[PATCH 3/3] os: Add AddClientOnOpenFD() to create a new client for an file descriptor

[Kristian Høgsberg]

On Wed, Mar 19, 2014 at 4:29 PM, Peter Hutterer
<peter.hutterer at who-t.net> wrote:
> On Tue, Mar 18, 2014 at 10:06:00PM -0700, Kristian Høgsberg wrote:
>> When the Xwayland server is socket-activated, we need to connect and
>> initialize the window manager before the activating client gets to
>> proceed with connecting.  We do this by passing a socket file
>> descriptor for the window manager connection to the Xwayland server,
>> which then uses this new function to set it up as an X client.
>>
>> Signed-off-by: Kristian Høgsberg <krh at bitplanet.net>
>> ---
>>  include/os.h    |  2 ++
>>  os/connection.c | 20 ++++++++++++++++++++
>>  2 files changed, 22 insertions(+)
>>
>> diff --git a/include/os.h b/include/os.h
>> index 90229e6..5be3bcc 100644
>> --- a/include/os.h
>> +++ b/include/os.h
>> @@ -168,6 +168,8 @@ extern _X_EXPORT void MakeClientGrabPervious(ClientPtr /*client */ );
>>
>>  extern _X_EXPORT void ListenOnOpenFD(int /* fd */ , int /* noxauth */ );
>>
>> +extern _X_EXPORT void AddClientOnOpenFD(int /* fd */ );
>> +
>>  extern _X_EXPORT CARD32 GetTimeInMillis(void);
>>  extern _X_EXPORT CARD64 GetTimeInMicros(void);
>>
>> diff --git a/os/connection.c b/os/connection.c
>> index b50f9e9..f1e7a25 100644
>> --- a/os/connection.c
>> +++ b/os/connection.c
>> @@ -1312,3 +1312,23 @@ ListenOnOpenFD(int fd, int noxauth)
>>      XdmcpReset();
>>  #endif
>>  }
>> +
>> +/* based on TRANS(SocketUNIXAccept) (XtransConnInfo ciptr, int *status) */
>> +void
>> +AddClientOnOpenFD(int fd)
>> +{
>> +    XtransConnInfo ciptr;
>> +    CARD32 connect_time;
>> +
>> +    ciptr = _XSERVTransReopenCOTSServer(5, fd, "@anonymous");
>> +
>> +    _XSERVTransSetOption(ciptr, TRANS_NONBLOCKING, 1);
>> +    ciptr->flags |= TRANS_NOXAUTH;
>> +
>> +    connect_time = GetTimeInMillis();
>> +
>> +    if (!AllocNewConnection(ciptr, fd, connect_time)) {
>> +        ErrorF(stderr, "failed to create client for wayland server\n");
>
> ErrorF doesn't take a stream, so you can drop the first argument.

It certainly doesn't... I used fprintf for some reason in the patch I
split into these three patches, that's where stderr came from.

> Plus, a
> more generic error message may be good, it's a bit confusing to see wayland
> here when the actual function seems generic enough.

Yup, true.  I'll send out a new patch that returns TRUE/FALSE for
success/failure and log a FatalError in Xwayland if it fails.

Thanks for reviewing,
Kristian

> Cheers,
>    Peter
>
>> +        return;
>> +    }
>> +}
>> --
>> 1.9.0
>>
>> _______________________________________________
>> xorg-devel at lists.x.org: X.Org development
>> Archives: http://lists.x.org/archives/xorg-devel
>> Info: http://lists.x.org/mailman/listinfo/xorg-devel
>>