Re: [PATCH v3 1/2] Btrfs: rework ulist with list+rb_tree

[Josef Bacik <jbacik@fb.com>]

On 01/28/2014 11:03 AM, Wang Shilong wrote:
> Hello Josef,
>
>> On 01/27/2014 04:59 AM, Wang Shilong wrote:
>>> We are really suffering from now ulist's implementation, some developers
>>> gave their try, and i just gave some of my ideas for things:
>>>
>>>   1. use list+rb_tree instead of arrary+rb_tree
>>>
>>>   2. add cur_list to iterator rather than ulist structure.
>>>
>>>   3. add seqnum into every node when they are added, this is
>>>   used to do selfcheck when iterating node.
>>>
>>> I noticed Zach Brown's comments before, long term is to kick off
>>> ulist implementation, however, for now, we need at least avoid
>>> arrary from ulist.
>>>
>>> Cc: Liu Bo <bo.li.liu@oracle.com>
>>> Cc: Josef Bacik <jbacik@fb.com>
>>> Cc: Zach Brown <zab@redhat.com>
>>> Signed-off-by: Wang Shilong <wangsl.fnst@cn.fujitsu.com>
>>> ---
>>> v2->v3:
>>> 	only do selfchecks with CONFIG_BTRFS_DEBUG enabled(Thanks to Josef!)
>>> 	update ulist's comments since they are out of date.
>>> v1->v2:
>>> 	add RFC title since this patch needs more reviews and comments.
>>> 	fix a used after free bug in ulist_fini().
>> I like the patch but it doesn't build since things like qgroups rely on ulist->nnodes.  You need to fix that in your patch and make sure this stuff compiles.  Thanks,
> Sorry about if it did not compile.
>
> but I really compiled and tested it  in my box, did you apply your qgroup patches?
> This patch is based on btrfs-next without your previous qgroup patches.
>
> Anyway i will double check it…..
>
Yeah I thought I was doing something wrong but I'm definitely on my 
master branch which doesn't have my qgroup patches in it.  If it's still 
working for you now just wait a bit for me to push out this next update 
and rebase onto it and resend.  Thanks,

Josef