From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 48) id A74273858C41; Mon, 6 Nov 2023 13:35:53 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org A74273858C41 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=sourceware.org; s=default; t=1699277753; bh=kTiV1AG2yqJjosPzbNM2ZnOUeGSSe6/kqJ3Mvp0lOhM=; h=From:To:Subject:Date:In-Reply-To:References:From; b=YCeV3wTKoPVcNdIgbWAONPjjDabj6ZFPubQD+5VnPgnDJStlLiRQ0UVl4jZNsL9pU +TYkzff84Q8CAFXqa6UIxbrmKT22mludiBKjLCsbjwb0vLAJIKDEr44pr8EWqUG4/L M/BusCwSASjXoXOdHCBdgJqjuADGugVGpodDekwA= From: "adhemerval.zanella at linaro dot org" To: glibc-bugs@sourceware.org Subject: [Bug libc/28310] Do not use affinity mask for sysconf (_SC_NPROCESSORS_CONF) Date: Mon, 06 Nov 2023 13:35:52 +0000 X-Bugzilla-Reason: CC X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: glibc X-Bugzilla-Component: libc X-Bugzilla-Version: unspecified X-Bugzilla-Keywords: X-Bugzilla-Severity: normal X-Bugzilla-Who: adhemerval.zanella at linaro dot org X-Bugzilla-Status: RESOLVED X-Bugzilla-Resolution: FIXED X-Bugzilla-Priority: P2 X-Bugzilla-Assigned-To: adhemerval.zanella at linaro dot org X-Bugzilla-Target-Milestone: 2.35 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://sourceware.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 List-Id: https://sourceware.org/bugzilla/show_bug.cgi?id=3D28310 --- Comment #3 from Adhemerval Zanella --- I proposed a fix [1], that now DJ has withdraw his dissent. For old re [1] https://patchwork.sourceware.org/project/glibc/patch/20231011164356.2218554= -1-adhemerval.zanella@linaro.org/(In reply to Rajalakshmi from comment #2) > Hi Adhemerval >=20 > Commit 342298278eabc75baabcaced110a11a02c3d3580 reverted the use=C2=A0of > sched_getaffinity=C2=A0on=C2=A0get_nprocs() > However, in commit 11a02b035b464ab6813676adfd19c4a59c36d907 we still use > __get_nprocs_sched=C2=A0() instead of __get_nprocs() in malloc/arena.c=20 >=20 > There is a performance degradation issue reported on POWER10 RHEL9(glibc > 2.34) compared to RHEL 8 *(glibc 2.28) when using IBM ESSL application w= ith > 80 omp threads. While debugging , figured out that using __get_nprocs() > instead of __get_nprocs_sched=C2=A0() in malloc/arena.c fixes the issue. = I don=E2=80=99t > have reduced test to replicate the issue.=20 >=20 > Do you have any comments/suggestions? I proposed a fix [1], that now DJ has withdraw his dissent. I will prob ins= tall this this week, and as mitigation you can tune the maximum number of arenas (GLIBC_TUNABLES=3Dglibc.malloc.arena_max=3Dnumber). [1] https://patchwork.sourceware.org/project/glibc/patch/20231011164356.2218554= -1-adhemerval.zanella@linaro.org/ --=20 You are receiving this mail because: You are on the CC list for the bug.=