Re: [PATCH v2 1/1] fs/writeback: skip AS_NO_DATA_INTEGRITY mappings in wait_sb_inodes()

["David Hildenbrand (Red Hat)" <david@kernel.org>]

On 1/19/26 01:40, Andrew Morton wrote:
> On Tue, 6 Jan 2026 18:54:47 +0100 "David Hildenbrand (Red Hat)" <david@kernel.org> wrote:
> 
>> On 1/6/26 17:05, Miklos Szeredi wrote:
>>> On Tue, 6 Jan 2026 at 16:41, David Hildenbrand (Red Hat)
>>> <david@kernel.org> wrote:
>>>
>>>> I assume the usual suspects, including mm/memory-failure.c.
>>>>
>>>> memory_failure() not only contains a folio_wait_writeback() but also a
>>>> folio_lock(), so twice the fun :)
>>>
>>> As long as it's run from a workqueue it shouldn't affect the rest of
>>> the system, right?  The wq thread will consume a nontrivial amount of
>>> resources, I suppose, so it would be better to implement those waits
>>> asynchronously.
>>
>> Good question. I know that memory_failure() can be triggered out of
>> various context, but I never traced it back to its origin.
> 
> I'm seeing unhappy okays from David and Jan, so I'll upstream this
> patch later in the week, unless someone stops me.

Just to clarify, I'm happy with this patch as is. The semantics make 
perfect sense.

-- 
Cheers

David