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

Re: [Libguestfs] [PATCH v2 08/13] resize: add function mbr_part_type



On Fri, Sep 19, 2014 at 02:21:28PM +0100, Richard W.M. Jones wrote:
> On Fri, Sep 19, 2014 at 03:39:10PM +0800, Hu Tao wrote:
> > Function mbr_part_type returns one of "primary", "extended" and
> > "logical". The type is used by parted when adding partitions.
> > 
> > Signed-off-by: Hu Tao <hutao cn fujitsu com>
> > ---
> >  resize/resize.ml | 12 +++++++++---
> >  1 file changed, 9 insertions(+), 3 deletions(-)
> > 
> > diff --git a/resize/resize.ml b/resize/resize.ml
> > index 2ffd26e..f57b2ff 100644
> > --- a/resize/resize.ml
> > +++ b/resize/resize.ml
> > @@ -50,6 +50,7 @@ type partition = {
> >    p_id : partition_id;           (* Partition (MBR/GPT) ID. *)
> >    p_type : partition_content;    (* Content type and content size. *)
> >    p_label : string option;       (* Label/name. *)
> > +  p_part_num: int;               (* partition number *)
> 
> I don't think it's necessary to store p_part_num in the main
> partitions struct.
> 
> Instead you can get to the partition number using:
> 
>   p.p_part.G.part_num
> 
> For example, mbr_part_type could be written:
> 
>   let mbr_part_type x =
>     if x.p_part.G.part_num <= 4 && x.p_type <> ContentExtendedPartition then "primary"
>   [etc]

Yes, this works, too.

> 
> > +  let mbr_part_type x =
> > +    if x.p_part_num <= 4 && x.p_type <> ContentExtendedPartition then "primary"
> > +    else if x.p_part_num <= 4 && x.p_type = ContentExtendedPartition then "extended"
> > +    else "logical" in
> > +
> >    (* Now partition the target disk. *)
> >    List.iter (
> >      fun p ->
> > -      g#part_add "/dev/sdb" "primary" p.p_target_start p.p_target_end
> > +      g#part_add "/dev/sdb" (mbr_part_type p) p.p_target_start p.p_target_end
> >    ) partitions;
> 
> The big problem with this is I think it doesn't work properly for GPT.
> GPT has no idea what "extended" is -- such a concept does not exist.
> So if you passed a GPT guest to virt-resize with >= 5 partitions in
> it, then virt-resize will not work.

The type becomes partition name if the partition table is GPT. So this
patch is not a problem.

However, I didn't test the series with GPT partitions, I'll test in v3.

Regards,
Hu

> 
> I think you need to check the partition table type somewhere.
> 
> Rich.
> 
> -- 
> Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
> Read my programming and virtualization blog: http://rwmj.wordpress.com
> Fedora Windows cross-compiler. Compile Windows programs, test, and
> build Windows installers. Over 100 libraries supported.
> http://fedoraproject.org/wiki/MinGW


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