From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 48) id 367813839C45; Wed, 23 Jun 2021 18:28:49 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 367813839C45 From: "vmakarov at gcc dot gnu.org" To: gcc-bugs@gcc.gnu.org Subject: [Bug rtl-optimization/100328] IRA doesn't model matching constraint well Date: Wed, 23 Jun 2021 18:28:48 +0000 X-Bugzilla-Reason: CC X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: gcc X-Bugzilla-Component: rtl-optimization X-Bugzilla-Version: 12.0 X-Bugzilla-Keywords: X-Bugzilla-Severity: normal X-Bugzilla-Who: vmakarov at gcc dot gnu.org X-Bugzilla-Status: UNCONFIRMED 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: 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, 23 Jun 2021 18:28:49 -0000 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D100328 --- Comment #2 from Vladimir Makarov --- (In reply to Kewen Lin from comment #1) > Created attachment 50715 [details] > ira:consider matching cstr in all alternatives >=20 > With little understanding on ira, I am not quite sure this patch is on the > reasonable direction. It aims to check the matching constraint in all > alternatives, if there is one alternative with matching constraint and > matches the current preferred regclass, it will record the output operand > number and further create one copy for it. Normally it can get the priori= ty > against shuffle copies and the matching constraint will get satisfied with > higher possibility, reload doesn't create extra copies to meet the matchi= ng > constraint or the desirable register class when it has to. >=20 > For FMA A,B,C,D, I think ideally copies A/B, A/C, A/D can firstly stay as > shuffle copies, and later any of A,B,C,D gets assigned by one hardware > register which is a VSX register but not a FP register, which means it has > to pay costs once we can NOT go with VSX alternatives, so at that time we > can increase the freq for the remaining copies related to this, once the > matching constraint gets satisfied further, there aren't any extra costs = to > pay. This idea seems a bit complicated in the current framework, so the > proposed patch aggressively emphasizes the matching constraint at the time > of creating copies. >=20 > FWIW bootstrapped/regtested on powerpc64le-linux-gnu P9. The evaluation w= ith > Power9 SPEC2017 all run shows 505.mcf_r +2.98%, 508.namd_r +3.37%, 519.lb= m_r > +2.51%, no remarkable degradation is observed. Thank you for working on this issue. The current implementation of ira_get_dup_out_num was specifically tuned for better register allocation for x86-64 div insns. Your patch definitely improves code for power9 and I would say significantly (congratulations!). The patch you proposed makes me think that it might wo= rk for major targets as well. I would prefer to avoid introducing new parameter because there are too man= y of them already and its description is cryptic. It would be nice if you benchmark the patch on x86-64 too, If there is no overall degradation with new behaviour we could remove the parameter and ma= ke the new behaviour as a default. If it is not, well we will keep the paramet= er. As for the patch itself, I don't like some variable names. Sorry. Could y= ou use op_regno, out_regno, and present_alt instead of op_no, out_no, tot.=20 Please, in general use longer variable names reflecting their purpose as GCC developers reads code in many times more than writing it.=