From nobody Sun Nov 24 12:47:53 2024 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 01C141E2018; Tue, 5 Nov 2024 17:26:39 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1730827600; cv=none; b=LqvAI1hhUWm/nAknN8BqVzkOp/QjxFeE7TInqqcImYJudLfJcVFcrRzUeef62YW8cjWyBCIcMs8JamOoxdwHt6Rc3yx34/chZ9wVNS6dt7CNxl8TI5lZPB3pChjmIjhBaOwNBizzfcEiPV6Yke0hjnz/OaVghJS+0feOIrjAY2I= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1730827600; c=relaxed/simple; bh=pyZ5l6v8C0oJhH7ls8q9xbjc6qgfiX0PBuXO5xZ4WWc=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=LNzRIaKZNxmWHNP0Jrtzb31H/donbtYKpLc/LwIyGSLF3ezD0wXhTqKRA1Nib2FuHw50apU73tWshE0aaC9kW6bq2yWk0TFCrvEYwQ7I5v+lG6VR7Nwcsq7kddqF8zrlw4YAMPwokiXpKechhteDdIkanHuppxF9je4gOpCO0uI= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=gYUuRW9/; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="gYUuRW9/" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 1F346C4CED5; Tue, 5 Nov 2024 17:26:39 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1730827599; bh=pyZ5l6v8C0oJhH7ls8q9xbjc6qgfiX0PBuXO5xZ4WWc=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=gYUuRW9/4pzD90KTibV861Zu0JILvmD5bPZtre9llO3oIG3850sVymQEILRb0JFU2 7RTEYbj5VXSmAThObrJubApbalYiVF+Y0uNFkr0tNPs5bXXWIJtJkx5HZIH30OTrhN wJe6n+kTW3A3uYmvdbtFyBtdacEnBV8puQvbhRioipJ8DQGkMJWVfLfxx7ZXlskksq BPdFNoo9GPzlLBv2rbKq7KL3bXKnzROOoA/+7t887LdMvYRqXGV6U7bcRqYuv3/1qY JCEWLdVJSi4uKjaX5+0ITL2SXAHEhXe4WP0Wkxg6yLcpxPQNC8a7G/+DbbUygcFvN6 MXXMcUxGFXvTw== From: Namhyung Kim To: Arnaldo Carvalho de Melo , Ian Rogers , Kan Liang Cc: Jiri Olsa , Adrian Hunter , Peter Zijlstra , Ingo Molnar , LKML , linux-perf-users@vger.kernel.org, Song Liu , bpf@vger.kernel.org, Stephane Eranian , Vlastimil Babka , Kees Cook , Roman Gushchin , Hyeonggon Yoo <42.hyeyoo@gmail.com> Subject: [PATCH 1/4] perf lock contention: Add and use LCB_F_TYPE_MASK Date: Tue, 5 Nov 2024 09:26:32 -0800 Message-ID: <20241105172635.2463800-2-namhyung@kernel.org> X-Mailer: git-send-email 2.47.0.199.ga7371fff76-goog In-Reply-To: <20241105172635.2463800-1-namhyung@kernel.org> References: <20241105172635.2463800-1-namhyung@kernel.org> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" This is a preparation for the later change. It'll use more bits in the flags so let's rename the type part and use the mask to extract the type. Signed-off-by: Namhyung Kim --- tools/perf/builtin-lock.c | 4 ++-- tools/perf/util/bpf_skel/lock_data.h | 3 ++- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/tools/perf/builtin-lock.c b/tools/perf/builtin-lock.c index 062e2b56a2ab570e..89ee2a2f78603906 100644 --- a/tools/perf/builtin-lock.c +++ b/tools/perf/builtin-lock.c @@ -1597,7 +1597,7 @@ static const struct { =20 static const char *get_type_str(unsigned int flags) { - flags &=3D LCB_F_MAX_FLAGS - 1; + flags &=3D LCB_F_TYPE_MASK; =20 for (unsigned int i =3D 0; i < ARRAY_SIZE(lock_type_table); i++) { if (lock_type_table[i].flags =3D=3D flags) @@ -1608,7 +1608,7 @@ static const char *get_type_str(unsigned int flags) =20 static const char *get_type_name(unsigned int flags) { - flags &=3D LCB_F_MAX_FLAGS - 1; + flags &=3D LCB_F_TYPE_MASK; =20 for (unsigned int i =3D 0; i < ARRAY_SIZE(lock_type_table); i++) { if (lock_type_table[i].flags =3D=3D flags) diff --git a/tools/perf/util/bpf_skel/lock_data.h b/tools/perf/util/bpf_ske= l/lock_data.h index de12892f992f8d43..4f0aae5483745dfa 100644 --- a/tools/perf/util/bpf_skel/lock_data.h +++ b/tools/perf/util/bpf_skel/lock_data.h @@ -32,7 +32,8 @@ struct contention_task_data { #define LCD_F_MMAP_LOCK (1U << 31) #define LCD_F_SIGHAND_LOCK (1U << 30) =20 -#define LCB_F_MAX_FLAGS (1U << 7) +#define LCB_F_TYPE_MAX (1U << 7) +#define LCB_F_TYPE_MASK 0x0000007FU =20 struct contention_data { u64 total_time; --=20 2.47.0.199.ga7371fff76-goog From nobody Sun Nov 24 12:47:53 2024 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id B5ACC1E2851; Tue, 5 Nov 2024 17:26:40 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1730827600; cv=none; b=XmRCSOlnl81h2NsLYrmGJ0orC7QawFa9oE65NU81P+D+hW/HC6jZtU8G/nt2H55uZZDD9CXcdxAMw5LBB/DRSTuJPAVopFRJufr34Xqphl+oLKUuOkfjqOocFlJViWLzeAbEz4OGj/z63iGOJ+hpuO5jwunf70Q3TZ8qlUD1qMA= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1730827600; c=relaxed/simple; bh=Q7PJcEHIX5DpXW3MulNYQgSzp8YB6bzZUwfUXY/+6Ic=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=XcPOTlv2zzXeLod7fTFKsm3bBkorFgOyUC4UoELP3STIm8XTt9c4CzwlR0ZNv3YM5uJoaqgOoAGr6UOQXETCkw+KSlRf/IilGCRZIzE2H71J1OiF0l7/lKD/3CEi6x2EzrUIxT+QjaXbDXGdoOlRUed5ksDm4w/M3x8YXeSFZRI= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=TBJDSAnk; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="TBJDSAnk" Received: by smtp.kernel.org (Postfix) with ESMTPSA id BADB5C4CED3; Tue, 5 Nov 2024 17:26:39 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1730827600; bh=Q7PJcEHIX5DpXW3MulNYQgSzp8YB6bzZUwfUXY/+6Ic=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=TBJDSAnklCnZmD0dVQkFVb1qVFRrVAOQTNb8UyrHxkMPZhF74wAEO5MpQysVpLZ0Q tcH101yUySYf61QfYPOrqxgzxQGcivcgqEfSmhOmH2Cy2f4Rao5+TTtN4BHMI+/5SG fECOdKsRnyzbpuO4rZGYUSxGjgyiH3PA9M9J5K+ZC2xoS8/Umq0J3HNLCYGHL+CjGm h765/Gjv8nq6uQYSBWVM0Zbc/XwWjCbuK7t/i0LTANNb3xsbI7Ooofp+D1ujA3BFXZ ce59Q2oFhGxAfgUda2bR6PRGQ3WhUg++3iDYmgB4r/Yke7P8CvJxuuPVw62fY3Lhwf DTiCnjC7qyhfQ== From: Namhyung Kim To: Arnaldo Carvalho de Melo , Ian Rogers , Kan Liang Cc: Jiri Olsa , Adrian Hunter , Peter Zijlstra , Ingo Molnar , LKML , linux-perf-users@vger.kernel.org, Song Liu , bpf@vger.kernel.org, Stephane Eranian , Vlastimil Babka , Kees Cook , Roman Gushchin , Hyeonggon Yoo <42.hyeyoo@gmail.com> Subject: [PATCH 2/4] perf lock contention: Run BPF slab cache iterator Date: Tue, 5 Nov 2024 09:26:33 -0800 Message-ID: <20241105172635.2463800-3-namhyung@kernel.org> X-Mailer: git-send-email 2.47.0.199.ga7371fff76-goog In-Reply-To: <20241105172635.2463800-1-namhyung@kernel.org> References: <20241105172635.2463800-1-namhyung@kernel.org> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Recently the kernel got the kmem_cache iterator to traverse metadata of slab objects. This can be used to symbolize dynamic locks in a slab. The new slab_caches hash map will have the pointer of the kmem_cache as a key and save the name and a id. The id will be saved in the flags part of the lock. Signed-off-by: Namhyung Kim --- tools/perf/util/bpf_lock_contention.c | 51 +++++++++++++++++++ .../perf/util/bpf_skel/lock_contention.bpf.c | 28 ++++++++++ tools/perf/util/bpf_skel/lock_data.h | 12 +++++ tools/perf/util/bpf_skel/vmlinux/vmlinux.h | 8 +++ 4 files changed, 99 insertions(+) diff --git a/tools/perf/util/bpf_lock_contention.c b/tools/perf/util/bpf_lo= ck_contention.c index 41a1ad08789511c3..a2efd40897bad316 100644 --- a/tools/perf/util/bpf_lock_contention.c +++ b/tools/perf/util/bpf_lock_contention.c @@ -12,12 +12,60 @@ #include #include #include +#include #include =20 #include "bpf_skel/lock_contention.skel.h" #include "bpf_skel/lock_data.h" =20 static struct lock_contention_bpf *skel; +static bool has_slab_iter; + +static void check_slab_cache_iter(struct lock_contention *con) +{ + struct btf *btf =3D btf__load_vmlinux_btf(); + s32 ret; + + ret =3D libbpf_get_error(btf); + if (ret) { + pr_debug("BTF loading failed: %d\n", ret); + return; + } + + ret =3D btf__find_by_name_kind(btf, "bpf_iter__kmem_cache", BTF_KIND_STRU= CT); + if (ret < 0) { + bpf_program__set_autoload(skel->progs.slab_cache_iter, false); + pr_debug("slab cache iterator is not available: %d\n", ret); + goto out; + } + + has_slab_iter =3D true; + + bpf_map__set_max_entries(skel->maps.slab_caches, con->map_nr_entries); +out: + btf__free(btf); +} + +static void run_slab_cache_iter(void) +{ + int fd; + char buf[256]; + + if (!has_slab_iter) + return; + + fd =3D bpf_iter_create(bpf_link__fd(skel->links.slab_cache_iter)); + if (fd < 0) { + pr_debug("cannot create slab cache iter: %d\n", fd); + return; + } + + /* This will run the bpf program */ + while (read(fd, buf, sizeof(buf)) > 0) + continue; + + close(fd); +} =20 int lock_contention_prepare(struct lock_contention *con) { @@ -109,6 +157,8 @@ int lock_contention_prepare(struct lock_contention *con) skel->rodata->use_cgroup_v2 =3D 1; } =20 + check_slab_cache_iter(con); + if (lock_contention_bpf__load(skel) < 0) { pr_err("Failed to load lock-contention BPF skeleton\n"); return -1; @@ -304,6 +354,7 @@ static void account_end_timestamp(struct lock_contentio= n *con) =20 int lock_contention_start(void) { + run_slab_cache_iter(); skel->bss->enabled =3D 1; return 0; } diff --git a/tools/perf/util/bpf_skel/lock_contention.bpf.c b/tools/perf/ut= il/bpf_skel/lock_contention.bpf.c index 1069bda5d733887f..fd24ccb00faec0ba 100644 --- a/tools/perf/util/bpf_skel/lock_contention.bpf.c +++ b/tools/perf/util/bpf_skel/lock_contention.bpf.c @@ -100,6 +100,13 @@ struct { __uint(max_entries, 1); } cgroup_filter SEC(".maps"); =20 +struct { + __uint(type, BPF_MAP_TYPE_HASH); + __uint(key_size, sizeof(long)); + __uint(value_size, sizeof(struct slab_cache_data)); + __uint(max_entries, 1); +} slab_caches SEC(".maps"); + struct rw_semaphore___old { struct task_struct *owner; } __attribute__((preserve_access_index)); @@ -136,6 +143,8 @@ int perf_subsys_id =3D -1; =20 __u64 end_ts; =20 +__u32 slab_cache_id; + /* error stat */ int task_fail; int stack_fail; @@ -563,4 +572,23 @@ int BPF_PROG(end_timestamp) return 0; } =20 +SEC("iter/kmem_cache") +int slab_cache_iter(struct bpf_iter__kmem_cache *ctx) +{ + struct kmem_cache *s =3D ctx->s; + struct slab_cache_data d; + + if (s =3D=3D NULL) + return 0; + + d.id =3D ++slab_cache_id << LCB_F_SLAB_ID_SHIFT; + bpf_probe_read_kernel_str(d.name, sizeof(d.name), s->name); + + if (d.id >=3D LCB_F_SLAB_ID_END) + return 0; + + bpf_map_update_elem(&slab_caches, &s, &d, BPF_NOEXIST); + return 0; +} + char LICENSE[] SEC("license") =3D "Dual BSD/GPL"; diff --git a/tools/perf/util/bpf_skel/lock_data.h b/tools/perf/util/bpf_ske= l/lock_data.h index 4f0aae5483745dfa..c15f734d7fc4aecb 100644 --- a/tools/perf/util/bpf_skel/lock_data.h +++ b/tools/perf/util/bpf_skel/lock_data.h @@ -32,9 +32,16 @@ struct contention_task_data { #define LCD_F_MMAP_LOCK (1U << 31) #define LCD_F_SIGHAND_LOCK (1U << 30) =20 +#define LCB_F_SLAB_ID_SHIFT 16 +#define LCB_F_SLAB_ID_START (1U << 16) +#define LCB_F_SLAB_ID_END (1U << 26) +#define LCB_F_SLAB_ID_MASK 0x03FF0000U + #define LCB_F_TYPE_MAX (1U << 7) #define LCB_F_TYPE_MASK 0x0000007FU =20 +#define SLAB_NAME_MAX 28 + struct contention_data { u64 total_time; u64 min_time; @@ -55,4 +62,9 @@ enum lock_class_sym { LOCK_CLASS_RQLOCK, }; =20 +struct slab_cache_data { + u32 id; + char name[SLAB_NAME_MAX]; +}; + #endif /* UTIL_BPF_SKEL_LOCK_DATA_H */ diff --git a/tools/perf/util/bpf_skel/vmlinux/vmlinux.h b/tools/perf/util/b= pf_skel/vmlinux/vmlinux.h index 4dcad7b682bdee9c..7b81d3173917fdb5 100644 --- a/tools/perf/util/bpf_skel/vmlinux/vmlinux.h +++ b/tools/perf/util/bpf_skel/vmlinux/vmlinux.h @@ -195,4 +195,12 @@ struct bpf_perf_event_data_kern { */ struct rq {}; =20 +struct kmem_cache { + const char *name; +} __attribute__((preserve_access_index)); + +struct bpf_iter__kmem_cache { + struct kmem_cache *s; +} __attribute__((preserve_access_index)); + #endif // __VMLINUX_H --=20 2.47.0.199.ga7371fff76-goog From nobody Sun Nov 24 12:47:53 2024 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 6AACC1E379F; Tue, 5 Nov 2024 17:26:41 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1730827601; cv=none; b=SVWxN9j7WRl5BkhYQ3z8/AkYxv5iWitSLnhUVQ1u+0thnwCuDojt5WgW9rSis/qjUyZw1EUgGpwLg5VRl24N0I1ittfcleqGIHkficVRZnDGckqMYEUMuqLsDr6AMYF4L1MzHk7XYQcgWhLa1WqJe9rBdZKjYBXcEKuaeu1zwZY= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1730827601; c=relaxed/simple; bh=rCrRH2SCyNzpCWakImf0vWTxXAc5rQV0bb/Ub7r5nWg=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=fe8lRsSQkHj2Bx7gdpLyATLUbwM6Gdp4Vs1KOYmdANy0UfW5ufyZ+jJP+abGDpS4GjQijGEtqzugslmQgwe1y0xgzFKiovUZbExvNVnDEwlUDHPRBKSBts1y+PKfaG9VsPwkOETRNIjZDy9B/nn3PV/vdLoKKsIY45j/70FaY1Y= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=ipRGSnc6; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="ipRGSnc6" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 6670CC4CEE4; Tue, 5 Nov 2024 17:26:40 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1730827600; bh=rCrRH2SCyNzpCWakImf0vWTxXAc5rQV0bb/Ub7r5nWg=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=ipRGSnc6C7vR56Zs1f1AZEN4VhugITS8P5E+iL0kahd5mYHMWiLarIUx1ON7xbEMa uhKH71KXi8tsklx1u/tI3uVbe63x30mMBkKm2v4M2z30XnEyoEtkXP89OtuPFebJUq FB5OqVb/Y+7HaBc3Jxq6bQ025p5EpXZWkeVmqH1Q6ilCbi3RBoSZ+RhbFfB2jeTpfW iH477Ln0WyaDlUEJbcyqhh/PNagxp91uo5cBtM+FzzOiXb6QMO4S+OMjcGgeV18eQU vyia6PMQtFTGB6vM1yYWdgQNj+kJwWYZP9GUnLR51t90R+zD0ge9/m0+m7X8oLzl0E jMxzH4ultBXzg== From: Namhyung Kim To: Arnaldo Carvalho de Melo , Ian Rogers , Kan Liang Cc: Jiri Olsa , Adrian Hunter , Peter Zijlstra , Ingo Molnar , LKML , linux-perf-users@vger.kernel.org, Song Liu , bpf@vger.kernel.org, Stephane Eranian , Vlastimil Babka , Kees Cook , Roman Gushchin , Hyeonggon Yoo <42.hyeyoo@gmail.com> Subject: [PATCH 3/4] perf lock contention: Resolve slab object name using BPF Date: Tue, 5 Nov 2024 09:26:34 -0800 Message-ID: <20241105172635.2463800-4-namhyung@kernel.org> X-Mailer: git-send-email 2.47.0.199.ga7371fff76-goog In-Reply-To: <20241105172635.2463800-1-namhyung@kernel.org> References: <20241105172635.2463800-1-namhyung@kernel.org> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" The bpf_get_kmem_cache() kfunc can return an address of the slab cache (kmem_cache). As it has the name of the slab cache from the iterator, we can use it to symbolize some dynamic kernel locks in a slab. Before: root@virtme-ng:/home/namhyung/project/linux# tools/perf/perf lock con -ab= l sleep 1 contended total wait max wait avg wait address sy= mbol 2 3.34 us 2.87 us 1.67 us ffff9d7800ad9600 (= mutex) 2 2.16 us 1.93 us 1.08 us ffff9d7804b992d8 (= mutex) 4 1.37 us 517 ns 343 ns ffff9d78036e6e00 (= mutex) 1 1.27 us 1.27 us 1.27 us ffff9d7804b99378 (= mutex) 2 845 ns 599 ns 422 ns ffffffff9e1c3620 de= layed_uprobe_lock (mutex) 1 845 ns 845 ns 845 ns ffffffff9da0b280 ji= ffies_lock (spinlock) 2 377 ns 259 ns 188 ns ffffffff9e1cf840 pc= pu_alloc_mutex (mutex) 1 305 ns 305 ns 305 ns ffffffff9e1b4cf8 tr= acepoint_srcu_srcu_usage (mutex) 1 295 ns 295 ns 295 ns ffffffff9e1c0940 pa= ck_mutex (mutex) 1 232 ns 232 ns 232 ns ffff9d7804b7d8d8 (= mutex) 1 180 ns 180 ns 180 ns ffffffff9e1b4c28 tr= acepoint_srcu_srcu_usage (mutex) 1 165 ns 165 ns 165 ns ffffffff9da8b3a0 te= xt_mutex (mutex) After: root@virtme-ng:/home/namhyung/project/linux# tools/perf/perf lock con -ab= l sleep 1 contended total wait max wait avg wait address sy= mbol 2 1.95 us 1.77 us 975 ns ffff9d5e852d3498 &t= ask_struct (mutex) 1 1.18 us 1.18 us 1.18 us ffff9d5e852d3538 &t= ask_struct (mutex) 4 1.12 us 354 ns 279 ns ffff9d5e841ca800 &k= malloc-cg-512 (mutex) 2 859 ns 617 ns 429 ns ffffffffa41c3620 de= layed_uprobe_lock (mutex) 3 691 ns 388 ns 230 ns ffffffffa41c0940 pa= ck_mutex (mutex) 3 421 ns 164 ns 140 ns ffffffffa3a8b3a0 te= xt_mutex (mutex) 1 409 ns 409 ns 409 ns ffffffffa41b4cf8 tr= acepoint_srcu_srcu_usage (mutex) 2 362 ns 239 ns 181 ns ffffffffa41cf840 pc= pu_alloc_mutex (mutex) 1 220 ns 220 ns 220 ns ffff9d5e82b534d8 &s= ignal_cache (mutex) 1 215 ns 215 ns 215 ns ffffffffa41b4c28 tr= acepoint_srcu_srcu_usage (mutex) Note that the name starts with '&' sign for slab objects to inform they are dynamic locks. It won't give the accurate lock or type names but it's still useful. We may add type info to the slab cache later to get the exact name of the lock in the type later. Signed-off-by: Namhyung Kim --- tools/perf/util/bpf_lock_contention.c | 52 +++++++++++++++++++ .../perf/util/bpf_skel/lock_contention.bpf.c | 21 +++++++- 2 files changed, 71 insertions(+), 2 deletions(-) diff --git a/tools/perf/util/bpf_lock_contention.c b/tools/perf/util/bpf_lo= ck_contention.c index a2efd40897bad316..50c3039c647d4d77 100644 --- a/tools/perf/util/bpf_lock_contention.c +++ b/tools/perf/util/bpf_lock_contention.c @@ -2,6 +2,7 @@ #include "util/cgroup.h" #include "util/debug.h" #include "util/evlist.h" +#include "util/hashmap.h" #include "util/machine.h" #include "util/map.h" #include "util/symbol.h" @@ -20,12 +21,25 @@ =20 static struct lock_contention_bpf *skel; static bool has_slab_iter; +static struct hashmap slab_hash; + +static size_t slab_cache_hash(long key, void *ctx __maybe_unused) +{ + return key; +} + +static bool slab_cache_equal(long key1, long key2, void *ctx __maybe_unuse= d) +{ + return key1 =3D=3D key2; +} =20 static void check_slab_cache_iter(struct lock_contention *con) { struct btf *btf =3D btf__load_vmlinux_btf(); s32 ret; =20 + hashmap__init(&slab_hash, slab_cache_hash, slab_cache_equal, /*ctx=3D*/NU= LL); + ret =3D libbpf_get_error(btf); if (ret) { pr_debug("BTF loading failed: %d\n", ret); @@ -50,6 +64,7 @@ static void run_slab_cache_iter(void) { int fd; char buf[256]; + long key, *prev_key; =20 if (!has_slab_iter) return; @@ -65,6 +80,34 @@ static void run_slab_cache_iter(void) continue; =20 close(fd); + + /* Read the slab cache map and build a hash with IDs */ + fd =3D bpf_map__fd(skel->maps.slab_caches); + prev_key =3D NULL; + while (!bpf_map_get_next_key(fd, prev_key, &key)) { + struct slab_cache_data *data; + + data =3D malloc(sizeof(*data)); + if (data =3D=3D NULL) + break; + + if (bpf_map_lookup_elem(fd, &key, data) < 0) + break; + + hashmap__add(&slab_hash, data->id, data); + prev_key =3D &key; + } +} + +static void exit_slab_cache_iter(void) +{ + struct hashmap_entry *cur; + unsigned bkt; + + hashmap__for_each_entry(&slab_hash, cur, bkt) + free(cur->pvalue); + + hashmap__clear(&slab_hash); } =20 int lock_contention_prepare(struct lock_contention *con) @@ -398,6 +441,7 @@ static const char *lock_contention_get_name(struct lock= _contention *con, =20 if (con->aggr_mode =3D=3D LOCK_AGGR_ADDR) { int lock_fd =3D bpf_map__fd(skel->maps.lock_syms); + struct slab_cache_data *slab_data; =20 /* per-process locks set upper bits of the flags */ if (flags & LCD_F_MMAP_LOCK) @@ -416,6 +460,12 @@ static const char *lock_contention_get_name(struct loc= k_contention *con, return "rq_lock"; } =20 + /* look slab_hash for dynamic locks in a slab object */ + if (hashmap__find(&slab_hash, flags & LCB_F_SLAB_ID_MASK, &slab_data)) { + snprintf(name_buf, sizeof(name_buf), "&%s", slab_data->name); + return name_buf; + } + return ""; } =20 @@ -590,5 +640,7 @@ int lock_contention_finish(struct lock_contention *con) cgroup__put(cgrp); } =20 + exit_slab_cache_iter(); + return 0; } diff --git a/tools/perf/util/bpf_skel/lock_contention.bpf.c b/tools/perf/ut= il/bpf_skel/lock_contention.bpf.c index fd24ccb00faec0ba..b5bc37955560a58e 100644 --- a/tools/perf/util/bpf_skel/lock_contention.bpf.c +++ b/tools/perf/util/bpf_skel/lock_contention.bpf.c @@ -123,6 +123,8 @@ struct mm_struct___new { struct rw_semaphore mmap_lock; } __attribute__((preserve_access_index)); =20 +extern struct kmem_cache *bpf_get_kmem_cache(u64 addr) __ksym __weak; + /* control flags */ const volatile int has_cpu; const volatile int has_task; @@ -496,8 +498,23 @@ int contention_end(u64 *ctx) }; int err; =20 - if (aggr_mode =3D=3D LOCK_AGGR_ADDR) - first.flags |=3D check_lock_type(pelem->lock, pelem->flags); + if (aggr_mode =3D=3D LOCK_AGGR_ADDR) { + first.flags |=3D check_lock_type(pelem->lock, + pelem->flags & LCB_F_TYPE_MASK); + + /* Check if it's from a slab object */ + if (bpf_get_kmem_cache) { + struct kmem_cache *s; + struct slab_cache_data *d; + + s =3D bpf_get_kmem_cache(pelem->lock); + if (s !=3D NULL) { + d =3D bpf_map_lookup_elem(&slab_caches, &s); + if (d !=3D NULL) + first.flags |=3D d->id; + } + } + } =20 err =3D bpf_map_update_elem(&lock_stat, &key, &first, BPF_NOEXIST); if (err < 0) { --=20 2.47.0.199.ga7371fff76-goog From nobody Sun Nov 24 12:47:53 2024 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 0900C1E572D; Tue, 5 Nov 2024 17:26:41 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1730827602; cv=none; b=ZezHTLKP8HPzBjQuVr+znx+/q2/IQCi0DJGB2hL/0F1GUi1GBunY2D8ZWIYdp0x6722LjtrFV7F/8zxvz2Mtr2xoeJ0CU+F6Lp9OancE4Q0O9lZ+O0++l+UkMTcmmmqHYGhW9cYG7QFWRf3vGFaWxctwTGF0pA8FovbHICM9ibE= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1730827602; c=relaxed/simple; bh=rklxV/2//4lWBAPsekffYOBgTTnrf7YeFUf2y+iWzfs=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=Phmxs1uH+9t8LG+J0vLv5W22piXmKJOznLYvO/zLj2++4cHFmsP1wWqXXCPG0zcm2/NF2isMXsLmlAd5MdCEX9qKAq92zlViE/D8EdWKENbwE/Pn3QNoLnR4I/F4/gKnq8y2bWdTDq7iQ0sMJGvxrAltCne06xW8lRUbTvHPthY= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=KhnfUAkN; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="KhnfUAkN" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 159DAC4AF0F; Tue, 5 Nov 2024 17:26:41 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1730827601; bh=rklxV/2//4lWBAPsekffYOBgTTnrf7YeFUf2y+iWzfs=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=KhnfUAkNd+26fCsREoBcuI2sinLC4XQnbVuXzUzDdNNNvzIv7MDvYSyjcOKLzsEDR StxuoHO+cfGPJS11GQS8aZP+2xdh8SLWWaXlOSp4yaQzVUiALtKI/fmuKBJ5zUtiRp XDwKUYpxh49/h1uPO/qItUXhKhfsuVC0tkOxev88lLGYv/ND9I2EEPn4svQnDTFsh3 EMIM5fw862gsLbxkFtUWS+tB2cNCR9xiBK5cioftotgR8feWTcJy3dw2JkLJFiU0QA OxaAA31qkgJJoSRMHka18wpH/bVpdO7Fiqjb9nyNZBzt6Uf48c0nQkWL3KuJnaStZl ZR4NJ5RGtwZEQ== From: Namhyung Kim To: Arnaldo Carvalho de Melo , Ian Rogers , Kan Liang Cc: Jiri Olsa , Adrian Hunter , Peter Zijlstra , Ingo Molnar , LKML , linux-perf-users@vger.kernel.org, Song Liu , bpf@vger.kernel.org, Stephane Eranian , Vlastimil Babka , Kees Cook , Roman Gushchin , Hyeonggon Yoo <42.hyeyoo@gmail.com> Subject: [PATCH 4/4] perf lock contention: Handle slab objects in -L/--lock-filter option Date: Tue, 5 Nov 2024 09:26:35 -0800 Message-ID: <20241105172635.2463800-5-namhyung@kernel.org> X-Mailer: git-send-email 2.47.0.199.ga7371fff76-goog In-Reply-To: <20241105172635.2463800-1-namhyung@kernel.org> References: <20241105172635.2463800-1-namhyung@kernel.org> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" This is to filter lock contention from specific slab objects only. Like in the lock symbol output, we can use '&' prefix to filter slab object names. root@virtme-ng:/home/namhyung/project/linux# tools/perf/perf lock con -ab= l sleep 1 contended total wait max wait avg wait address sy= mbol 3 14.99 us 14.44 us 5.00 us ffffffff851c0940 pa= ck_mutex (mutex) 2 2.75 us 2.56 us 1.38 us ffff98d7031fb498 &t= ask_struct (mutex) 4 1.42 us 557 ns 355 ns ffff98d706311400 &k= malloc-cg-512 (mutex) 2 953 ns 714 ns 476 ns ffffffff851c3620 de= layed_uprobe_lock (mutex) 1 929 ns 929 ns 929 ns ffff98d7031fb538 &t= ask_struct (mutex) 3 561 ns 210 ns 187 ns ffffffff84a8b3a0 te= xt_mutex (mutex) 1 479 ns 479 ns 479 ns ffffffff851b4cf8 tr= acepoint_srcu_srcu_usage (mutex) 2 320 ns 195 ns 160 ns ffffffff851cf840 pc= pu_alloc_mutex (mutex) 1 212 ns 212 ns 212 ns ffff98d7031784d8 &s= ignal_cache (mutex) 1 177 ns 177 ns 177 ns ffffffff851b4c28 tr= acepoint_srcu_srcu_usage (mutex) With the filter, it can show contentions from the task_struct only. root@virtme-ng:/home/namhyung/project/linux# tools/perf/perf lock con -ab= l -L '&task_struct' sleep 1 contended total wait max wait avg wait address sy= mbol 2 1.97 us 1.71 us 987 ns ffff98d7032fd658 &t= ask_struct (mutex) 1 1.20 us 1.20 us 1.20 us ffff98d7032fd6f8 &t= ask_struct (mutex) It can work with other aggregation mode: root@virtme-ng:/home/namhyung/project/linux# tools/perf/perf lock con -ab= -L '&task_struct' sleep 1 contended total wait max wait avg wait type caller 1 25.10 us 25.10 us 25.10 us mutex perf_eve= nt_exit_task+0x39 1 21.60 us 21.60 us 21.60 us mutex futex_ex= it_release+0x21 1 5.56 us 5.56 us 5.56 us mutex futex_ex= ec_release+0x21 Signed-off-by: Namhyung Kim --- tools/perf/builtin-lock.c | 35 ++++++++++++++++ tools/perf/util/bpf_lock_contention.c | 40 ++++++++++++++++++- .../perf/util/bpf_skel/lock_contention.bpf.c | 21 +++++++++- tools/perf/util/lock-contention.h | 2 + 4 files changed, 95 insertions(+), 3 deletions(-) diff --git a/tools/perf/builtin-lock.c b/tools/perf/builtin-lock.c index 89ee2a2f78603906..405e95666257b7fe 100644 --- a/tools/perf/builtin-lock.c +++ b/tools/perf/builtin-lock.c @@ -1646,6 +1646,12 @@ static void lock_filter_finish(void) =20 zfree(&filters.cgrps); filters.nr_cgrps =3D 0; + + for (int i =3D 0; i < filters.nr_slabs; i++) + free(filters.slabs[i]); + + zfree(&filters.slabs); + filters.nr_slabs =3D 0; } =20 static void sort_contention_result(void) @@ -2412,6 +2418,27 @@ static bool add_lock_sym(char *name) return true; } =20 +static bool add_lock_slab(char *name) +{ + char **tmp; + char *sym =3D strdup(name); + + if (sym =3D=3D NULL) { + pr_err("Memory allocation failure\n"); + return false; + } + + tmp =3D realloc(filters.slabs, (filters.nr_slabs + 1) * sizeof(*filters.s= labs)); + if (tmp =3D=3D NULL) { + pr_err("Memory allocation failure\n"); + return false; + } + + tmp[filters.nr_slabs++] =3D sym; + filters.slabs =3D tmp; + return true; +} + static int parse_lock_addr(const struct option *opt __maybe_unused, const = char *str, int unset __maybe_unused) { @@ -2435,6 +2462,14 @@ static int parse_lock_addr(const struct option *opt = __maybe_unused, const char * continue; } =20 + if (*tok =3D=3D '&') { + if (!add_lock_slab(tok + 1)) { + ret =3D -1; + break; + } + continue; + } + /* * At this moment, we don't have kernel symbols. Save the symbols * in a separate list and resolve them to addresses later. diff --git a/tools/perf/util/bpf_lock_contention.c b/tools/perf/util/bpf_lo= ck_contention.c index 50c3039c647d4d77..2891a81380204b1d 100644 --- a/tools/perf/util/bpf_lock_contention.c +++ b/tools/perf/util/bpf_lock_contention.c @@ -113,7 +113,7 @@ static void exit_slab_cache_iter(void) int lock_contention_prepare(struct lock_contention *con) { int i, fd; - int ncpus =3D 1, ntasks =3D 1, ntypes =3D 1, naddrs =3D 1, ncgrps =3D 1; + int ncpus =3D 1, ntasks =3D 1, ntypes =3D 1, naddrs =3D 1, ncgrps =3D 1, = nslabs =3D 1; struct evlist *evlist =3D con->evlist; struct target *target =3D con->target; =20 @@ -202,6 +202,13 @@ int lock_contention_prepare(struct lock_contention *co= n) =20 check_slab_cache_iter(con); =20 + if (con->filters->nr_slabs && has_slab_iter) { + skel->rodata->has_slab =3D 1; + nslabs =3D con->filters->nr_slabs; + } + + bpf_map__set_max_entries(skel->maps.slab_filter, nslabs); + if (lock_contention_bpf__load(skel) < 0) { pr_err("Failed to load lock-contention BPF skeleton\n"); return -1; @@ -272,6 +279,36 @@ int lock_contention_prepare(struct lock_contention *co= n) bpf_program__set_autoload(skel->progs.collect_lock_syms, false); =20 lock_contention_bpf__attach(skel); + + /* run the slab iterator after attaching */ + run_slab_cache_iter(); + + if (con->filters->nr_slabs) { + u8 val =3D 1; + int cache_fd; + long key, *prev_key; + + fd =3D bpf_map__fd(skel->maps.slab_filter); + + /* Read the slab cache map and build a hash with its address */ + cache_fd =3D bpf_map__fd(skel->maps.slab_caches); + prev_key =3D NULL; + while (!bpf_map_get_next_key(cache_fd, prev_key, &key)) { + struct slab_cache_data data; + + if (bpf_map_lookup_elem(cache_fd, &key, &data) < 0) + break; + + for (i =3D 0; i < con->filters->nr_slabs; i++) { + if (!strcmp(con->filters->slabs[i], data.name)) { + bpf_map_update_elem(fd, &key, &val, BPF_ANY); + break; + } + } + prev_key =3D &key; + } + } + return 0; } =20 @@ -397,7 +434,6 @@ static void account_end_timestamp(struct lock_contentio= n *con) =20 int lock_contention_start(void) { - run_slab_cache_iter(); skel->bss->enabled =3D 1; return 0; } diff --git a/tools/perf/util/bpf_skel/lock_contention.bpf.c b/tools/perf/ut= il/bpf_skel/lock_contention.bpf.c index b5bc37955560a58e..048a04fc3a7fc27d 100644 --- a/tools/perf/util/bpf_skel/lock_contention.bpf.c +++ b/tools/perf/util/bpf_skel/lock_contention.bpf.c @@ -100,6 +100,13 @@ struct { __uint(max_entries, 1); } cgroup_filter SEC(".maps"); =20 +struct { + __uint(type, BPF_MAP_TYPE_HASH); + __uint(key_size, sizeof(long)); + __uint(value_size, sizeof(__u8)); + __uint(max_entries, 1); +} slab_filter SEC(".maps"); + struct { __uint(type, BPF_MAP_TYPE_HASH); __uint(key_size, sizeof(long)); @@ -131,6 +138,7 @@ const volatile int has_task; const volatile int has_type; const volatile int has_addr; const volatile int has_cgroup; +const volatile int has_slab; const volatile int needs_callstack; const volatile int stack_skip; const volatile int lock_owner; @@ -213,7 +221,7 @@ static inline int can_record(u64 *ctx) __u64 addr =3D ctx[0]; =20 ok =3D bpf_map_lookup_elem(&addr_filter, &addr); - if (!ok) + if (!ok && !has_slab) return 0; } =20 @@ -226,6 +234,17 @@ static inline int can_record(u64 *ctx) return 0; } =20 + if (has_slab && bpf_get_kmem_cache) { + __u8 *ok; + __u64 addr =3D ctx[0]; + long kmem_cache_addr; + + kmem_cache_addr =3D (long)bpf_get_kmem_cache(addr); + ok =3D bpf_map_lookup_elem(&slab_filter, &kmem_cache_addr); + if (!ok) + return 0; + } + return 1; } =20 diff --git a/tools/perf/util/lock-contention.h b/tools/perf/util/lock-conte= ntion.h index 1a7248ff388947e1..95331b6ec062410d 100644 --- a/tools/perf/util/lock-contention.h +++ b/tools/perf/util/lock-contention.h @@ -10,10 +10,12 @@ struct lock_filter { int nr_addrs; int nr_syms; int nr_cgrps; + int nr_slabs; unsigned int *types; unsigned long *addrs; char **syms; u64 *cgrps; + char **slabs; }; =20 struct lock_stat { --=20 2.47.0.199.ga7371fff76-goog