From patchwork Thu Jan 17 23:17:28 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Song Liu X-Patchwork-Id: 1027050 X-Patchwork-Delegate: bpf@iogearbox.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=pass (p=none dis=none) header.from=fb.com Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=fb.com header.i=@fb.com header.b="MMmupVfK"; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 43gg3m1hKbz9sDn for ; Fri, 18 Jan 2019 10:18:12 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727076AbfAQXSL (ORCPT ); Thu, 17 Jan 2019 18:18:11 -0500 Received: from mx0b-00082601.pphosted.com ([67.231.153.30]:52596 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1726611AbfAQXSK (ORCPT ); Thu, 17 Jan 2019 18:18:10 -0500 Received: from pps.filterd (m0001303.ppops.net [127.0.0.1]) by m0001303.ppops.net (8.16.0.27/8.16.0.27) with SMTP id x0HNFrE9018365 for ; Thu, 17 Jan 2019 15:18:08 -0800 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=fb.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : mime-version : content-type; s=facebook; bh=sIJR6OuveyCpeaMtzqutR2EuyW3qcGB5lVGtd/HzzR8=; b=MMmupVfK1we+suFaOH/7t/zA1JAjEuNGH7qQh4yoEWjvT/oLv+pmfc9co9Wz8JPaZ6eR +wWljt3rOgY0CLyVtypEmyvYOfPR0OhXPbEY9AoTtw2xgKj28XFv/P1jNy5u6RPvgDaE jiD/BayoFuh5J9xvTyNmIyCxMPcAWEH5v/Y= Received: from mail.thefacebook.com ([199.201.64.23]) by m0001303.ppops.net with ESMTP id 2q300h0njb-14 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-SHA384 bits=256 verify=NOT) for ; Thu, 17 Jan 2019 15:18:08 -0800 Received: from mx-out.facebook.com (2620:10d:c081:10::13) by mail.thefacebook.com (2620:10d:c081:35::125) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA) id 15.1.1531.3; Thu, 17 Jan 2019 15:17:40 -0800 Received: by devbig006.ftw2.facebook.com (Postfix, from userid 4523) id 26ED062E2162; Thu, 17 Jan 2019 15:17:32 -0800 (PST) Smtp-Origin-Hostprefix: devbig From: Song Liu Smtp-Origin-Hostname: devbig006.ftw2.facebook.com To: , CC: Peter Zijlstra , , , , , Song Liu Smtp-Origin-Cluster: ftw2c04 Subject: [PATCH kallsyms, bpf 1/3] rbtree_latch: Introduce latch_tree_first() and latch_tree_next() Date: Thu, 17 Jan 2019 15:17:28 -0800 Message-ID: <20190117231730.2413466-2-songliubraving@fb.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20190117231730.2413466-1-songliubraving@fb.com> References: <20190117231730.2413466-1-songliubraving@fb.com> X-FB-Internal: Safe MIME-Version: 1.0 X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:, , definitions=2019-01-17_08:, , signatures=0 X-Proofpoint-Spam-Reason: safe X-FB-Internal: Safe Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org From: Peter Zijlstra These two functions will be used by kallsym_tree for dynamic symbols. Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Song Liu --- include/linux/rbtree_latch.h | 54 ++++++++++++++++++++++++++++++++++++ 1 file changed, 54 insertions(+) diff --git a/include/linux/rbtree_latch.h b/include/linux/rbtree_latch.h index 7d012faa509a..d0001a136d3e 100644 --- a/include/linux/rbtree_latch.h +++ b/include/linux/rbtree_latch.h @@ -211,4 +211,58 @@ latch_tree_find(void *key, struct latch_tree_root *root, return node; } +/** + * latch_tree_first() - return the first node in @root per sort order + * @root: trees to search + * + * Does a lockless lookup in the trees @root for the first node. + */ +static __always_inline struct latch_tree_node * +latch_tree_first(struct latch_tree_root *root) +{ + struct latch_tree_node *ltn = NULL; + struct rb_node *node; + unsigned int seq; + + do { + struct rb_root *rbr; + + seq = raw_read_seqcount_latch(&root->seq); + rbr = &root->tree[seq & 1]; + node = rb_first(rbr); + } while (read_seqcount_retry(&root->seq, seq)); + + if (node) + ltn = __lt_from_rb(node, seq & 1); + + return ltn; +} + +/** + * latch_tree_next() - find the next @ltn in @root per sort order + * @root: trees to which @ltn belongs + * @ltn: nodes to start from + * + * Does a lockless lookup in the trees @root for the next node starting at + * @ltn. + * + * Using this function outside of the write side lock is rather dodgy but given + * latch_tree_erase() doesn't re-init the nodes and the whole iteration is done + * under a single RCU critical section, it should be non-fatal and generate some + * semblance of order - albeit possibly missing chunks of the tree. + */ +static __always_inline struct latch_tree_node * +latch_tree_next(struct latch_tree_root *root, struct latch_tree_node *ltn) +{ + struct rb_node *node; + unsigned int seq; + + do { + seq = raw_read_seqcount_latch(&root->seq); + node = rb_next(<n->node[seq & 1]); + } while (read_seqcount_retry(&root->seq, seq)); + + return __lt_from_rb(node, seq & 1); +} + #endif /* RB_TREE_LATCH_H */ From patchwork Thu Jan 17 23:17:29 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Song Liu X-Patchwork-Id: 1027048 X-Patchwork-Delegate: bpf@iogearbox.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=pass (p=none dis=none) header.from=fb.com Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=fb.com header.i=@fb.com header.b="RXr2JzIr"; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 43gg3S3t7Bz9sDr for ; Fri, 18 Jan 2019 10:17:56 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727303AbfAQXRz (ORCPT ); Thu, 17 Jan 2019 18:17:55 -0500 Received: from mx0b-00082601.pphosted.com ([67.231.153.30]:36168 "EHLO mx0b-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727076AbfAQXRz (ORCPT ); Thu, 17 Jan 2019 18:17:55 -0500 Received: from pps.filterd (m0109332.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.27/8.16.0.27) with SMTP id x0HNFN42012735 for ; Thu, 17 Jan 2019 15:17:53 -0800 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=fb.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : mime-version : content-type; s=facebook; bh=C2DagUQrToVBOlXFmrmNKwWs8EUdFsVTIbbGz7MLRyY=; b=RXr2JzIrEJVk1JyahBFtaXCmiayRsuoupVoIPSER0V0JY9KjjBXm3wLA5vEw4vx8GAmf boNKH7AIYMu5fGN8LnFixXyIvHHdlIPLjTJn83Ls5ogcAJJPN4IvQ2x4RZf9kl/qQwgF Zi4/pS8bNgvjVh0VCbXzyH2cfNsItKRJhk0= Received: from maileast.thefacebook.com ([199.201.65.23]) by mx0a-00082601.pphosted.com with ESMTP id 2q32y5g1vc-19 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-SHA384 bits=256 verify=NOT) for ; Thu, 17 Jan 2019 15:17:53 -0800 Received: from mx-out.facebook.com (2620:10d:c0a1:3::13) by mail.thefacebook.com (2620:10d:c021:18::173) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA) id 15.1.1531.3; Thu, 17 Jan 2019 15:17:32 -0800 Received: by devbig006.ftw2.facebook.com (Postfix, from userid 4523) id 3449262E2174; Thu, 17 Jan 2019 15:17:32 -0800 (PST) Smtp-Origin-Hostprefix: devbig From: Song Liu Smtp-Origin-Hostname: devbig006.ftw2.facebook.com To: , CC: Peter Zijlstra , , , , , Song Liu Smtp-Origin-Cluster: ftw2c04 Subject: [PATCH kallsyms, bpf 2/3] kallsyms: Introduce kallsym_tree for dynamic symbols Date: Thu, 17 Jan 2019 15:17:29 -0800 Message-ID: <20190117231730.2413466-3-songliubraving@fb.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20190117231730.2413466-1-songliubraving@fb.com> References: <20190117231730.2413466-1-songliubraving@fb.com> X-FB-Internal: Safe MIME-Version: 1.0 X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:, , definitions=2019-01-17_09:, , signatures=0 X-Proofpoint-Spam-Reason: safe X-FB-Internal: Safe Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org From: Peter Zijlstra kallsym_tree is based on rbtree_latch. It is designed to hold dynamic kernel symbols like bpf program, ftrace kallsyms, etc. Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Song Liu --- include/linux/kallsyms.h | 16 ++++ kernel/extable.c | 2 + kernel/kallsyms.c | 188 ++++++++++++++++++++++++++++++++++++++- 3 files changed, 205 insertions(+), 1 deletion(-) diff --git a/include/linux/kallsyms.h b/include/linux/kallsyms.h index 657a83b943f0..be83ac3d8228 100644 --- a/include/linux/kallsyms.h +++ b/include/linux/kallsyms.h @@ -11,6 +11,8 @@ #include #include #include +#include +#include #include @@ -20,6 +22,20 @@ struct module; +struct kallsym_node +{ + struct latch_tree_node kn_node; + unsigned long kn_addr; + unsigned long kn_len; + enum perf_record_ksymbol_type ksym_type; + void (*kn_names)(struct kallsym_node *kn, char *sym_name, char **mod_name); +}; + +extern void kallsym_tree_add(struct kallsym_node *kn); +extern void kallsym_tree_del(struct kallsym_node *kn); + +extern bool is_kallsym_tree_text_address(unsigned long addr); + static inline int is_kernel_inittext(unsigned long addr) { if (addr >= (unsigned long)_sinittext diff --git a/kernel/extable.c b/kernel/extable.c index 6a5b61ebc66c..5271e9b649b1 100644 --- a/kernel/extable.c +++ b/kernel/extable.c @@ -145,6 +145,8 @@ int kernel_text_address(unsigned long addr) if (is_module_text_address(addr)) goto out; + if (is_kallsym_tree_text_address(addr)) + goto out; if (is_ftrace_trampoline(addr)) goto out; if (is_kprobe_optinsn_slot(addr) || is_kprobe_insn_slot(addr)) diff --git a/kernel/kallsyms.c b/kernel/kallsyms.c index 14934afa9e68..30611a5379fd 100644 --- a/kernel/kallsyms.c +++ b/kernel/kallsyms.c @@ -24,6 +24,8 @@ #include #include #include +#include +#include /* * These will be re-linked against their real values @@ -48,6 +50,165 @@ extern const u16 kallsyms_token_index[] __weak; extern const unsigned int kallsyms_markers[] __weak; +static DEFINE_SPINLOCK(kallsym_lock); +static struct latch_tree_root kallsym_tree __cacheline_aligned; + +static __always_inline unsigned long +kallsym_node_addr(struct latch_tree_node *node) +{ + struct kallsym_node *kn; + + kn = container_of(node, struct kallsym_node, kn_node); + return kn->kn_addr; +} + +static __always_inline bool kallsym_tree_less(struct latch_tree_node *a, + struct latch_tree_node *b) +{ + return kallsym_node_addr(a) < kallsym_node_addr(b); +} + +static __always_inline int kallsym_tree_comp(void *key, + struct latch_tree_node *n) +{ + unsigned long val = (unsigned long)key; + unsigned long sym_start, sym_end; + const struct kallsym_node *kn; + + kn = container_of(n, struct kallsym_node, kn_node); + sym_start = kn->kn_addr; + sym_end = sym_start + kn->kn_len; + + if (val < sym_start) + return -1; + if (val >= sym_end) + return 1; + + return 0; +} + +static const struct latch_tree_ops kallsym_tree_ops = { + .less = kallsym_tree_less, + .comp = kallsym_tree_comp, +}; + +void kallsym_tree_add(struct kallsym_node *kn) +{ + char namebuf[KSYM_NAME_LEN] = ""; + char *modname = NULL; + + spin_lock_irq(&kallsym_lock); + latch_tree_insert(&kn->kn_node, &kallsym_tree, &kallsym_tree_ops); + spin_unlock_irq(&kallsym_lock); + + kn->kn_names(kn, namebuf, &modname); + + if (modname) { + int len = strlen(namebuf); + + snprintf(namebuf + len, sizeof(namebuf) - len, " [%s]", modname); + } + + perf_event_ksymbol(kn->ksym_type, kn->kn_addr, kn->kn_len, false, namebuf); +} + +void kallsym_tree_del(struct kallsym_node *kn) +{ + char namebuf[KSYM_NAME_LEN] = ""; + char *modname = NULL; + + kn->kn_names(kn, namebuf, &modname); + + if (modname) { + int len = strlen(namebuf); + + snprintf(namebuf + len, sizeof(namebuf) - len, " [%s]", modname); + } + + perf_event_ksymbol(kn->ksym_type, kn->kn_addr, kn->kn_len, true, namebuf); + + spin_lock_irq(&kallsym_lock); + latch_tree_erase(&kn->kn_node, &kallsym_tree, &kallsym_tree_ops); + spin_unlock_irq(&kallsym_lock); +} + +static struct kallsym_node *kallsym_tree_find(unsigned long addr) +{ + struct kallsym_node *kn = NULL; + struct latch_tree_node *n; + + n = latch_tree_find((void *)addr, &kallsym_tree, &kallsym_tree_ops); + if (n) + kn = container_of(n, struct kallsym_node, kn_node); + + return kn; +} + +static char *kallsym_tree_address_lookup(unsigned long addr, unsigned long *size, + unsigned long *off, char **modname, char *sym) +{ + struct kallsym_node *kn; + char *ret = NULL; + + rcu_read_lock(); + kn = kallsym_tree_find(addr); + if (kn) { + kn->kn_names(kn, sym, modname); + + ret = sym; + if (size) + *size = kn->kn_len; + if (off) + *off = addr - kn->kn_addr; + } + rcu_read_unlock(); + + return ret; +} + +bool is_kallsym_tree_text_address(unsigned long addr) +{ + bool ret; + + rcu_read_lock(); + ret = kallsym_tree_find(addr) != NULL; + rcu_read_unlock(); + + return ret; +} + +static int kallsym_tree_kallsym(unsigned int symnum, unsigned long *value, char *type, + char *sym, char *modname, int *exported) +{ + struct latch_tree_node *ltn; + int i, ret = -ERANGE; + + rcu_read_lock(); + for (i = 0, ltn = latch_tree_first(&kallsym_tree); i < symnum && ltn; + i++, ltn = latch_tree_next(&kallsym_tree, ltn)) + ; + + if (ltn) { + struct kallsym_node *kn; + char *mod; + + kn = container_of(ltn, struct kallsym_node, kn_node); + + kn->kn_names(kn, sym, &mod); + if (mod) + strlcpy(modname, mod, MODULE_NAME_LEN); + else + modname[0] = '\0'; + + *type = 't'; + *exported = 0; + ret = 0; + } + rcu_read_unlock(); + + return ret; +} + /* * Expand a compressed symbol data into the resulting uncompressed string, * if uncompressed string is too long (>= maxlen), it will be truncated, @@ -265,6 +426,7 @@ int kallsyms_lookup_size_offset(unsigned long addr, unsigned long *symbolsize, if (is_ksym_addr(addr)) return !!get_symbol_pos(addr, symbolsize, offset); return !!module_address_lookup(addr, symbolsize, offset, NULL, namebuf) || + !!kallsym_tree_address_lookup(addr, symbolsize, offset, NULL, namebuf) || !!__bpf_address_lookup(addr, symbolsize, offset, namebuf); } @@ -300,6 +462,10 @@ const char *kallsyms_lookup(unsigned long addr, /* See if it's in a module or a BPF JITed image. */ ret = module_address_lookup(addr, symbolsize, offset, modname, namebuf); + if (!ret) + ret = kallsym_tree_address_lookup(addr, symbolsize, + offset, modname, namebuf); + if (!ret) ret = bpf_address_lookup(addr, symbolsize, offset, modname, namebuf); @@ -434,6 +600,7 @@ struct kallsym_iter { loff_t pos; loff_t pos_arch_end; loff_t pos_mod_end; + loff_t pos_tree_end; loff_t pos_ftrace_mod_end; unsigned long value; unsigned int nameoff; /* If iterating in core kernel symbols. */ @@ -478,9 +645,24 @@ static int get_ksymbol_mod(struct kallsym_iter *iter) return 1; } +static int get_ksymbol_tree(struct kallsym_iter *iter) +{ + int ret = kallsym_tree_kallsym(iter->pos - iter->pos_mod_end, + &iter->value, &iter->type, + iter->name, iter->module_name, + &iter->exported); + + if (ret < 0) { + iter->pos_tree_end = iter->pos; + return 0; + } + + return 1; +} + static int get_ksymbol_ftrace_mod(struct kallsym_iter *iter) { - int ret = ftrace_mod_get_kallsym(iter->pos - iter->pos_mod_end, + int ret = ftrace_mod_get_kallsym(iter->pos - iter->pos_tree_end, &iter->value, &iter->type, iter->name, iter->module_name, &iter->exported); @@ -545,6 +727,10 @@ static int update_iter_mod(struct kallsym_iter *iter, loff_t pos) get_ksymbol_mod(iter)) return 1; + if ((!iter->pos_tree_end || iter->pos_tree_end > pos) && + get_ksymbol_tree(iter)) + return 1; + if ((!iter->pos_ftrace_mod_end || iter->pos_ftrace_mod_end > pos) && get_ksymbol_ftrace_mod(iter)) return 1; From patchwork Thu Jan 17 23:17:30 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Song Liu X-Patchwork-Id: 1027051 X-Patchwork-Delegate: bpf@iogearbox.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=pass (p=none dis=none) header.from=fb.com Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=fb.com header.i=@fb.com header.b="JZAwXmoY"; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 43gg4J2XwTz9sDn for ; Fri, 18 Jan 2019 10:18:40 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727018AbfAQXSi (ORCPT ); Thu, 17 Jan 2019 18:18:38 -0500 Received: from mx0a-00082601.pphosted.com ([67.231.145.42]:35634 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1725982AbfAQXSi (ORCPT ); Thu, 17 Jan 2019 18:18:38 -0500 Received: from pps.filterd (m0109334.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.27/8.16.0.27) with SMTP id x0HNDW9B020143 for ; Thu, 17 Jan 2019 15:18:36 -0800 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=fb.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : mime-version : content-type; s=facebook; bh=VxKTSBhcMiX7Im063mreaOqNtB4COubIUusgpLvH83c=; b=JZAwXmoYTT328kP9dyJAwSkTxDN0MfsKrC2MD+5+RHyn8N6W0eCPlCOrKhXtF9eB5Ltv K7bM1xU7x8Wo65iaWf6txkthmaOfdJB+x5wpDQg99CuduVL5W0f1GiRtjg5yRbFKqBmY JJdLhDXzxJ7NK+RD3S3wW7XblbemwoRlH1I= Received: from mail.thefacebook.com ([199.201.64.23]) by mx0a-00082601.pphosted.com with ESMTP id 2q31vh89bu-18 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-SHA384 bits=256 verify=NOT) for ; Thu, 17 Jan 2019 15:18:36 -0800 Received: from mx-out.facebook.com (2620:10d:c081:10::13) by mail.thefacebook.com (2620:10d:c081:35::128) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_CBC_SHA) id 15.1.1531.3; Thu, 17 Jan 2019 15:17:40 -0800 Received: by devbig006.ftw2.facebook.com (Postfix, from userid 4523) id 4195562E214C; Thu, 17 Jan 2019 15:17:32 -0800 (PST) Smtp-Origin-Hostprefix: devbig From: Song Liu Smtp-Origin-Hostname: devbig006.ftw2.facebook.com To: , CC: Peter Zijlstra , , , , , Song Liu Smtp-Origin-Cluster: ftw2c04 Subject: [PATCH kallsyms, bpf 3/3] bpf: migrate symbols for BPF programs to kallsym_tree Date: Thu, 17 Jan 2019 15:17:30 -0800 Message-ID: <20190117231730.2413466-4-songliubraving@fb.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20190117231730.2413466-1-songliubraving@fb.com> References: <20190117231730.2413466-1-songliubraving@fb.com> X-FB-Internal: Safe MIME-Version: 1.0 X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:, , definitions=2019-01-17_09:, , signatures=0 X-Proofpoint-Spam-Reason: safe X-FB-Internal: Safe Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org From: Peter Zijlstra This patch migrates BPF program symbols from bpf_tree to kallsym_tree. Since kallsym_tree_add/del already calls perf_event_ksymbol(), this patch also removes unnecessary perf_event_bpf_emit_ksymbols(). Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Song Liu --- include/linux/bpf.h | 7 +- include/linux/filter.h | 42 ----------- kernel/bpf/core.c | 167 +++++------------------------------------ kernel/events/core.c | 35 --------- kernel/extable.c | 4 +- kernel/kallsyms.c | 19 +---- 6 files changed, 25 insertions(+), 249 deletions(-) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index e734f163bd0b..403e1f88a1fa 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -13,7 +13,7 @@ #include #include #include -#include +#include #include #include @@ -307,8 +307,9 @@ struct bpf_prog_aux { bool offload_requested; struct bpf_prog **func; void *jit_data; /* JIT specific data. arch dependent */ - struct latch_tree_node ksym_tnode; - struct list_head ksym_lnode; + + struct kallsym_node ktn; + const struct bpf_prog_ops *ops; struct bpf_map **used_maps; struct bpf_prog *prog; diff --git a/include/linux/filter.h b/include/linux/filter.h index d531d4250bff..61264a3f944c 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -932,23 +932,6 @@ static inline bool bpf_jit_kallsyms_enabled(void) return false; } -const char *__bpf_address_lookup(unsigned long addr, unsigned long *size, - unsigned long *off, char *sym); -bool is_bpf_text_address(unsigned long addr); -int bpf_get_kallsym(unsigned int symnum, unsigned long *value, char *type, - char *sym); - -static inline const char * -bpf_address_lookup(unsigned long addr, unsigned long *size, - unsigned long *off, char **modname, char *sym) -{ - const char *ret = __bpf_address_lookup(addr, size, off, sym); - - if (ret && modname) - *modname = NULL; - return ret; -} - void bpf_prog_kallsyms_add(struct bpf_prog *fp); void bpf_prog_kallsyms_del(struct bpf_prog *fp); void bpf_get_prog_name(const struct bpf_prog *prog, char *sym); @@ -975,31 +958,6 @@ static inline bool bpf_jit_kallsyms_enabled(void) return false; } -static inline const char * -__bpf_address_lookup(unsigned long addr, unsigned long *size, - unsigned long *off, char *sym) -{ - return NULL; -} - -static inline bool is_bpf_text_address(unsigned long addr) -{ - return false; -} - -static inline int bpf_get_kallsym(unsigned int symnum, unsigned long *value, - char *type, char *sym) -{ - return -ERANGE; -} - -static inline const char * -bpf_address_lookup(unsigned long addr, unsigned long *size, - unsigned long *off, char **modname, char *sym) -{ - return NULL; -} - static inline void bpf_prog_kallsyms_add(struct bpf_prog *fp) { } diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 19c49313c709..e53912234a91 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -30,7 +30,6 @@ #include #include #include -#include #include #include #include @@ -100,8 +99,6 @@ struct bpf_prog *bpf_prog_alloc(unsigned int size, gfp_t gfp_extra_flags) fp->aux->prog = fp; fp->jit_requested = ebpf_jit_enabled(); - INIT_LIST_HEAD_RCU(&fp->aux->ksym_lnode); - return fp; } EXPORT_SYMBOL_GPL(bpf_prog_alloc); @@ -530,86 +527,38 @@ void bpf_get_prog_name(const struct bpf_prog *prog, char *sym) *sym = 0; } -static __always_inline unsigned long -bpf_get_prog_addr_start(struct latch_tree_node *n) -{ - unsigned long symbol_start, symbol_end; - const struct bpf_prog_aux *aux; - - aux = container_of(n, struct bpf_prog_aux, ksym_tnode); - bpf_get_prog_addr_region(aux->prog, &symbol_start, &symbol_end); - - return symbol_start; -} - -static __always_inline bool bpf_tree_less(struct latch_tree_node *a, - struct latch_tree_node *b) -{ - return bpf_get_prog_addr_start(a) < bpf_get_prog_addr_start(b); -} - -static __always_inline int bpf_tree_comp(void *key, struct latch_tree_node *n) -{ - unsigned long val = (unsigned long)key; - unsigned long symbol_start, symbol_end; - const struct bpf_prog_aux *aux; - - aux = container_of(n, struct bpf_prog_aux, ksym_tnode); - bpf_get_prog_addr_region(aux->prog, &symbol_start, &symbol_end); - - if (val < symbol_start) - return -1; - if (val >= symbol_end) - return 1; - - return 0; -} - -static const struct latch_tree_ops bpf_tree_ops = { - .less = bpf_tree_less, - .comp = bpf_tree_comp, -}; - -static DEFINE_SPINLOCK(bpf_lock); -static LIST_HEAD(bpf_kallsyms); -static struct latch_tree_root bpf_tree __cacheline_aligned; - -static void bpf_prog_ksym_node_add(struct bpf_prog_aux *aux) -{ - WARN_ON_ONCE(!list_empty(&aux->ksym_lnode)); - list_add_tail_rcu(&aux->ksym_lnode, &bpf_kallsyms); - latch_tree_insert(&aux->ksym_tnode, &bpf_tree, &bpf_tree_ops); -} - -static void bpf_prog_ksym_node_del(struct bpf_prog_aux *aux) -{ - if (list_empty(&aux->ksym_lnode)) - return; - - latch_tree_erase(&aux->ksym_tnode, &bpf_tree, &bpf_tree_ops); - list_del_rcu(&aux->ksym_lnode); -} static bool bpf_prog_kallsyms_candidate(const struct bpf_prog *fp) { return fp->jited && !bpf_prog_was_classic(fp); } -static bool bpf_prog_kallsyms_verify_off(const struct bpf_prog *fp) +static void bpf_kn_names(struct kallsym_node *kn, char *sym, char **modname) { - return list_empty(&fp->aux->ksym_lnode) || - fp->aux->ksym_lnode.prev == LIST_POISON2; + struct bpf_prog_aux *aux = container_of(kn, struct bpf_prog_aux, ktn); + + *modname = "eBPF-jit"; + bpf_get_prog_name(aux->prog, sym); } void bpf_prog_kallsyms_add(struct bpf_prog *fp) { + unsigned long sym_start, sym_end; + if (!bpf_prog_kallsyms_candidate(fp) || !capable(CAP_SYS_ADMIN)) return; - spin_lock_bh(&bpf_lock); - bpf_prog_ksym_node_add(fp->aux); - spin_unlock_bh(&bpf_lock); + bpf_get_prog_addr_region(fp, &sym_start, &sym_end); + + fp->aux->ktn.kn_addr = sym_start; + fp->aux->ktn.kn_len = sym_end - sym_start; + fp->aux->ktn.kn_names = bpf_kn_names; +#ifdef CONFIG_PERF_EVENTS + fp->aux->ktn.ksym_type = PERF_RECORD_KSYMBOL_TYPE_BPF; +#endif + + kallsym_tree_add(&fp->aux->ktn); } void bpf_prog_kallsyms_del(struct bpf_prog *fp) @@ -617,85 +566,7 @@ void bpf_prog_kallsyms_del(struct bpf_prog *fp) if (!bpf_prog_kallsyms_candidate(fp)) return; - spin_lock_bh(&bpf_lock); - bpf_prog_ksym_node_del(fp->aux); - spin_unlock_bh(&bpf_lock); -} - -static struct bpf_prog *bpf_prog_kallsyms_find(unsigned long addr) -{ - struct latch_tree_node *n; - - if (!bpf_jit_kallsyms_enabled()) - return NULL; - - n = latch_tree_find((void *)addr, &bpf_tree, &bpf_tree_ops); - return n ? - container_of(n, struct bpf_prog_aux, ksym_tnode)->prog : - NULL; -} - -const char *__bpf_address_lookup(unsigned long addr, unsigned long *size, - unsigned long *off, char *sym) -{ - unsigned long symbol_start, symbol_end; - struct bpf_prog *prog; - char *ret = NULL; - - rcu_read_lock(); - prog = bpf_prog_kallsyms_find(addr); - if (prog) { - bpf_get_prog_addr_region(prog, &symbol_start, &symbol_end); - bpf_get_prog_name(prog, sym); - - ret = sym; - if (size) - *size = symbol_end - symbol_start; - if (off) - *off = addr - symbol_start; - } - rcu_read_unlock(); - - return ret; -} - -bool is_bpf_text_address(unsigned long addr) -{ - bool ret; - - rcu_read_lock(); - ret = bpf_prog_kallsyms_find(addr) != NULL; - rcu_read_unlock(); - - return ret; -} - -int bpf_get_kallsym(unsigned int symnum, unsigned long *value, char *type, - char *sym) -{ - struct bpf_prog_aux *aux; - unsigned int it = 0; - int ret = -ERANGE; - - if (!bpf_jit_kallsyms_enabled()) - return ret; - - rcu_read_lock(); - list_for_each_entry_rcu(aux, &bpf_kallsyms, ksym_lnode) { - if (it++ != symnum) - continue; - - bpf_get_prog_name(aux->prog, sym); - - *value = (unsigned long)aux->prog->bpf_func; - *type = BPF_SYM_ELF_TYPE; - - ret = 0; - break; - } - rcu_read_unlock(); - - return ret; + kallsym_tree_del(&fp->aux->ktn); } static atomic_long_t bpf_jit_current; @@ -806,8 +677,6 @@ void __weak bpf_jit_free(struct bpf_prog *fp) bpf_jit_binary_unlock_ro(hdr); bpf_jit_binary_free(hdr); - - WARN_ON_ONCE(!bpf_prog_kallsyms_verify_off(fp)); } bpf_prog_unlock_free(fp); diff --git a/kernel/events/core.c b/kernel/events/core.c index 0a8dab322111..5f6ab55f77cf 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -7790,31 +7790,6 @@ static void perf_event_bpf_output(struct perf_event *event, void *data) perf_output_end(&handle); } -static void perf_event_bpf_emit_ksymbols(struct bpf_prog *prog, - enum perf_bpf_event_type type) -{ - bool unregister = type == PERF_BPF_EVENT_PROG_UNLOAD; - char sym[KSYM_NAME_LEN]; - int i; - - if (prog->aux->func_cnt == 0) { - bpf_get_prog_name(prog, sym); - perf_event_ksymbol(PERF_RECORD_KSYMBOL_TYPE_BPF, - (u64)(unsigned long)prog->bpf_func, - prog->jited_len, unregister, sym); - } else { - for (i = 0; i < prog->aux->func_cnt; i++) { - struct bpf_prog *subprog = prog->aux->func[i]; - - bpf_get_prog_name(subprog, sym); - perf_event_ksymbol( - PERF_RECORD_KSYMBOL_TYPE_BPF, - (u64)(unsigned long)subprog->bpf_func, - subprog->jited_len, unregister, sym); - } - } -} - void perf_event_bpf_event(struct bpf_prog *prog, enum perf_bpf_event_type type, u16 flags) @@ -7825,16 +7800,6 @@ void perf_event_bpf_event(struct bpf_prog *prog, type >= PERF_BPF_EVENT_MAX) return; - switch (type) { - case PERF_BPF_EVENT_PROG_LOAD: - case PERF_BPF_EVENT_PROG_UNLOAD: - if (atomic_read(&nr_ksymbol_events)) - perf_event_bpf_emit_ksymbols(prog, type); - break; - default: - break; - } - if (!atomic_read(&nr_bpf_events)) return; diff --git a/kernel/extable.c b/kernel/extable.c index 5271e9b649b1..c7b7bd8e24f6 100644 --- a/kernel/extable.c +++ b/kernel/extable.c @@ -135,7 +135,7 @@ int kernel_text_address(unsigned long addr) * coming back from idle, or cpu on or offlining. * * is_module_text_address() as well as the kprobe slots - * and is_bpf_text_address() require RCU to be watching. + * and is_kallsym_tree_text_address() require RCU to be watching. */ no_rcu = !rcu_is_watching(); @@ -151,8 +151,6 @@ int kernel_text_address(unsigned long addr) goto out; if (is_kprobe_optinsn_slot(addr) || is_kprobe_insn_slot(addr)) goto out; - if (is_bpf_text_address(addr)) - goto out; ret = 0; out: if (no_rcu) diff --git a/kernel/kallsyms.c b/kernel/kallsyms.c index 30611a5379fd..17bcc6815cf3 100644 --- a/kernel/kallsyms.c +++ b/kernel/kallsyms.c @@ -426,8 +426,7 @@ int kallsyms_lookup_size_offset(unsigned long addr, unsigned long *symbolsize, if (is_ksym_addr(addr)) return !!get_symbol_pos(addr, symbolsize, offset); return !!module_address_lookup(addr, symbolsize, offset, NULL, namebuf) || - !!kallsym_tree_address_lookup(addr, symbolsize, offset, NULL, namebuf) || - !!__bpf_address_lookup(addr, symbolsize, offset, namebuf); + !!kallsym_tree_address_lookup(addr, symbolsize, offset, NULL, namebuf); } /* @@ -465,11 +464,6 @@ const char *kallsyms_lookup(unsigned long addr, if (!ret) ret = kallsym_tree_address_lookup(addr, symbolsize, offset, modname, namebuf); - - if (!ret) - ret = bpf_address_lookup(addr, symbolsize, - offset, modname, namebuf); - if (!ret) ret = ftrace_mod_address_lookup(addr, symbolsize, offset, modname, namebuf); @@ -674,15 +668,6 @@ static int get_ksymbol_ftrace_mod(struct kallsym_iter *iter) return 1; } -static int get_ksymbol_bpf(struct kallsym_iter *iter) -{ - strlcpy(iter->module_name, "bpf", MODULE_NAME_LEN); - iter->exported = 0; - return bpf_get_kallsym(iter->pos - iter->pos_ftrace_mod_end, - &iter->value, &iter->type, - iter->name) < 0 ? 0 : 1; -} - /* Returns space to next name. */ static unsigned long get_ksymbol_core(struct kallsym_iter *iter) { @@ -735,7 +720,7 @@ static int update_iter_mod(struct kallsym_iter *iter, loff_t pos) get_ksymbol_ftrace_mod(iter)) return 1; - return get_ksymbol_bpf(iter); + return 0; } /* Returns false if pos at or past end of file. */