Re: [PATCH v2 09/11] block/qcow2: implement blockdev-amend

[Maxim Levitsky]

On Fri, 2019-10-04 at 21:03 +0200, Max Reitz wrote:
> On 13.09.19 00:30, Maxim Levitsky wrote:
> > Currently only for changing crypto parameters
> 
> Yep, that elegantly avoids most of the problems we’d have otherwise. :-)
> 
> > Signed-off-by: Maxim Levitsky <address@hidden>
> > ---
> >  block/qcow2.c        | 71 ++++++++++++++++++++++++++++++++++++++++++++
> >  qapi/block-core.json |  6 ++--
> >  2 files changed, 75 insertions(+), 2 deletions(-)
> > 
> > diff --git a/block/qcow2.c b/block/qcow2.c
> > index 26f83aeb44..c8847ec6e2 100644
> > --- a/block/qcow2.c
> > +++ b/block/qcow2.c
> > @@ -3079,6 +3079,18 @@ qcow2_co_create(BlockdevCreateOptions 
> > *create_options, Error **errp)
> >      assert(create_options->driver == BLOCKDEV_DRIVER_QCOW2);
> >      qcow2_opts = &create_options->u.qcow2;
> >  
> > +    if (!qcow2_opts->has_size) {
> > +        error_setg(errp, "Size is manadatory for image creation");
> > +        return -EINVAL;
> > +
> > +    }
> > +
> > +    if (!qcow2_opts->has_file) {
> > +        error_setg(errp, "'file' is manadatory for image creation");
> > +        return -EINVAL;
> > +
> > +    }
> > +
> >      bs = bdrv_open_blockdev_ref(qcow2_opts->file, errp);
> >      if (bs == NULL) {
> >          return -EIO;
> > @@ -5111,6 +5123,64 @@ static int qcow2_amend_options(BlockDriverState *bs, 
> > QemuOpts *opts,
> >      return 0;
> >  }
> >  
> > +
> > +static int coroutine_fn qcow2_co_amend(BlockDriverState *bs,
> > +                                       BlockdevCreateOptions *opts,
> > +                                       bool force,
> > +                                       Error **errp)
> > +{
> > +    BlockdevCreateOptionsQcow2 *qopts = &opts->u.qcow2;
> > +    BDRVQcow2State *s = bs->opaque;
> > +    int ret;
> > +
> > +    /*
> > +     * This is ugly as hell, in later versions of this patch
> > +     * something has to be done about this
> 
> Well, at least the language of the comment needs adjustment. :-)
Thats for sure :-)

BTW, if I opt for having a separate amend parameter struct,
this will fix this problem as well, as all unsupported
fields will just not be there.

[..]

Best regards,
        Maxim Levitsky