On 2020/6/8 5:17, Jaegeuk Kim wrote:
> On 06/05, Chao Yu wrote:
>>     71.07%     0.01%  kworker/u256:1+  [kernel.kallsyms]  [k] wb_writeback
>>             |
>>              --71.06%--wb_writeback
>>                        |
>>                        |--68.96%--__writeback_inodes_wb
>>                        |          |
>>                        |           --68.95%--writeback_sb_inodes
>>                        |                     |
>>                        |                     
>> |--65.08%--__writeback_single_inode
>>                        |                     |          |
>>                        |                     |           
>> --64.35%--do_writepages
>>                        |                     |                     |
>>                        |                     |                     
>> |--59.83%--f2fs_write_node_pages
>>                        |                     |                     |         
>>  |
>>                        |                     |                     |         
>>   --59.74%--f2fs_sync_node_pages
>>                        |                     |                     |         
>>             |
>>                        |                     |                     |         
>>             |--27.91%--pagevec_lookup_range_tag
>>                        |                     |                     |         
>>             |          |
>>                        |                     |                     |         
>>             |           --27.90%--find_get_pages_range_tag
>>
>> If filesystem was injected w/ checkpoint errror, before umount, kworker
>> will always hold one core in order to writeback a large number of node
>> pages, that looks not reasonable, to avoid that, we can allow data/node
>> write in such case, since we can force all data/node writes with OPU mode,
>> and clear recovery flag on node, and checkpoint is not allowed as well,
>> so we don't need to worry about writeback's effect on data/node in
>> previous checkpoint, then with this way, it can decrease memory footprint
>> cost by node/data pages and avoid looping into data/node writeback
>> process.
> 
> This patch breaks fault injection test with filesystem corruption. Please 
> check.

My bad, will check soon.

Thanks,

> 
>>
>> Signed-off-by: Chao Yu <yuch...@huawei.com>
>> ---
>>  fs/f2fs/data.c | 19 ++++++++++++-------
>>  fs/f2fs/node.c |  7 +++++--
>>  2 files changed, 17 insertions(+), 9 deletions(-)
>>
>> diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
>> index 9d40db50cd65..2b3c846181bb 100644
>> --- a/fs/f2fs/data.c
>> +++ b/fs/f2fs/data.c
>> @@ -2519,6 +2519,8 @@ bool f2fs_should_update_outplace(struct inode *inode, 
>> struct f2fs_io_info *fio)
>>  {
>>      struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
>>  
>> +    if (unlikely(f2fs_cp_error(sbi)))
>> +            return true;
>>      if (f2fs_lfs_mode(sbi))
>>              return true;
>>      if (S_ISDIR(inode->i_mode))
>> @@ -2702,13 +2704,16 @@ int f2fs_write_single_data_page(struct page *page, 
>> int *submitted,
>>      /* we should bypass data pages to proceed the kworkder jobs */
>>      if (unlikely(f2fs_cp_error(sbi))) {
>>              mapping_set_error(page->mapping, -EIO);
>> -            /*
>> -             * don't drop any dirty dentry pages for keeping lastest
>> -             * directory structure.
>> -             */
>> -            if (S_ISDIR(inode->i_mode))
>> -                    goto redirty_out;
>> -            goto out;
>> +
>> +            if (has_not_enough_free_secs(sbi, 0, 0)) {
>> +                    /*
>> +                     * don't drop any dirty dentry pages for keeping lastest
>> +                     * directory structure.
>> +                     */
>> +                    if (S_ISDIR(inode->i_mode))
>> +                            goto redirty_out;
>> +                    goto out;
>> +            }
>>      }
>>  
>>      if (unlikely(is_sbi_flag_set(sbi, SBI_POR_DOING)))
>> diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
>> index 03e24df1c84f..372c04efad38 100644
>> --- a/fs/f2fs/node.c
>> +++ b/fs/f2fs/node.c
>> @@ -1527,7 +1527,10 @@ static int __write_node_page(struct page *page, bool 
>> atomic, bool *submitted,
>>                      unlock_page(page);
>>                      return 0;
>>              }
>> -            goto redirty_out;
>> +            if (has_not_enough_free_secs(sbi, 0, 0))
>> +                    goto redirty_out;
>> +            set_fsync_mark(page, 0);
>> +            set_dentry_mark(page, 0);
>>      }
>>  
>>      if (unlikely(is_sbi_flag_set(sbi, SBI_POR_DOING)))
>> @@ -1568,7 +1571,7 @@ static int __write_node_page(struct page *page, bool 
>> atomic, bool *submitted,
>>              goto redirty_out;
>>      }
>>  
>> -    if (atomic && !test_opt(sbi, NOBARRIER))
>> +    if (atomic && !test_opt(sbi, NOBARRIER) && !f2fs_cp_error(sbi))
>>              fio.op_flags |= REQ_PREFLUSH | REQ_FUA;
>>  
>>      /* should add to global list before clearing PAGECACHE status */
>> -- 
>> 2.18.0.rc1
> .
> 

Reply via email to