[Pharo-project] UUID and Cog?

Levente Uzonyi leves at elte.hu
Tue Oct 5 22:35:32 EDT 2010


On Tue, 5 Oct 2010, Philippe Marschall wrote:

> On 05.10.2010 18:23, Levente Uzonyi wrote:
>> On Tue, 5 Oct 2010, Philippe Marschall wrote:
>>
>>> On 05.10.2010 16:41, David T. Lewis wrote:
>>>> If anyone has any specific information as to the root cause of the
>>>> libuuid
>>>> bug, I'd appreciate if you can update the bug report here:
>>>>
>>>>   http://bugs.squeak.org/view.php?id=7358
>>>>
>>>> I had previously tried to summarize as follows:
>>>>
>>>>> The bug exists in some versions of libuuid on some Linux distributions.
>>>>> The workaround is to compile the plugin internal. The latest versions
>>>>> of Unix VM have the plugin compiled internally, although some users may
>>>>> still experience problems if they have a leftover external plugin from
>>>>> a prior installation.
>>>>
>>>> But it's not clear if this is accurate given the recent reports here.
>>>> In particular, if a precompiled Cog VM with the UUID plugin compiled
>>>> internally leads to a crash, then the above is not a sufficient
>>>> workaround (at least not on Cog).
>>>>
>>>> If we are seeing crashes on Cog with an internal UUID plugin, but
>>>> not on the traditional VM with an internal UUID plugin, then I think
>>>> I smell a pthread related issue.
>>>
>>> How can I find out whether the plugin is compiled internally or not?
>>
>> The crash report lists the loaded plugins. If the plugin name ends with
>> (i), then it's internal, if it ends with (e), then it's external. If
>> you're using Eliot's build, then it's external.
>
> Yeah, that's what I'm using, it's external.
>
>> If you just want a working VM and the quality of UUIDs doesn't matter,
>> then you can safely delete the UUIDPlugin from the lib/squeak/3.9-7/
>> directory.
>
> Uhm, I'm making Monticello commits and good UUIDs would help.
>

I think the UUIDGenerator in the image produces UUIDs which are good 
enough for MC.

>>>
>>> What I see in the console is the following:
>>>
>>> Segmentation fault
>>>
>>>
>>>
>>> Smalltalk stack dump:
>>> 0xff79f344 M UUID>initialize -1045926312: a(n) UUID
>>> 0xff79f35c M UUID class(Behavior)>new: -1246728604: a(n) UUID class
>>> 0xff79f380 I UUID class>new -1246728604: a(n) UUID class
>>>
>>> ...
>>>
>>> Most recent primitives
>>> ...
>>> withArgs:executeMethod:
>>> basicNew:
>>> primMakeUUID
>>>
>>> Is there a way to provide more information?
>>
>> The stack dump looks the same as what other people have. The version of
>> your linux might be helpful, so others can reproduce the issue if you
>> don't want to debug it yourself.
>
> Gentoo 64bit: 2.6.35-gentoo-r8
> util-linux (where libuuid afaik comes from): 2.18-r1

Do you have a 32-bit version of libuuid installed? If not, it won't work.
What does "ldd UUIDPlugin" print?


Levente

>
> Cheers
> Philippe
>
>
> _______________________________________________
> Pharo-project mailing list
> Pharo-project at lists.gforge.inria.fr
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>




More information about the Pharo-dev mailing list