From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:55091) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aia57-0005m8-Mn for qemu-devel@nongnu.org; Wed, 23 Mar 2016 00:07:42 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1aia56-0003cy-Nb for qemu-devel@nongnu.org; Wed, 23 Mar 2016 00:07:41 -0400 Date: Wed, 23 Mar 2016 12:07:22 +0800 From: Peter Xu Message-ID: <20160323040722.GF28183@pxdev.xzpeter.org> References: <1458271654-23706-1-git-send-email-peterx@redhat.com> <1458271654-23706-3-git-send-email-peterx@redhat.com> <56F19222.2050003@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <56F19222.2050003@redhat.com> Subject: Re: [Qemu-devel] [PATCH v5 2/5] arm: qmp: add query-gic-capabilities interface List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Eric Blake Cc: wei@redhat.com, peter.maydell@linaro.org, drjones@redhat.com, armbru@redhat.com, mdroth@linux.vnet.ibm.com, qemu-devel@nongnu.org, qemu-arm@nongnu.org, abologna@redhat.com On Tue, Mar 22, 2016 at 12:42:42PM -0600, Eric Blake wrote: > On 03/17/2016 09:27 PM, Peter Xu wrote: > > This patch adds the command "query-gic-capabilities" but not implemnet > > s/not implemnet/does not implement/ Yep, again. Thanks. > > > it. The command is ARM-only. Return of the command is a list of > > GICCapability struct that describes all GIC versions that current QEMU > > and system support. > > > > Signed-off-by: Peter Xu > > --- > > > +++ b/qapi-schema.json > > @@ -4156,3 +4156,14 @@ > > 'data': { 'version': 'int', > > 'emulated': 'bool', > > 'kernel': 'bool' } } > > + > > +## > > +# @query-gic-capabilities: > > +# > > +# Return a list of supported GIC version capabilities. > > +# > > +# Returns: a list of GICCapability. > > +# > > +# Since: 2.6 > > +## > > +{ 'command': 'query-gic-capabilities', 'returns': ['GICCapability'] } > > On the surface, this seems okay. As mentioned before, I would have > squashed 1 and 2 into a single patch. The GICCapability type is > extensible, and introspection is sufficient at seeing what the type is > currently capable of exposing. > > On the other hand... > > > +++ b/scripts/qapi.py > > @@ -46,6 +46,7 @@ returns_whitelist = [ > > 'query-tpm-models', > > 'query-tpm-types', > > 'ringbuf-read', > > + 'query-gic-capability', > > ...it required a whitelist, because you are violating the usual > convention of returning a dict. If you DO need the whitelist, your > addition should have been kept sorted. But you don't need it, if you > would modify your QAPI to return a dict: > > { 'struct': 'GICCapabilitiesReturn', > 'data': { 'capabilities': ['GICCapability'] } } > { 'command': 'query-gic-capabilities', > 'returns': 'GICCapabilitiesReturn' } > > Yes, the dict has only a single key, and that key points to the same > list; but now you have future extensibility: in the future, we could > return any future global data as a sibling to the array, without having > to modify every element of the array to repeat redundant information. Yes, I think this is better solution. Will adopt this in next version. Thanks for the comments! -- peterx