On 2018-09-20 08:07, Markus Armbruster wrote:
> Thomas Huth <thuth@xxxxxxxxxx> writes:
>
>> In early times, network backends were specified by a "vlan" and "name"
>> tuple. With the introduction of netdevs, the "name" was replaced by an
>> "id" (which is supposed to be unique), but the "name" parameter stayed
>> as an alias which could be used instead of "id". Unfortunately, we miss
>> the duplication check for "name":
>>
>> $ qemu-system-x86_64 -net user,name=n1 -net user,name=n1
>>
>> ... starts without an error, while "id" correctly complains:
>>
>> $ qemu-system-x86_64 -net user,id=n1 -net user,id=n1
>> qemu-system-x86_64: -net user,id=n1: Duplicate ID 'n1' for net
>>
>> Instead of trying to fix the code for the legacy "name" parameter, let's
>> rather get rid of this old interface and deprecate the "name" parameter
>> now - this will also be less confusing for the users in the long run.
>>
>> While we're at it, also deprecate the old syntax for the hostfwd_add
>> and hostfwd_remove commands that still work with this legacy "hub"
>> plus "name" tuple. It is enough to specify the netdev id there instead.
>>
>> Also add a missing dependency to the Makefile to make sure that the
>> docs get correctly regenerated when qemu-deprecated.texi is changed.
>
> Sure sounds like three patches to me.
Since the changes are really small and slightly related, one compact
patch sounded fine for me, too. But ok, I just discovered yet another
unrelated thing that should be added to the deprecation chapter
("enforce-config-section"), so I'll turn this into a small patch series
instead.
>> @@ -99,6 +104,14 @@ The ``query-cpus'' command is replaced by the ``query-cpus-fast'' command.
>> The ``arch'' output member of the ``query-cpus-fast'' command is
>> replaced by the ``target'' output member.
>>
>> +@section System emulator human monitor commands
>> +
>> +@subsection The hub_id parameter of 'hostfwd_add' / 'hostfwd_remove' (since 3.1)
>> +
>> +The @option{hub_id} parameter of the 'hostfwd_add' and 'hostfwd_remove' HMP
>> +commands is redundant. It is enough to specify the netdev ID of the backend
>> +that should be changed.
>
> Hmm. What's redundant is the [hub_id name] alternative in
>
> .params = "[hub_id name]|[netdev_id] ..."
>
> The [netdev_id] alternative was added in January (commit 93653066445).
> Suggest to word it more like other sections do:
>
> Parameters @option{hub_id} and @option{name} have been replaced by
> @option{netdev_id}.
Ok. The funny thing is that IIUC "name" and "netdev_id" are the very
same identifier nowadays (I was not aware of this in the past), so the
descriptions could also have been written as "[[hub_id] netdev_id]"
instead. ==> This "name" vs. "id" thing is really confusing, it's really
a good idea to remove it soon.
Thomas
--
libvir-list mailing list
libvir-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/libvir-list
