From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugzilla@gcc.gnu.org>
Received: by sourceware.org (Postfix, from userid 48)
	id DD80C3858418; Sat,  3 Sep 2022 16:08:48 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org DD80C3858418
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1662221328;
	bh=EoPNL+XxyzUWrB4YLIjmfKWYIXY34YpJH7n/DWdGi+E=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=XJXVUZR0IaLsDJuQeRmSP340VG+KOb4MmmJP9Vvas+ifjm0wVV5wa3RqbXnu+AKWG
	 7f8qZj9sENwi5eduYb+7ZveyH71KEFkmgIt0b4qD7haHuWshuBu6j93G2JgaT9BDEE
	 3XxfVe4ltYXYzcewwaTdklOuc6YyWqZglA2GQhnI=
From: "aldyh at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/91645] Missed optimization with sqrt(x*x)
Date: Sat, 03 Sep 2022 16:08:48 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: tree-optimization
X-Bugzilla-Version: 9.2.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: normal
X-Bugzilla-Who: aldyh at gcc dot gnu.org
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: cc
Message-ID: <bug-91645-4-R6zaNRdV2q@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-91645-4@http.gcc.gnu.org/bugzilla/>
References: <bug-91645-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
List-Id: <gcc-bugs.sourceware.org>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D91645

Aldy Hernandez <aldyh at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |aldyh at gcc dot gnu.org,
                   |                            |amacleod at redhat dot com,
                   |                            |jakub at gcc dot gnu.org,
                   |                            |law at gcc dot gnu.org

--- Comment #7 from Aldy Hernandez <aldyh at gcc dot gnu.org> ---
> Now, the problem is that GCC doesn't seem to optimize away the call to sq=
rtf
> based on some surrounding code. As an example, it would be neat to have t=
his
> (or something similar) to get compiled into the same mulss-sqrtss-ret:
>=20
> float test (float x)=20
> {
>     float y =3D x*x;
>     if (y >=3D 0.f)
>         return std::sqrt(y);
>     __builtin_unreachable();
> }
>=20
> If I understand it correctly, the 'y >=3D 0.f' excludes 'y' being NaN and=
 'y'
> being negative (though this is excluded by 'y =3D x*x'), so there is no n=
eed
> to check if the argument to `std::sqrt` is any bad, enabling to just do
> 'sqrtss' and return.

I'm not an FP expert, but I think we have enough information to do this rig=
ht
now.  The evrp dump now has:

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D BB 2 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
Imports: y_2=20=20
Exports: y_2=20=20
    <bb 2> :
    y_2 =3D x_1(D) * x_1(D);
    if (y_2 >=3D 0.0)
      goto <bb 3>; [INV]
    else
      goto <bb 4>; [INV]

2->3  (T) y_2 :         [frange] float [0.0,  Inf] !NAN=20
2->4  (F) y_2 :         [frange] float [ -Inf, 0.0]=20


=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D BB 3 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
y_2     [frange] float [0.0,  Inf] !NAN=20
    <bb 3> :
    _6 =3D __builtin_sqrtf (y_2);
    return _6;

Which means that y_2 is known to be [0.0, Inf] excluding a NAN.

What needs to happen for the call to __builtin_sqrtf to be optimized to sqr=
tss?=