From: Zdenek Dvorak <rakdver@kam.mff.cuni.cz>
To: Bingfeng Mei <bmei@broadcom.com>
Cc: Jean Christophe Beyler <jean.christophe.beyler@gmail.com>,
"gcc@gcc.gnu.org" <gcc@gcc.gnu.org>
Subject: Re: Turning off unrolling to certain loops
Date: Thu, 15 Oct 2009 14:41:00 -0000 [thread overview]
Message-ID: <20091015100020.GA21790@kam.mff.cuni.cz> (raw)
In-Reply-To: <7FB04A5C213E9943A72EE127DB74F0AD93CCBE4AD6@SJEXCHCCR02.corp.ad.broadcom.com>
Hi,
> I faced a similar issue a while ago. I filed a bug report
> (http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36712) In the end,
> I implemented a simple tree-level unrolling pass in our port
> which uses all the existing infrastructure. It works quite well for
> our purpose, but I hesitated to submit the patch because it contains
> our not-very-elegannt #prgama unroll implementation.
could you please do so anyway? Even if there are some issues with the
#prgama unroll implementation, it could serve as a basis of a usable
patch.
> /* Perfect unrolling of a loop */
> static void tree_unroll_perfect_loop (struct loop *loop, unsigned factor,
> edge exit)
> {
> ...
> }
>
>
>
> /* Go through all the loops:
> 1. Determine unrolling factor
> 2. Unroll loops in different conditions
> -- perfect loop: no extra copy of original loop
> -- other loops: the original version of loops to execute the remaining iterations
> */
> static unsigned int rest_of_tree_unroll (void)
> {
...
> tree niters = number_of_exit_cond_executions(loop);
>
> bool perfect_unrolling = false;
> if(niters != NULL_TREE && niters!= chrec_dont_know && TREE_CODE(niters) == INTEGER_CST){
> int num_iters = tree_low_cst(niters, 1);
> if((num_iters % unroll_factor) == 0)
> perfect_unrolling = true;
> }
>
> /* If no. of iterations can be divided by unrolling factor, we have perfect unrolling */
> if(perfect_unrolling){
> tree_unroll_perfect_loop(loop, unroll_factor, single_dom_exit(loop));
> }
> else{
> tree_unroll_loop (loop, unroll_factor, single_dom_exit (loop), &desc);
> }
It would be better to move this test to tree_unroll_loop, and not
duplicate its code in tree_unroll_perfect_loop.
Zdenek
next prev parent reply other threads:[~2009-10-15 10:00 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-10-05 20:46 Jean Christophe Beyler
2009-10-06 6:54 ` Zdenek Dvorak
2009-10-06 13:34 ` Jean Christophe Beyler
[not found] ` <20091006135624.GA18714@kam.mff.cuni.cz>
[not found] ` <c568a2600910060756m96adf86mb1507c6717fdfdb6@mail.gmail.com>
[not found] ` <20091006150918.GA19277@kam.mff.cuni.cz>
[not found] ` <c568a2600910060828l24ec5ecct926eb2624a1d2157@mail.gmail.com>
2009-10-08 16:18 ` Jean Christophe Beyler
2009-10-08 16:23 ` Zdenek Dvorak
2009-10-08 18:52 ` Jean Christophe Beyler
2009-10-14 18:56 ` Jean Christophe Beyler
2009-10-15 0:45 ` Zdenek Dvorak
2009-10-15 10:00 ` Bingfeng Mei
2009-10-15 14:41 ` Zdenek Dvorak [this message]
2009-10-15 15:56 ` Jean Christophe Beyler
2009-10-15 16:27 ` Bingfeng Mei
2009-10-15 17:16 ` Jean Christophe Beyler
2009-10-16 12:17 ` Bingfeng Mei
2009-10-16 13:51 ` Jean Christophe Beyler
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=20091015100020.GA21790@kam.mff.cuni.cz \
--to=rakdver@kam.mff.cuni.cz \
--cc=bmei@broadcom.com \
--cc=gcc@gcc.gnu.org \
--cc=jean.christophe.beyler@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).