This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Improving alias analysis
- From: "Sanjiv Kumar Gupta, Noida" <sanjivg at noida dot hcltech dot com>
- To: gcc at gcc dot gnu dot org
- Cc: dnovillo at redhat dot com
- Date: Tue, 26 Nov 2002 13:34:26 +0530
- Subject: Re: Improving alias analysis
On Thu, 17 Oct 2002, Sanjiv Kumar Gupta, Noida wrote:
> I intend to improve the alias analysis so that it can handle
> such cases. For that, I am currently thinking of possible
> solutions. One solution could be backward traversing of the
> list of RTXs and finding out if the two pointers differ.
> The other could be maintaining symoblic values for regsiters
> at each instruction level.
>
>>Whatever you do, please try and use published algorithms. Alias
>>analysis is a very well known technique and many solutions exist
>>in the literature.
>>Diego.
I've decided to implement the following paper to solve the alias
analysis problem
http://citeseer.nj.nec.com/cache/papers/cs/1318/ftp:zSzzSzftp.cs.arizona.edu
zSzreportszSz1997zSzTR97-13.pdf/debray98alias.pdf
(long and broken URL, copy paste the complete URL in your browser)
This approach is capable of representing the address arithmatic and
can distinguish between pointer pseudos that have been computed
using distinct displacements from a base.
To see the problem description, refer to my earlier postings
http://gcc.gnu.org/ml/gcc/2002-10/msg01558.html
http://gcc.gnu.org/ml/gcc/2002-10/msg01032.html
--Sanjiv