From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 48) id 4D6C338582B6; Wed, 6 Mar 2024 09:12:08 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 4D6C338582B6 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org; s=default; t=1709716328; bh=XsB0XpgUdRmUtcaTTftzuEWIcz01Yk1/nTdptDBCycs=; h=From:To:Subject:Date:In-Reply-To:References:From; b=jT7yYukRIxn1CQijC2hT1+iT8bDNI6mJRTlKGb91fv1wZ6jEBcbRHHenS+8wjea8Y 5Y/4IKaoHsniW8YNwugXQ6UWnHt566BmX841aw3vZUgiC537Mp87t6uzt7JVsPSuX4 8ylCKEjOTBml1Iw/7inHKndiPVPF9uSKaoOQBE/U= From: "chenglulu at loongson dot cn" To: gcc-bugs@gcc.gnu.org Subject: [Bug target/112919] LoongArch: Alignments in tune parameters are not precise and they regress performance Date: Wed, 06 Mar 2024 09:12:06 +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: 14.0 X-Bugzilla-Keywords: X-Bugzilla-Severity: normal X-Bugzilla-Who: chenglulu at loongson dot cn 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 List-Id: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D112919 --- Comment #14 from chenglulu --- (In reply to chenglulu from comment #13) > (In reply to Xi Ruoyao from comment #9) > > (In reply to chenglulu from comment #8) > > > (In reply to Xi Ruoyao from comment #7) > > > > Any update? :) > > >=20 > > > Well, I haven't run it yet. Since this does not have a big impact on = the > > > spec score, I am currently testing it on a single-channel machine, so= the > > > test time will be longer. > > > I will reply here as soon as the results are available. > >=20 > > Can we determine on LA664 if the current default alignment is better th= an > > not aligning at all? Coremarks results suggest the current default is = even > > worse than not aligning, but arguably Coremarks is far different from r= eal > > workloads. However if the current default is not better than not aligni= ng > > (or the difference is only marginal and is likely covered up by some ra= ndom > > fluctuation) we can disable the aligning for LA664. > >=20 > > (Maybe we and the HW engineers have done some repetitive work or even s= ome > > work cancelling each other out :(. ) >=20 > The results of spec2006 on 3A6000 were obtained, I removed the more volat= ile > test items, '-falign-loops=3D8 -falign-functions=3D8 -falign-jumps=3D32 > -falign-lables=3D4' this set of parameters got the highest score. This is= the > same combination of parameters as the coremark tested by Xu Chenghua. Hi=EF=BC=8CRuoyao: The test of the 3a5000 will also be completed around the 15th of this month= , so I want to change the code after the test results of the 3a5000 are out. What do you think?(In reply to Xi Ruoyao from comment #9) > (In reply to chenglulu from comment #8) > > (In reply to Xi Ruoyao from comment #7) > > > Any update? :) > >=20 > > Well, I haven't run it yet. Since this does not have a big impact on the > > spec score, I am currently testing it on a single-channel machine, so t= he > > test time will be longer. > > I will reply here as soon as the results are available. >=20 > Can we determine on LA664 if the current default alignment is better than > not aligning at all? Coremarks results suggest the current default is ev= en > worse than not aligning, but arguably Coremarks is far different from real > workloads. However if the current default is not better than not aligning > (or the difference is only marginal and is likely covered up by some rand= om > fluctuation) we can disable the aligning for LA664. Hi=EF=BC=8CRuoyao: The results of spec2006 on 3A6000 were obtained, I removed the more volati= le test items, '-falign-loops=3D8 -falign-functions=3D8 -falign-jumps=3D32 -falign-lables=3D4' this set of parameters got the highest score. This is t= he same combination of parameters as the coremark tested by Xu Chenghua. The test of the 3A5000 will also be completed around the 15th of this month= , so I want to change the code after the test results of the 3a5000 are out. What do you think?=