No, this patch is a workaround in fact.
It is a bug of libhal on Solaris.
I just check libhal's code on solaris and on community.
In function libhal_device_add_property_watch(),
On Solaris:
http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/lib/hal/libhal/common/libhal.c
*if* (dbus_error_is_set
<http://src.opensolaris.org/source/s?defs=dbus_error_is_set&project=/onnv>
(error
<http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/lib/hal/libhal/common/libhal.c#error>))
{
return FALSE;
}
On Community:
http://cgit.freedesktop.org/hal/tree/libhal/libhal.c
| if (error != NULL&& dbus_error_is_set (error)) {
return FALSE;
}
|
I have asked hal team to fix this bug.
After they fixed this, we can delete our patch internal.
On 01/11/10 07:36 PM, Padraig O'Briain wrote:
> Should this patch be upstream?
>
> Padraig
>
> On 01/11/10 09:40, Harry Lu wrote:
>> It is caused by the community patch of
>> https://bugzilla.gnome.org/show_bug.cgi?id=597827
>>
>> which was in gvfs 1.4.2.
>>
>> We have gvfs 1.4.1 in B130 and 1.4.3 in B131.
>>
>> Harry
>>
>> On Mon, 2010-01-11 at 09:02 +0000, Padraig O'Briain wrote:
>>
>>> What changed to cause this problem to appear in the current build?
>>>
>>> Padraig
>>>
>>> On 01/11/10 08:52, Jerry Tan wrote:
>>>
>>>> The bug for 13755 : gvfs-hal-volume-monitor core
>>>>
>>>> Reason: when call libhal function, it passed a NULL as DbusError,
>>>> but libhal can not handle it , so it core dump.
>>>>
>>>> add dbus_init_error() to avoid this core dump.
>>>>
>>>> please see the attachment.
>>>>
>>>>
>>>>
>>>> Thanks.
>>>>
>>>>
>>>>
>>>>
>>>>
>>
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/jds-review/attachments/20100112/1ab3a8f7/attachment.html>
