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 AF138384F489; Fri, 18 Nov 2022 11:46:40 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org AF138384F489
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1668772000;
	bh=0vfDHZPth+2S8GDNjowYfDj0dubWjVdUtq5Q1u926YQ=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=NHKjLXukUTo9Haqi/LJaACxD1cTxU7FIRc2X2EDd403ZjQ7Rf7j8Rbzj9iS3nJW3i
	 YDyBi5omIfsPdkpnkkyX53CTnwvMhFiQSTRKkJze+RAkJcmOKJhdfxF/uUElbG7X0U
	 3LN0Eb2tBEnSMnenGyeCE2ftzBNVJJM1XlG1gB8w=
From: "jakub at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/107748] [13 Regression] Isn't _mm_cvtsbh_ss incorrect?
Date: Fri, 18 Nov 2022 11:46:40 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: target
X-Bugzilla-Version: 13.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: jakub at gcc dot gnu.org
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P1
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: 13.0
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-107748-4-BwAk8vg0CF@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-107748-4@http.gcc.gnu.org/bugzilla/>
References: <bug-107748-4@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=3D107748

--- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
(In reply to Hongtao.liu from comment #2)
> float
> _mm_cvtsbh_ss (__bf16 __A)
> {
>   union{ float sf; __bf16 bf[2];} __tmp;
>   __tmp.sf =3D 0.0f;
>   __tmp.bf[1] =3D __A;
>   return __tmp.sf;
> }
>=20
> Looks like gcc can optimize it to
>=20
> _mm_cvtsbh_ss(bool _Accum):
>         movd    %xmm0, %eax
>         sall    $16, %eax
>         movd    %eax, %xmm0
>         ret

That is an option too, but please uglify with __ the sf and bf identifiers
above.
Also, not just for this but more importantly for the __bf16 -> float
conversions
gcc emits for -ffast-math or for cstorebf4 or cbranchcc4, it would be nice =
if
we optimized those so that if the source and destination are in SSE registe=
rs
that we don't convert from SSE to GPR, shift and convert back from GPR to S=
SE,
while we could do it through some permutation of the SSE register that just
pretends it is a V*HImode and moves the first element to second and zeros t=
he
first (and perhaps all elements above second too, or not, whatever is faste=
r).
Dunno if it could be done as a peephole2, or something different.
Just try:
__attribute__((optimize ("fast-math")))
float foo (__bf16 x) { return x; }
int bar (__bf16 x, __bf16 y) { return x =3D=3D y; }
void baz (void);
void qux (__bf16 x, __bf16 y) { if (x =3D=3D y) baz (); }
Oh, and one more thing, for -mavx512bf16 -mavx512vl -ffast-math it would be
nice
to use the AVX512BF16 instruction for float -> __bf16 conversions rather th=
an
library routine.  But that instruction doesn't handle sNaNs properly and
flushes subnormals to 0, so I think we shouldn't do it if HONORS_NANS (BFmo=
de)
or
!flag_unsafe_math_optimizations.=