public inbox for gcc@gcc.gnu.org
 help / color / mirror / Atom feed
From: Sebastian Pop <sebpop@gmail.com>
To: Richard Guenther <richard.guenther@gmail.com>
Cc: Li Feng <nemokingdom@gmail.com>, GCC <gcc@gcc.gnu.org>
Subject: Re: How could I get alias set information from data_reference_p
Date: Tue, 14 Jul 2009 16:03:00 -0000	[thread overview]
Message-ID: <cb9d34b20907140903y3a9e067g81d35a7f4ab3d58@mail.gmail.com> (raw)
In-Reply-To: <84fc9c000907140826t645135efp72d196d04ebb9095@mail.gmail.com>

On Tue, Jul 14, 2009 at 10:26, Richard
Guenther<richard.guenther@gmail.com> wrote:
> What do you mean by 'different number for each alias set'?

An alias set numbering maps alias sets to integer numbers,
and that map is one-to-one.

>  If you want to have a number that is the same for all conflicting
> memory references then you have to build the full conflict map and
> partition it.

> Likely not what you want?

maybe

> Why do you need alias-set numbers?

We want to represent the alias set information as an extra subscript
on memory accesses: for example,

if we have A[10] and supposing that A is in alias set 6, this would be
represented as "memory_access[6][10]".

For B[3][4] with base array B in alias set 7 and 8, we would represent
this as "memory_access[7][3][4] or memory_access[8][3][4]".

The data dependence test that we currently have in the graphite branch
would work with alias sets represented by an extra dimension to the
array dimensions.

Sebastian

  reply	other threads:[~2009-07-14 16:03 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-07-14  6:01 Li Feng
2009-07-14  8:54 ` Richard Guenther
2009-07-14  9:12   ` Li Feng
2009-07-14  9:40     ` Richard Guenther
2009-07-14 15:14       ` Sebastian Pop
2009-07-14 15:26         ` Richard Guenther
2009-07-14 16:03           ` Sebastian Pop [this message]
2009-07-14 16:09             ` Sebastian Pop
2009-07-14 21:34               ` Richard Guenther
2009-07-15  7:59                 ` Li Feng
2009-07-15 11:02                 ` Tobias Grosser
2009-07-15 11:26                   ` Richard Guenther
2009-07-15 19:16                     ` Tobias Grosser
2009-07-15 20:46                       ` Richard Guenther
2009-07-15 20:49                         ` Richard Guenther
     [not found]                         ` <15137_1247690941_4A5E40BC_15137_586_1_84fc9c000907151348s41395cc5u6cfacb60cde78bfa@mail.gmail.com>
2009-07-15 23:16                           ` Tobias Grosser
2009-07-16  8:39                             ` Richard Guenther
2009-07-16  9:00                               ` Li Feng
2009-07-16  9:16                                 ` Richard Guenther
2009-07-16 15:45                                 ` Daniel Berlin
2009-07-16 16:04                                   ` Sebastian Pop
2009-07-17  1:36                                   ` Li Feng

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=cb9d34b20907140903y3a9e067g81d35a7f4ab3d58@mail.gmail.com \
    --to=sebpop@gmail.com \
    --cc=gcc@gcc.gnu.org \
    --cc=nemokingdom@gmail.com \
    --cc=richard.guenther@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).