On 9/19/25 02:42, Chris Mason wrote:
> On Mon, 8 Sep 2025 10:04:38 +1000 Balbir Singh <[email protected]> wrote:
>
>> Extend migrate_vma_collect_pmd() to handle partially mapped large
>> folios that require splitting before migration can proceed.
>>
>> During PTE walk in the collection phase, if a large folio is only
>> partially mapped in the migration range, it must be split to ensure
>> the folio is correctly migrated.
>>
>> Signed-off-by: Balbir Singh <[email protected]>
>> ---
>> mm/migrate_device.c | 94 +++++++++++++++++++++++++++++++++++++++++++++
>> 1 file changed, 94 insertions(+)
>>
>> diff --git a/mm/migrate_device.c b/mm/migrate_device.c
>> index abd9f6850db6..f45ef182287d 100644
>> --- a/mm/migrate_device.c
>> +++ b/mm/migrate_device.c
>> @@ -54,6 +54,53 @@ static int migrate_vma_collect_hole(unsigned long start,
>> return 0;
>> }
>>
>> +/**
>> + * migrate_vma_split_folio() - Helper function to split a THP folio
>> + * @folio: the folio to split
>> + * @fault_page: struct page associated with the fault if any
>> + *
>> + * Returns 0 on success
>> + */
>> +static int migrate_vma_split_folio(struct folio *folio,
>> + struct page *fault_page)
>> +{
>> + int ret;
>> + struct folio *fault_folio = fault_page ? page_folio(fault_page) : NULL;
>> + struct folio *new_fault_folio = NULL;
>> +
>> + if (folio != fault_folio) {
>> + folio_get(folio);
>> + folio_lock(folio);
>> + }
>
> Can fault_folio ever be non-null and different from folio? Apologies for
> not knowing the lock ordering rules but this jumps out.
>
Yes, migration can occur in fault context or be driver driven
>> +
>> + ret = split_folio(folio);
>> + if (ret) {
>> + if (folio != fault_folio) {
>> + folio_unlock(folio);
>> + folio_put(folio);
>> + }
>> + return ret;
>> + }
>> +
>> + new_fault_folio = fault_page ? page_folio(fault_page) : NULL;
>> +
>> + /*
>> + * Ensure the lock is held on the correct
>> + * folio after the split
>> + */
>> + if (!new_fault_folio) {
>> + folio_unlock(folio);
>> + folio_put(folio);
>> + } else if (folio != new_fault_folio) {
>> + folio_get(new_fault_folio);
>> + folio_lock(new_fault_folio);
>> + folio_unlock(folio);
>> + folio_put(folio);
>> + }
>
> Same question here, do we need trylocks?
>
Since we had the folio lock before, the assumption is that we can
still grab the lock after split and it's OK to wait, since this
is not a hot-path.
Balbir Singh
