qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Marc-André Lureau" <marcandre.lureau@redhat.com>
To: Eric Blake <eblake@redhat.com>
Cc: qemu-devel <qemu-devel@nongnu.org>,
	"P. Berrange, Daniel" <berrange@redhat.com>,
	"Armbruster, Markus" <armbru@redhat.com>,
	"Bonzini, Paolo" <pbonzini@redhat.com>
Subject: Re: [Qemu-devel] [PATCH v3 4/4] qobject: modify qobject_ref() to assert on NULL and return obj
Date: Fri, 13 Apr 2018 18:05:42 +0200	[thread overview]
Message-ID: <CAMxuvazz_Zzget3-oVD_Fa4-G4YT+HLXg9Psh0bjpD9tbPxO3Q@mail.gmail.com> (raw)
In-Reply-To: <49475649-b8d8-e09a-8e95-7ce36c7396f2@redhat.com>

Hi

On Thu, Mar 29, 2018 at 6:10 PM, Eric Blake <eblake@redhat.com> wrote:
> On 03/29/2018 10:48 AM, Marc-André Lureau wrote:
>>
>> Following a discussion on the mailing list: while it may be convenient
>> to accept NULL value in qobject_unref() (for similar reasons as free()
>> accepts NULL), it is a probably a bad idea to accept NULL argument in
>> qobject_ref().
>>
>> Furthermore, it is convenient and more clear to call qobject_ref() at
>> the time when the reference is associated with a variable, or
>> argument. For this reason, make qobject_ref() return the same pointer
>> as given.
>>
>> Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
>> ---
>
>
>> @@ -101,13 +101,18 @@ static inline void qobject_unref(QObject *obj)
>>     /**
>>    * qobject_ref(): Increment QObject's reference count
>> + * @obj: a #QObject or children type instance (must not be NULL)
>
>
> s/children/child/
>
>> + *
>> + * Returns: the same @obj. The type of @obj will be propagated to the
>> + * return type.
>>    */
>>   #define qobject_ref(obj)                        \
>> -    qobject_ref(QOBJECT(obj))
>> +    (typeof(obj)) qobject_ref(QOBJECT(obj))
>
>
> You're missing outer ().  There are cases where '(cast) (expr)' and
> '((cast)(expr))' can behave differently when combined with surrounding
> syntax; for example, -> has higher precedence than cast.  Consider:
>
> qobject_ref(my_qdict)->size;
>
> As you wrote it, you would attempt to dereference the 'size' member of void*
> (compile error), prior to casting that result to QDict*; with the
> parenthesis, you get the intended dereference of the size member of the
> QDict pointer.
>

ok

>> +++ b/block.c
>> @@ -5134,8 +5134,8 @@ static bool append_open_options(QDict *d,
>> BlockDriverState *bs)
>>               continue;
>>           }
>>   -        qobject_ref(qdict_entry_value(entry));
>> -        qdict_put_obj(d, qdict_entry_key(entry),
>> qdict_entry_value(entry));
>> +        qdict_put_obj(d, qdict_entry_key(entry),
>> +                      qobject_ref(qdict_entry_value(entry)));
>
>
> However, I like this simplification that your patch enables.  How did you
> find candidate spots? Manually, or via Coccinelle?

Manual review (there are not that many qobject_ref())

thanks

  reply	other threads:[~2018-04-13 16:05 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-03-29 15:48 [Qemu-devel] [PATCH v3 0/4] RFC: simplify qobject refcount Marc-André Lureau
2018-03-29 15:48 ` [Qemu-devel] [PATCH v3 1/4] qobject: ensure base is at offset 0 Marc-André Lureau
2018-03-29 15:48 ` [Qemu-devel] [PATCH v3 2/4] qobject: introduce QObjectCommon Marc-André Lureau
2018-03-29 16:15   ` Eric Blake
2018-03-29 15:48 ` [Qemu-devel] [PATCH v3 3/4] qobject: replace qobject_incref/QINCREF qobject_decref/QDECREF Marc-André Lureau
2018-03-29 16:23   ` Eric Blake
2018-04-13 16:13     ` Paolo Bonzini
2018-03-29 15:48 ` [Qemu-devel] [PATCH v3 4/4] qobject: modify qobject_ref() to assert on NULL and return obj Marc-André Lureau
2018-03-29 16:10   ` Eric Blake
2018-04-13 16:05     ` Marc-André Lureau [this message]
2018-03-31  6:26 ` [Qemu-devel] [PATCH v3 0/4] RFC: simplify qobject refcount no-reply

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=CAMxuvazz_Zzget3-oVD_Fa4-G4YT+HLXg9Psh0bjpD9tbPxO3Q@mail.gmail.com \
    --to=marcandre.lureau@redhat.com \
    --cc=armbru@redhat.com \
    --cc=berrange@redhat.com \
    --cc=eblake@redhat.com \
    --cc=pbonzini@redhat.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).