[LLVMdev] poolalloc and steensgaard

Daniel Guo realdaniel2013 at gmail.com
Thu Feb 12 10:43:38 PST 2015


Yes, i also got that reason from web, then i need to build something from
scratch now...
On Feb 12, 2015 5:21 AM, "Yogesh Chavan" <cs13m1012 at iith.ac.in> wrote:

> Hi Daniel,
>
> Please have a look at the mail forwarded to you. It seems there are some
> concerns about patents.
>
> Thanks,
> -Yogesh
>
> ---------- Forwarded message ----------
> From: Philip Reames <listmail at philipreames.com>
> Date: Thu, Feb 12, 2015 at 1:36 AM
> Subject: Re: [LLVMdev] Importing Andersen and Steensgaard Alias Analysis
> in Latest LLVM
> To: Yogesh Chavan <cs13m1012 at iith.ac.in>, llvmdev at cs.uiuc.edu
> Cc: George Burgess IV <george.burgess.iv at gmail.com>
>
>
> When questions like this one were raised in the past, there were concerns
> about patents.  I think the issues were summarized in that thread, you
> might try searching for it.
>
> (As always, not a lawyer, etc...)
>
>
>
> On Thu, Feb 12, 2015 at 1:21 AM, Daniel Guo <realdaniel2013 at gmail.com>
> wrote:
>
>> Hi Yogesh,
>> Thanks for the reply.
>> Sounds like i have two choices:
>> (1) use llvm2.7 poolalloc (including steensgaard), but it is buggy?
>> (2) adjust steensgaard based on lastest poolalloc, but it needs too much
>> modification? and also the correctness can not be promised?
>>
>> Daniel.
>>
>> On Wed, Feb 11, 2015 at 2:21 PM, Yogesh Chavan <cs13m1012 at iith.ac.in>
>> wrote:
>>
>>> Hi Daniel,
>>>
>>> To my knowledge, there is no other implementation available for
>>> Steensgaard AA except poolalloc.
>>> I have tried Steensgaard AA under LLVM 2.7 poolalloc, but there are lots
>>> of bugs which need to be fixed. It will be a good idea to use previous
>>> implementation, but you will have to modify (a lot) it with respect to the
>>> latest version of LLVM poolalloc.
>>>
>>> Thanks,
>>> -Yogesh
>>>
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20150212/57f9a176/attachment.html>


More information about the llvm-dev mailing list