qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Eric Blake <eblake@redhat.com>
To: Markus Armbruster <armbru@redhat.com>
Cc: Michael Roth <mdroth@linux.vnet.ibm.com>,
	qemu-devel@nongnu.org, Luiz Capitulino <lcapitulino@redhat.com>
Subject: Re: [Qemu-devel] [PATCH v13 03/14] qapi: Convert QType into QAPI built-in enum type
Date: Thu, 26 Nov 2015 22:09:29 -0700	[thread overview]
Message-ID: <5657E589.1060901@redhat.com> (raw)
In-Reply-To: <87ziy0stjx.fsf@blackfin.pond.sub.org>

[-- Attachment #1: Type: text/plain, Size: 2275 bytes --]

On 11/26/2015 07:51 AM, Markus Armbruster wrote:
> Eric Blake <eblake@redhat.com> writes:
> 
>> What's more meta than using qapi to define qapi? :)
>>
>> Convert QType into a full-fledged[*] builtin qapi enum type, so
>> that a subsequent patch can then use it as the discriminator
>> type of qapi alternate types.  Fortunately, the judicious use of
>> 'prefix' in the qapi definition avoids churn to the spelling of
>> the enum constants.
>>

>>  #include <stddef.h>
>>  #include <assert.h>
>> +#include "qapi-types.h"
> 
> Needed for QType.  Risk for circular inclusion.  We're currently fine,
> because generated qapi-types.h includes only "qemu/typedefs.h" (visible
> below).  Should we add a comment to qapi-types.py?

How about this:
 @@ -323,6 +319,7 @@ fdef.write(mcgen('''
 ''',
                   prefix=prefix))

+# To avoid circular headers, use only typedefs.h here, not qobject.h
 fdecl.write(mcgen('''
 #include <stdbool.h>
 #include <stdint.h>
-#include "qapi/qmp/qobject.h"
+#include "qemu/typedefs.h"
 '''))

 schema = QAPISchema(input_file)


>>      def visit_enum_type(self, name, info, values, prefix):
>> -        self._fwdecl += gen_enum(name, values, prefix)
>> -        self._fwdefn += gen_enum_lookup(name, values, prefix)
>> +        # Special case for our lone builtin enum type
>> +        # TODO use something cleaner than existence of info
>> +        if not info:
>> +            self._btin += gen_enum(name, values, prefix)
>> +            if do_builtins:
>> +                self.defn += gen_enum_lookup(name, values, prefix)
>> +        else:
>> +            self._fwdecl += gen_enum(name, values, prefix)
>> +            self._fwdefn += gen_enum_lookup(name, values, prefix)
> 
> Odd: gen_enum_lookup() goes into .defn for built-ins, but ._fwdefn for
> user-defineds.  Makes me suspect it ._fwdefn isn't needed anymore.  A
> quick test compile is happy with .defn for both.
> 
> If we want to keep ._fwdefn for some reason, we should use for built-ins
> as well.

We need it for gen_alternate_qtypes().  But that disappears in 4/14, so
I'll add a patch to clean it up.

-- 
Eric Blake   eblake redhat com    +1-919-301-3266
Libvirt virtualization library http://libvirt.org


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 604 bytes --]

  reply	other threads:[~2015-11-27  5:09 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-20 17:24 [Qemu-devel] [PATCH v13 00/14] qapi member collision (post-introspection cleanups, subset D) Eric Blake
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 01/14] qobject: Simplify QObject Eric Blake
2015-11-26 14:27   ` Markus Armbruster
2015-11-26 15:06   ` Markus Armbruster
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 02/14] qobject: Rename qtype_code to QType Eric Blake
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 03/14] qapi: Convert QType into QAPI built-in enum type Eric Blake
2015-11-26 14:51   ` Markus Armbruster
2015-11-27  5:09     ` Eric Blake [this message]
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 04/14] qapi: Simplify visiting of alternate types Eric Blake
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 05/14] qapi: Inline _make_implicit_tag() Eric Blake
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 06/14] qapi: Fix alternates that accept 'number' but not 'int' Eric Blake
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 07/14] qapi: Simplify visits of optional fields Eric Blake
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 08/14] qapi: Shorter " Eric Blake
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 09/14] qapi: Prepare new QAPISchemaMember base class Eric Blake
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 10/14] qapi: Track enum values by QAPISchemaMember, not string Eric Blake
2015-11-26 16:29   ` Markus Armbruster
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 11/14] qapi: Populate info['name'] for each entity Eric Blake
2015-11-26 16:46   ` Markus Armbruster
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 12/14] qapi: Enforce (or whitelist) case conventions on qapi members Eric Blake
2015-11-27  9:03   ` Markus Armbruster
2015-12-01 22:35     ` Eric Blake
2015-12-02  8:20       ` Markus Armbruster
2015-12-02 15:47         ` Eric Blake
2015-11-27  9:42   ` Markus Armbruster
2015-12-01 16:12     ` Eric Blake
2015-12-02 10:55       ` Markus Armbruster
2015-11-20 17:24 ` [Qemu-devel] [PATCH v13 13/14] qapi: Move duplicate collision checks to schema check() Eric Blake
2015-11-20 17:25 ` [Qemu-devel] [PATCH v13 14/14] qapi: Detect base class loops Eric Blake
2015-11-27  9:56 ` [Qemu-devel] [PATCH v13 00/14] qapi member collision (post-introspection cleanups, subset D) Markus Armbruster
2015-12-01 16:28   ` Eric Blake

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5657E589.1060901@redhat.com \
    --to=eblake@redhat.com \
    --cc=armbru@redhat.com \
    --cc=lcapitulino@redhat.com \
    --cc=mdroth@linux.vnet.ibm.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).