From mboxrd@z Thu Jan  1 00:00:00 1970
From: Andrew Cooper <andrew.cooper3@citrix.com>
Subject: Re: [PATCH XEN v8 02/29] tools: Refactor
 /dev/xen/evtchn wrappers into libxenevtchn.
Date: Mon, 25 Jan 2016 14:49:47 +0000
Message-ID: <56A6360B.5090707@citrix.com>
References: <1452864168.32341.97.camel@citrix.com>
	<1452864188-2417-1-git-send-email-ian.campbell@citrix.com>
	<1452864188-2417-3-git-send-email-ian.campbell@citrix.com>
	<56A262FF.8060806@oracle.com> <1453715326.4320.112.camel@citrix.com>
	<22182.12970.424449.937569@mariner.uk.xensource.com>
	<1453733273.4320.179.camel@citrix.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Return-path: <xen-devel-bounces@lists.xen.org>
In-Reply-To: <1453733273.4320.179.camel@citrix.com>
List-Unsubscribe: <http://lists.xen.org/cgi-bin/mailman/options/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xen.org>
List-Help: <mailto:xen-devel-request@lists.xen.org?subject=help>
List-Subscribe: <http://lists.xen.org/cgi-bin/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=subscribe>
Sender: xen-devel-bounces@lists.xen.org
Errors-To: xen-devel-bounces@lists.xen.org
To: Ian Campbell <ian.campbell@citrix.com>, Ian Jackson <Ian.Jackson@eu.citrix.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>, wei.liu2@citrix.com, xen-devel@lists.xen.org
List-Id: xen-devel@lists.xenproject.org

On 25/01/16 14:47, Ian Campbell wrote:
> On Mon, 2016-01-25 at 14:35 +0000, Ian Jackson wrote:
>> Ian Campbell writes ("Re: [Xen-devel] [PATCH XEN v8 02/29] tools:
>> Refactor /dev/xen/evtchn wrappers into libxenevtchn."):
>>> Various of the tools/libs/*/include/*.h have a
>>>
>>>     /* Callers who don't care don't need to #include <xentoollog.h> */
>>>     typedef struct xentoollog_logger xentoollog_logger;
>>>
>>> but since that typedef matches in all cases I think it is allowed to be
>>> repeated like this, isn't it?
>> No, I'm afraid not.
> :-/
>
>>   It is permitted to repeatedly mention `struct
>> xentoollog_logger', but the typedef must only occur once in any
>> translation unit.
> For some reason I thought the typedef's could be repeated so long as they
> were identical.
>
> Seems at least Debian's gcc tolerates such things (which makes me wonder
> how many more instances of this might have snuck in, I guess folks like
> Boris would have spotted them pretty quick).
>
>> If it is desirable to let callers avoid including xentoollog.h, then
>> all those headers need to say:
>>
>>   struct xentoollog_logger;
>>   int some_function(..., struct xentoollog_logger *lg, ...);
>>
>> So in your patches something like:
>>
>>   struct xentoollog_logger;
>>
>>   xenevtchn_handle *xenevtchn_open(struct xentoollog_logger *logger,
>>                                    unsigned open_flags);
> I'll do something like this for tools/libs/*/include.
>
> FYI the underlying patches are all in since Friday.

How about doing the header strictness check on all headers at one?  If
repeating typedefs is a gccism, that should catch it.

~Andrew