Re: [PATCH] ntfs: wait for bitmap folio writeback before modifying it

[liubaolin <liubaolin12138@163.com>]

在 2026/10/29 14:17, Shengzhuo Wei 写道:
> On 2026-10-29 10:07, liubaolin wrote:
>   
>> Hi Shengzhuo,
>>     Could this deadlock during MFT bitmap
>> expansion？ntfs_mft_bitmap_extend_allocation_nolock() can recurse into MFT
>> record allocation while holding mftbmp_ni->runlist.lock.
>>
>>    With L being that lock, could we get the following interleaving?
>> A and B are $MFT::$BITMAP folios; B is outside the cached iomap.
>> Time flows downwards.
>>
>>    Writeback                        Bitmap expansion
>>    ---------                        ----------------
>>    lock A
>>    set PG_writeback(A)
>>    map A (take/release L)
>>    add A to wpc->wb_ctx
>>    unlock A
>>    [BIO still unsubmitted]
>>                                     take L
>>                                     extend bitmap mapping pairs
>>                                     need another extent record
>>                                     ntfs_mft_record_alloc()
>>                                       mft_bitmap_alloc_free_rec()
>>                                         lock A
>>                                         folio_wait_writeback(A)
>>                                     [wait for A's I/O, holding L]
>>
>>    lock B
>>    set PG_writeback(B)
>>    ntfs_writeback_range(B)
>>      __ntfs_write_iomap_begin()
>>        down_write(L)
>>        [blocked on L]
>>
>>    Cannot reach iomap_add_to_ioend()
>>    to submit A's pending BIO.        Cannot release L until
>>                                     A's writeback completes.
>>
>>     ntfs_writeback_range() obtains the next mapping before
>> iomap_add_to_ioend() can submit the pending BIO.
>>     Would we need to submit that BIO before taking the mapping locks?
>>
>> Thanks,
>> Baolin.
>>
> 
> Hi Baolin,
> 
> Thanks for the analysis. I checked the paths and agree with you.
> 
> Following your suggestion, I prepared a change to submit the pending
> ioend before obtaining a new mapping in ntfs_writeback_range(), clearing
> wpc->wb_ctx even on error to avoid resubmission. I plan to make this a
> prerequisite for v2, retaining all six waits.
> 
> One concern is that submitting at mapping boundaries may reduce BIO
> aggregation, including for ordinary file writeback. The change has
> passed object builds, but has not been runtime- or performance-tested.
> 
> Does this approach look reasonable to you?
> 
> Thanks,
> Shengzhuo

Hi Shengzhuo,

    That sounds reasonable to me.
    Clearing wpc->wb_ctx on error also makes sense, since the submit 
helper calls bio_endio() on that path.
    You could run some tests to check the performance impact.A 
before/after comparison of throughput and BIO sizes would help, 
especially where adjacent mappings are physically contiguous. Testing 
bitmap expansion alongside writeback would also help check the locking 
change.

Thanks,
Baolin