From nobody Thu Dec 11 19:17:07 2025 Received: from mail-pg1-f202.google.com (mail-pg1-f202.google.com [209.85.215.202]) (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 553A02C11DF for ; Tue, 2 Dec 2025 17:53:59 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.215.202 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764698041; cv=none; b=J4a91Nbdv0yHT2xB0mQHzPr6rEOO+zbx2XfOt+LanwU4lLoVWAwdixtZqTYs9al0D5BDsvqac4bk8I/c5qfyP786q6RF1KjuEuDOY5wjcr0pOiEXJ7D6mkXMY8BUwW/jTXGa87miUX3pJaNvCEJg6exYuTrXSisLt7YgZXxJt5c= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764698041; c=relaxed/simple; bh=szV5c/klxqwi85v6hDhJdpF+W1mEdRaVxUIT+ScWzVI=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Content-Type; b=Rq2G9uwO5WPtQoVO8AIJw2aIJzI0tr2oDhuoEIg3jBxN1qj4X7pbEiuGPmvvlKIndB4fCYXPppvFsqgc4c+z6wipix+y49fRLA43eLOHNts6TkX3IBasEetm3LYs1gduc7fvf2y8gKELTYgeGVqt4uSMYEIISfOu4GBVNxtpfSM= 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=XXBp8Q9c; arc=none smtp.client-ip=209.85.215.202 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="XXBp8Q9c" Received: by mail-pg1-f202.google.com with SMTP id 41be03b00d2f7-b5edecdf94eso7940579a12.2 for ; Tue, 02 Dec 2025 09:53:59 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1764698039; x=1765302839; 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=R3fmXtA0VS/bmYEg2BbSjjVcEoYGv2SDd9KAsTbvG/8=; b=XXBp8Q9c2Hl/DqiMY8GTNmYeloHQ3pxQ69BT4M2+fVC+REfQFTxrpOfFxBmSk4lqJA Phg/bjhCFWGRxElbfCJKit6UFa+AwpgQXWIyroIIxZA1E3NpDM5iXogkXsxaK04yu6E1 uDzRE4wXetTbnWmML5v1CLXVHUKX+jFTOrNK0vTSBXUNopwKpJZyd2EDn9T7hr7oQGYU LSXCTlldlj8FqB45VPh8xAp8RA73hxtWICwa/pZm4/w+dYwDzCLEJbuTpyJ8Mq7Rzlra 4n6Do1Qr6uMP87FzKxdQ2/oCppxvrtZMVCaWBD32vu7umRCGZFPLXns5s6YSbxxN/ysY vRQQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1764698039; x=1765302839; 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=R3fmXtA0VS/bmYEg2BbSjjVcEoYGv2SDd9KAsTbvG/8=; b=ef/nQLsr9kk2028SRxlQrpfuRYNvw28pUirT13tDWpBnMu2mQtQpQQp3aWgm142HO5 jYhB5mpY0imzFp0nGF+FoKCnm/qeyirw6jcdslJVXcIObOayp5kvuXEd8FuW1vmxQcVN G0DULAhd7kvjOrZEhzxSzKJerC1NW6Q4rdl8f83CS1g4rKfDmVCD41uifoEv/IrJM9Ns B71chqIlnABRh0f/WTbv6VjLdGRRB4yj5w2fVhvIYysg70N4N9SAwf5JQUh6+pMvLV16 sjX6qsMcvVk5bZJKM4qyfSYB5WyQXGq70zQTHtSEpEqooqnbcGhzlaWqr0R6V2Vsf4t8 FJMw== X-Forwarded-Encrypted: i=1; AJvYcCVba4Ktfp8cHe5x03IFI9UKe88d1wF/eJYa0uEEVKHeBSDOipLaAG+ScIHTBCE8yZwA9LIRYt/cWgFM2w4=@vger.kernel.org X-Gm-Message-State: AOJu0YwclOKnPH7iOKxTW5ZjWMMb9uGb55sap87ZcCQT3nRl0EoD8bAo 0eKp+lUQ45aizz+wZw8OPXUbCo7ho5G+EEG7jqvB7RUX3CC6O4fI8vUJIW3ql7gownFOcKZI6Ss oULcamskQUw== X-Google-Smtp-Source: AGHT+IG+9ilXz5gVmVUIGvBMb1Evs62040WJAwi5pvbFDXSLe38ZR9z1a5GSthOPddBy2hZfZJJtgWL7cxY0 X-Received: from dlbqd22.prod.google.com ([2002:a05:7023:b16:b0:11a:5223:14ab]) (user=irogers job=prod-delivery.src-stubby-dispatcher) by 2002:a05:7022:7f0b:b0:11b:ca88:c4f5 with SMTP id a92af1059eb24-11c9d860f40mr18393276c88.35.1764698038740; Tue, 02 Dec 2025 09:53:58 -0800 (PST) Date: Tue, 2 Dec 2025 09:50:28 -0800 In-Reply-To: <20251202175043.623597-1-irogers@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20251202175043.623597-1-irogers@google.com> X-Mailer: git-send-email 2.52.0.158.g65b55ccf14-goog Message-ID: <20251202175043.623597-34-irogers@google.com> Subject: [PATCH v9 33/48] perf jevents: Add L2 metrics for Intel 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" Give a breakdown of various L2 counters as metrics, including totals, reads, hardware prefetcher, RFO, code and evictions. Signed-off-by: Ian Rogers Tested-by: Thomas Falcon --- tools/perf/pmu-events/intel_metrics.py | 170 +++++++++++++++++++++++++ 1 file changed, 170 insertions(+) diff --git a/tools/perf/pmu-events/intel_metrics.py b/tools/perf/pmu-events= /intel_metrics.py index 7fcc0a1c544d..d190d97f4aff 100755 --- a/tools/perf/pmu-events/intel_metrics.py +++ b/tools/perf/pmu-events/intel_metrics.py @@ -263,6 +263,175 @@ def IntelBr(): description=3D"breakdown of retired branch instruct= ions") =20 =20 +def IntelL2() -> Optional[MetricGroup]: + try: + DC_HIT =3D Event("L2_RQSTS.DEMAND_DATA_RD_HIT") + except: + return None + try: + DC_MISS =3D Event("L2_RQSTS.DEMAND_DATA_RD_MISS") + l2_dmnd_miss =3D DC_MISS + l2_dmnd_rd_all =3D DC_MISS + DC_HIT + except: + DC_ALL =3D Event("L2_RQSTS.ALL_DEMAND_DATA_RD") + l2_dmnd_miss =3D DC_ALL - DC_HIT + l2_dmnd_rd_all =3D DC_ALL + l2_dmnd_mrate =3D d_ratio(l2_dmnd_miss, interval_sec) + l2_dmnd_rrate =3D d_ratio(l2_dmnd_rd_all, interval_sec) + + DC_PFH =3D None + DC_PFM =3D None + l2_pf_all =3D None + l2_pf_mrate =3D None + l2_pf_rrate =3D None + try: + DC_PFH =3D Event("L2_RQSTS.PF_HIT") + DC_PFM =3D Event("L2_RQSTS.PF_MISS") + l2_pf_all =3D DC_PFH + DC_PFM + l2_pf_mrate =3D d_ratio(DC_PFM, interval_sec) + l2_pf_rrate =3D d_ratio(l2_pf_all, interval_sec) + except: + pass + + DC_RFOH =3D None + DC_RFOM =3D None + l2_rfo_all =3D None + l2_rfo_mrate =3D None + l2_rfo_rrate =3D None + try: + DC_RFOH =3D Event("L2_RQSTS.RFO_HIT") + DC_RFOM =3D Event("L2_RQSTS.RFO_MISS") + l2_rfo_all =3D DC_RFOH + DC_RFOM + l2_rfo_mrate =3D d_ratio(DC_RFOM, interval_sec) + l2_rfo_rrate =3D d_ratio(l2_rfo_all, interval_sec) + except: + pass + + DC_CH =3D None + try: + DC_CH =3D Event("L2_RQSTS.CODE_RD_HIT") + except: + pass + DC_CM =3D Event("L2_RQSTS.CODE_RD_MISS") + DC_IN =3D Event("L2_LINES_IN.ALL") + DC_OUT_NS =3D None + DC_OUT_S =3D None + l2_lines_out =3D None + l2_out_rate =3D None + wbn =3D None + isd =3D None + try: + DC_OUT_NS =3D Event("L2_LINES_OUT.NON_SILENT", + "L2_LINES_OUT.DEMAND_DIRTY", + "L2_LINES_IN.S") + DC_OUT_S =3D Event("L2_LINES_OUT.SILENT", + "L2_LINES_OUT.DEMAND_CLEAN", + "L2_LINES_IN.I") + if DC_OUT_S.name =3D=3D "L2_LINES_OUT.SILENT" and ( + args.model.startswith("skylake") or + args.model =3D=3D "cascadelakex"): + DC_OUT_S.name =3D "L2_LINES_OUT.SILENT/any/" + # bring is back to per-CPU + l2_s =3D Select(DC_OUT_S / 2, Literal("#smt_on"), DC_OUT_S) + l2_ns =3D DC_OUT_NS + l2_lines_out =3D l2_s + l2_ns + l2_out_rate =3D d_ratio(l2_lines_out, interval_sec) + nlr =3D max(l2_ns - DC_WB_U - DC_WB_D, 0) + wbn =3D d_ratio(nlr, interval_sec) + isd =3D d_ratio(l2_s, interval_sec) + except: + pass + DC_OUT_U =3D None + l2_pf_useless =3D None + l2_useless_rate =3D None + try: + DC_OUT_U =3D Event("L2_LINES_OUT.USELESS_HWPF") + l2_pf_useless =3D DC_OUT_U + l2_useless_rate =3D d_ratio(l2_pf_useless, interval_sec) + except: + pass + DC_WB_U =3D None + DC_WB_D =3D None + wbu =3D None + wbd =3D None + try: + DC_WB_U =3D Event("IDI_MISC.WB_UPGRADE") + DC_WB_D =3D Event("IDI_MISC.WB_DOWNGRADE") + wbu =3D d_ratio(DC_WB_U, interval_sec) + wbd =3D d_ratio(DC_WB_D, interval_sec) + except: + pass + + l2_lines_in =3D DC_IN + l2_code_all =3D (DC_CH + DC_CM) if DC_CH else None + l2_code_rate =3D d_ratio(l2_code_all, interval_sec) if DC_CH else None + l2_code_miss_rate =3D d_ratio(DC_CM, interval_sec) + l2_in_rate =3D d_ratio(l2_lines_in, interval_sec) + + return MetricGroup("lpm_l2", [ + MetricGroup("lpm_l2_totals", [ + Metric("lpm_l2_totals_in", "L2 cache total in per second", + l2_in_rate, "In/s"), + Metric("lpm_l2_totals_out", "L2 cache total out per second", + l2_out_rate, "Out/s") if l2_out_rate else None, + ]), + MetricGroup("lpm_l2_rd", [ + Metric("lpm_l2_rd_hits", "L2 cache data read hits", + d_ratio(DC_HIT, l2_dmnd_rd_all), "100%"), + Metric("lpm_l2_rd_hits", "L2 cache data read hits", + d_ratio(l2_dmnd_miss, l2_dmnd_rd_all), "100%"), + Metric("lpm_l2_rd_requests", "L2 cache data read requests per = second", + l2_dmnd_rrate, "requests/s"), + Metric("lpm_l2_rd_misses", "L2 cache data read misses per seco= nd", + l2_dmnd_mrate, "misses/s"), + ]), + MetricGroup("lpm_l2_hwpf", [ + Metric("lpm_l2_hwpf_hits", "L2 cache hardware prefetcher hits", + d_ratio(DC_PFH, l2_pf_all), "100%"), + Metric("lpm_l2_hwpf_misses", "L2 cache hardware prefetcher mis= ses", + d_ratio(DC_PFM, l2_pf_all), "100%"), + Metric("lpm_l2_hwpf_useless", "L2 cache hardware prefetcher us= eless prefetches per second", + l2_useless_rate, "100%") if l2_useless_rate else None, + Metric("lpm_l2_hwpf_requests", "L2 cache hardware prefetcher r= equests per second", + l2_pf_rrate, "100%"), + Metric("lpm_l2_hwpf_misses", "L2 cache hardware prefetcher mis= ses per second", + l2_pf_mrate, "100%"), + ]) if DC_PFH else None, + MetricGroup("lpm_l2_rfo", [ + Metric("lpm_l2_rfo_hits", "L2 cache request for ownership (RFO= ) hits", + d_ratio(DC_RFOH, l2_rfo_all), "100%"), + Metric("lpm_l2_rfo_misses", "L2 cache request for ownership (R= FO) misses", + d_ratio(DC_RFOM, l2_rfo_all), "100%"), + Metric("lpm_l2_rfo_requests", "L2 cache request for ownership = (RFO) requests per second", + l2_rfo_rrate, "requests/s"), + Metric("lpm_l2_rfo_misses", "L2 cache request for ownership (R= FO) misses per second", + l2_rfo_mrate, "misses/s"), + ]) if DC_RFOH else None, + MetricGroup("lpm_l2_code", [ + Metric("lpm_l2_code_hits", "L2 cache code hits", + d_ratio(DC_CH, l2_code_all), "100%") if DC_CH else None, + Metric("lpm_l2_code_misses", "L2 cache code misses", + d_ratio(DC_CM, l2_code_all), "100%") if DC_CH else None, + Metric("lpm_l2_code_requests", "L2 cache code requests per sec= ond", + l2_code_rate, "requests/s") if DC_CH else None, + Metric("lpm_l2_code_misses", "L2 cache code misses per second", + l2_code_miss_rate, "misses/s"), + ]), + MetricGroup("lpm_l2_evict", [ + MetricGroup("lpm_l2_evict_mef_lines", [ + Metric("lpm_l2_evict_mef_lines_l3_hot_lru", "L2 evictions = M/E/F lines L3 hot LRU per second", + wbu, "HotLRU/s") if wbu else None, + Metric("lpm_l2_evict_mef_lines_l3_norm_lru", "L2 evictions= M/E/F lines L3 normal LRU per second", + wbn, "NormLRU/s") if wbn else None, + Metric("lpm_l2_evict_mef_lines_dropped", "L2 evictions M/E= /F lines dropped per second", + wbd, "dropped/s") if wbd else None, + Metric("lpm_l2_evict_is_lines_dropped", "L2 evictions I/S = lines dropped per second", + isd, "dropped/s") if isd else None, + ]), + ]), + ], description=3D"L2 data cache analysis") + + def IntelPorts() -> Optional[MetricGroup]: pipeline_events =3D json.load( open(f"{_args.events_path}/x86/{_args.model}/pipeline.json")) @@ -386,6 +555,7 @@ def main() -> None: Smi(), Tsx(), IntelBr(), + IntelL2(), IntelPorts(), IntelSwpf(), ]) --=20 2.52.0.158.g65b55ccf14-goog