From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 48) id E3DFD3857C63; Wed, 19 Jan 2022 08:48:44 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org E3DFD3857C63 From: "rguenther at suse dot de" To: gcc-bugs@gcc.gnu.org Subject: [Bug target/103771] [12 Regression] Missed vectorization under -mavx512f -mavx512vl after r12-5489 Date: Wed, 19 Jan 2022 08:48:44 +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: 12.0 X-Bugzilla-Keywords: missed-optimization X-Bugzilla-Severity: normal X-Bugzilla-Who: rguenther at suse dot de 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: 12.0 X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: Message-ID: In-Reply-To: References: 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 X-BeenThere: gcc-bugs@gcc.gnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Gcc-bugs mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 19 Jan 2022 08:48:45 -0000 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D103771 --- Comment #30 from rguenther at suse dot de --- On Wed, 19 Jan 2022, crazylht at gmail dot com wrote: > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D103771 >=20 > --- Comment #29 from Hongtao.liu --- > (In reply to Richard Biener from comment #28) > > (In reply to Hongtao.liu from comment #25) > > > in fold_unary_loc > > > ---cut from fold-const.cc----- > > > 9276 else if (TREE_CODE (arg0) =3D=3D COND_EXPR) > > > 9277 { > > > 9278 tree arg01 =3D TREE_OPERAND (arg0, 1); > > > 9279 tree arg02 =3D TREE_OPERAND (arg0, 2); > > > 9280 if (! VOID_TYPE_P (TREE_TYPE (arg01))) > > > 9281 arg01 =3D fold_build1_loc (loc, code, type, > > > 9282 fold_convert_loc (loc, > > > 9283 TREE_TYPE (op= 0), > > > arg01)); > > > 9284 if (! VOID_TYPE_P (TREE_TYPE (arg02))) > > > 9285 arg02 =3D fold_build1_loc (loc, code, type, > > > 9286 fold_convert_loc (loc, > > > 9287 TREE_TYPE (op= 0), > > > arg02)); > > > 9288=3D> tem =3D fold_build3_loc (loc, COND_EXPR, type, TREE_= OPERAND > > > (arg0, 0), > > > 9289 arg01, arg02); > > >=20 > > > -----------end--------------- > > >=20 > > > gcc always tries to simplify (convert (cond (cmp a b) c d) ---- > (co= nd (cmp > > > a b) (convert c) (convert d))=EF=BC=8C exactly the opposite of what t= his case wants. > >=20 > > It also then undos this if the result didn't simplify and plays trick to > > avoid > > recursions. > >=20 > > I think this particular transform ought to be specialized, maybe to > > (T)p?(T')a:(T')b or maybe done during gimplification or RTL expansion o= nly. > >=20 > > The "cheap" way of avoiding a conflict is to wrap the match.pd pattern > > with opposite logic in > >=20 > > #if GIMPLE > > #endif > >=20 > It doesn't work,=20 > > (with a comment explaining this) > >=20 > > Note that we can move a conversion out only if the sources of the conve= rsions > > have compatible types but we always can move a conversion in. > >=20 > > Alternatively this transform can also be done in a vectorizer pattern b= ased > > on vector compatibility of the ?: predicate with the data. > yes, I'm thinking of doing this in fold_build_cond_expr which is only use= d by > pass_ifcvt to generate cond_expr. That's also a reasonable place but the vectorizer pattern recog phase might have more contextual information to determine the best types to use. fold_build_cond_expr is probably easiest to adjust though.=