[LLVMdev] GVN PRE algorithms in LLVM
dberlin at dberlin.org
Tue Mar 10 09:20:21 PDT 2015
The GVN algorithm used in LLVM currently (I'm rewriting it) is the basic
hash based RPO algorithm.
The new one i'm writing is based on http://dl.acm.org/citation.cfm?id=512536
LLVM has different algorithms for both scalar PRE and load PRE.
They are basically variants of standard PRE algorithms transformed into
SSA, but with some deliberate limitations made as implementation speed/code
size trade offs.
At some point, i plan on implement reimplementing GVNPRE that will handle
both scalars and loads (as GCC's does)
If you have interest in working on stuff, i'm happy to chat/help point out
whatever I can. Depends on what you want to do.
On Tue, Mar 10, 2015 at 4:25 AM, Rahul Jain <1989.rahuljain at gmail.com>
> This landed in my spam folder so resending it to the list!
> On Mar 9, 2015 9:56 PM, "Rakshit Singla" <cs12b1029 at iith.ac.in> wrote:
>> Hello everyone
>> I am Rakshit Singla, a third year undergrad at IIT Hyderabad, India. I
>> finished a basic compilers course last semester and am currently doing a
>> compiler optimizations course. I have been exploring LLVM for the past few
>> months (wrote a front-end for the Classroom Object Oriented Language and
>> have been studying pieces of code.) I would like to work with LLVM and
>> contribute to the community.
>> For starters, I have a couple of questions.
>> What is the GVN algorithm used in LLVM? Is it the one by Alpern, Wegman,
>> Zadeck or Briggs, Cooper or some other?
>> How much of PRE is done in LLVM? Are any of the well known algorithms for
>> PRE used in LLVM?
>> Thanks and regards,
>> Rakshit Singla
>> Third Year Undergrad
>> Indian Institute of Technology Hyderabad
>> LLVM Developers mailing list
>> LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the llvm-dev