On 2021/6/24 上午6:37, David Sterba wrote:

On Wed, Jun 23, 2021 at 01:55:21PM +0800, Qu Wenruo wrote:

quoted

Qu Wenruo (8):
   btrfs: don't pass compressed pages to
     btrfs_writepage_endio_finish_ordered()
   btrfs: make btrfs_subpage_end_and_test_writer() to handle pages not
     locked by btrfs_page_start_writer_lock()
   btrfs: make compress_file_range() to be subpage compatible
   btrfs: make btrfs_submit_compressed_write() to be subpage compatible
   btrfs: use async_chunk::async_cow to replace the confusing pending
     pointer
   btrfs: make end_compressed_bio_writeback() to be subpage compatble
   btrfs: make extent_write_locked_range() to be subpage compatible
   btrfs: only allow subpage compression if the range fully covers the
     first page

Some of the patches seem independent and potentially could be taken into
5.14 now but I guess the whole subpage could go in one big batch to 5.15
so it won't help you much anyway. The significant change is the last
patch and so far I think it's acceptable.

Nonononono, please don't merge any patche from the series.

As the patches are not fully stress tested, and I have already seen some 
ASSERT()s triggered, thus there must be something wrong in the 
preparation part.

Sorry I should have added some warning like "WIP, don't merge any patch 
in the series".

This is mostly for the discussion on the last patch, not really to get 
any patch merged.

Thanks,
Qu