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

Re: [lvm-devel] [PATCH] Refactor lib/ code to allow deferred PV labelling



On 04/27/2011 08:24 PM +0100, Alasdair G Kergon wrote:
>> @@ -51,6 +51,9 @@ struct physical_volume {
> 
>> +	/* NB. Only useful/used when status & UNLABELLED_PV! */
>> +	int64_t label_sector;
> 
>> @@ -59,6 +65,13 @@ struct volume_group {
>> +	struct dm_list pvs_to_create;
> 
> These are really just temporary places to store things - not used as an intrinsic
> part of the structs themselves.  How difficult is it to handle them
> separately outside the structs?

Well, normally, we store that in PV-based format instance context
(the 'struct text_fid_pv_context' that is stored in 'private' field in
'struct format_instance). But if that PV is part of the VG, we use a
common VG-based format instance - so we can't store that PV info there
separately for each PV...

So how about using 'struct pvcreate_params' we attach to 'struct pv_to_create'
in add_pv_to_vg fn:

  if (pv->status & UNLABELLED_PV) {
                if (!(pvc = dm_pool_zalloc(mem, sizeof(*pvc)))) {
                        log_error("pv_to_create allocation for '%s' failed", pv_name);
                        return 0;
                }    
                pvc->pv = pv;
                pvc->pp = pp;
                ...
  ...


Can we use that to store that label sector info? Would that work?

If we don't find a place for that label sector other than the struct PV
itself as proposed in this patch, we can also remove the 'struct text_fid_pv_context'.
The text_fid_pv_context is here to store the label_sector info only. At least for now.
I wanted to use the PV struct for that as well the other day, like you do here... ;)
It's useless to store that in two places at the same time.

Peter

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