From mboxrd@z Thu Jan 1 00:00:00 1970 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 bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 5DBACCFA441 for ; Thu, 20 Nov 2025 21:25:46 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender:List-Subscribe:List-Help :List-Post:List-Archive:List-Unsubscribe:List-Id:Content-Transfer-Encoding: MIME-Version:References:In-Reply-To:Message-ID:Date:Subject:Cc:To:From: Reply-To:Content-Type:Content-ID:Content-Description:Resent-Date:Resent-From: Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Owner; bh=Pk3sYj4Vzr/Cj6+lYsqjDDXoCljQqc04IL8G1nhA0BY=; b=WorYLafWbSDw32KYDFwIx3EbiO Ka0FgRLv7ClbSGyQaljNrNF0Hrav0yvdzlpULghP2gbphl3kJK/WPWSNM/opb/XqLjZLEhvSnt0EL TRe4G84gj1Tfx4SRvE2v4x1nKry/BOMGcnNv5as3ljmZgIFneVHelkH4CxwtXCrRiph8d9LVU1VnE Cj12eOIfbFwB9S+lVP/hcm0Hq46fij+8Zau0Lg9b46ajiklaXfok4sox6dxADfJ5Qpd+wZ3kZB1Uj k4HvQO3/RBPOGaP76EzxGnVuG2/5rMOHOQfW8cHFxqvrY8a1nlQLqVoAkYAlE3QDLvP/Ur/L6nR/R zkVsfGEQ==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.98.2 #2 (Red Hat Linux)) id 1vMC9t-00000007S9i-10fB; Thu, 20 Nov 2025 21:25:41 +0000 Received: from sea.source.kernel.org ([2600:3c0a:e001:78e:0:1991:8:25]) by bombadil.infradead.org with esmtps (Exim 4.98.2 #2 (Red Hat Linux)) id 1vMC9p-00000007S6V-0Ejt for linux-arm-kernel@lists.infradead.org; Thu, 20 Nov 2025 21:25:40 +0000 Received: from smtp.kernel.org (transwarp.subspace.kernel.org [100.75.92.58]) by sea.source.kernel.org (Postfix) with ESMTP id B660644339; Thu, 20 Nov 2025 21:25:36 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 940F6C4CEF1; Thu, 20 Nov 2025 21:25:33 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1763673936; bh=tWsIX28eUU09Y+4XW8T9/1bbDivP+CjAgctxdhSf0nM=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=G5bdWYXbUX7EtiNCcmeVlhB9VsPW+pdJ9kuu2dZxAM15q+VCh9jqvzzI3YvzYjLV7 WyY5d3z1KexwMY14vA19/UaPPX15FsSDX6FpNXEsDEYCyG/sQDY3zicRRnWS1pLzhO aa2Prrs92VBCtpZ5ZYadyKGTB1kn2Kdq5FDNueFlsU3swjzCOtWaAMhcorraHx1kUa xqZnOLkg3uejk5h7l1ZbwWY7UqtphvcIemA1lUQRCHV7vCxoLeBbJpI5ROaqleSApi fr0/L4gigkYqmCUnAguD0B8uXObigT8WS7R/VGoIxChs3B0sKAU+g4gqY0YVpIl8y1 DJ9a8t03RKUGA== From: Jiri Olsa To: Steven Rostedt , Florent Revest , Mark Rutland Cc: bpf@vger.kernel.org, linux-kernel@vger.kernel.org, linux-trace-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org, Alexei Starovoitov , Daniel Borkmann , Andrii Nakryiko , Menglong Dong , Song Liu Subject: [PATCHv3 bpf-next 8/8] bpf, x86: Use single ftrace_ops for direct calls Date: Thu, 20 Nov 2025 22:24:02 +0100 Message-ID: <20251120212402.466524-9-jolsa@kernel.org> X-Mailer: git-send-email 2.51.1 In-Reply-To: <20251120212402.466524-1-jolsa@kernel.org> References: <20251120212402.466524-1-jolsa@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20251120_132538_929004_36B3F21E X-CRM114-Status: GOOD ( 20.87 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org Using single ftrace_ops for direct calls update instead of allocating ftrace_ops object for each trampoline. With single ftrace_ops object we can use update_ftrace_direct_* api that allows multiple ip sites updates on single ftrace_ops object. Adding HAVE_SINGLE_FTRACE_DIRECT_OPS config option to be enabled on each arch that supports this. At the moment we can enable this only on x86 arch, because arm relies on ftrace_ops object representing just single trampoline image (stored in ftrace_ops::direct_call). Ach that do not support this will continue to use *_ftrace_direct api. Signed-off-by: Jiri Olsa --- arch/x86/Kconfig | 1 + kernel/bpf/trampoline.c | 172 +++++++++++++++++++++++++++++++++++----- kernel/trace/Kconfig | 3 + kernel/trace/ftrace.c | 7 +- 4 files changed, 164 insertions(+), 19 deletions(-) diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index fa3b616af03a..65a2fc279b46 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -332,6 +332,7 @@ config X86 select SCHED_SMT if SMP select ARCH_SUPPORTS_SCHED_CLUSTER if SMP select ARCH_SUPPORTS_SCHED_MC if SMP + select HAVE_SINGLE_FTRACE_DIRECT_OPS if X86_64 && DYNAMIC_FTRACE_WITH_DIRECT_CALLS config INSTRUCTION_DECODER def_bool y diff --git a/kernel/bpf/trampoline.c b/kernel/bpf/trampoline.c index d8762f7990ed..9cc4de215df7 100644 --- a/kernel/bpf/trampoline.c +++ b/kernel/bpf/trampoline.c @@ -33,12 +33,40 @@ static DEFINE_MUTEX(trampoline_mutex); #ifdef CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS static int bpf_trampoline_update(struct bpf_trampoline *tr, bool lock_direct_mutex); +#ifdef CONFIG_HAVE_SINGLE_FTRACE_DIRECT_OPS +static struct bpf_trampoline *direct_ops_ip_lookup(struct ftrace_ops *ops, unsigned long ip) +{ + struct hlist_head *head_ip; + struct bpf_trampoline *tr; + + mutex_lock(&trampoline_mutex); + head_ip = &trampoline_ip_table[hash_64(ip, TRAMPOLINE_HASH_BITS)]; + hlist_for_each_entry(tr, head_ip, hlist_ip) { + if (tr->ip == ip) + goto out; + } + tr = NULL; +out: + mutex_unlock(&trampoline_mutex); + return tr; +} +#else +static struct bpf_trampoline *direct_ops_ip_lookup(struct ftrace_ops *ops, unsigned long ip) +{ + return ops->private; +} +#endif /* CONFIG_HAVE_SINGLE_FTRACE_DIRECT_OPS */ + static int bpf_tramp_ftrace_ops_func(struct ftrace_ops *ops, unsigned long ip, enum ftrace_ops_cmd cmd) { - struct bpf_trampoline *tr = ops->private; + struct bpf_trampoline *tr; int ret = 0; + tr = direct_ops_ip_lookup(ops, ip); + if (!tr) + return -EINVAL; + if (cmd == FTRACE_OPS_CMD_ENABLE_SHARE_IPMODIFY_SELF) { /* This is called inside register_ftrace_direct_multi(), so * tr->mutex is already locked. @@ -137,6 +165,126 @@ void bpf_image_ksym_del(struct bpf_ksym *ksym) PAGE_SIZE, true, ksym->name); } +#ifdef CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS +#ifdef CONFIG_HAVE_SINGLE_FTRACE_DIRECT_OPS +/* + * We have only single direct_ops which contains all the direct call + * sites and is the only global ftrace_ops for all trampolines. + * + * We use 'update_ftrace_direct_*' api for attachment. + */ +struct ftrace_ops direct_ops = { + .ops_func = bpf_tramp_ftrace_ops_func, +}; + +static int direct_ops_alloc(struct bpf_trampoline *tr) +{ + tr->fops = &direct_ops; + return 0; +} + +static void direct_ops_free(struct bpf_trampoline *tr) { } + +static struct ftrace_hash *hash_from(unsigned long ip, void *addr) +{ + struct ftrace_hash *hash; + + ip = ftrace_location(ip); + if (!ip) + return NULL; + hash = alloc_ftrace_hash(FTRACE_HASH_DEFAULT_BITS); + if (!hash) + return NULL; + if (!add_hash_entry_direct(hash, ip, (unsigned long) addr)) { + free_ftrace_hash(hash); + return NULL; + } + return hash; +} + +static int direct_ops_add(struct ftrace_ops *ops, unsigned long ip, void *addr) +{ + struct ftrace_hash *hash = hash_from(ip, addr); + int err = -ENOMEM; + + if (hash) + err = update_ftrace_direct_add(ops, hash); + free_ftrace_hash(hash); + return err; +} + +static int direct_ops_del(struct ftrace_ops *ops, unsigned long ip, void *addr) +{ + struct ftrace_hash *hash = hash_from(ip, addr); + int err = -ENOMEM; + + if (hash) + err = update_ftrace_direct_del(ops, hash); + free_ftrace_hash(hash); + return err; +} + +static int direct_ops_mod(struct ftrace_ops *ops, unsigned long ip, void *addr, bool lock_direct_mutex) +{ + struct ftrace_hash *hash = hash_from(ip, addr); + int err = -ENOMEM; + + if (hash) + err = update_ftrace_direct_mod(ops, hash, lock_direct_mutex); + free_ftrace_hash(hash); + return err; +} +#else +/* + * We allocate ftrace_ops object for each trampoline and it contains + * call site specific for that trampoline. + * + * We use *_ftrace_direct api for attachment. + */ +static int direct_ops_alloc(struct bpf_trampoline *tr) +{ + tr->fops = kzalloc(sizeof(struct ftrace_ops), GFP_KERNEL); + if (!tr->fops) + return -ENOMEM; + tr->fops->private = tr; + tr->fops->ops_func = bpf_tramp_ftrace_ops_func; + return 0; +} + +static void direct_ops_free(struct bpf_trampoline *tr) +{ + if (tr->fops) { + ftrace_free_filter(tr->fops); + kfree(tr->fops); + } +} + +static int direct_ops_add(struct ftrace_ops *ops, unsigned long ip, void *addr) +{ + int ret; + + ret = ftrace_set_filter_ip(ops, (unsigned long)ip, 0, 1); + if (ret) + return ret; + return register_ftrace_direct(ops, (long)addr); +} + +static int direct_ops_del(struct ftrace_ops *ops, unsigned long ip, void *addr) +{ + return unregister_ftrace_direct(ops, (long)addr, false); +} + +static int direct_ops_mod(struct ftrace_ops *ops, unsigned long ip, void *addr, bool lock_direct_mutex) +{ + if (lock_direct_mutex) + return modify_ftrace_direct(ops, (long)addr); + return modify_ftrace_direct_nolock(ops, (long)addr); +} +#endif /* CONFIG_HAVE_SINGLE_FTRACE_DIRECT_OPS */ +#else +static void direct_ops_free(struct bpf_trampoline *tr) { } +#endif /* CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS */ + static struct bpf_trampoline *bpf_trampoline_lookup(u64 key, unsigned long ip) { struct bpf_trampoline *tr; @@ -155,14 +303,11 @@ static struct bpf_trampoline *bpf_trampoline_lookup(u64 key, unsigned long ip) if (!tr) goto out; #ifdef CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS - tr->fops = kzalloc(sizeof(struct ftrace_ops), GFP_KERNEL); - if (!tr->fops) { + if (direct_ops_alloc(tr)) { kfree(tr); tr = NULL; goto out; } - tr->fops->private = tr; - tr->fops->ops_func = bpf_tramp_ftrace_ops_func; #endif tr->key = key; @@ -187,7 +332,7 @@ static int unregister_fentry(struct bpf_trampoline *tr, void *old_addr) int ret; if (tr->func.ftrace_managed) - ret = unregister_ftrace_direct(tr->fops, (long)old_addr, false); + ret = direct_ops_del(tr->fops, tr->ip, old_addr); else ret = bpf_arch_text_poke(ip, BPF_MOD_CALL, old_addr, NULL); @@ -201,10 +346,7 @@ static int modify_fentry(struct bpf_trampoline *tr, void *old_addr, void *new_ad int ret; if (tr->func.ftrace_managed) { - if (lock_direct_mutex) - ret = modify_ftrace_direct(tr->fops, (long)new_addr); - else - ret = modify_ftrace_direct_nolock(tr->fops, (long)new_addr); + ret = direct_ops_mod(tr->fops, tr->ip, new_addr, lock_direct_mutex); } else { ret = bpf_arch_text_poke(ip, BPF_MOD_CALL, old_addr, new_addr); } @@ -226,10 +368,7 @@ static int register_fentry(struct bpf_trampoline *tr, void *new_addr) } if (tr->func.ftrace_managed) { - ret = ftrace_set_filter_ip(tr->fops, (unsigned long)ip, 0, 1); - if (ret) - return ret; - ret = register_ftrace_direct(tr->fops, (long)new_addr); + ret = direct_ops_add(tr->fops, tr->ip, new_addr); } else { ret = bpf_arch_text_poke(ip, BPF_MOD_CALL, NULL, new_addr); } @@ -865,10 +1004,7 @@ void bpf_trampoline_put(struct bpf_trampoline *tr) */ hlist_del(&tr->hlist_key); hlist_del(&tr->hlist_ip); - if (tr->fops) { - ftrace_free_filter(tr->fops); - kfree(tr->fops); - } + direct_ops_free(tr); kfree(tr); out: mutex_unlock(&trampoline_mutex); diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index d2c79da81e4f..4bf5beb04a5b 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -50,6 +50,9 @@ config HAVE_DYNAMIC_FTRACE_WITH_REGS config HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS bool +config HAVE_SINGLE_FTRACE_DIRECT_OPS + bool + config HAVE_DYNAMIC_FTRACE_WITH_CALL_OPS bool diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index b0e4d7a5d47c..4380839d2422 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -2605,8 +2605,13 @@ unsigned long ftrace_find_rec_direct(unsigned long ip) static void call_direct_funcs(unsigned long ip, unsigned long pip, struct ftrace_ops *ops, struct ftrace_regs *fregs) { - unsigned long addr = READ_ONCE(ops->direct_call); + unsigned long addr; +#ifdef CONFIG_HAVE_SINGLE_FTRACE_DIRECT_OPS + addr = ftrace_find_rec_direct(ip); +#else + addr = READ_ONCE(ops->direct_call); +#endif if (!addr) return; -- 2.51.1