt at linux.ibm.com>
>>>
>>> When unused memory map is freed the preserved part of the memory map is
>>> extended to match pageblock boundaries because lots of core mm
>>> functionality relies on homogeneity of the memory map within pageblock
>>> boundaries.
>>>
>>> Since pfn_valid() is used to check whether there is a valid memory map
>>> entry for a PFN, make it return true also for PFNs that have memory map
>>> entries even if there is no actual memory populated there.
>>>
>>> Signed-off-by: Mike Rapoport <rppt at linux.ibm.com>
>>> ---
>>>    arch/arm/mm/init.c | 15 ++++++++++++++-
>>>    1 file changed, 14 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/arch/arm/mm/init.c b/arch/arm/mm/init.c
>>> index 9d4744a632c6..bb678c0ba143 100644
>>> --- a/arch/arm/mm/init.c
>>> +++ b/arch/arm/mm/init.c
>>> @@ -125,11 +125,24 @@ static void __init zone_sizes_init(unsigned long min, unsigned long max_low,
>>>    int pfn_valid(unsigned long pfn)
>>>    {
>>>    	phys_addr_t addr = __pfn_to_phys(pfn);
>>> +	unsigned long pageblock_size = PAGE_SIZE * pageblock_nr_pages;
>>>    	if (__phys_to_pfn(addr) != pfn)
>>>    		return 0;
>>> -	return memblock_is_map_memory(addr);
>>> +	if (memblock_is_map_memory(addr))
>>> +		return 1;
>>> +
>>> +	/*
>>> +	 * If address less than pageblock_size bytes away from a present
>>> +	 * memory chunk there still will be a memory map entry for it
>>> +	 * because we round freed memory map to the pageblock boundaries
>>> +	 */
>>> +	if (memblock_is_map_memory(ALIGN(addr + 1, pageblock_size)) ||
>>> +	    memblock_is_map_memory(ALIGN_DOWN(addr, pageblock_size)))
>>> +		return 1;
>>
>> Hi Mike, with patch3, the system won't boot.
> 
> Hmm, apparently I've miscalculated the ranges...
> 
> Can you please check with the below patch on top of this series:

Yes, it works,

On node 0 totalpages: 311551
   Normal zone: 1230 pages used for memmap
   Normal zone: 0 pages reserved
   Normal zone: 157440 pages, LIFO batch:31
   Normal zone: 17152 pages in unavailable ranges
   HighMem zone: 154111 pages, LIFO batch:31
   HighMem zone: 513 pages in unavailable ranges

and the oom testcase could pass.

Tested-by: Kefeng Wang <wangkefeng.wang at huawei.com>


There is memblock_is_region_reserved(check if a region intersects 
reserved memory), it also checks the size, should we add a similar func?

> 
> diff --git a/arch/arm/mm/init.c b/arch/arm/mm/init.c
> index bb678c0ba143..2fafbbc8e73b 100644
> --- a/arch/arm/mm/init.c
> +++ b/arch/arm/mm/init.c
> @@ -138,8 +138,9 @@ int pfn_valid(unsigned long pfn)
>   	 * memory chunk there still will be a memory map entry for it
>   	 * because we round freed memory map to the pageblock boundaries
>   	 */
> -	if (memblock_is_map_memory(ALIGN(addr + 1, pageblock_size)) ||
> -	    memblock_is_map_memory(ALIGN_DOWN(addr, pageblock_size)))
> +	if (memblock_overlaps_region(&memblock.memory,
> +				     ALIGN_DOWN(addr, pageblock_size),
> +				     pageblock_size);
>   		return 1;
>   
>   	return 0;
>