From nobody Tue Dec 9 03:43:37 2025 Received: from mail-pl1-f201.google.com (mail-pl1-f201.google.com [209.85.214.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 2880230C63B for ; Thu, 13 Nov 2025 03:21:31 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.214.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1763004092; cv=none; b=n8ZNlaBwvFQxpgzjSvGeKio95eUieVF5Rq9CTv5iG0FjgAHwaLDePGH5xaeb6VVSk8TQT3kqYnuC76B56Zhq5Qh6pFqa1LvJ7aYIAS8Sq9nBtUtlDEVEmPUefe0lU0nBk4l6UqZ8xZCct6ytZ6NDPtb3fdMWoBaoC+VG1Mzf3t0= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1763004092; c=relaxed/simple; bh=H32Kmb4s4EUbi5GaQqYvBoOMfwS8An4o87KjgWVGEGk=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Content-Type; b=Qdw7oqMyo2RRSvI4qujdem3Y1N+zWmj1cEJNmbGw/pxTnC4hng0ymIyXlIu18QDBrmfcUKXd84jSRmAOdU9mRZXTC60itPOw+z470i+EXMhISN4DKnxnTo6UTbsWMN2FwrMkr8fSyf+ngSzIEsqr8EwUvfaInHXOY6OW81Fmd7U= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--irogers.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=k/kISEgw; arc=none smtp.client-ip=209.85.214.201 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--irogers.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="k/kISEgw" Received: by mail-pl1-f201.google.com with SMTP id d9443c01a7336-2956f09f382so3586975ad.1 for ; Wed, 12 Nov 2025 19:21:30 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1763004090; x=1763608890; darn=vger.kernel.org; h=to:from:subject:message-id:references:mime-version:in-reply-to:date :from:to:cc:subject:date:message-id:reply-to; bh=Q/xav7sGweATI1gLemlCd0Y/xoi8YTcN9Imv7Ww6cug=; b=k/kISEgwtPlECbgBwAeO6OesdCKOoRq4bygVdXTItN5ZJHU9IXxTi1UouxleTgXiHG f43UKQHU9aY8nMHhJRkrWtwkTniE/lOppzWVsHMs939MPdcrk413ayYaKXdehLUZFcPo +oyrlfgo67t1pNDULvLecOAMxTvW8YE9OkTn3Ha9m6NV+pV/2qdG2FI0zFxMdLWB6GVm 67qoRDeZcO2jAO/ohGOzUEIkGVd8JRRXOt+unWXgglQKTwE497VUy3uPGGhnd/BCDqZ+ DvWbNF0TDqPWuHeSuX+URIeHu+ejINXcHxO9ng+d0bb5WyFVsrDsbHZnUY4u/41MWoDZ 0lHg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1763004090; x=1763608890; h=to:from:subject:message-id:references:mime-version:in-reply-to:date :x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=Q/xav7sGweATI1gLemlCd0Y/xoi8YTcN9Imv7Ww6cug=; b=PlTJ7RRmX3NOAU946mhuIrcr8EvYvHLOr2Cv05DS7mTQqTW2mNxCPErQQzjD1LHkBI qKmnAj2JIzV54/LfnH3FHOuwVuHgdPhxbsDO2WyGhNNgKQi3EnYOqVGCulP6EE+w5PCP ExM4VfK7LM9dIn2hwH/Vxmatu/pEk1dlgA051Kv5BWvgLmgfoLWi0f7srEOYfTqAWGZ6 LvmUk6Q6ClSyNys5YPxdTBccWTnE+w7dxJTGBPMgCw+b8wPrPEXVFQZBXENtM8Ao6fgD dLqd9t7Bah6SNe8tr9+/2speGzJyiwbYXotSfRXdbuPZYGmGS5RmxynVws6oQEwONMzz n3UQ== X-Forwarded-Encrypted: i=1; AJvYcCVV71x3KzLFZgDYv9z98y3nGxjcI3RjebnXJiKDDBcNMhGo15prc+JEoSiQymzEXkcm5e2C7WJvUCx0Xzg=@vger.kernel.org X-Gm-Message-State: AOJu0YwDf1JKLjBwXT5MEb9J7BgfhtjBLtRppfTbmaCUwdhUMcNhrGhz z9zu8fgvGPkyBer7t/32d8Ku+POYITwat6mkdisVYzJcViLEd7OWcJZeuWutXuFmDGyEqMipc4R jojxbDk9Llw== X-Google-Smtp-Source: AGHT+IFVI3FgTE1LB1tkcrGzvX8YL/V7HJxOSDGDYv3mvmQNq20sUUzAhDgaaL0APymsq+Uiq0+1iYBcv0e2 X-Received: from dlbdz8.prod.google.com ([2002:a05:7022:ec8:b0:119:b185:ea72]) (user=irogers job=prod-delivery.src-stubby-dispatcher) by 2002:a17:902:f60c:b0:297:ec44:56f with SMTP id d9443c01a7336-2985a52b15amr20494665ad.14.1763004090460; Wed, 12 Nov 2025 19:21:30 -0800 (PST) Date: Wed, 12 Nov 2025 19:20:07 -0800 In-Reply-To: <20251113032040.1994090-1-irogers@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20251113032040.1994090-1-irogers@google.com> X-Mailer: git-send-email 2.51.2.1041.gc1ab5b90ca-goog Message-ID: <20251113032040.1994090-20-irogers@google.com> Subject: [PATCH v8 19/52] perf jevents: Add software prefetch (swpf) metric group for AMD From: Ian Rogers To: Adrian Hunter , Alexander Shishkin , Arnaldo Carvalho de Melo , Benjamin Gray , Caleb Biggers , Edward Baker , Ian Rogers , Ingo Molnar , James Clark , Jing Zhang , Jiri Olsa , John Garry , Leo Yan , Namhyung Kim , Perry Taylor , Peter Zijlstra , Samantha Alt , Sandipan Das , Thomas Falcon , Weilin Wang , Xu Yang , linux-kernel@vger.kernel.org, linux-perf-users@vger.kernel.org Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Add metrics that give the utility of software prefetches on zen2, zen3 and zen4. Signed-off-by: Ian Rogers --- tools/perf/pmu-events/amd_metrics.py | 101 +++++++++++++++++++++++++++ 1 file changed, 101 insertions(+) diff --git a/tools/perf/pmu-events/amd_metrics.py b/tools/perf/pmu-events/a= md_metrics.py index 1880ccf9c6fc..06cb56cbd617 100755 --- a/tools/perf/pmu-events/amd_metrics.py +++ b/tools/perf/pmu-events/amd_metrics.py @@ -121,6 +121,106 @@ def AmdBr(): description=3D"breakdown of retired branch instruct= ions") =20 =20 +def AmdSwpf() -> Optional[MetricGroup]: + """Returns a MetricGroup representing AMD software prefetch metrics.""" + global _zen_model + if _zen_model <=3D 1: + return None + + swp_ld =3D Event("ls_dispatch.ld_dispatch") + swp_t0 =3D Event("ls_pref_instr_disp.prefetch") + swp_w =3D Event("ls_pref_instr_disp.prefetch_w") # Missing on Zen1 + swp_nt =3D Event("ls_pref_instr_disp.prefetch_nta") + swp_mab =3D Event("ls_inef_sw_pref.mab_mch_cnt") + swp_l2 =3D Event("ls_sw_pf_dc_fills.local_l2", + "ls_sw_pf_dc_fills.lcl_l2", + "ls_sw_pf_dc_fill.ls_mabresp_lcl_l2") + swp_lc =3D Event("ls_sw_pf_dc_fills.local_ccx", + "ls_sw_pf_dc_fills.int_cache", + "ls_sw_pf_dc_fill.ls_mabresp_lcl_cache") + swp_lm =3D Event("ls_sw_pf_dc_fills.dram_io_near", + "ls_sw_pf_dc_fills.mem_io_local", + "ls_sw_pf_dc_fill.ls_mabresp_lcl_dram") + swp_rc =3D Event("ls_sw_pf_dc_fills.far_cache", + "ls_sw_pf_dc_fills.ext_cache_remote", + "ls_sw_pf_dc_fill.ls_mabresp_rmt_cache") + swp_rm =3D Event("ls_sw_pf_dc_fills.dram_io_far", + "ls_sw_pf_dc_fills.mem_io_remote", + "ls_sw_pf_dc_fill.ls_mabresp_rmt_dram") + + # All the swpf that were satisfied beyond L1D are good. + all_pf =3D swp_t0 + swp_w + swp_nt + good_pf =3D swp_l2 + swp_lc + swp_lm + swp_rc + swp_rm + bad_pf =3D max(all_pf - good_pf, 0) + + loc_pf =3D swp_l2 + swp_lc + swp_lm + rem_pf =3D swp_rc + swp_rm + + req_pend =3D max(0, bad_pf - swp_mab) + + r1 =3D d_ratio(ins, all_pf) + r2 =3D d_ratio(swp_ld, all_pf) + r3 =3D d_ratio(swp_t0, interval_sec) + r4 =3D d_ratio(swp_w, interval_sec) + r5 =3D d_ratio(swp_nt, interval_sec) + overview =3D MetricGroup("lpm_swpf_overview", [ + Metric("lpm_swpf_ov_insn_bt_swpf", "Insn between SWPF", r1, "insns= "), + Metric("lpm_swpf_ov_loads_bt_swpf", "Loads between SWPF", r2, "loa= ds"), + Metric("lpm_swpf_ov_rate_prefetch_t0_t1_t2", "Rate prefetch TO_T1_= T2", r3, + "insns/sec"), + Metric("lpm_swpf_ov_rate_prefetch_w", + "Rate prefetch W", r4, "insns/sec"), + Metric("lpm_swpf_ov_rate_preftech_nta", + "Rate prefetch NTA", r5, "insns/sec"), + ]) + + r1 =3D d_ratio(swp_mab, all_pf) + r2 =3D d_ratio(req_pend, all_pf) + usefulness_bad =3D MetricGroup("lpm_swpf_usefulness_bad", [ + Metric("lpm_swpf_use_bad_hit_l1", "Usefulness bad hit L1", r1, "10= 0%"), + Metric("lpm_swpf_use_bad_req_pend", + "Usefulness bad req pending", r2, "100%"), + ]) + + r1 =3D d_ratio(good_pf, all_pf) + usefulness_good =3D MetricGroup("lpm_swpf_usefulness_good", [ + Metric("lpm_swpf_use_good_other_src", "Usefulness good other src",= r1, + "100%"), + ]) + + usefulness =3D MetricGroup("lpm_swpf_usefulness", [ + usefulness_bad, + usefulness_good, + ]) + + r1 =3D d_ratio(swp_l2, good_pf) + r2 =3D d_ratio(swp_lc, good_pf) + r3 =3D d_ratio(swp_lm, good_pf) + data_src_local =3D MetricGroup("lpm_swpf_data_src_local", [ + Metric("lpm_swpf_data_src_local_l2", + "Data source local l2", r1, "100%"), + Metric("lpm_swpf_data_src_local_ccx_l3_loc_ccx", + "Data source local ccx l3 loc ccx", r2, "100%"), + Metric("lpm_swpf_data_src_local_memory_or_io", + "Data source local memory or IO", r3, "100%"), + ]) + + r1 =3D d_ratio(swp_rc, good_pf) + r2 =3D d_ratio(swp_rm, good_pf) + data_src_remote =3D MetricGroup("lpm_swpf_data_src_remote", [ + Metric("lpm_swpf_data_src_remote_cache", "Data source remote cache= ", r1, + "100%"), + Metric("lpm_swpf_data_src_remote_memory_or_io", + "Data source remote memory or IO", r2, "100%"), + ]) + + data_src =3D MetricGroup("lpm_swpf_data_src", [ + data_src_local, data_src_remote]) + + return MetricGroup("lpm_swpf", [overview, usefulness, data_src], + description=3D"Software prefetch breakdown (CCX L3 = =3D L3 of current thread, Loc CCX =3D CCX cache on some socket)") + + def AmdUpc() -> Metric: ops =3D Event("ex_ret_ops", "ex_ret_cops") upc =3D d_ratio(ops, smt_cycles) @@ -187,6 +287,7 @@ def main() -> None: =20 all_metrics =3D MetricGroup("", [ AmdBr(), + AmdSwpf(), AmdUpc(), Idle(), Rapl(), --=20 2.51.2.1041.gc1ab5b90ca-goog