public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "aldyh at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/24021] VRP does not work with floating points
Date: Mon, 01 Aug 2022 13:27:04 +0000	[thread overview]
Message-ID: <bug-24021-4-lDrcTs4Xgk@http.gcc.gnu.org/bugzilla/> (raw)
In-Reply-To: <bug-24021-4@http.gcc.gnu.org/bugzilla/>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=24021

--- Comment #25 from Aldy Hernandez <aldyh at gcc dot gnu.org> ---
Adding some notes here as I work through this PR...

Even with floating aware VRP, we won't be able to do much because SCEV (which
ranger and VRP use) does not work with non-integers.

At EVRP time we see:

double BG_SplineLength ()
{
  double i;
  double lastPoint;

  <bb 2> :
  goto <bb 6>; [INV]

  <bb 3> :
  if (i_3 == 0.0)
    goto <bb 5>; [INV]
  else
    goto <bb 4>; [INV]

  <bb 4> :

  <bb 5> :
  # lastPoint_1 = PHI <i_3(3), 2.0e+0(4)>
  i_10 = i_3 + 1.00000001490116119384765625e-1;

  <bb 6> :
  # lastPoint_2 = PHI <lastPoint_5(D)(2), lastPoint_1(5)>
  # i_3 = PHI
<1.00000000000000002081668171172168513294309377670288085938e-2(2), i_10(5)>
  if (i_3 <= 1.0e+0)
    goto <bb 3>; [INV]
  else
    goto <bb 7>; [INV]

  <bb 7> :
  return lastPoint_2;

}

On the 6->3 edge we know that i_3 is [-INF, 1.0], so even adding a range-op
entry for the PLUS_EXPR, we'd know that i_3 is [-INF, 1.1], which is not enough
to fold the i_3 == 0.0 conditional.  If you convert this testcase to integer
and turn off SCEV (and unrolling and FRE, etc), you'll notice that VRP doesn't
do much here (in either legacy nor ranger mode).

We would need SCEV to give us [1.0, 1.1] in order to fold the i_3 == 0.0
conditional.  For that matter, if I hack SCEV to give us the expected end
points, I can get evrp to fold the conditional.  I think once VRP is FP aware,
we can close this PR because this particular testcase is SCEV specific.  So for
this particular testcase, perhaps we could open a SCEV+FP PR?

As an aside, the second conditional will not be folded by VRP (legacy or
ranger), even with SCEV, even if you convert the above testcase to
integer...it's either gotten later by unrolling+FRE for FP, or cddce for
integers.

What I'm trying to say is that even with FP VRP, which we'll have shortly, the
first conditional won't be folded because SCEV doesn't do floats, and the
second one won't be, because it's not VRP's job.

  parent reply	other threads:[~2022-08-01 13:27 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <bug-24021-4@http.gcc.gnu.org/bugzilla/>
2021-08-25 20:29 ` amacleod at redhat dot com
2022-02-07 16:07 ` aldyh at gcc dot gnu.org
2022-02-09 16:35 ` aldyh at gcc dot gnu.org
2022-02-09 17:04 ` law at gcc dot gnu.org
2022-02-09 17:27 ` aldyh at gcc dot gnu.org
2022-02-09 17:29 ` vincent-gcc at vinc17 dot net
2022-02-09 17:31 ` vincent-gcc at vinc17 dot net
2022-02-09 18:30 ` jakub at gcc dot gnu.org
2022-02-09 18:31 ` amacleod at redhat dot com
2022-02-09 18:35 ` amacleod at redhat dot com
2022-02-09 19:05 ` vincent-gcc at vinc17 dot net
2022-02-10  9:33 ` rguenth at gcc dot gnu.org
2022-02-10 12:02 ` jakub at gcc dot gnu.org
2022-02-10 12:21 ` rguenther at suse dot de
2022-02-10 14:17 ` vincent-gcc at vinc17 dot net
2022-04-05 19:34 ` amacleod at redhat dot com
2022-04-06  9:34 ` aldyh at gcc dot gnu.org
2022-04-06  9:46 ` jakub at gcc dot gnu.org
2022-04-06  9:57 ` aldyh at gcc dot gnu.org
2022-04-06 10:30 ` aldyh at gcc dot gnu.org
2022-04-06 10:47 ` jakub at gcc dot gnu.org
2022-08-01 13:27 ` aldyh at gcc dot gnu.org [this message]
2022-08-01 14:30 ` aldyh at gcc dot gnu.org
2022-11-08 15:54 ` cvs-commit at gcc dot gnu.org
2022-11-08 15:57 ` aldyh at gcc dot gnu.org
2022-11-28 22:15 ` pinskia at gcc dot gnu.org
     [not found] <bug-24021-6528@http.gcc.gnu.org/bugzilla/>
2005-10-23 23:43 ` pinskia at gcc dot gnu dot org

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=bug-24021-4-lDrcTs4Xgk@http.gcc.gnu.org/bugzilla/ \
    --to=gcc-bugzilla@gcc.gnu.org \
    --cc=gcc-bugs@gcc.gnu.org \
    /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).