On Wed, 2015-12-23 at 12:28 -0800, Dan Williams wrote:
> On Sun, Dec 20, 2015 at 1:18 AM, <vishal@xxxxxxxxxx> wrote:
> > From: Vishal Verma <vishal.l.verma@xxxxxxxxx>
> >
> > Enable the gendisk badblocks feature for pmem namespaces.
> > If the pmem namespace being created has any known poison associated
> > with
> > its physical address space, convert the poison ranges to bad sectors
> > exposed using the badblocks interface.
> >
> > Signed-off-by: Vishal Verma <vishal.l.verma@xxxxxxxxx>
> > ---
> > drivers/nvdimm/pmem.c | 124
> > ++++++++++++++++++++++++++++++++++++++++++++++++++
> > 1 file changed, 124 insertions(+)
> >
> > diff --git a/drivers/nvdimm/pmem.c b/drivers/nvdimm/pmem.c
> > index 8ee7989..462570f 100644
> > --- a/drivers/nvdimm/pmem.c
> > +++ b/drivers/nvdimm/pmem.c
> [..]
> > @@ -196,6 +311,15 @@ static int pmem_attach_disk(struct device *dev,
> > set_capacity(disk, (pmem->size - pmem->data_offset) / 512);
> > pmem->pmem_disk = disk;
> >
> > + ret = disk_alloc_badblocks(disk);
> > + if (ret)
> > + return ret;
>
> I think we should skip allocating a bad block list in the case we find
> no poison. Then we can do a simple "if (disk->bb)" to check if there
> are any bad blocks.
Sounds good - I'll move this allocation into the (what will be a core)
routine that traverses the list, and make it such that we allocate
disk->bb iff a poison entry is found for _this_ pmem range.
It will mean a check at every list node, but that shouldn't matter as
this is just init-time overhead.
-Vishal��.n��������+%������w��{.n�����{�����ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f
