From patchwork Fri Feb 4 23:17:06 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexei Starovoitov X-Patchwork-Id: 12735822 X-Patchwork-Delegate: bpf@iogearbox.net 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 89FDEC433F5 for ; Fri, 4 Feb 2022 23:17:19 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1378006AbiBDXRS (ORCPT ); Fri, 4 Feb 2022 18:17:18 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40264 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S234586AbiBDXRS (ORCPT ); Fri, 4 Feb 2022 18:17:18 -0500 Received: from mail-pg1-x52f.google.com (mail-pg1-x52f.google.com [IPv6:2607:f8b0:4864:20::52f]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 23E62DFDA6D7 for ; Fri, 4 Feb 2022 15:17:16 -0800 (PST) Received: by mail-pg1-x52f.google.com with SMTP id g20so6205267pgn.10 for ; Fri, 04 Feb 2022 15:17:16 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=Ym0egyZ6wQ+6Evxd/TMe29RTUH+h4dLheTJVj8yT+qk=; b=YU4FMY3Kr/luNSXm0vvEEiaMXgmjVLEVvTvcMrKxiofJpjjZoYilUwuqPz+aAdhJXN bOUmsAobNoA/Lv+pYccvuGn8jHPEG57KVmeElqFxvV+rGcRRIm2CVfBTtNTCFgeJoJ9P 7NpJAynyKs6v2ySsTX1CBYg+6GuTqmkukvLRHX9lMU6T0NKZ++NI/GBccNi5YlCy2H9q rl4dZT+P1azc2Gum2gl6bmstBCmgZLtAz2mdj9zlyveGlramIKseieMzVduBOMVJBGMy 09wYiBAPXA2RLfX7CJZ/vR8x9+wLFRmKTkoS9snQyvDxgGZb2JcDjP+iT0F8rQP/Kp4x KGSA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=Ym0egyZ6wQ+6Evxd/TMe29RTUH+h4dLheTJVj8yT+qk=; b=XADbshv2wWVBPVIKSbykQr4rVo9Gn2DSVRQcg+6BAKhA6NvKdSKJKwVU8PIqb6IDMj RKu08lrxI9IMKigkEdqAA83zMU30SWwbuwXYLLqjbNfuLf12asU63Ck7a7qYnFkhwF5M 89fzbIP5vlRUleeJRf5CLTQpOsUz/srunTkki74FBPXGZAnQ9RsV0tO600qz9sYtwIr5 RfZjn2Fwq7KGKAdh13A3tIdZmeBQ8DK7oPXkJ2s0sRC4FbaMwYhKKx5ulG07JkpvwzNE sfGHbisEemtcmi4+oFUwvFpBM0b0gGPqKH9NPsbkR6biQi15mB8fskH8mYaZLL6PLhR6 uyHQ== X-Gm-Message-State: AOAM532niKDvtAVBEXcgwK9gMbsRfK7Aa3BKTSfZYEZGiK1Cs24qMxAO T/ix82pqMhkhutiuMjUiybo= X-Google-Smtp-Source: ABdhPJzVrw0aXFLmSq72gRuRlybsX27Wtg84wRaMHnfcZ8wUhbZ0PauDgxEEc82k3Kp2yyUAP+48Og== X-Received: by 2002:a65:60d2:: with SMTP id r18mr1058668pgv.444.1644016635619; Fri, 04 Feb 2022 15:17:15 -0800 (PST) Received: from ast-mbp.thefacebook.com ([2620:10d:c090:400::5:e4ee]) by smtp.gmail.com with ESMTPSA id il18sm16014726pjb.27.2022.02.04.15.17.14 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 04 Feb 2022 15:17:15 -0800 (PST) From: Alexei Starovoitov To: davem@davemloft.net Cc: daniel@iogearbox.net, andrii@kernel.org, bpf@vger.kernel.org, kernel-team@fb.com Subject: [PATCH bpf-next 1/5] bpf: Extend sys_bpf commands for bpf_syscall programs. Date: Fri, 4 Feb 2022 15:17:06 -0800 Message-Id: <20220204231710.25139-2-alexei.starovoitov@gmail.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220204231710.25139-1-alexei.starovoitov@gmail.com> References: <20220204231710.25139-1-alexei.starovoitov@gmail.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org X-Patchwork-Delegate: bpf@iogearbox.net From: Alexei Starovoitov bpf_sycall programs can be used directly by the kernel modules to load programs and create maps via kernel skeleton. . Export bpf_sys_bpf syscall wrapper to be used in kernel skeleton. . Export bpf_map_get to be used in kernel skeleton. . Allow prog_run cmd for bpf_syscall programs with recursion check. . Enable link_create and raw_tp_open cmds. Signed-off-by: Alexei Starovoitov Acked-by: Yonghong Song --- kernel/bpf/syscall.c | 38 ++++++++++++++++++++++++++++++++++---- 1 file changed, 34 insertions(+), 4 deletions(-) diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 72ce1edde950..08ac01acc51e 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -985,6 +985,7 @@ struct bpf_map *bpf_map_get(u32 ufd) return map; } +EXPORT_SYMBOL(bpf_map_get); struct bpf_map *bpf_map_get_with_uref(u32 ufd) { @@ -4756,23 +4757,52 @@ static bool syscall_prog_is_valid_access(int off, int size, return true; } -BPF_CALL_3(bpf_sys_bpf, int, cmd, void *, attr, u32, attr_size) +BPF_CALL_3(bpf_sys_bpf, int, cmd, union bpf_attr *, attr, u32, attr_size) { + struct bpf_prog *prog; + switch (cmd) { case BPF_MAP_CREATE: case BPF_MAP_UPDATE_ELEM: case BPF_MAP_FREEZE: case BPF_PROG_LOAD: case BPF_BTF_LOAD: + case BPF_LINK_CREATE: + case BPF_RAW_TRACEPOINT_OPEN: break; - /* case BPF_PROG_TEST_RUN: - * is not part of this list to prevent recursive test_run - */ +#ifdef CONFIG_BPF_JIT /* __bpf_prog_enter_sleepable used by trampoline and JIT */ + case BPF_PROG_TEST_RUN: + if (attr->test.data_in || attr->test.data_out || + attr->test.ctx_out || attr->test.duration || + attr->test.repeat || attr->test.flags) + return -EINVAL; + + prog = bpf_prog_get_type(attr->test.prog_fd, BPF_PROG_TYPE_SYSCALL); + if (IS_ERR(prog)) + return PTR_ERR(prog); + + if (attr->test.ctx_size_in < prog->aux->max_ctx_offset || + attr->test.ctx_size_in > U16_MAX) { + bpf_prog_put(prog); + return -EINVAL; + } + + if (!__bpf_prog_enter_sleepable(prog)) { + /* recursion detected */ + bpf_prog_put(prog); + return -EBUSY; + } + attr->test.retval = bpf_prog_run(prog, (void *) (long) attr->test.ctx_in); + __bpf_prog_exit_sleepable(prog, 0 /* bpf_prog_run does runtime stats */); + bpf_prog_put(prog); + return 0; +#endif default: return -EINVAL; } return __sys_bpf(cmd, KERNEL_BPFPTR(attr), attr_size); } +EXPORT_SYMBOL(bpf_sys_bpf); static const struct bpf_func_proto bpf_sys_bpf_proto = { .func = bpf_sys_bpf, From patchwork Fri Feb 4 23:17:07 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexei Starovoitov X-Patchwork-Id: 12735823 X-Patchwork-Delegate: bpf@iogearbox.net 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 4272EC433EF for ; Fri, 4 Feb 2022 23:17:21 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1378005AbiBDXRU (ORCPT ); Fri, 4 Feb 2022 18:17:20 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40282 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S234586AbiBDXRU (ORCPT ); Fri, 4 Feb 2022 18:17:20 -0500 Received: from mail-pl1-x632.google.com (mail-pl1-x632.google.com [IPv6:2607:f8b0:4864:20::632]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 517D7DFDA6E4 for ; Fri, 4 Feb 2022 15:17:19 -0800 (PST) Received: by mail-pl1-x632.google.com with SMTP id z5so6390862plg.8 for ; Fri, 04 Feb 2022 15:17:19 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=UYODaeY9R2B2JnqLrZ6rCIiYeLQeCLssHwMEt39uLGQ=; b=kkQ8o1cMtYF1Mtu4hB6IlDgzH4UUPQqz4/Wa4yfh38pVYanYZboRqOSM1P7LATYiUA WIYt1Y7C0aKJ09aUVopzdzlF0Ex3cBjMQpEMQTuZyjYv7o+z4wWpvIcQZ7mBmbWHPSht kxkp98oM9Zd98NUG0H2iIy+AumVNa1TEyRJgtYLZtJskGK12WQZjuU5Z7TgpNjTeL6aQ Syj6ur2p0vPxnxv1I+h8G6rhPr2FhvA+ZyFtuwyRnqch96/zCWDap+opUl5LBZJTYrVG eXtvSD8ngvkOmwxMfSpyD0iL9+lfCEuyPSB2yJ2ik1Wwpoeo0J5mUvl/NTvdwWWSLwWC NEhQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=UYODaeY9R2B2JnqLrZ6rCIiYeLQeCLssHwMEt39uLGQ=; b=yWfmu03mlJhIALeYfc0wBuRCZsRRgFCshpQJ8UzNFmZeHPzgXIdS+dsX5/vSaH2NgM GpvbjgcQwdaw/D4uQSQb+kAYigIYpphfkNytkrEZP6rda3iq8OYCP3qH5xtdb603q9Oh cPRbDBdUiLSnCCtTZAmmYdZojPAVa2AcvuTcaIyH5Zx/mmkd+4/PpixJnhrUR5+Haov6 17894foJV4nuUFOP7ehW1oa0iNZ9pfg9mMCv4xD/hvqO16rfk9604+FDuRBRm9Tqrzw7 RomQtQKRK5UKs7uesndE5JH5Ze1wi/h/W3ASPp+i0UKJEj1rhb4MfFgyIBXZAwoOXgsE 0gBw== X-Gm-Message-State: AOAM5325QTZ39kc4Urb1BXVmNPxuV/l+iT1qvXrw34FlqTiIfFsltfV/ WBTAxGhxRugF5+yz1yUs7iM= X-Google-Smtp-Source: ABdhPJx+AQ0OTFnHKCktERXpv3sQALbbV0ZUgxe5S0RGNDLsWwhlxKl4sp7NPwOXMu03zDZCZ7v1zg== X-Received: by 2002:a17:90b:4d8c:: with SMTP id oj12mr1507831pjb.10.1644016638793; Fri, 04 Feb 2022 15:17:18 -0800 (PST) Received: from ast-mbp.thefacebook.com ([2620:10d:c090:400::5:e4ee]) by smtp.gmail.com with ESMTPSA id f15sm3505842pfn.19.2022.02.04.15.17.17 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 04 Feb 2022 15:17:18 -0800 (PST) From: Alexei Starovoitov To: davem@davemloft.net Cc: daniel@iogearbox.net, andrii@kernel.org, bpf@vger.kernel.org, kernel-team@fb.com Subject: [PATCH bpf-next 2/5] libbpf: Prepare light skeleton for the kernel. Date: Fri, 4 Feb 2022 15:17:07 -0800 Message-Id: <20220204231710.25139-3-alexei.starovoitov@gmail.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220204231710.25139-1-alexei.starovoitov@gmail.com> References: <20220204231710.25139-1-alexei.starovoitov@gmail.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org X-Patchwork-Delegate: bpf@iogearbox.net From: Alexei Starovoitov Prepare light skeleton to be used in the kernel module and in the user space. The look and feel of lskel.h is mostly the same with the difference that for user space the skel->rodata is the same pointer before and after skel_load operation, while in the kernel the skel->rodata after skel_open and the skel->rodata after skel_load are different pointers. Typical usage of skeleton remains the same for kernel and user space: skel = my_bpf__open(); skel->rodata->my_global_var = init_val; err = my_bpf__load(skel); err = my_bpf__attach(skel); // access skel->rodata->my_global_var; // access skel->bss->another_var; Signed-off-by: Alexei Starovoitov --- tools/lib/bpf/skel_internal.h | 168 +++++++++++++++++++++++++++++++--- 1 file changed, 153 insertions(+), 15 deletions(-) diff --git a/tools/lib/bpf/skel_internal.h b/tools/lib/bpf/skel_internal.h index dcd3336512d4..4d99ef8cbbba 100644 --- a/tools/lib/bpf/skel_internal.h +++ b/tools/lib/bpf/skel_internal.h @@ -3,9 +3,19 @@ #ifndef __SKEL_INTERNAL_H #define __SKEL_INTERNAL_H +#ifdef __KERNEL__ +#include +#include +#include +#include +#include +#else #include #include #include +#include +#include "bpf.h" +#endif #ifndef __NR_bpf # if defined(__mips__) && defined(_ABIO32) @@ -25,16 +35,12 @@ * requested during loader program generation. */ struct bpf_map_desc { - union { - /* input for the loader prog */ - struct { - __aligned_u64 initial_value; - __u32 max_entries; - }; + struct { /* output of the loader prog */ - struct { - int map_fd; - }; + int map_fd; + /* input for the loader prog */ + __u32 max_entries; + __aligned_u64 initial_value; }; }; struct bpf_prog_desc { @@ -57,11 +63,135 @@ struct bpf_load_and_run_opts { const char *errstr; }; +long bpf_sys_bpf(__u32 cmd, void *attr, __u32 attr_size); + static inline int skel_sys_bpf(enum bpf_cmd cmd, union bpf_attr *attr, unsigned int size) { +#ifdef __KERNEL__ + return bpf_sys_bpf(cmd, attr, size); +#else return syscall(__NR_bpf, cmd, attr, size); +#endif +} + +#ifdef __KERNEL__ +static inline int close(int fd) +{ + return close_fd(fd); +} +static inline void *skel_alloc(size_t size) +{ + return kcalloc(1, size, GFP_KERNEL); +} +static inline void skel_free(const void *p) +{ + kfree(p); +} +static inline void skel_free_map_data(void *p, __u64 addr, size_t sz) +{ + if (addr && addr != ~0ULL) + vm_munmap(addr, sz); + if (addr != ~0ULL) + kvfree(p); +} +/* skel->bss/rodata maps are populated in three steps. + * + * For kernel use: + * skel_prep_map_data() allocates kernel memory that kernel module can directly access. + * skel_prep_init_value() allocates a region in user space process and copies + * potentially modified initial map value into it. + * The loader program will perform copy_from_user() from maps.rodata.initial_value. + * skel_finalize_map_data() sets skel->rodata to point to actual value in a bpf map. + * + * For user space: + * skel_prep_map_data() mmaps anon memory into skel->rodata that can be accessed directly. + * skel_prep_init_value() copies rodata pointer into map.rodata.initial_value. + * The loader program will perform copy_from_user() from maps.rodata.initial_value. + * skel_finalize_map_data() remaps bpf array map value from the kernel memory into + * skel->rodata address. + */ +static inline void *skel_prep_map_data(const void *val, size_t mmap_sz, size_t val_sz) +{ + void *addr; + + addr = kvmalloc(val_sz, GFP_KERNEL); + if (!addr) + return NULL; + memcpy(addr, val, val_sz); + return addr; +} +static inline __u64 skel_prep_init_value(void **addr, size_t mmap_sz, size_t val_sz) +{ + __u64 ret = 0; + void *uaddr; + + uaddr = (void *) vm_mmap(NULL, 0, mmap_sz, PROT_READ | PROT_WRITE, + MAP_SHARED | MAP_ANONYMOUS, 0); + if (IS_ERR(uaddr)) + goto out; + if (copy_to_user(uaddr, *addr, val_sz)) { + vm_munmap((long) uaddr, mmap_sz); + goto out; + } + ret = (__u64) (long) uaddr; +out: + kvfree(*addr); + *addr = NULL; + return ret; } +static inline void *skel_finalize_map_data(__u64 *addr, size_t mmap_sz, int flags, int fd) +{ + struct bpf_map *map; + void *ptr = NULL; + + vm_munmap(*addr, mmap_sz); + *addr = ~0ULL; + + map = bpf_map_get(fd); + if (IS_ERR(map)) + return NULL; + if (map->map_type != BPF_MAP_TYPE_ARRAY) + goto out; + ptr = ((struct bpf_array *)map)->value; + /* the ptr stays valid, since FD is not closed */ +out: + bpf_map_put(map); + return ptr; +} +#else +static inline void *skel_alloc(size_t size) +{ + return calloc(1, size); +} +static inline void skel_free(void *p) +{ + free(p); +} +static inline void skel_free_map_data(void *p, __u64 addr, size_t sz) +{ + munmap(p, sz); +} +static inline void *skel_prep_map_data(const void *val, size_t mmap_sz, size_t val_sz) +{ + void *addr; + + addr = mmap(NULL, mmap_sz, PROT_READ | PROT_WRITE, + MAP_SHARED | MAP_ANONYMOUS, -1, 0); + if (addr == (void *) -1) + return NULL; + memcpy(addr, val, val_sz); + return addr; +} +static inline __u64 skel_prep_init_value(void **addr, size_t mmap_sz, size_t val_sz) +{ + return (__u64) (long) *addr; +} +static inline void *skel_finalize_map_data(__u64 *addr, size_t mmap_sz, int flags, int fd) +{ + return mmap((void *)*addr, mmap_sz, flags, MAP_SHARED | MAP_FIXED, fd, 0); +} +#endif static inline int skel_closenz(int fd) { @@ -136,22 +266,28 @@ static inline int skel_link_create(int prog_fd, int target_fd, return skel_sys_bpf(BPF_LINK_CREATE, &attr, attr_sz); } +#ifdef __KERNEL__ +#define set_err +#else +#define set_err err = -errno +#endif + static inline int bpf_load_and_run(struct bpf_load_and_run_opts *opts) { int map_fd = -1, prog_fd = -1, key = 0, err; union bpf_attr attr; - map_fd = skel_map_create(BPF_MAP_TYPE_ARRAY, "__loader.map", 4, opts->data_sz, 1); + err = map_fd = skel_map_create(BPF_MAP_TYPE_ARRAY, "__loader.map", 4, opts->data_sz, 1); if (map_fd < 0) { opts->errstr = "failed to create loader map"; - err = -errno; + set_err; goto out; } err = skel_map_update_elem(map_fd, &key, opts->data, 0); if (err < 0) { opts->errstr = "failed to update loader map"; - err = -errno; + set_err; goto out; } @@ -166,10 +302,10 @@ static inline int bpf_load_and_run(struct bpf_load_and_run_opts *opts) attr.log_size = opts->ctx->log_size; attr.log_buf = opts->ctx->log_buf; attr.prog_flags = BPF_F_SLEEPABLE; - prog_fd = skel_sys_bpf(BPF_PROG_LOAD, &attr, sizeof(attr)); + err = prog_fd = skel_sys_bpf(BPF_PROG_LOAD, &attr, sizeof(attr)); if (prog_fd < 0) { opts->errstr = "failed to load loader prog"; - err = -errno; + set_err; goto out; } @@ -181,10 +317,12 @@ static inline int bpf_load_and_run(struct bpf_load_and_run_opts *opts) if (err < 0 || (int)attr.test.retval < 0) { opts->errstr = "failed to execute loader prog"; if (err < 0) { - err = -errno; + set_err; } else { err = (int)attr.test.retval; +#ifndef __KERNEL__ errno = -err; +#endif } goto out; } From patchwork Fri Feb 4 23:17:08 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexei Starovoitov X-Patchwork-Id: 12735824 X-Patchwork-Delegate: bpf@iogearbox.net 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 2B4FCC433EF for ; Fri, 4 Feb 2022 23:17:24 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S234586AbiBDXRX (ORCPT ); Fri, 4 Feb 2022 18:17:23 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40298 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1378004AbiBDXRW (ORCPT ); Fri, 4 Feb 2022 18:17:22 -0500 Received: from mail-pj1-x1029.google.com (mail-pj1-x1029.google.com [IPv6:2607:f8b0:4864:20::1029]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 61FA7DFDA6F3 for ; Fri, 4 Feb 2022 15:17:22 -0800 (PST) Received: by mail-pj1-x1029.google.com with SMTP id v4so1124319pjh.2 for ; Fri, 04 Feb 2022 15:17:22 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=Imcqa94K6qmkHgNev/2e7jXS6AGjkc6K2IMNqaaw6eM=; b=FfqwNDqQSwCQ1YJXqcggO0CtYlh8W/0NWfQ3PtOoP7W0u8nXCrE9yug1UCALNCgndq vndh0fwOAvgmoJPZ2ktncpMGdv0odWK59PtMdX7Izdfsfli56WdDSXbyOlB4z/R5lCrf ZUuV284nVZzRI204VEa9FwSLUulk3t/dWypuvfKYQ2fTELPp+SHrwr8vVk9sP5jD00HQ qg5yDvBpmh0ow20CDvcOnO5W/5BLhkdKw1xcEz/lZ/HjaAl9gIIpU6GnCTKbPYw/X63B GHLw2bTzmtlYVCddcZQM9thtpHGKJQ03gTcwNV6e7DNGNAx6U1/5ksbO+YdGY2+mz+qe qs6w== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=Imcqa94K6qmkHgNev/2e7jXS6AGjkc6K2IMNqaaw6eM=; b=g3CBJfFM8+sL/BvtbbJGujM11AaTe8yXLN9KerfJyWndQ21cUjI4QZjKiKk6Iu/nLP 7vHCR1uEEPX/CDfyWhtWvOocxnffwyMlhQTx6D8HXZUEEnM66B5aACDqnq6ZkGi30bSj DKGEOVK5K5Rky3B4N2n/jx3FoxbvA4W/91Pj6H2upU+5BmQ0N7lIVUcYz2L9WDMQilLs Xs3dWKmVgVftFqZawhwgh8zmoyrz4QyNInZ9eLH8v8lgwZ3Q+CS0sILccCwaNm7beNTj LQO8SosL8YGuFtc0DbMEtZ/PA3R9RxFXLXIxOG3ydqZorORwSvMxmxoHDvIFJa9nV1C2 YQHw== X-Gm-Message-State: AOAM531CQOuvFN5er0hRDjrEH0eE2QnscutFV8LXeL64gPyJgzvXlAez lH+mrExj488EYTURqc0RY9M= X-Google-Smtp-Source: ABdhPJxWOCTR44LY1VeUBnzj2LZbbJWMDPGZDILUY567eHNNFoC6suHBGn3iP3r83dxHArAeXA8bnw== X-Received: by 2002:a17:90b:1187:: with SMTP id gk7mr3760759pjb.199.1644016641774; Fri, 04 Feb 2022 15:17:21 -0800 (PST) Received: from ast-mbp.thefacebook.com ([2620:10d:c090:400::5:e4ee]) by smtp.gmail.com with ESMTPSA id lj14sm7931576pjb.45.2022.02.04.15.17.20 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 04 Feb 2022 15:17:21 -0800 (PST) From: Alexei Starovoitov To: davem@davemloft.net Cc: daniel@iogearbox.net, andrii@kernel.org, bpf@vger.kernel.org, kernel-team@fb.com Subject: [PATCH bpf-next 3/5] bpftool: Generalize light skeleton generation. Date: Fri, 4 Feb 2022 15:17:08 -0800 Message-Id: <20220204231710.25139-4-alexei.starovoitov@gmail.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220204231710.25139-1-alexei.starovoitov@gmail.com> References: <20220204231710.25139-1-alexei.starovoitov@gmail.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org X-Patchwork-Delegate: bpf@iogearbox.net From: Alexei Starovoitov Generealize light skeleton by hiding mmap details in skel_internal.h In this form generated lskel.h is usable both by user space and by the kernel. Signed-off-by: Alexei Starovoitov --- tools/bpf/bpftool/gen.c | 45 ++++++++++++++++++++++++----------------- 1 file changed, 27 insertions(+), 18 deletions(-) diff --git a/tools/bpf/bpftool/gen.c b/tools/bpf/bpftool/gen.c index eacfc6a2060d..903abbf077ce 100644 --- a/tools/bpf/bpftool/gen.c +++ b/tools/bpf/bpftool/gen.c @@ -472,7 +472,7 @@ static void codegen_destroy(struct bpf_object *obj, const char *obj_name) continue; if (bpf_map__is_internal(map) && (bpf_map__map_flags(map) & BPF_F_MMAPABLE)) - printf("\tmunmap(skel->%1$s, %2$zd);\n", + printf("\tskel_free_map_data(skel->%1$s, skel->maps.%1$s.initial_value, %2$zd);\n", ident, bpf_map_mmap_sz(map)); codegen("\ \n\ @@ -481,7 +481,7 @@ static void codegen_destroy(struct bpf_object *obj, const char *obj_name) } codegen("\ \n\ - free(skel); \n\ + skel_free(skel); \n\ } \n\ ", obj_name); @@ -525,7 +525,7 @@ static int gen_trace(struct bpf_object *obj, const char *obj_name, const char *h { \n\ struct %1$s *skel; \n\ \n\ - skel = calloc(sizeof(*skel), 1); \n\ + skel = skel_alloc(sizeof(*skel)); \n\ if (!skel) \n\ goto cleanup; \n\ skel->ctx.sz = (void *)&skel->links - (void *)skel; \n\ @@ -544,18 +544,12 @@ static int gen_trace(struct bpf_object *obj, const char *obj_name, const char *h codegen("\ \n\ - skel->%1$s = \n\ - mmap(NULL, %2$zd, PROT_READ | PROT_WRITE,\n\ - MAP_SHARED | MAP_ANONYMOUS, -1, 0); \n\ - if (skel->%1$s == (void *) -1) \n\ - goto cleanup; \n\ - memcpy(skel->%1$s, (void *)\"\\ \n\ - ", ident, bpf_map_mmap_sz(map)); + skel->%1$s = skel_prep_map_data((void *)\"\\ \n\ + ", ident); mmap_data = bpf_map__initial_value(map, &mmap_size); print_hex(mmap_data, mmap_size); - printf("\", %2$zd);\n" - "\tskel->maps.%1$s.initial_value = (__u64)(long)skel->%1$s;\n", - ident, mmap_size); + printf("\", %1$zd, %2$zd);\n", + bpf_map_mmap_sz(map), mmap_size); } codegen("\ \n\ @@ -592,6 +586,24 @@ static int gen_trace(struct bpf_object *obj, const char *obj_name, const char *h codegen("\ \n\ \"; \n\ + "); + bpf_object__for_each_map(map, obj) { + size_t mmap_size = 0; + + if (!get_map_ident(map, ident, sizeof(ident))) + continue; + + if (!bpf_map__is_internal(map) || + !(bpf_map__map_flags(map) & BPF_F_MMAPABLE)) + continue; + + bpf_map__initial_value(map, &mmap_size); + printf("\tskel->maps.%1$s.initial_value =" + " skel_prep_init_value((void **)&skel->%1$s, %2$zd, %3$zd);\n", + ident, bpf_map_mmap_sz(map), mmap_size); + } + codegen("\ + \n\ err = bpf_load_and_run(&opts); \n\ if (err < 0) \n\ return err; \n\ @@ -611,9 +623,8 @@ static int gen_trace(struct bpf_object *obj, const char *obj_name, const char *h else mmap_flags = "PROT_READ | PROT_WRITE"; - printf("\tskel->%1$s =\n" - "\t\tmmap(skel->%1$s, %2$zd, %3$s, MAP_SHARED | MAP_FIXED,\n" - "\t\t\tskel->maps.%1$s.map_fd, 0);\n", + printf("\tskel->%1$s = skel_finalize_map_data(&skel->maps.%1$s.initial_value,\n" + "\t\t\t%2$zd, %3$s, skel->maps.%1$s.map_fd);\n", ident, bpf_map_mmap_sz(map), mmap_flags); } codegen("\ @@ -751,8 +762,6 @@ static int do_skeleton(int argc, char **argv) #ifndef %2$s \n\ #define %2$s \n\ \n\ - #include \n\ - #include \n\ #include \n\ \n\ struct %1$s { \n\ From patchwork Fri Feb 4 23:17:09 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexei Starovoitov X-Patchwork-Id: 12735825 X-Patchwork-Delegate: bpf@iogearbox.net 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 178E6C433F5 for ; Fri, 4 Feb 2022 23:17:28 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1378004AbiBDXR1 (ORCPT ); Fri, 4 Feb 2022 18:17:27 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40314 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1378008AbiBDXR0 (ORCPT ); Fri, 4 Feb 2022 18:17:26 -0500 Received: from mail-pg1-x52b.google.com (mail-pg1-x52b.google.com [IPv6:2607:f8b0:4864:20::52b]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 62D2BDFDA6FE for ; Fri, 4 Feb 2022 15:17:25 -0800 (PST) Received: by mail-pg1-x52b.google.com with SMTP id 132so6230913pga.5 for ; Fri, 04 Feb 2022 15:17:25 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=ENmm2RIFAUVug724BIM6PFSnweNoJNbJmLqC9jMYHHM=; b=igGEkooKYmgL+I9ppLM3DeExBKo1TKYz4WByVXdzM4hyKY5WmLrx1AI4Hbuxhgnzav PwxdyqWEYwmeLkdZk86P6i1ogqxtAMA/Jqjthc/X5t2J5VHUich21s0xNLDcmgfm+e7+ wxaVYrThka3QJyhpfA1E0bEJuJg/kOrpLL4X1Iei194Ohq+SpHJACyvS3r/lCF8l/7ML W4H08SfhJ0B0pFQ1Tf9IUugtTnbkajcpCrdF3C77JkZMUcgMe5k4gf0PNIf5Mcc4IIYD 1cqFo+XnA2EKKd/KjCVlWf3/jvx4I+9Ptt0numCoOpRH2KFGfb8wL9rWDZzw8ypiRsAP 69HQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=ENmm2RIFAUVug724BIM6PFSnweNoJNbJmLqC9jMYHHM=; b=Co5jXznTskqxczW223oMEuOUmk/srXi+wp9nBIrzF3uwV8dDK5w3dbQpjmzWXkRM74 MkrLJpwzgL9tWl+8BiErR3OswnaZty4WYKyiTpOTZQZv5hVIxSxnNC6EV/T0OgNtKLdQ 0uc+uh50ftPCluCM75cXD6p1NKB5UTwsRJJuVKeYLwkcjJWdsPDevXQaYd/Qz9xbxU5K KiCAkS7ZrYLG5MciiGJpOCTt/a5CivbNpl5KhE/Kmwc+sA37YMwtQ26kQp38kekhY7Th BF/+SqCbdyS6NBu9ffX9cFMPKkOASiYZLnsB8SVNfsqIS/T907DSbEFjjZK3czSu+BmG W42A== X-Gm-Message-State: AOAM5338bmvyCsXQdBgi5XmvpIRKwlLYncqqDLn6MM5FuXvsPIeypwNC ARNGQjOgERmDI+dD8xahKN/+RZ1KmHU= X-Google-Smtp-Source: ABdhPJxPDch2VPAqJuYYR4aTFGVes2y/1zwIBZAQw8SzYlUl/Dd2gJZmky0iQ0dG1GcZTE43tvDiNQ== X-Received: by 2002:a63:e94e:: with SMTP id q14mr1067207pgj.376.1644016644837; Fri, 04 Feb 2022 15:17:24 -0800 (PST) Received: from ast-mbp.thefacebook.com ([2620:10d:c090:400::5:e4ee]) by smtp.gmail.com with ESMTPSA id q12sm4019591pfk.199.2022.02.04.15.17.23 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 04 Feb 2022 15:17:24 -0800 (PST) From: Alexei Starovoitov To: davem@davemloft.net Cc: daniel@iogearbox.net, andrii@kernel.org, bpf@vger.kernel.org, kernel-team@fb.com Subject: [PATCH bpf-next 4/5] bpf: Update iterators.lskel.h. Date: Fri, 4 Feb 2022 15:17:09 -0800 Message-Id: <20220204231710.25139-5-alexei.starovoitov@gmail.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220204231710.25139-1-alexei.starovoitov@gmail.com> References: <20220204231710.25139-1-alexei.starovoitov@gmail.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org X-Patchwork-Delegate: bpf@iogearbox.net From: Alexei Starovoitov Light skeleton and skel_internal.h have changed. Update iterators.lskel.h. Signed-off-by: Alexei Starovoitov --- .../bpf/preload/iterators/iterators.lskel.h | 28 +++++++------------ 1 file changed, 10 insertions(+), 18 deletions(-) diff --git a/kernel/bpf/preload/iterators/iterators.lskel.h b/kernel/bpf/preload/iterators/iterators.lskel.h index d90562d672d2..3e45237f59f4 100644 --- a/kernel/bpf/preload/iterators/iterators.lskel.h +++ b/kernel/bpf/preload/iterators/iterators.lskel.h @@ -3,8 +3,6 @@ #ifndef __ITERATORS_BPF_SKEL_H__ #define __ITERATORS_BPF_SKEL_H__ -#include -#include #include struct iterators_bpf { @@ -70,31 +68,25 @@ iterators_bpf__destroy(struct iterators_bpf *skel) iterators_bpf__detach(skel); skel_closenz(skel->progs.dump_bpf_map.prog_fd); skel_closenz(skel->progs.dump_bpf_prog.prog_fd); - munmap(skel->rodata, 4096); + skel_free_map_data(skel->rodata, skel->maps.rodata.initial_value, 4096); skel_closenz(skel->maps.rodata.map_fd); - free(skel); + skel_free(skel); } static inline struct iterators_bpf * iterators_bpf__open(void) { struct iterators_bpf *skel; - skel = calloc(sizeof(*skel), 1); + skel = skel_alloc(sizeof(*skel)); if (!skel) goto cleanup; skel->ctx.sz = (void *)&skel->links - (void *)skel; - skel->rodata = - mmap(NULL, 4096, PROT_READ | PROT_WRITE, - MAP_SHARED | MAP_ANONYMOUS, -1, 0); - if (skel->rodata == (void *) -1) - goto cleanup; - memcpy(skel->rodata, (void *)"\ + skel->rodata = skel_prep_map_data((void *)"\ \x20\x20\x69\x64\x20\x6e\x61\x6d\x65\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\ \x20\x20\x20\x6d\x61\x78\x5f\x65\x6e\x74\x72\x69\x65\x73\x0a\0\x25\x34\x75\x20\ \x25\x2d\x31\x36\x73\x25\x36\x64\x0a\0\x20\x20\x69\x64\x20\x6e\x61\x6d\x65\x20\ \x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x61\x74\x74\x61\x63\x68\x65\ -\x64\x0a\0\x25\x34\x75\x20\x25\x2d\x31\x36\x73\x20\x25\x73\x20\x25\x73\x0a\0", 98); - skel->maps.rodata.initial_value = (__u64)(long)skel->rodata; +\x64\x0a\0\x25\x34\x75\x20\x25\x2d\x31\x36\x73\x20\x25\x73\x20\x25\x73\x0a\0", 4096, 98); return skel; cleanup: iterators_bpf__destroy(skel); @@ -343,11 +335,11 @@ iterators_bpf__load(struct iterators_bpf *skel) \0\0\x18\x62\0\0\0\0\0\0\0\0\0\0\x30\x0e\0\0\xb7\x03\0\0\x1c\0\0\0\x85\0\0\0\ \xa6\0\0\0\xbf\x07\0\0\0\0\0\0\xc5\x07\xd4\xff\0\0\0\0\x63\x7a\x78\xff\0\0\0\0\ \x61\xa0\x78\xff\0\0\0\0\x18\x61\0\0\0\0\0\0\0\0\0\0\x80\x0e\0\0\x63\x01\0\0\0\ -\0\0\0\x61\x60\x20\0\0\0\0\0\x15\0\x03\0\0\0\0\0\x18\x61\0\0\0\0\0\0\0\0\0\0\ +\0\0\0\x61\x60\x1c\0\0\0\0\0\x15\0\x03\0\0\0\0\0\x18\x61\0\0\0\0\0\0\0\0\0\0\ \x5c\x0e\0\0\x63\x01\0\0\0\0\0\0\xb7\x01\0\0\0\0\0\0\x18\x62\0\0\0\0\0\0\0\0\0\ \0\x50\x0e\0\0\xb7\x03\0\0\x48\0\0\0\x85\0\0\0\xa6\0\0\0\xbf\x07\0\0\0\0\0\0\ \xc5\x07\xc3\xff\0\0\0\0\x18\x61\0\0\0\0\0\0\0\0\0\0\0\0\0\0\x63\x71\0\0\0\0\0\ -\0\x79\x63\x18\0\0\0\0\0\x15\x03\x04\0\0\0\0\0\x18\x61\0\0\0\0\0\0\0\0\0\0\x98\ +\0\x79\x63\x20\0\0\0\0\0\x15\x03\x04\0\0\0\0\0\x18\x61\0\0\0\0\0\0\0\0\0\0\x98\ \x0e\0\0\xb7\x02\0\0\x62\0\0\0\x85\0\0\0\x94\0\0\0\x18\x62\0\0\0\0\0\0\0\0\0\0\ \0\0\0\0\x61\x20\0\0\0\0\0\0\x18\x61\0\0\0\0\0\0\0\0\0\0\x08\x0f\0\0\x63\x01\0\ \0\0\0\0\0\x18\x60\0\0\0\0\0\0\0\0\0\0\0\x0f\0\0\x18\x61\0\0\0\0\0\0\0\0\0\0\ @@ -401,12 +393,12 @@ iterators_bpf__load(struct iterators_bpf *skel) \x28\0\0\0\0\0\x61\xa0\x84\xff\0\0\0\0\x63\x06\x2c\0\0\0\0\0\x18\x61\0\0\0\0\0\ \0\0\0\0\0\0\0\0\0\x61\x10\0\0\0\0\0\0\x63\x06\x18\0\0\0\0\0\xb7\0\0\0\0\0\0\0\ \x95\0\0\0\0\0\0\0"; + skel->maps.rodata.initial_value = skel_prep_init_value((void **)&skel->rodata, 4096, 98); err = bpf_load_and_run(&opts); if (err < 0) return err; - skel->rodata = - mmap(skel->rodata, 4096, PROT_READ, MAP_SHARED | MAP_FIXED, - skel->maps.rodata.map_fd, 0); + skel->rodata = skel_finalize_map_data(&skel->maps.rodata.initial_value, + 4096, PROT_READ, skel->maps.rodata.map_fd); return 0; } From patchwork Fri Feb 4 23:17:10 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexei Starovoitov X-Patchwork-Id: 12735826 X-Patchwork-Delegate: bpf@iogearbox.net 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 0823FC433EF for ; Fri, 4 Feb 2022 23:17:32 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S243113AbiBDXRb (ORCPT ); Fri, 4 Feb 2022 18:17:31 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40346 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1378008AbiBDXRa (ORCPT ); Fri, 4 Feb 2022 18:17:30 -0500 Received: from mail-pj1-x102f.google.com (mail-pj1-x102f.google.com [IPv6:2607:f8b0:4864:20::102f]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 81251C06664A for ; Fri, 4 Feb 2022 15:17:28 -0800 (PST) Received: by mail-pj1-x102f.google.com with SMTP id v4so1124488pjh.2 for ; Fri, 04 Feb 2022 15:17:28 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=WhO24zbRwspkmY3e6DNmHdAVkmwrh2weKhKJ9ZwWK/M=; b=lwMuXhaWuzFPR+wTP2KVdxr1MxraV+efHd7/Q7Xv5YhCZvkAlmrk/ggJCKyqQjjb9n WFA8vOhWZJ2QvR/fIrPAN3aD7e2hMLLCeBKKFL2vc7AMeriQqIZhHT7faFjt30NV4+fa /lbrXaX8VJ+qYCCVwJrAmTkWVhHSEYg/R65hi7CCDE7lfjtodiygU8i62BemX+639+aY yRI59jHt5GpBdtmkkh3JUf2+p5SJI0GZHWvS8Moo56XNVRI3MY6O9ofZ2PW57WK4kq9r 7ZrT14D3qwTeQNB1Tf/3UGimx43ExnddfXAEV0HW/+D7YXJe2btasCtH+WAFleqHbGWo c8rg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=WhO24zbRwspkmY3e6DNmHdAVkmwrh2weKhKJ9ZwWK/M=; b=bcFRGF1qi4xG6rGSyYQ2AfhCOZWVjueaWoUqZ7W0Dgdg3p9LCcVwVibWr2EGCv+Cma 3eY2BdvTfOZarT6oAp9XWJEZ71yQ9ZDAbXi+T83w+1FVMhqtI0ZkPZwfRONv6e0woFkz rpFReLdTT0qzFwXhpB7er/gOgHi/dKytaBvwTzyE1oscQBOU+9x5WdMM0oosRNmPR8NI ZDMIX4Y1XqmWGzZj8Ao7ceW2hTT8WQ9sKfeAzJPQLikwDyE+u60HBWsoB7gxGrgOrloR biAx5Rry3r0F8dSyusvJ2mrkUOG7nKj0IjrzW8J+8SYk//O1Jy91J7ciizEIpNUYJVYy 3uww== X-Gm-Message-State: AOAM5323SZZaSUlMpeSThpg4Bm74TdO5USXg/t6TItciYF4XzG7sV8ch TXea6Xucl9QrhJdHQ0xwYmbExCnk37U= X-Google-Smtp-Source: ABdhPJxDcR/vnEmwXL5UbKKldUnJGtYguSsFcMQf2D9xllc0tStKAkTbva2WNQlIvlBhVp9V3WArzg== X-Received: by 2002:a17:902:d48e:: with SMTP id c14mr5429156plg.79.1644016647874; Fri, 04 Feb 2022 15:17:27 -0800 (PST) Received: from ast-mbp.thefacebook.com ([2620:10d:c090:400::5:e4ee]) by smtp.gmail.com with ESMTPSA id q11sm4079798pfk.149.2022.02.04.15.17.26 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 04 Feb 2022 15:17:27 -0800 (PST) From: Alexei Starovoitov To: davem@davemloft.net Cc: daniel@iogearbox.net, andrii@kernel.org, bpf@vger.kernel.org, kernel-team@fb.com Subject: [PATCH bpf-next 5/5] bpf: Convert bpf_preload.ko to use light skeleton. Date: Fri, 4 Feb 2022 15:17:10 -0800 Message-Id: <20220204231710.25139-6-alexei.starovoitov@gmail.com> X-Mailer: git-send-email 2.32.0 In-Reply-To: <20220204231710.25139-1-alexei.starovoitov@gmail.com> References: <20220204231710.25139-1-alexei.starovoitov@gmail.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org X-Patchwork-Delegate: bpf@iogearbox.net From: Alexei Starovoitov The main change is a move of the single line #include "iterators.lskel.h" from iterators/iterators.c to bpf_preload_kern.c. Which means that generated light skeleton can be used from user space or user mode driver like iterators.c or from the kernel module. The direct use of light skeleton from the kernel module simplifies the code, since UMD is no longer necessary. The libbpf.a required user space and UMD. The CO-RE in the kernel and generated "loader bpf program" used by the light skeleton are capable to perform complex loading operations traditionally provided by libbpf. In addition UMD approach was launching UMD process every time bpffs has to be mounted. With light skeleton in the kernel the bpf_preload kernel module loads bpf iterators once and pins them multiple times into different bpffs mounts. Note the light skeleton cannot be used during early boot or out of kthread since light skeleton needs a valid mm. This limitation could be lifted in the future. Signed-off-by: Alexei Starovoitov --- kernel/bpf/inode.c | 39 ++---- kernel/bpf/preload/Kconfig | 9 +- kernel/bpf/preload/Makefile | 14 +-- kernel/bpf/preload/bpf_preload.h | 8 +- kernel/bpf/preload/bpf_preload_kern.c | 119 ++++++++---------- kernel/bpf/preload/bpf_preload_umd_blob.S | 7 -- .../preload/iterators/bpf_preload_common.h | 13 -- kernel/bpf/preload/iterators/iterators.c | 108 ---------------- kernel/bpf/syscall.c | 2 + 9 files changed, 72 insertions(+), 247 deletions(-) delete mode 100644 kernel/bpf/preload/bpf_preload_umd_blob.S delete mode 100644 kernel/bpf/preload/iterators/bpf_preload_common.h delete mode 100644 kernel/bpf/preload/iterators/iterators.c diff --git a/kernel/bpf/inode.c b/kernel/bpf/inode.c index 5a8d9f7467bf..4f841e16779e 100644 --- a/kernel/bpf/inode.c +++ b/kernel/bpf/inode.c @@ -710,11 +710,10 @@ static DEFINE_MUTEX(bpf_preload_lock); static int populate_bpffs(struct dentry *parent) { struct bpf_preload_info objs[BPF_PRELOAD_LINKS] = {}; - struct bpf_link *links[BPF_PRELOAD_LINKS] = {}; int err = 0, i; /* grab the mutex to make sure the kernel interactions with bpf_preload - * UMD are serialized + * are serialized */ mutex_lock(&bpf_preload_lock); @@ -722,40 +721,22 @@ static int populate_bpffs(struct dentry *parent) if (!bpf_preload_mod_get()) goto out; - if (!bpf_preload_ops->info.tgid) { - /* preload() will start UMD that will load BPF iterator programs */ - err = bpf_preload_ops->preload(objs); - if (err) + err = bpf_preload_ops->preload(objs); + if (err) + goto out_put; + for (i = 0; i < BPF_PRELOAD_LINKS; i++) { + bpf_link_inc(objs[i].link); + err = bpf_iter_link_pin_kernel(parent, + objs[i].link_name, objs[i].link); + if (err) { + bpf_link_put(objs[i].link); goto out_put; - for (i = 0; i < BPF_PRELOAD_LINKS; i++) { - links[i] = bpf_link_by_id(objs[i].link_id); - if (IS_ERR(links[i])) { - err = PTR_ERR(links[i]); - goto out_put; - } } - for (i = 0; i < BPF_PRELOAD_LINKS; i++) { - err = bpf_iter_link_pin_kernel(parent, - objs[i].link_name, links[i]); - if (err) - goto out_put; - /* do not unlink successfully pinned links even - * if later link fails to pin - */ - links[i] = NULL; - } - /* finish() will tell UMD process to exit */ - err = bpf_preload_ops->finish(); - if (err) - goto out_put; } out_put: bpf_preload_mod_put(); out: mutex_unlock(&bpf_preload_lock); - for (i = 0; i < BPF_PRELOAD_LINKS && err; i++) - if (!IS_ERR_OR_NULL(links[i])) - bpf_link_put(links[i]); return err; } diff --git a/kernel/bpf/preload/Kconfig b/kernel/bpf/preload/Kconfig index 26bced262473..9de6cfa5dbb1 100644 --- a/kernel/bpf/preload/Kconfig +++ b/kernel/bpf/preload/Kconfig @@ -18,10 +18,11 @@ menuconfig BPF_PRELOAD if BPF_PRELOAD config BPF_PRELOAD_UMD - tristate "bpf_preload kernel module with user mode driver" - depends on CC_CAN_LINK - depends on m || CC_CAN_LINK_STATIC + tristate "bpf_preload kernel module" + # light skeleton cannot run out of kthread without mm + depends on m default m help - This builds bpf_preload kernel module with embedded user mode driver. + This builds bpf_preload kernel module with embedded BPF programs for + introspection in bpffs. endif diff --git a/kernel/bpf/preload/Makefile b/kernel/bpf/preload/Makefile index baf47d9c7557..167534e3b0b4 100644 --- a/kernel/bpf/preload/Makefile +++ b/kernel/bpf/preload/Makefile @@ -3,16 +3,6 @@ LIBBPF_SRCS = $(srctree)/tools/lib/bpf/ LIBBPF_INCLUDE = $(LIBBPF_SRCS)/.. -userccflags += -I $(srctree)/tools/include/ -I $(srctree)/tools/include/uapi \ - -I $(LIBBPF_INCLUDE) -Wno-unused-result - -userprogs := bpf_preload_umd - -bpf_preload_umd-objs := iterators/iterators.o - -$(obj)/bpf_preload_umd: - -$(obj)/bpf_preload_umd_blob.o: $(obj)/bpf_preload_umd - obj-$(CONFIG_BPF_PRELOAD_UMD) += bpf_preload.o -bpf_preload-objs += bpf_preload_kern.o bpf_preload_umd_blob.o +CFLAGS_bpf_preload_kern.o += -I $(LIBBPF_INCLUDE) +bpf_preload-objs += bpf_preload_kern.o diff --git a/kernel/bpf/preload/bpf_preload.h b/kernel/bpf/preload/bpf_preload.h index 2f9932276f2e..f065c91213a0 100644 --- a/kernel/bpf/preload/bpf_preload.h +++ b/kernel/bpf/preload/bpf_preload.h @@ -2,13 +2,13 @@ #ifndef _BPF_PRELOAD_H #define _BPF_PRELOAD_H -#include -#include "iterators/bpf_preload_common.h" +struct bpf_preload_info { + char link_name[16]; + struct bpf_link *link; +}; struct bpf_preload_ops { - struct umd_info info; int (*preload)(struct bpf_preload_info *); - int (*finish)(void); struct module *owner; }; extern struct bpf_preload_ops *bpf_preload_ops; diff --git a/kernel/bpf/preload/bpf_preload_kern.c b/kernel/bpf/preload/bpf_preload_kern.c index 53736e52c1df..30207c048d36 100644 --- a/kernel/bpf/preload/bpf_preload_kern.c +++ b/kernel/bpf/preload/bpf_preload_kern.c @@ -2,101 +2,80 @@ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt #include #include -#include -#include -#include #include "bpf_preload.h" +#include "iterators/iterators.lskel.h" -extern char bpf_preload_umd_start; -extern char bpf_preload_umd_end; +static struct bpf_link *maps_link, *progs_link; +static struct iterators_bpf *skel; -static int preload(struct bpf_preload_info *obj); -static int finish(void); +static void free_links_and_skel(void) +{ + if (!IS_ERR_OR_NULL(maps_link)) + bpf_link_put(maps_link); + if (!IS_ERR_OR_NULL(progs_link)) + bpf_link_put(progs_link); + iterators_bpf__destroy(skel); +} + +static int preload(struct bpf_preload_info *obj) +{ + strlcpy(obj[0].link_name, "maps.debug", sizeof(obj[0].link_name)); + obj[0].link = maps_link; + strlcpy(obj[1].link_name, "progs.debug", sizeof(obj[1].link_name)); + obj[1].link = progs_link; + return 0; +} -static struct bpf_preload_ops umd_ops = { - .info.driver_name = "bpf_preload", +static struct bpf_preload_ops ops = { .preload = preload, - .finish = finish, .owner = THIS_MODULE, }; -static int preload(struct bpf_preload_info *obj) +static int load_skel(void) { - int magic = BPF_PRELOAD_START; - loff_t pos = 0; - int i, err; - ssize_t n; + int err; - err = fork_usermode_driver(&umd_ops.info); + skel = iterators_bpf__open(); + if (!skel) + return -ENOMEM; + err = iterators_bpf__load(skel); if (err) - return err; - - /* send the start magic to let UMD proceed with loading BPF progs */ - n = kernel_write(umd_ops.info.pipe_to_umh, - &magic, sizeof(magic), &pos); - if (n != sizeof(magic)) - return -EPIPE; - - /* receive bpf_link IDs and names from UMD */ - pos = 0; - for (i = 0; i < BPF_PRELOAD_LINKS; i++) { - n = kernel_read(umd_ops.info.pipe_from_umh, - &obj[i], sizeof(*obj), &pos); - if (n != sizeof(*obj)) - return -EPIPE; + goto out; + err = iterators_bpf__attach(skel); + if (err) + goto out; + maps_link = bpf_link_get_from_fd(skel->links.dump_bpf_map_fd); + if (IS_ERR(maps_link)) { + err = PTR_ERR(maps_link); + goto out; } - return 0; -} - -static int finish(void) -{ - int magic = BPF_PRELOAD_END; - struct pid *tgid; - loff_t pos = 0; - ssize_t n; - - /* send the last magic to UMD. It will do a normal exit. */ - n = kernel_write(umd_ops.info.pipe_to_umh, - &magic, sizeof(magic), &pos); - if (n != sizeof(magic)) - return -EPIPE; - - tgid = umd_ops.info.tgid; - if (tgid) { - wait_event(tgid->wait_pidfd, thread_group_exited(tgid)); - umd_cleanup_helper(&umd_ops.info); + progs_link = bpf_link_get_from_fd(skel->links.dump_bpf_prog_fd); + if (IS_ERR(progs_link)) { + err = PTR_ERR(progs_link); + goto out; } return 0; +out: + free_links_and_skel(); + return err; } -static int __init load_umd(void) +static int __init load(void) { int err; - err = umd_load_blob(&umd_ops.info, &bpf_preload_umd_start, - &bpf_preload_umd_end - &bpf_preload_umd_start); + err = load_skel(); if (err) return err; - bpf_preload_ops = &umd_ops; + bpf_preload_ops = &ops; return err; } -static void __exit fini_umd(void) +static void __exit fini(void) { - struct pid *tgid; - bpf_preload_ops = NULL; - - /* kill UMD in case it's still there due to earlier error */ - tgid = umd_ops.info.tgid; - if (tgid) { - kill_pid(tgid, SIGKILL, 1); - - wait_event(tgid->wait_pidfd, thread_group_exited(tgid)); - umd_cleanup_helper(&umd_ops.info); - } - umd_unload_blob(&umd_ops.info); + free_links_and_skel(); } -late_initcall(load_umd); -module_exit(fini_umd); +late_initcall(load); +module_exit(fini); MODULE_LICENSE("GPL"); diff --git a/kernel/bpf/preload/bpf_preload_umd_blob.S b/kernel/bpf/preload/bpf_preload_umd_blob.S deleted file mode 100644 index f1f40223b5c3..000000000000 --- a/kernel/bpf/preload/bpf_preload_umd_blob.S +++ /dev/null @@ -1,7 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ - .section .init.rodata, "a" - .global bpf_preload_umd_start -bpf_preload_umd_start: - .incbin "kernel/bpf/preload/bpf_preload_umd" - .global bpf_preload_umd_end -bpf_preload_umd_end: diff --git a/kernel/bpf/preload/iterators/bpf_preload_common.h b/kernel/bpf/preload/iterators/bpf_preload_common.h deleted file mode 100644 index 8464d1a48c05..000000000000 --- a/kernel/bpf/preload/iterators/bpf_preload_common.h +++ /dev/null @@ -1,13 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _BPF_PRELOAD_COMMON_H -#define _BPF_PRELOAD_COMMON_H - -#define BPF_PRELOAD_START 0x5555 -#define BPF_PRELOAD_END 0xAAAA - -struct bpf_preload_info { - char link_name[16]; - int link_id; -}; - -#endif diff --git a/kernel/bpf/preload/iterators/iterators.c b/kernel/bpf/preload/iterators/iterators.c deleted file mode 100644 index 4dafe0f4f2b2..000000000000 --- a/kernel/bpf/preload/iterators/iterators.c +++ /dev/null @@ -1,108 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* Copyright (c) 2020 Facebook */ -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include "iterators.lskel.h" -#include "bpf_preload_common.h" - -int to_kernel = -1; -int from_kernel = 0; - -static int __bpf_obj_get_info_by_fd(int bpf_fd, void *info, __u32 *info_len) -{ - union bpf_attr attr; - int err; - - memset(&attr, 0, sizeof(attr)); - attr.info.bpf_fd = bpf_fd; - attr.info.info_len = *info_len; - attr.info.info = (long) info; - - err = skel_sys_bpf(BPF_OBJ_GET_INFO_BY_FD, &attr, sizeof(attr)); - if (!err) - *info_len = attr.info.info_len; - return err; -} - -static int send_link_to_kernel(int link_fd, const char *link_name) -{ - struct bpf_preload_info obj = {}; - struct bpf_link_info info = {}; - __u32 info_len = sizeof(info); - int err; - - err = __bpf_obj_get_info_by_fd(link_fd, &info, &info_len); - if (err) - return err; - obj.link_id = info.id; - if (strlen(link_name) >= sizeof(obj.link_name)) - return -E2BIG; - strcpy(obj.link_name, link_name); - if (write(to_kernel, &obj, sizeof(obj)) != sizeof(obj)) - return -EPIPE; - return 0; -} - -int main(int argc, char **argv) -{ - struct iterators_bpf *skel; - int err, magic; - int debug_fd; - - debug_fd = open("/dev/console", O_WRONLY | O_NOCTTY | O_CLOEXEC); - if (debug_fd < 0) - return 1; - to_kernel = dup(1); - close(1); - dup(debug_fd); - /* now stdin and stderr point to /dev/console */ - - read(from_kernel, &magic, sizeof(magic)); - if (magic != BPF_PRELOAD_START) { - printf("bad start magic %d\n", magic); - return 1; - } - /* libbpf opens BPF object and loads it into the kernel */ - skel = iterators_bpf__open_and_load(); - if (!skel) { - /* iterators.skel.h is little endian. - * libbpf doesn't support automatic little->big conversion - * of BPF bytecode yet. - * The program load will fail in such case. - */ - printf("Failed load could be due to wrong endianness\n"); - return 1; - } - err = iterators_bpf__attach(skel); - if (err) - goto cleanup; - - /* send two bpf_link IDs with names to the kernel */ - err = send_link_to_kernel(skel->links.dump_bpf_map_fd, "maps.debug"); - if (err) - goto cleanup; - err = send_link_to_kernel(skel->links.dump_bpf_prog_fd, "progs.debug"); - if (err) - goto cleanup; - - /* The kernel will proceed with pinnging the links in bpffs. - * UMD will wait on read from pipe. - */ - read(from_kernel, &magic, sizeof(magic)); - if (magic != BPF_PRELOAD_END) { - printf("bad final magic %d\n", magic); - err = -EINVAL; - } -cleanup: - iterators_bpf__destroy(skel); - - return err != 0; -} diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 08ac01acc51e..54859f584115 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -2490,6 +2490,7 @@ void bpf_link_put(struct bpf_link *link) bpf_link_free(link); } } +EXPORT_SYMBOL(bpf_link_put); static int bpf_link_release(struct inode *inode, struct file *filp) { @@ -2632,6 +2633,7 @@ struct bpf_link *bpf_link_get_from_fd(u32 ufd) return link; } +EXPORT_SYMBOL(bpf_link_get_from_fd); struct bpf_tracing_link { struct bpf_link link;