Re: [RESEND PATCH 1/2] btrfs: Read stripe len directly in btrfs_rmap_block

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 




On 10.04.20 г. 22:29 ч., Josef Bacik wrote:
> On 4/3/20 9:40 AM, Nikolay Borisov wrote:
>> extent_map::orig_block_len contains the size of a physical stripe when
>> it's used to describe block groups (calculated in read_one_chunk via
>> calc_stripe_length or calculated in decide_stripe_size and then
>> assigned to
>> extent_map::orig_block_len in create_chunk). Exploit this fact to get the
>> size directly rather than opencoding the calculations. No functional
>> changes.
>>
>> Signed-off-by: Nikolay Borisov <nborisov@xxxxxxxx>
>> ---
>>
>> Hello David,
>>
>> You had some reservations for this patch but now I've expanded the
>> changelog to
>> explain why it's safe to do so.
>>
>>
>>   fs/btrfs/block-group.c | 13 +++----------
>>   1 file changed, 3 insertions(+), 10 deletions(-)
>>
>> diff --git a/fs/btrfs/block-group.c b/fs/btrfs/block-group.c
>> index 786849fcc319..d0dbaa470b88 100644
>> --- a/fs/btrfs/block-group.c
>> +++ b/fs/btrfs/block-group.c
>> @@ -1628,19 +1628,12 @@ int btrfs_rmap_block(struct btrfs_fs_info
>> *fs_info, u64 chunk_start,
>>           return -EIO;
>>
>>       map = em->map_lookup;
>> -    data_stripe_length = em->len;
>> +    data_stripe_length = em->orig_block_len;
>>       io_stripe_size = map->stripe_len;
>>
>> -    if (map->type & BTRFS_BLOCK_GROUP_RAID10)
>> -        data_stripe_length = div_u64(data_stripe_length,
>> -                         map->num_stripes / map->sub_stripes);
>> -    else if (map->type & BTRFS_BLOCK_GROUP_RAID0)
>> -        data_stripe_length = div_u64(data_stripe_length,
>> map->num_stripes);
>> -    else if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) {
>> -        data_stripe_length = div_u64(data_stripe_length,
>> -                         nr_data_stripes(map));
>> +    /* For raid5/6 adjust to a full IO stripe length */
>> +    if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK)
>>           io_stripe_size = map->stripe_len * nr_data_stripes(map);
>> -    }
>>
> 
> So now data_stripe_length is different in the RAID1 case and the RAID1C*
> case, right?  Is that ok?  I *think* it is, but I'm a little drunk and
> can't really reason it out well.  Thanks,

The stripe for RAID1C* is calculated in decide_stripe_size_regular based
on the data in btrfs_raid_array. This patch only gets whatever was
calculated at chunk allocation time.

> 
> Josef



[Index of Archives]     [Linux Filesystem Development]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux