[Soot-list] Seeking proposals for Google Summer Of Code projects

[Marc-Andre Laverdiere-Papineau]

Really? Interesting.
First, we could multithread it - one class body loaded per thread would 
be a good start.

There has been many ideas thrown around for having things cached and the 
like. But I haven't profiled it to see the hotspots - has anybody done that?

I think that would be a nice GSOC.

Regards,

Marc-André Laverdière-Papineau
Doctorant - PhD Candidate

On 13-02-27 10:47 AM, Patrick Lam wrote:
> I think that most of the processing time for Spark is not Spark itself
> but jimplification of its input. Faster jimplification could be a
> potential SoC project, but I'm not really sure about how to do it.
>
> pat
>
> On 02/27/13 10:46, Marc-Andre Laverdiere-Papineau wrote:
>> Thanks for your input. Do you think we can isolate a subset?
>> Spark comes to mind as probably where people would like to see speedups :)
>>
>> Marc-André Laverdière-Papineau
>> Doctorant - PhD Candidate
>>
>> On 13-02-27 09:47 AM, Patrick Lam wrote:
>>> (From personal experience) I think that's too hard as a summer project
>>> for a student.
>>>
>>> On 02/27/13 08:12, Marc-Andre Laverdiere-Papineau wrote:
>>>> I was forgetting the obvious: parallelization of the the code base.
>>>>
>>>> Right now, Soot works in single-threaded mode. That gives a long long
>>>> processing time for Spark and other phases. We should have a way to
>>>> parallelize things.
>>>>
>>>> Marc-André Laverdière-Papineau
>>>> Doctorant - PhD Candidate
>>>>
>>>> On 13-02-14 01:36 PM, Eric Bodden wrote:
>>>>> Hi all.
>>>>>
>>>>> Google has just put up  their webpage for the 2013 Summer of Code:
>>>>> http://www.google-melange.com/gsoc/homepage/google/gsoc2013
>>>>>
>>>>> I would like to solicit a few ideas for projects that summer students
>>>>> could work on. Please fill in your ideas here:
>>>>> https://docs.google.com/document/d/19CVbaPeqJHQm8XwSGBnK0RKn1XMRwmBoxODJIOIvD5U/edit#
>>>>> Also indicate if you would be willing to supervise the project.
>>>>>
>>>>> My plan is to see if we get a good set of proposals and if we do then
>>>>> apply for participation.
>>>>>
>>>>> Cheers,
>>>>> Eric
>>>>>
>>>> _______________________________________________
>>>> Soot-list mailing list
>>>> Soot-list at sable.mcgill.ca
>>>> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>>>
>>> _______________________________________________
>>> Soot-list mailing list
>>> Soot-list at sable.mcgill.ca
>>> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>>>
>> _______________________________________________
>> Soot-list mailing list
>> Soot-list at sable.mcgill.ca
>> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>
> _______________________________________________
> Soot-list mailing list
> Soot-list at sable.mcgill.ca
> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>