From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 48) id D0E623858412; Wed, 17 Nov 2021 13:03:05 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org D0E623858412 From: "hubicka at gcc dot gnu.org" To: gcc-bugs@gcc.gnu.org Subject: [Bug ipa/103246] [12 Regression] 416.gamess miscompare with -O2 -g -flto=auto since r12-5223-gecdf414bd89e6ba251f6b3f494407139b4dbae0e Date: Wed, 17 Nov 2021 13:03:05 +0000 X-Bugzilla-Reason: CC X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: gcc X-Bugzilla-Component: ipa X-Bugzilla-Version: 12.0 X-Bugzilla-Keywords: lto, wrong-code X-Bugzilla-Severity: normal X-Bugzilla-Who: hubicka at gcc dot gnu.org X-Bugzilla-Status: NEW X-Bugzilla-Resolution: X-Bugzilla-Priority: P1 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, 17 Nov 2021 13:03:05 -0000 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D103246 --- Comment #16 from Jan Hubicka --- Building with LTO only ecpder.fppized.o and ecplib.fppized.o I get the fail= ure going away with --dbg-cnt=3Dipa_cp_values:10 and :9 works. Different ipa-cp decision is +***dbgcnt: upper limit 10 reached for ipa_cp_values.*** + - Creating a specialized node of formdr/4 for all known contexts. + replacing param #5 int & restrict with const &ecpidx.iamin + replacing param #6 int & restrict with const &ecpidx.iamax + Removed a reference from formii.constprop/284 to ecpidx/11. + ...and replaced it with LOAD one. + Removed a reference from ecp2d/1 to ecpidx/11. + Removed a reference from ecp1d/0 to ecpidx/11. + replacing param #9 logical(kind=3D4) & restrict with const &ecpidx.nor= m - forcing load reference In modref the propagation goes same way and we update signature as follows: +Updating summary for formdr.constprop/285 from: + loads: + Limits: 32 bases, 16 refs + Every base + stores: + Limits: 32 bases, 16 refs + Every base + Side effects + Nondeterministic + parm 0 flags: no_direct_clobber no_direct_escape no_indirect_escape + parm 1 flags: no_direct_clobber no_direct_escape no_indirect_escape + parm 2 flags: no_direct_escape no_indirect_escape + parm 3 flags: no_direct_escape no_indirect_escape + parm 4 flags: no_direct_escape no_indirect_escape + parm 5 flags: no_direct_clobber no_indirect_clobber no_direct_escape no_indirect_escape no_indirect_read + parm 6 flags: no_direct_clobber no_indirect_clobber no_direct_escape no_indirect_escape no_indirect_read + parm 7 flags: no_direct_clobber no_indirect_clobber no_direct_escape no_indirect_escape no_indirect_read + parm 8 flags: no_direct_clobber no_indirect_clobber no_direct_escape no_indirect_escape no_indirect_read + parm 9 flags: no_direct_clobber no_indirect_clobber no_direct_escape no_indirect_escape no_indirect_read +to: + loads: + Limits: 32 bases, 16 refs + Every base + stores: + Limits: 32 bases, 16 refs + Every base + Side effects + Nondeterministic + parm 0 flags: no_direct_clobber no_direct_escape no_indirect_escape + parm 1 flags: no_direct_clobber no_direct_escape no_indirect_escape + parm 2 flags: no_direct_escape no_indirect_escape + parm 3 flags: no_direct_escape no_indirect_escape + parm 4 flags: no_direct_escape no_indirect_escape + parm 5 flags: no_direct_clobber no_indirect_clobber no_direct_escape no_indirect_escape no_indirect_read + parm 6 flags: no_direct_clobber no_indirect_clobber no_direct_escape no_indirect_escape no_indirect_read this changes partitioning decisions which makes quite a lot of fuzz in latr dump files.=