From nobody Mon Jun 15 16:14:24 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id E0AF0C433F5 for ; Tue, 26 Apr 2022 14:02:40 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1351396AbiDZOFq (ORCPT ); Tue, 26 Apr 2022 10:05:46 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:44350 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1351375AbiDZODx (ORCPT ); Tue, 26 Apr 2022 10:03:53 -0400 Received: from smtp-fw-80007.amazon.com (smtp-fw-80007.amazon.com [99.78.197.218]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 42C2E1A82C; Tue, 26 Apr 2022 07:00:37 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amazon.com; i=@amazon.com; q=dns/txt; s=amazon201209; t=1650981647; x=1682517647; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=2kGoDhwDDU4SUv0yonPrLgPdKoDB23uSHWFWBT2qSyo=; b=ZyvvWUzNOp8iKFNsAEIBTB2cipSq4YEOVRY/9f57a1+CXJV75tgoo00b OKQpQexRYzEiFE1/1sTKFOks4IH/tQdhlaKTJIXgGolUDWb56P6zwFovm bCRnKLAdESugIo+3xLQ9FCvl+GuoBzVzmv9YzY9QKwiDaDHcjph4oQb1I Y=; X-IronPort-AV: E=Sophos;i="5.90,291,1643673600"; d="scan'208";a="83172134" Received: from pdx4-co-svc-p1-lb2-vlan2.amazon.com (HELO email-inbound-relay-iad-1a-4ba5c7da.us-east-1.amazon.com) ([10.25.36.210]) by smtp-border-fw-80007.pdx80.corp.amazon.com with ESMTP; 26 Apr 2022 14:00:20 +0000 Received: from EX13MTAUWB001.ant.amazon.com (iad12-ws-svc-p26-lb9-vlan2.iad.amazon.com [10.40.163.34]) by email-inbound-relay-iad-1a-4ba5c7da.us-east-1.amazon.com (Postfix) with ESMTPS id 4F20630008C; Tue, 26 Apr 2022 14:00:14 +0000 (UTC) Received: from EX13D02UWB001.ant.amazon.com (10.43.161.240) by EX13MTAUWB001.ant.amazon.com (10.43.161.207) with Microsoft SMTP Server (TLS) id 15.0.1497.32; Tue, 26 Apr 2022 13:59:47 +0000 Received: from EX13MTAUEA002.ant.amazon.com (10.43.61.77) by EX13D02UWB001.ant.amazon.com (10.43.161.240) with Microsoft SMTP Server (TLS) id 15.0.1497.32; Tue, 26 Apr 2022 13:59:47 +0000 Received: from dev-dsk-alisaidi-1d-b9a0e636.us-east-1.amazon.com (172.19.181.128) by mail-relay.amazon.com (10.43.61.169) with Microsoft SMTP Server id 15.0.1497.32 via Frontend Transport; Tue, 26 Apr 2022 13:59:46 +0000 Received: by dev-dsk-alisaidi-1d-b9a0e636.us-east-1.amazon.com (Postfix, from userid 5131138) id 5192F20ED; Tue, 26 Apr 2022 13:59:45 +0000 (UTC) From: Ali Saidi To: , , , , , CC: , , , , , , , , , , , , , , , Subject: [PATCH v7 1/5] perf: Add SNOOP_PEER flag to perf mem data struct Date: Tue, 26 Apr 2022 13:59:33 +0000 Message-ID: <20220426135937.18497-2-alisaidi@amazon.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220426135937.18497-1-alisaidi@amazon.com> References: <20220426135937.18497-1-alisaidi@amazon.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Add a flag to the perf mem data struct to signal that a request caused a cache-to-cache transfer of a line from a peer of the requestor and wasn't sourced from a lower cache level. The line being moved from one peer cache to another has latency and performance implications. On Arm64 Neoverse systems the data source can indicate a cache-to-cache transfer but not if the line is dirty or clean, so instead of overloading HITM define a new flag that indicates this type of transfer. Signed-off-by: Ali Saidi Reviewed-by: Leo Yan --- include/uapi/linux/perf_event.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/uapi/linux/perf_event.h b/include/uapi/linux/perf_even= t.h index d37629dbad72..7b88bfd097dc 100644 --- a/include/uapi/linux/perf_event.h +++ b/include/uapi/linux/perf_event.h @@ -1310,7 +1310,7 @@ union perf_mem_data_src { #define PERF_MEM_SNOOP_SHIFT 19 =20 #define PERF_MEM_SNOOPX_FWD 0x01 /* forward */ -/* 1 free */ +#define PERF_MEM_SNOOPX_PEER 0x02 /* xfer from peer */ #define PERF_MEM_SNOOPX_SHIFT 38 =20 /* locked instruction */ --=20 2.32.0 From nobody Mon Jun 15 16:14:24 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id E5FD5C433F5 for ; Tue, 26 Apr 2022 13:59:59 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1351348AbiDZODF (ORCPT ); Tue, 26 Apr 2022 10:03:05 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:41414 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1351323AbiDZODC (ORCPT ); Tue, 26 Apr 2022 10:03:02 -0400 Received: from smtp-fw-2101.amazon.com (smtp-fw-2101.amazon.com [72.21.196.25]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 2BCAD1906A4; Tue, 26 Apr 2022 06:59:54 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amazon.com; i=@amazon.com; q=dns/txt; s=amazon201209; t=1650981595; x=1682517595; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=g7YWa84Ggc9PbBc2yEEwZ6f7hUrSoU1ylmKnDzAKmLc=; b=IlLlD5D2LcOFJm5d3IPKQcxs7zgwIxbnWkG+rse82zgbwH5qLlEEGfy0 xrCkZ95LLvEfntZPnqjy2IdGa1VEjUcq1bi1p4y8DSVsdV5kC1S9/8wzH w8C7pgUX9K5x576kXip5LrRdcfyVuWb0l7fFntGZ/MC608YbxCB+sHaDg M=; X-IronPort-AV: E=Sophos;i="5.90,291,1643673600"; d="scan'208";a="193119156" Received: from iad12-co-svc-p1-lb1-vlan2.amazon.com (HELO email-inbound-relay-iad-1d-10222bbc.us-east-1.amazon.com) ([10.43.8.2]) by smtp-border-fw-2101.iad2.amazon.com with ESMTP; 26 Apr 2022 13:59:53 +0000 Received: from EX13MTAUWB001.ant.amazon.com (iad12-ws-svc-p26-lb9-vlan3.iad.amazon.com [10.40.163.38]) by email-inbound-relay-iad-1d-10222bbc.us-east-1.amazon.com (Postfix) with ESMTPS id 664101A0008; Tue, 26 Apr 2022 13:59:47 +0000 (UTC) Received: from EX13D02UWB002.ant.amazon.com (10.43.161.160) by EX13MTAUWB001.ant.amazon.com (10.43.161.249) with Microsoft SMTP Server (TLS) id 15.0.1497.32; Tue, 26 Apr 2022 13:59:46 +0000 Received: from EX13MTAUWB001.ant.amazon.com (10.43.161.207) by EX13D02UWB002.ant.amazon.com (10.43.161.160) with Microsoft SMTP Server (TLS) id 15.0.1497.32; Tue, 26 Apr 2022 13:59:46 +0000 Received: from dev-dsk-alisaidi-1d-b9a0e636.us-east-1.amazon.com (172.19.181.128) by mail-relay.amazon.com (10.43.161.249) with Microsoft SMTP Server id 15.0.1497.32 via Frontend Transport; Tue, 26 Apr 2022 13:59:46 +0000 Received: by dev-dsk-alisaidi-1d-b9a0e636.us-east-1.amazon.com (Postfix, from userid 5131138) id 4E9A920EC; Tue, 26 Apr 2022 13:59:45 +0000 (UTC) From: Ali Saidi To: , , , , , CC: , , , , , , , , , , , , , , , Subject: [PATCH v7 2/5] perf tools: sync addition of PERF_MEM_SNOOPX_PEER Date: Tue, 26 Apr 2022 13:59:34 +0000 Message-ID: <20220426135937.18497-3-alisaidi@amazon.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220426135937.18497-1-alisaidi@amazon.com> References: <20220426135937.18497-1-alisaidi@amazon.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Add a flag to the perf mem data struct to signal that a request caused a cache-to-cache transfer of a line from a peer of the requestor and wasn't sourced from a lower cache level. The line being moved from one peer cache to another has latency and performance implications. On Arm64 Neoverse systems the data source can indicate a cache-to-cache transfer but not if the line is dirty or clean, so instead of overloading HITM define a new flag that indicates this type of transfer. Signed-off-by: Ali Saidi Reviewed-by: Leo Yan --- tools/include/uapi/linux/perf_event.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/include/uapi/linux/perf_event.h b/tools/include/uapi/lin= ux/perf_event.h index d37629dbad72..7b88bfd097dc 100644 --- a/tools/include/uapi/linux/perf_event.h +++ b/tools/include/uapi/linux/perf_event.h @@ -1310,7 +1310,7 @@ union perf_mem_data_src { #define PERF_MEM_SNOOP_SHIFT 19 =20 #define PERF_MEM_SNOOPX_FWD 0x01 /* forward */ -/* 1 free */ +#define PERF_MEM_SNOOPX_PEER 0x02 /* xfer from peer */ #define PERF_MEM_SNOOPX_SHIFT 38 =20 /* locked instruction */ --=20 2.32.0 From nobody Mon Jun 15 16:14:24 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id E60B5C4332F for ; Tue, 26 Apr 2022 14:01:08 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1351406AbiDZOD6 (ORCPT ); Tue, 26 Apr 2022 10:03:58 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:43810 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1351368AbiDZODu (ORCPT ); Tue, 26 Apr 2022 10:03:50 -0400 Received: from smtp-fw-80007.amazon.com (smtp-fw-80007.amazon.com [99.78.197.218]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id DF759186D1; Tue, 26 Apr 2022 07:00:36 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amazon.com; i=@amazon.com; q=dns/txt; s=amazon201209; t=1650981643; x=1682517643; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=IC8noDBK4X51feo0Vpeik5XnaP2Xw/g4NSKwFbVg0tA=; b=NP3ek2pwqwhRffJ2VNm/OFqCFEx8v8oWLcwV90vGZMudJtqC4Vpn6/Gj AQZlnkZJ4yi+AJgpjbJRT3KDx3JsCFAYOBsOOlyLHIDO1Bckp3BzFrgji xAL+T6I6Cgq4EMtlh8YO65d6++SnVGZzlkccLe4xmdupfCBsy2EzKTqzJ s=; X-IronPort-AV: E=Sophos;i="5.90,291,1643673600"; d="scan'208";a="83172116" Received: from pdx4-co-svc-p1-lb2-vlan2.amazon.com (HELO email-inbound-relay-iad-1a-b27d4a00.us-east-1.amazon.com) ([10.25.36.210]) by smtp-border-fw-80007.pdx80.corp.amazon.com with ESMTP; 26 Apr 2022 14:00:18 +0000 Received: from EX13MTAUWB001.ant.amazon.com (iad12-ws-svc-p26-lb9-vlan3.iad.amazon.com [10.40.163.38]) by email-inbound-relay-iad-1a-b27d4a00.us-east-1.amazon.com (Postfix) with ESMTPS id EF84281111; Tue, 26 Apr 2022 14:00:11 +0000 (UTC) Received: from EX13D02UWB003.ant.amazon.com (10.43.161.48) by EX13MTAUWB001.ant.amazon.com (10.43.161.207) with Microsoft SMTP Server (TLS) id 15.0.1497.32; Tue, 26 Apr 2022 13:59:47 +0000 Received: from EX13MTAUEA002.ant.amazon.com (10.43.61.77) by EX13D02UWB003.ant.amazon.com (10.43.161.48) with Microsoft SMTP Server (TLS) id 15.0.1497.32; Tue, 26 Apr 2022 13:59:46 +0000 Received: from dev-dsk-alisaidi-1d-b9a0e636.us-east-1.amazon.com (172.19.181.128) by mail-relay.amazon.com (10.43.61.169) with Microsoft SMTP Server id 15.0.1497.32 via Frontend Transport; Tue, 26 Apr 2022 13:59:46 +0000 Received: by dev-dsk-alisaidi-1d-b9a0e636.us-east-1.amazon.com (Postfix, from userid 5131138) id 4F93D20E9; Tue, 26 Apr 2022 13:59:45 +0000 (UTC) From: Ali Saidi To: , , , , , CC: , , , , , , , , , , , , , , , Subject: [PATCH v7 3/5] perf mem: Print snoop peer flag Date: Tue, 26 Apr 2022 13:59:35 +0000 Message-ID: <20220426135937.18497-4-alisaidi@amazon.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220426135937.18497-1-alisaidi@amazon.com> References: <20220426135937.18497-1-alisaidi@amazon.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" From: Leo Yan Since PERF_MEM_SNOOPX_PEER flag is a new snoop type, print this flag if it is set. Before: memstress 3603 [020] 122.463754: 1 l1d-miss: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP N/A|TLB Walker hit|LCK No|BL= K N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) memstress 3603 [020] 122.463754: 1 l1d-access: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP N/A|TLB Walker hit|LCK No|BL= K N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) memstress 3603 [020] 122.463754: 1 llc-miss: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP N/A|TLB Walker hit|LCK No|BL= K N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) memstress 3603 [020] 122.463754: 1 llc-access: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP N/A|TLB Walker hit|LCK No|BL= K N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) memstress 3603 [020] 122.463754: 1 tlb-access: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP N/A|TLB Walker hit|LCK No|BL= K N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) memstress 3603 [020] 122.463754: 1 memory: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP N/A|TLB Walker hit|LCK No|BL= K N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) After: memstress 3603 [020] 122.463754: 1 l1d-miss: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP Peer|TLB Walker hit|LCK No|B= LK N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) memstress 3603 [020] 122.463754: 1 l1d-access: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP Peer|TLB Walker hit|LCK No|B= LK N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) memstress 3603 [020] 122.463754: 1 llc-miss: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP Peer|TLB Walker hit|LCK No|B= LK N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) memstress 3603 [020] 122.463754: 1 llc-access: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP Peer|TLB Walker hit|LCK No|B= LK N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) memstress 3603 [020] 122.463754: 1 tlb-access: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP Peer|TLB Walker hit|LCK No|B= LK N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) memstress 3603 [020] 122.463754: 1 memory: = 8688000842 |OP LOAD|LVL L3 or L3 hit|SNP Peer|TLB Walker hit|LCK No|B= LK N/A aaaac17c3e88 [unknown] (/home/ubuntu/memstress) Signed-off-by: Leo Yan Reviewed-by: Ali Saidi Tested-by: Ali Saidi --- tools/perf/util/mem-events.c | 18 +++++++++++++++--- 1 file changed, 15 insertions(+), 3 deletions(-) diff --git a/tools/perf/util/mem-events.c b/tools/perf/util/mem-events.c index efaf263464b9..db5225caaabe 100644 --- a/tools/perf/util/mem-events.c +++ b/tools/perf/util/mem-events.c @@ -410,6 +410,11 @@ static const char * const snoop_access[] =3D { "HitM", }; =20 +static const char * const snoopx_access[] =3D { + "Fwd", + "Peer", +}; + int perf_mem__snp_scnprintf(char *out, size_t sz, struct mem_info *mem_inf= o) { size_t i, l =3D 0; @@ -430,13 +435,20 @@ int perf_mem__snp_scnprintf(char *out, size_t sz, str= uct mem_info *mem_info) } l +=3D scnprintf(out + l, sz - l, snoop_access[i]); } - if (mem_info && - (mem_info->data_src.mem_snoopx & PERF_MEM_SNOOPX_FWD)) { + + m =3D 0; + if (mem_info) + m =3D mem_info->data_src.mem_snoopx; + + for (i =3D 0; m && i < ARRAY_SIZE(snoopx_access); i++, m >>=3D 1) { + if (!(m & 0x1)) + continue; + if (l) { strcat(out, " or "); l +=3D 4; } - l +=3D scnprintf(out + l, sz - l, "Fwd"); + l +=3D scnprintf(out + l, sz - l, snoopx_access[i]); } =20 if (*out =3D=3D '\0') --=20 2.32.0 From nobody Mon Jun 15 16:14:24 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id AACE7C433EF for ; Tue, 26 Apr 2022 14:00:42 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1351363AbiDZODs (ORCPT ); Tue, 26 Apr 2022 10:03:48 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:44028 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1351355AbiDZODj (ORCPT ); Tue, 26 Apr 2022 10:03:39 -0400 Received: from smtp-fw-6001.amazon.com (smtp-fw-6001.amazon.com [52.95.48.154]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 6AFCE17A93; Tue, 26 Apr 2022 07:00:16 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amazon.com; i=@amazon.com; q=dns/txt; s=amazon201209; t=1650981623; x=1682517623; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=GDPYyfBxVot3TWjo8mSYj5TESCgNz4m4lu/DlfqWZvQ=; b=I559ZMHqccp7SqxGe1aEvACpaBAv/hc/5202WpXvmDVY0gFqTZs/djtf V2y//BOOKf3mKKhxJn+fAvsGwkgXeJBhNtzWhQvdVAvhljf0nOV6QsKWw ngxrbvYl1Oe+oP/SD3jrdjUvjU8HJ5gWv8HRmsHoDpvpQSJAT5o/7P9oU o=; X-IronPort-AV: E=Sophos;i="5.90,291,1643673600"; d="scan'208";a="197784770" Received: from iad12-co-svc-p1-lb1-vlan3.amazon.com (HELO email-inbound-relay-iad-1a-b27d4a00.us-east-1.amazon.com) ([10.43.8.6]) by smtp-border-fw-6001.iad6.amazon.com with ESMTP; 26 Apr 2022 14:00:15 +0000 Received: from EX13MTAUWB001.ant.amazon.com (iad12-ws-svc-p26-lb9-vlan3.iad.amazon.com [10.40.163.38]) by email-inbound-relay-iad-1a-b27d4a00.us-east-1.amazon.com (Postfix) with ESMTPS id EBE948113C; Tue, 26 Apr 2022 14:00:09 +0000 (UTC) Received: from EX13D02UWB001.ant.amazon.com (10.43.161.240) by EX13MTAUWB001.ant.amazon.com (10.43.161.207) with Microsoft SMTP Server (TLS) id 15.0.1497.32; Tue, 26 Apr 2022 13:59:46 +0000 Received: from EX13MTAUWB001.ant.amazon.com (10.43.161.207) by EX13D02UWB001.ant.amazon.com (10.43.161.240) with Microsoft SMTP Server (TLS) id 15.0.1497.32; Tue, 26 Apr 2022 13:59:46 +0000 Received: from dev-dsk-alisaidi-1d-b9a0e636.us-east-1.amazon.com (172.19.181.128) by mail-relay.amazon.com (10.43.161.249) with Microsoft SMTP Server id 15.0.1497.32 via Frontend Transport; Tue, 26 Apr 2022 13:59:46 +0000 Received: by dev-dsk-alisaidi-1d-b9a0e636.us-east-1.amazon.com (Postfix, from userid 5131138) id 4D8DC1788; Tue, 26 Apr 2022 13:59:45 +0000 (UTC) From: Ali Saidi To: , , , , , CC: , , , , , , , , , , , , , , , Subject: [PATCH v7 4/5] perf arm-spe: Don't set data source if it's not a memory operation Date: Tue, 26 Apr 2022 13:59:36 +0000 Message-ID: <20220426135937.18497-5-alisaidi@amazon.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220426135937.18497-1-alisaidi@amazon.com> References: <20220426135937.18497-1-alisaidi@amazon.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" From: Leo Yan Except memory load and store operations, Arm SPE records also can support other operation types, bug when set the data source field the current code assumes a record is a either load operation or store operation, this leads to wrongly synthesize memory samples. This patch strictly checks the record operation type, it only sets data source only for the operation types ARM_SPE_LD and ARM_SPE_ST, otherwise, returns zero for data source. Therefore, we can synthesize memory samples only when data source is a non-zero value, the function arm_spe__is_memory_event() is useless and removed. Signed-off-by: Leo Yan Reviewed-by: Ali Saidi Tested-by: Ali Saidi Reviewed-by: German Gomez --- tools/perf/util/arm-spe.c | 22 ++++++++-------------- 1 file changed, 8 insertions(+), 14 deletions(-) diff --git a/tools/perf/util/arm-spe.c b/tools/perf/util/arm-spe.c index d2b64e3f588b..e032efc03274 100644 --- a/tools/perf/util/arm-spe.c +++ b/tools/perf/util/arm-spe.c @@ -387,26 +387,16 @@ static int arm_spe__synth_instruction_sample(struct a= rm_spe_queue *speq, return arm_spe_deliver_synth_event(spe, speq, event, &sample); } =20 -#define SPE_MEM_TYPE (ARM_SPE_L1D_ACCESS | ARM_SPE_L1D_MISS | \ - ARM_SPE_LLC_ACCESS | ARM_SPE_LLC_MISS | \ - ARM_SPE_REMOTE_ACCESS) - -static bool arm_spe__is_memory_event(enum arm_spe_sample_type type) -{ - if (type & SPE_MEM_TYPE) - return true; - - return false; -} - static u64 arm_spe__synth_data_source(const struct arm_spe_record *record) { union perf_mem_data_src data_src =3D { 0 }; =20 if (record->op =3D=3D ARM_SPE_LD) data_src.mem_op =3D PERF_MEM_OP_LOAD; - else + else if (record->op =3D=3D ARM_SPE_ST) data_src.mem_op =3D PERF_MEM_OP_STORE; + else + return 0; =20 if (record->type & (ARM_SPE_LLC_ACCESS | ARM_SPE_LLC_MISS)) { data_src.mem_lvl =3D PERF_MEM_LVL_L3; @@ -510,7 +500,11 @@ static int arm_spe_sample(struct arm_spe_queue *speq) return err; } =20 - if (spe->sample_memory && arm_spe__is_memory_event(record->type)) { + /* + * When data_src is zero it means the record is not a memory operation, + * skip to synthesize memory sample for this case. + */ + if (spe->sample_memory && data_src) { err =3D arm_spe__synth_mem_sample(speq, spe->memory_id, data_src); if (err) return err; --=20 2.32.0 From nobody Mon Jun 15 16:14:24 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id C0E54C433F5 for ; Tue, 26 Apr 2022 14:00:17 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S235262AbiDZODV (ORCPT ); Tue, 26 Apr 2022 10:03:21 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:42098 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1344484AbiDZODM (ORCPT ); Tue, 26 Apr 2022 10:03:12 -0400 Received: from smtp-fw-80007.amazon.com (smtp-fw-80007.amazon.com [99.78.197.218]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 98AE7191C60; Tue, 26 Apr 2022 07:00:01 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amazon.com; i=@amazon.com; q=dns/txt; s=amazon201209; t=1650981603; x=1682517603; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=vkW2dzHpr+6WZ25bEoY3oyUmALpzBn7pHUAUs7mBpFQ=; b=gmvgv4S/qOmOoEozCJaNy6Yc3YRl5oUUaehQuPr/WjfSotSOfcyvf4Ht SfG66KhQ2Vy7BRhDMz5h/xGRCs0sf5Xad0+MABbkTEV5RUPwneaOBxJcu Y53K6xoASK96sq0Qs/VUrEhrNTZ/ICCdBaZHVESpbo578+S2JfpZsJcR9 4=; X-IronPort-AV: E=Sophos;i="5.90,291,1643673600"; d="scan'208";a="83171972" Received: from pdx4-co-svc-p1-lb2-vlan2.amazon.com (HELO email-inbound-relay-pdx-2c-51ba86d8.us-west-2.amazon.com) ([10.25.36.210]) by smtp-border-fw-80007.pdx80.corp.amazon.com with ESMTP; 26 Apr 2022 13:59:49 +0000 Received: from EX13MTAUWC002.ant.amazon.com (pdx1-ws-svc-p6-lb9-vlan2.pdx.amazon.com [10.236.137.194]) by email-inbound-relay-pdx-2c-51ba86d8.us-west-2.amazon.com (Postfix) with ESMTPS id 08D2F838F8; Tue, 26 Apr 2022 13:59:47 +0000 (UTC) Received: from EX13D02UWC004.ant.amazon.com (10.43.162.236) by EX13MTAUWC002.ant.amazon.com (10.43.162.240) with Microsoft SMTP Server (TLS) id 15.0.1497.32; Tue, 26 Apr 2022 13:59:47 +0000 Received: from EX13MTAUEB002.ant.amazon.com (10.43.60.12) by EX13D02UWC004.ant.amazon.com (10.43.162.236) with Microsoft SMTP Server (TLS) id 15.0.1497.32; Tue, 26 Apr 2022 13:59:47 +0000 Received: from dev-dsk-alisaidi-1d-b9a0e636.us-east-1.amazon.com (172.19.181.128) by mail-relay.amazon.com (10.43.60.234) with Microsoft SMTP Server id 15.0.1497.32 via Frontend Transport; Tue, 26 Apr 2022 13:59:46 +0000 Received: by dev-dsk-alisaidi-1d-b9a0e636.us-east-1.amazon.com (Postfix, from userid 5131138) id 5098420EB; Tue, 26 Apr 2022 13:59:45 +0000 (UTC) From: Ali Saidi To: , , , , , CC: , , , , , , , , , , , , , , , Subject: [PATCH v7 5/5] perf arm-spe: Use SPE data source for neoverse cores Date: Tue, 26 Apr 2022 13:59:37 +0000 Message-ID: <20220426135937.18497-6-alisaidi@amazon.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220426135937.18497-1-alisaidi@amazon.com> References: <20220426135937.18497-1-alisaidi@amazon.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" When synthesizing data from SPE, augment the type with source information for Arm Neoverse cores. The field is IMPLDEF but the Neoverse cores all use the same encoding. I can't find encoding information for any other SPE implementations to unify their choices with Arm's thus that is left for future work. This change populates the mem_lvl_num for Neoverse cores as well as the deprecated mem_lvl namespace. Signed-off-by: Ali Saidi Reviewed-by: German Gomez Reviewed-by: Leo Yan Tested-by: Leo Yan --- .../util/arm-spe-decoder/arm-spe-decoder.c | 1 + .../util/arm-spe-decoder/arm-spe-decoder.h | 12 ++ tools/perf/util/arm-spe.c | 130 +++++++++++++++--- 3 files changed, 127 insertions(+), 16 deletions(-) diff --git a/tools/perf/util/arm-spe-decoder/arm-spe-decoder.c b/tools/perf= /util/arm-spe-decoder/arm-spe-decoder.c index 5e390a1a79ab..091987dd3966 100644 --- a/tools/perf/util/arm-spe-decoder/arm-spe-decoder.c +++ b/tools/perf/util/arm-spe-decoder/arm-spe-decoder.c @@ -220,6 +220,7 @@ static int arm_spe_read_record(struct arm_spe_decoder *= decoder) =20 break; case ARM_SPE_DATA_SOURCE: + decoder->record.source =3D payload; break; case ARM_SPE_BAD: break; diff --git a/tools/perf/util/arm-spe-decoder/arm-spe-decoder.h b/tools/perf= /util/arm-spe-decoder/arm-spe-decoder.h index 69b31084d6be..46a61df1145b 100644 --- a/tools/perf/util/arm-spe-decoder/arm-spe-decoder.h +++ b/tools/perf/util/arm-spe-decoder/arm-spe-decoder.h @@ -29,6 +29,17 @@ enum arm_spe_op_type { ARM_SPE_ST =3D 1 << 1, }; =20 +enum arm_spe_neoverse_data_source { + ARM_SPE_NV_L1D =3D 0x0, + ARM_SPE_NV_L2 =3D 0x8, + ARM_SPE_NV_PEER_CORE =3D 0x9, + ARM_SPE_NV_LOCAL_CLUSTER =3D 0xa, + ARM_SPE_NV_SYS_CACHE =3D 0xb, + ARM_SPE_NV_PEER_CLUSTER =3D 0xc, + ARM_SPE_NV_REMOTE =3D 0xd, + ARM_SPE_NV_DRAM =3D 0xe, +}; + struct arm_spe_record { enum arm_spe_sample_type type; int err; @@ -40,6 +51,7 @@ struct arm_spe_record { u64 virt_addr; u64 phys_addr; u64 context_id; + u16 source; }; =20 struct arm_spe_insn; diff --git a/tools/perf/util/arm-spe.c b/tools/perf/util/arm-spe.c index e032efc03274..4065fa180140 100644 --- a/tools/perf/util/arm-spe.c +++ b/tools/perf/util/arm-spe.c @@ -34,6 +34,7 @@ #include "arm-spe-decoder/arm-spe-decoder.h" #include "arm-spe-decoder/arm-spe-pkt-decoder.h" =20 +#include "../../arch/arm64/include/asm/cputype.h" #define MAX_TIMESTAMP (~0ULL) =20 struct arm_spe { @@ -45,6 +46,7 @@ struct arm_spe { struct perf_session *session; struct machine *machine; u32 pmu_type; + u64 midr; =20 struct perf_tsc_conversion tc; =20 @@ -387,35 +389,128 @@ static int arm_spe__synth_instruction_sample(struct = arm_spe_queue *speq, return arm_spe_deliver_synth_event(spe, speq, event, &sample); } =20 -static u64 arm_spe__synth_data_source(const struct arm_spe_record *record) +static const struct midr_range neoverse_spe[] =3D { + MIDR_ALL_VERSIONS(MIDR_NEOVERSE_N1), + MIDR_ALL_VERSIONS(MIDR_NEOVERSE_N2), + MIDR_ALL_VERSIONS(MIDR_NEOVERSE_V1), + {}, +}; + +static void arm_spe__synth_data_source_neoverse(const struct arm_spe_recor= d *record, + union perf_mem_data_src *data_src) { - union perf_mem_data_src data_src =3D { 0 }; + /* + * Even though four levels of cache hierarchy are possible, no known + * production Neoverse systems currently include more than three levels + * so for the time being we assume three exist. If a production system + * is built with four the this function would have to be changed to + * detect the number of levels for reporting. + */ =20 - if (record->op =3D=3D ARM_SPE_LD) - data_src.mem_op =3D PERF_MEM_OP_LOAD; - else if (record->op =3D=3D ARM_SPE_ST) - data_src.mem_op =3D PERF_MEM_OP_STORE; - else - return 0; + /* + * We have no data on the hit level or data source for stores in the + * Neoverse SPE records. + */ + if (record->op & ARM_SPE_ST) { + data_src->mem_lvl =3D PERF_MEM_LVL_NA; + data_src->mem_lvl_num =3D PERF_MEM_LVLNUM_ANY_CACHE; + data_src->mem_snoop =3D PERF_MEM_SNOOP_NA; + return; + } + + switch (record->source) { + case ARM_SPE_NV_L1D: + data_src->mem_lvl =3D PERF_MEM_LVL_L1 | PERF_MEM_LVL_HIT; + data_src->mem_lvl_num =3D PERF_MEM_LVLNUM_L1; + data_src->mem_snoop =3D PERF_MEM_SNOOP_NONE; + break; + case ARM_SPE_NV_L2: + data_src->mem_lvl =3D PERF_MEM_LVL_L2 | PERF_MEM_LVL_HIT; + data_src->mem_lvl_num =3D PERF_MEM_LVLNUM_L2; + data_src->mem_snoop =3D PERF_MEM_SNOOP_NONE; + break; + case ARM_SPE_NV_PEER_CORE: + data_src->mem_lvl =3D PERF_MEM_LVL_L2 | PERF_MEM_LVL_HIT; + data_src->mem_lvl_num =3D PERF_MEM_LVLNUM_L2; + data_src->mem_snoopx =3D PERF_MEM_SNOOPX_PEER; + break; + /* + * We don't know if this is L1, L2 but we do know it was a cache-2-cache + * transfer, so set SNOOPX_PEER + */ + case ARM_SPE_NV_LOCAL_CLUSTER: + case ARM_SPE_NV_PEER_CLUSTER: + data_src->mem_lvl =3D PERF_MEM_LVL_L3 | PERF_MEM_LVL_HIT; + data_src->mem_lvl_num =3D PERF_MEM_LVLNUM_L3; + data_src->mem_snoopx =3D PERF_MEM_SNOOPX_PEER; + break; + /* + * System cache is assumed to be L3 + */ + case ARM_SPE_NV_SYS_CACHE: + data_src->mem_lvl =3D PERF_MEM_LVL_L3 | PERF_MEM_LVL_HIT; + data_src->mem_lvl_num =3D PERF_MEM_LVLNUM_L3; + data_src->mem_snoop =3D PERF_MEM_SNOOP_HIT; + break; + /* + * We don't know what level it hit in, except it came from the other + * socket + */ + case ARM_SPE_NV_REMOTE: + data_src->mem_lvl =3D PERF_MEM_LVL_REM_RAM1 | PERF_MEM_LVL_HIT; + data_src->mem_lvl_num =3D PERF_MEM_LVLNUM_RAM; + data_src->mem_remote =3D PERF_MEM_REMOTE_REMOTE; + data_src->mem_snoop =3D PERF_MEM_SNOOP_NA; + break; + case ARM_SPE_NV_DRAM: + data_src->mem_lvl =3D PERF_MEM_LVL_LOC_RAM | PERF_MEM_LVL_HIT; + data_src->mem_lvl_num =3D PERF_MEM_LVLNUM_RAM; + data_src->mem_snoop =3D PERF_MEM_SNOOP_NONE; + break; + default: + break; + } +} =20 +static void arm_spe__synth_data_source_generic(const struct arm_spe_record= *record, + union perf_mem_data_src *data_src) +{ if (record->type & (ARM_SPE_LLC_ACCESS | ARM_SPE_LLC_MISS)) { - data_src.mem_lvl =3D PERF_MEM_LVL_L3; + data_src->mem_lvl =3D PERF_MEM_LVL_L3; =20 if (record->type & ARM_SPE_LLC_MISS) - data_src.mem_lvl |=3D PERF_MEM_LVL_MISS; + data_src->mem_lvl |=3D PERF_MEM_LVL_MISS; else - data_src.mem_lvl |=3D PERF_MEM_LVL_HIT; + data_src->mem_lvl |=3D PERF_MEM_LVL_HIT; } else if (record->type & (ARM_SPE_L1D_ACCESS | ARM_SPE_L1D_MISS)) { - data_src.mem_lvl =3D PERF_MEM_LVL_L1; + data_src->mem_lvl =3D PERF_MEM_LVL_L1; =20 if (record->type & ARM_SPE_L1D_MISS) - data_src.mem_lvl |=3D PERF_MEM_LVL_MISS; + data_src->mem_lvl |=3D PERF_MEM_LVL_MISS; else - data_src.mem_lvl |=3D PERF_MEM_LVL_HIT; + data_src->mem_lvl |=3D PERF_MEM_LVL_HIT; } =20 if (record->type & ARM_SPE_REMOTE_ACCESS) - data_src.mem_lvl |=3D PERF_MEM_LVL_REM_CCE1; + data_src->mem_lvl |=3D PERF_MEM_LVL_REM_CCE1; +} + +static u64 arm_spe__synth_data_source(const struct arm_spe_record *record,= u64 midr) +{ + union perf_mem_data_src data_src =3D { 0 }; + bool is_neoverse =3D is_midr_in_range(midr, neoverse_spe); + + if (record->op =3D=3D ARM_SPE_LD) + data_src.mem_op =3D PERF_MEM_OP_LOAD; + else if (record->op =3D=3D ARM_SPE_ST) + data_src.mem_op =3D PERF_MEM_OP_STORE; + else + return 0; + + if (is_neoverse) + arm_spe__synth_data_source_neoverse(record, &data_src); + else + arm_spe__synth_data_source_generic(record, &data_src); =20 if (record->type & (ARM_SPE_TLB_ACCESS | ARM_SPE_TLB_MISS)) { data_src.mem_dtlb =3D PERF_MEM_TLB_WK; @@ -436,7 +531,7 @@ static int arm_spe_sample(struct arm_spe_queue *speq) u64 data_src; int err; =20 - data_src =3D arm_spe__synth_data_source(record); + data_src =3D arm_spe__synth_data_source(record, spe->midr); =20 if (spe->sample_flc) { if (record->type & ARM_SPE_L1D_MISS) { @@ -1177,6 +1272,8 @@ int arm_spe_process_auxtrace_info(union perf_event *e= vent, struct perf_record_auxtrace_info *auxtrace_info =3D &event->auxtrace_info; size_t min_sz =3D sizeof(u64) * ARM_SPE_AUXTRACE_PRIV_MAX; struct perf_record_time_conv *tc =3D &session->time_conv; + const char *cpuid =3D perf_env__cpuid(session->evlist->env); + u64 midr =3D strtol(cpuid, NULL, 16); struct arm_spe *spe; int err; =20 @@ -1196,6 +1293,7 @@ int arm_spe_process_auxtrace_info(union perf_event *e= vent, spe->machine =3D &session->machines.host; /* No kvm support */ spe->auxtrace_type =3D auxtrace_info->type; spe->pmu_type =3D auxtrace_info->priv[ARM_SPE_PMU_TYPE]; + spe->midr =3D midr; =20 spe->timeless_decoding =3D arm_spe__is_timeless_decoding(spe); =20 --=20 2.32.0