On 10/11/2015 11:12, Kevin Wolf wrote:
> > For full mirroring, this strategy will probably make the first
> > incremental iteration more expensive.
>
> You mean because we issue smaller, interleaved write and write_zeroes
> requests now instead of only large writes? That's probably right, but
> g
Am 10.11.2015 um 10:01 hat Paolo Bonzini geschrieben:
>
>
> On 10/11/2015 07:14, Fam Zheng wrote:
> > On Mon, 11/09 17:29, Kevin Wolf wrote:
> >> Am 09.11.2015 um 17:18 hat Paolo Bonzini geschrieben:
> >>>
> >>>
> >>> On 09/11/2015 17:04, Kevin Wolf wrote:
> Am 06.11.2015 um 11:22 hat Fam Zh
On 10/11/2015 07:14, Fam Zheng wrote:
> On Mon, 11/09 17:29, Kevin Wolf wrote:
>> Am 09.11.2015 um 17:18 hat Paolo Bonzini geschrieben:
>>>
>>>
>>> On 09/11/2015 17:04, Kevin Wolf wrote:
Am 06.11.2015 um 11:22 hat Fam Zheng geschrieben:
> The "pnum < nb_sectors" condition in deciding whe
On Mon, 11/09 17:29, Kevin Wolf wrote:
> Am 09.11.2015 um 17:18 hat Paolo Bonzini geschrieben:
> >
> >
> > On 09/11/2015 17:04, Kevin Wolf wrote:
> > > Am 06.11.2015 um 11:22 hat Fam Zheng geschrieben:
> > >> The "pnum < nb_sectors" condition in deciding whether to actually copy
> > >> data is un
Am 09.11.2015 um 17:18 hat Paolo Bonzini geschrieben:
>
>
> On 09/11/2015 17:04, Kevin Wolf wrote:
> > Am 06.11.2015 um 11:22 hat Fam Zheng geschrieben:
> >> The "pnum < nb_sectors" condition in deciding whether to actually copy
> >> data is unnecessarily strict, and the qiov initialization is
>
On 09/11/2015 17:04, Kevin Wolf wrote:
> Am 06.11.2015 um 11:22 hat Fam Zheng geschrieben:
>> The "pnum < nb_sectors" condition in deciding whether to actually copy
>> data is unnecessarily strict, and the qiov initialization is
>> unnecessarily too, for both bdrv_aio_write_zeroes and bdrv_aio_di
Am 06.11.2015 um 11:22 hat Fam Zheng geschrieben:
> The "pnum < nb_sectors" condition in deciding whether to actually copy
> data is unnecessarily strict, and the qiov initialization is
> unnecessarily too, for both bdrv_aio_write_zeroes and bdrv_aio_discard
> branches.
>
> Reorganize mirror_itera
On Fri, 11/06 19:36, Max Reitz wrote:
> > +next_sector = sector_num;
> > +next_chunk = sector_num / sectors_per_chunk;
>
> @next_sector and @next_chunk set here...
>
> > hbitmap_next_sector = s->sector_num;
> > -sector_num = s->sector_num;
> > -sectors_per_chunk = s->granular
On 06.11.2015 11:22, Fam Zheng wrote:
> The "pnum < nb_sectors" condition in deciding whether to actually copy
> data is unnecessarily strict, and the qiov initialization is
> unnecessarily too, for both bdrv_aio_write_zeroes and bdrv_aio_discard
> branches.
>
> Reorganize mirror_iteration flow so
The "pnum < nb_sectors" condition in deciding whether to actually copy
data is unnecessarily strict, and the qiov initialization is
unnecessarily too, for both bdrv_aio_write_zeroes and bdrv_aio_discard
branches.
Reorganize mirror_iteration flow so that we:
1) Find the contiguous zero/discard
10 matches
Mail list logo