Re: [Qemu-devel] [Qemu-block] [PATCH 2/2] qapi: deprecate implicit filters

[Maxim Levitsky <mlevitsk@redhat.com>]

On Wed, 2019-08-14 at 15:27 -0400, John Snow wrote:
> 
> On 8/14/19 6:07 AM, Vladimir Sementsov-Ogievskiy wrote:
> > To get rid of implicit filters related workarounds in future let's
> > deprecate them now.
> > 
> > Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
> > ---
> >  qemu-deprecated.texi      |  7 +++++++
> >  qapi/block-core.json      |  6 ++++--
> >  include/block/block_int.h | 10 +++++++++-
> >  blockdev.c                | 10 ++++++++++
> >  4 files changed, 30 insertions(+), 3 deletions(-)
> > 
> > diff --git a/qemu-deprecated.texi b/qemu-deprecated.texi
> > index 2753fafd0b..8222440148 100644
> > --- a/qemu-deprecated.texi
> > +++ b/qemu-deprecated.texi
> > @@ -183,6 +183,13 @@ the 'wait' field, which is only applicable to sockets in server mode
> >  
> >  Use blockdev-mirror and blockdev-backup instead.
> >  
> > +@subsection implicit filters (since 4.2)
> > +
> > +Mirror and commit jobs inserts filters, which becomes implicit if user
> > +omitted filter-node-name parameter. So omitting it is deprecated, set it
> > +always. Note, that drive-mirror don't have this parameter, so it will
> > +create implicit filter anyway, but drive-mirror is deprecated itself too.
> > +
> >  @section Human Monitor Protocol (HMP) commands
> >  
> >  @subsection The hub_id parameter of 'hostfwd_add' / 'hostfwd_remove' (since 3.1)
> > diff --git a/qapi/block-core.json b/qapi/block-core.json
> > index 4e35526634..0505ac9d8b 100644
> > --- a/qapi/block-core.json
> > +++ b/qapi/block-core.json
> > @@ -1596,7 +1596,8 @@
> >  # @filter-node-name: the node name that should be assigned to the
> >  #                    filter driver that the commit job inserts into the graph
> >  #                    above @top. If this option is not given, a node name is
> > -#                    autogenerated. (Since: 2.9)
> > +#                    autogenerated. Omitting this option is deprecated, it will
> > +#                    be required in future. (Since: 2.9)
> >  #
> >  # @auto-finalize: When false, this job will wait in a PENDING state after it has
> >  #                 finished its work, waiting for @block-job-finalize before
> > @@ -2249,7 +2250,8 @@
> >  # @filter-node-name: the node name that should be assigned to the
> >  #                    filter driver that the mirror job inserts into the graph
> >  #                    above @device. If this option is not given, a node name is
> > -#                    autogenerated. (Since: 2.9)
> > +#                    autogenerated. Omitting this option is deprecated, it will
> > +#                    be required in future. (Since: 2.9)
> >  #
> >  # @copy-mode: when to copy data to the destination; defaults to 'background'
> >  #             (Since: 3.0)
> > diff --git a/include/block/block_int.h b/include/block/block_int.h
> > index 3aa1e832a8..624da0b4a2 100644
> > --- a/include/block/block_int.h
> > +++ b/include/block/block_int.h
> > @@ -762,7 +762,15 @@ struct BlockDriverState {
> >      bool sg;        /* if true, the device is a /dev/sg* */
> >      bool probed;    /* if true, format was probed rather than specified */
> >      bool force_share; /* if true, always allow all shared permissions */
> > -    bool implicit;  /* if true, this filter node was automatically inserted */
> > +
> > +    /*
> > +     * @implicit field is deprecated, don't set it to true for new filters.
> > +     * If true, this filter node was automatically inserted and user don't
> > +     * know about it and unprepared for any effects of it. So, implicit
> > +     * filters are workarounded and skipped in many places of the block
> > +     * layer code.
> > +     */
> > +    bool implicit;
> >  
> >      BlockDriver *drv; /* NULL means no media */
> >      void *opaque;
> > diff --git a/blockdev.c b/blockdev.c
> > index 36e9368e01..b3cfaccce1 100644
> > --- a/blockdev.c
> > +++ b/blockdev.c
> > @@ -3292,6 +3292,11 @@ void qmp_block_commit(bool has_job_id, const char *job_id, const char *device,
> >      BlockdevOnError on_error = BLOCKDEV_ON_ERROR_REPORT;
> >      int job_flags = JOB_DEFAULT;
> >  
> > +    if (!has_filter_node_name) {
> > +        warn_report("Omitting filter-node-name parameter is deprecated, it "
> > +                    "will be required in future");
> > +    }
> > +
> >      if (!has_speed) {
> >          speed = 0;
> >      }
> > @@ -3990,6 +3995,11 @@ void qmp_blockdev_mirror(bool has_job_id, const char *job_id,
> >      Error *local_err = NULL;
> >      int ret;
> >  
> > +    if (!has_filter_node_name) {
> > +        warn_report("Omitting filter-node-name parameter is deprecated, it "
> > +                    "will be required in future");
> > +    }
> > +
> >      bs = qmp_get_root_bs(device, errp);
> >      if (!bs) {
> >          return;
> > 
> 
> This might be OK to do right away, though.
> 
> I asked Markus this not too long ago; do we want to amend the QAPI
> schema specification to allow commands to return with "Warning" strings,
> or "Deprecated" stings to allow in-band deprecation notices for cases
> like these?
> 
> example:
> 
> { "return": {},
>   "deprecated": True,
>   "warning": "Omitting filter-node-name parameter is deprecated, it will
> be required in the future"
> }
> 
> There's no "error" key, so this should be recognized as success by
> compatible clients, but they'll definitely see the extra information.
> 
> Part of my motivation is to facilitate a more aggressive deprecation of
> legacy features by ensuring that we are able to rigorously notify users
> through any means that they need to adjust their scripts.
> 
> --js
> 
This is a very good idea IMHO.

Best regards,
	Maxim Levitsky