From patchwork Thu Mar 18 06:31:04 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 404620 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id A7A31C433E0 for ; Thu, 18 Mar 2021 06:32:22 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 801AC64F24 for ; Thu, 18 Mar 2021 06:32:22 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229748AbhCRGbs convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:31:48 -0400 Received: from mx0b-00082601.pphosted.com ([67.231.153.30]:36418 "EHLO mx0b-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229704AbhCRGb0 (ORCPT ); Thu, 18 Mar 2021 02:31:26 -0400 Received: from pps.filterd (m0148460.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.43/8.16.0.43) with SMTP id 12I6QAmK004646 for ; Wed, 17 Mar 2021 23:31:25 -0700 Received: from mail.thefacebook.com ([163.114.132.120]) by mx0a-00082601.pphosted.com with ESMTP id 37bs29j9ch-17 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:25 -0700 Received: from intmgw001.05.ash7.facebook.com (2620:10d:c085:208::11) by mail.thefacebook.com (2620:10d:c085:11d::7) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:23 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id D96952ED24FA; Wed, 17 Mar 2021 23:31:18 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , Subject: [PATCH v3 bpf-next 01/12] libbpf: expose btf_type_by_id() internally Date: Wed, 17 Mar 2021 23:31:04 -0700 Message-ID: <20210318063115.49613-2-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 lowpriorityscore=0 bulkscore=0 spamscore=0 clxscore=1015 malwarescore=0 mlxscore=0 mlxlogscore=841 adultscore=0 suspectscore=0 priorityscore=1501 phishscore=0 impostorscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org btf_type_by_id() is internal-only convenience API returning non-const pointer to struct btf_type. Expose it outside of btf.c for re-use. Signed-off-by: Andrii Nakryiko --- tools/lib/bpf/btf.c | 2 +- tools/lib/bpf/libbpf_internal.h | 5 +++++ 2 files changed, 6 insertions(+), 1 deletion(-) diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index 3aa58f2ac183..e0b0a78b04fe 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -435,7 +435,7 @@ const struct btf *btf__base_btf(const struct btf *btf) } /* internal helper returning non-const pointer to a type */ -static struct btf_type *btf_type_by_id(struct btf *btf, __u32 type_id) +struct btf_type *btf_type_by_id(struct btf *btf, __u32 type_id) { if (type_id == 0) return &btf_void; diff --git a/tools/lib/bpf/libbpf_internal.h b/tools/lib/bpf/libbpf_internal.h index 343f6eb05637..d09860e435c8 100644 --- a/tools/lib/bpf/libbpf_internal.h +++ b/tools/lib/bpf/libbpf_internal.h @@ -107,6 +107,11 @@ static inline void *libbpf_reallocarray(void *ptr, size_t nmemb, size_t size) return realloc(ptr, total); } +struct btf; +struct btf_type; + +struct btf_type *btf_type_by_id(struct btf *btf, __u32 type_id); + void *btf_add_mem(void **data, size_t *cap_cnt, size_t elem_sz, size_t cur_cnt, size_t max_cnt, size_t add_cnt); int btf_ensure_mem(void **data, size_t *cap_cnt, size_t elem_sz, size_t need_cnt); From patchwork Thu Mar 18 06:31:05 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 404622 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, URIBL_BLOCKED, USER_AGENT_GIT autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 4BCE6C43332 for ; Thu, 18 Mar 2021 06:32:22 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 219AD64F10 for ; Thu, 18 Mar 2021 06:32:22 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229735AbhCRGbr convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:31:47 -0400 Received: from mx0a-00082601.pphosted.com ([67.231.145.42]:19962 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229559AbhCRGbY (ORCPT ); Thu, 18 Mar 2021 02:31:24 -0400 Received: from pps.filterd (m0109334.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.43/8.16.0.43) with SMTP id 12I6OQ9A011894 for ; Wed, 17 Mar 2021 23:31:24 -0700 Received: from mail.thefacebook.com ([163.114.132.120]) by mx0a-00082601.pphosted.com with ESMTP id 37bdyy6h96-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:24 -0700 Received: from intmgw001.37.frc1.facebook.com (2620:10d:c085:108::8) by mail.thefacebook.com (2620:10d:c085:11d::4) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:23 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id 0E42A2ED24FA; Wed, 17 Mar 2021 23:31:21 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , Subject: [PATCH v3 bpf-next 02/12] libbpf: generalize BTF and BTF.ext type ID and strings iteration Date: Wed, 17 Mar 2021 23:31:05 -0700 Message-ID: <20210318063115.49613-3-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 priorityscore=1501 bulkscore=0 lowpriorityscore=0 clxscore=1015 malwarescore=0 mlxlogscore=999 spamscore=0 phishscore=0 mlxscore=0 impostorscore=0 suspectscore=0 adultscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Extract and generalize the logic to iterate BTF type ID and string offset fields within BTF types and .BTF.ext data. Expose this internally in libbpf for re-use by bpf_linker. Additionally, complete strings deduplication handling for BTF.ext (e.g., CO-RE access strings), which was previously missing. There previously was no case of deduplicating .BTF.ext data, but bpf_linker is going to use it. Signed-off-by: Andrii Nakryiko --- tools/lib/bpf/btf.c | 393 ++++++++++++++++++-------------- tools/lib/bpf/libbpf_internal.h | 7 + 2 files changed, 228 insertions(+), 172 deletions(-) diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index e0b0a78b04fe..e137781f9bc6 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -3155,95 +3155,28 @@ static struct btf_dedup *btf_dedup_new(struct btf *btf, struct btf_ext *btf_ext, return d; } -typedef int (*str_off_fn_t)(__u32 *str_off_ptr, void *ctx); - /* * Iterate over all possible places in .BTF and .BTF.ext that can reference * string and pass pointer to it to a provided callback `fn`. */ -static int btf_for_each_str_off(struct btf_dedup *d, str_off_fn_t fn, void *ctx) +static int btf_for_each_str_off(struct btf_dedup *d, str_off_visit_fn fn, void *ctx) { - void *line_data_cur, *line_data_end; - int i, j, r, rec_size; - struct btf_type *t; + int i, r; for (i = 0; i < d->btf->nr_types; i++) { - t = btf_type_by_id(d->btf, d->btf->start_id + i); - r = fn(&t->name_off, ctx); + struct btf_type *t = btf_type_by_id(d->btf, d->btf->start_id + i); + + r = btf_type_visit_str_offs(t, fn, ctx); if (r) return r; - - switch (btf_kind(t)) { - case BTF_KIND_STRUCT: - case BTF_KIND_UNION: { - struct btf_member *m = btf_members(t); - __u16 vlen = btf_vlen(t); - - for (j = 0; j < vlen; j++) { - r = fn(&m->name_off, ctx); - if (r) - return r; - m++; - } - break; - } - case BTF_KIND_ENUM: { - struct btf_enum *m = btf_enum(t); - __u16 vlen = btf_vlen(t); - - for (j = 0; j < vlen; j++) { - r = fn(&m->name_off, ctx); - if (r) - return r; - m++; - } - break; - } - case BTF_KIND_FUNC_PROTO: { - struct btf_param *m = btf_params(t); - __u16 vlen = btf_vlen(t); - - for (j = 0; j < vlen; j++) { - r = fn(&m->name_off, ctx); - if (r) - return r; - m++; - } - break; - } - default: - break; - } } if (!d->btf_ext) return 0; - line_data_cur = d->btf_ext->line_info.info; - line_data_end = d->btf_ext->line_info.info + d->btf_ext->line_info.len; - rec_size = d->btf_ext->line_info.rec_size; - - while (line_data_cur < line_data_end) { - struct btf_ext_info_sec *sec = line_data_cur; - struct bpf_line_info_min *line_info; - __u32 num_info = sec->num_info; - - r = fn(&sec->sec_name_off, ctx); - if (r) - return r; - - line_data_cur += sizeof(struct btf_ext_info_sec); - for (i = 0; i < num_info; i++) { - line_info = line_data_cur; - r = fn(&line_info->file_name_off, ctx); - if (r) - return r; - r = fn(&line_info->line_off, ctx); - if (r) - return r; - line_data_cur += rec_size; - } - } + r = btf_ext_visit_str_offs(d->btf_ext, fn, ctx); + if (r) + return r; return 0; } @@ -4498,15 +4431,18 @@ static int btf_dedup_compact_types(struct btf_dedup *d) * then mapping it to a deduplicated type ID, stored in btf_dedup->hypot_map, * which is populated during compaction phase. */ -static int btf_dedup_remap_type_id(struct btf_dedup *d, __u32 type_id) +static int btf_dedup_remap_type_id(__u32 *type_id, void *ctx) { + struct btf_dedup *d = ctx; __u32 resolved_type_id, new_type_id; - resolved_type_id = resolve_type_id(d, type_id); + resolved_type_id = resolve_type_id(d, *type_id); new_type_id = d->hypot_map[resolved_type_id]; if (new_type_id > BTF_MAX_NR_TYPES) return -EINVAL; - return new_type_id; + + *type_id = new_type_id; + return 0; } /* @@ -4519,109 +4455,25 @@ static int btf_dedup_remap_type_id(struct btf_dedup *d, __u32 type_id) * referenced from any BTF type (e.g., struct fields, func proto args, etc) to * their final deduped type IDs. */ -static int btf_dedup_remap_type(struct btf_dedup *d, __u32 type_id) +static int btf_dedup_remap_types(struct btf_dedup *d) { - struct btf_type *t = btf_type_by_id(d->btf, type_id); int i, r; - switch (btf_kind(t)) { - case BTF_KIND_INT: - case BTF_KIND_ENUM: - case BTF_KIND_FLOAT: - break; - - case BTF_KIND_FWD: - case BTF_KIND_CONST: - case BTF_KIND_VOLATILE: - case BTF_KIND_RESTRICT: - case BTF_KIND_PTR: - case BTF_KIND_TYPEDEF: - case BTF_KIND_FUNC: - case BTF_KIND_VAR: - r = btf_dedup_remap_type_id(d, t->type); - if (r < 0) - return r; - t->type = r; - break; - - case BTF_KIND_ARRAY: { - struct btf_array *arr_info = btf_array(t); - - r = btf_dedup_remap_type_id(d, arr_info->type); - if (r < 0) - return r; - arr_info->type = r; - r = btf_dedup_remap_type_id(d, arr_info->index_type); - if (r < 0) - return r; - arr_info->index_type = r; - break; - } - - case BTF_KIND_STRUCT: - case BTF_KIND_UNION: { - struct btf_member *member = btf_members(t); - __u16 vlen = btf_vlen(t); - - for (i = 0; i < vlen; i++) { - r = btf_dedup_remap_type_id(d, member->type); - if (r < 0) - return r; - member->type = r; - member++; - } - break; - } - - case BTF_KIND_FUNC_PROTO: { - struct btf_param *param = btf_params(t); - __u16 vlen = btf_vlen(t); + for (i = 0; i < d->btf->nr_types; i++) { + struct btf_type *t = btf_type_by_id(d->btf, d->btf->start_id + i); - r = btf_dedup_remap_type_id(d, t->type); - if (r < 0) + r = btf_type_visit_type_ids(t, btf_dedup_remap_type_id, d); + if (r) return r; - t->type = r; - - for (i = 0; i < vlen; i++) { - r = btf_dedup_remap_type_id(d, param->type); - if (r < 0) - return r; - param->type = r; - param++; - } - break; - } - - case BTF_KIND_DATASEC: { - struct btf_var_secinfo *var = btf_var_secinfos(t); - __u16 vlen = btf_vlen(t); - - for (i = 0; i < vlen; i++) { - r = btf_dedup_remap_type_id(d, var->type); - if (r < 0) - return r; - var->type = r; - var++; - } - break; - } - - default: - return -EINVAL; } - return 0; -} + if (!d->btf_ext) + return 0; -static int btf_dedup_remap_types(struct btf_dedup *d) -{ - int i, r; + r = btf_ext_visit_type_ids(d->btf_ext, btf_dedup_remap_type_id, d); + if (r) + return r; - for (i = 0; i < d->btf->nr_types; i++) { - r = btf_dedup_remap_type(d, d->btf->start_id + i); - if (r < 0) - return r; - } return 0; } @@ -4675,3 +4527,200 @@ struct btf *libbpf_find_kernel_btf(void) pr_warn("failed to find valid kernel BTF\n"); return ERR_PTR(-ESRCH); } + +int btf_type_visit_type_ids(struct btf_type *t, type_id_visit_fn visit, void *ctx) +{ + int i, n, err; + + switch (btf_kind(t)) { + case BTF_KIND_INT: + case BTF_KIND_FLOAT: + case BTF_KIND_ENUM: + return 0; + + case BTF_KIND_FWD: + case BTF_KIND_CONST: + case BTF_KIND_VOLATILE: + case BTF_KIND_RESTRICT: + case BTF_KIND_PTR: + case BTF_KIND_TYPEDEF: + case BTF_KIND_FUNC: + case BTF_KIND_VAR: + return visit(&t->type, ctx); + + case BTF_KIND_ARRAY: { + struct btf_array *a = btf_array(t); + + err = visit(&a->type, ctx); + err = err ?: visit(&a->index_type, ctx); + return err; + } + + case BTF_KIND_STRUCT: + case BTF_KIND_UNION: { + struct btf_member *m = btf_members(t); + + for (i = 0, n = btf_vlen(t); i < n; i++, m++) { + err = visit(&m->type, ctx); + if (err) + return err; + } + return 0; + } + + case BTF_KIND_FUNC_PROTO: { + struct btf_param *m = btf_params(t); + + err = visit(&t->type, ctx); + if (err) + return err; + for (i = 0, n = btf_vlen(t); i < n; i++, m++) { + err = visit(&m->type, ctx); + if (err) + return err; + } + return 0; + } + + case BTF_KIND_DATASEC: { + struct btf_var_secinfo *m = btf_var_secinfos(t); + + for (i = 0, n = btf_vlen(t); i < n; i++, m++) { + err = visit(&m->type, ctx); + if (err) + return err; + } + return 0; + } + + default: + return -EINVAL; + } +} + +int btf_type_visit_str_offs(struct btf_type *t, str_off_visit_fn visit, void *ctx) +{ + int i, n, err; + + err = visit(&t->name_off, ctx); + if (err) + return err; + + switch (btf_kind(t)) { + case BTF_KIND_STRUCT: + case BTF_KIND_UNION: { + struct btf_member *m = btf_members(t); + + for (i = 0, n = btf_vlen(t); i < n; i++, m++) { + err = visit(&m->name_off, ctx); + if (err) + return err; + } + break; + } + case BTF_KIND_ENUM: { + struct btf_enum *m = btf_enum(t); + + for (i = 0, n = btf_vlen(t); i < n; i++, m++) { + err = visit(&m->name_off, ctx); + if (err) + return err; + } + break; + } + case BTF_KIND_FUNC_PROTO: { + struct btf_param *m = btf_params(t); + + for (i = 0, n = btf_vlen(t); i < n; i++, m++) { + err = visit(&m->name_off, ctx); + if (err) + return err; + } + break; + } + default: + break; + } + + return 0; +} + +int btf_ext_visit_type_ids(struct btf_ext *btf_ext, type_id_visit_fn visit, void *ctx) +{ + const struct btf_ext_info *seg; + struct btf_ext_info_sec *sec; + int i, err; + + seg = &btf_ext->func_info; + for_each_btf_ext_sec(seg, sec) { + struct bpf_func_info_min *rec; + + for_each_btf_ext_rec(seg, sec, i, rec) { + err = visit(&rec->type_id, ctx); + if (err < 0) + return err; + } + } + + seg = &btf_ext->core_relo_info; + for_each_btf_ext_sec(seg, sec) { + struct bpf_core_relo *rec; + + for_each_btf_ext_rec(seg, sec, i, rec) { + err = visit(&rec->type_id, ctx); + if (err < 0) + return err; + } + } + + return 0; +} + +int btf_ext_visit_str_offs(struct btf_ext *btf_ext, str_off_visit_fn visit, void *ctx) +{ + const struct btf_ext_info *seg; + struct btf_ext_info_sec *sec; + int i, err; + + seg = &btf_ext->func_info; + for_each_btf_ext_sec(seg, sec) { + err = visit(&sec->sec_name_off, ctx); + if (err) + return err; + } + + seg = &btf_ext->line_info; + for_each_btf_ext_sec(seg, sec) { + struct bpf_line_info_min *rec; + + err = visit(&sec->sec_name_off, ctx); + if (err) + return err; + + for_each_btf_ext_rec(seg, sec, i, rec) { + err = visit(&rec->file_name_off, ctx); + if (err) + return err; + err = visit(&rec->line_off, ctx); + if (err) + return err; + } + } + + seg = &btf_ext->core_relo_info; + for_each_btf_ext_sec(seg, sec) { + struct bpf_core_relo *rec; + + err = visit(&sec->sec_name_off, ctx); + if (err) + return err; + + for_each_btf_ext_rec(seg, sec, i, rec) { + err = visit(&rec->access_str_off, ctx); + if (err) + return err; + } + } + + return 0; +} diff --git a/tools/lib/bpf/libbpf_internal.h b/tools/lib/bpf/libbpf_internal.h index d09860e435c8..97b6b9cc9839 100644 --- a/tools/lib/bpf/libbpf_internal.h +++ b/tools/lib/bpf/libbpf_internal.h @@ -356,4 +356,11 @@ struct bpf_core_relo { enum bpf_core_relo_kind kind; }; +typedef int (*type_id_visit_fn)(__u32 *type_id, void *ctx); +typedef int (*str_off_visit_fn)(__u32 *str_off, void *ctx); +int btf_type_visit_type_ids(struct btf_type *t, type_id_visit_fn visit, void *ctx); +int btf_type_visit_str_offs(struct btf_type *t, str_off_visit_fn visit, void *ctx); +int btf_ext_visit_type_ids(struct btf_ext *btf_ext, type_id_visit_fn visit, void *ctx); +int btf_ext_visit_str_offs(struct btf_ext *btf_ext, str_off_visit_fn visit, void *ctx); + #endif /* __LIBBPF_LIBBPF_INTERNAL_H */ From patchwork Thu Mar 18 06:31:06 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 405374 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 04021C433E9 for ; Thu, 18 Mar 2021 06:32:52 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id BCB5964F30 for ; Thu, 18 Mar 2021 06:32:51 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229721AbhCRGcV convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:32:21 -0400 Received: from mx0b-00082601.pphosted.com ([67.231.153.30]:6268 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S229812AbhCRGbr (ORCPT ); Thu, 18 Mar 2021 02:31:47 -0400 Received: from pps.filterd (m0089730.ppops.net [127.0.0.1]) by m0089730.ppops.net (8.16.0.43/8.16.0.43) with SMTP id 12I6QRcn032153 for ; Wed, 17 Mar 2021 23:31:46 -0700 Received: from mail.thefacebook.com ([163.114.132.120]) by m0089730.ppops.net with ESMTP id 37bs1h29fv-9 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:46 -0700 Received: from intmgw001.05.ash9.facebook.com (2620:10d:c085:108::4) by mail.thefacebook.com (2620:10d:c085:21d::7) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:35 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id 387442ED24FA; Wed, 17 Mar 2021 23:31:23 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , Subject: [PATCH v3 bpf-next 03/12] libbpf: rename internal memory-management helpers Date: Wed, 17 Mar 2021 23:31:06 -0700 Message-ID: <20210318063115.49613-4-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 lowpriorityscore=0 mlxscore=0 mlxlogscore=999 bulkscore=0 priorityscore=1501 clxscore=1015 phishscore=0 impostorscore=0 adultscore=0 suspectscore=0 spamscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Rename btf_add_mem() and btf_ensure_mem() helpers that abstract away details of dynamically resizable memory to use libbpf_ prefix, as they are not BTF-specific. No functional changes. Signed-off-by: Andrii Nakryiko --- tools/lib/bpf/btf.c | 24 ++++++++++++------------ tools/lib/bpf/btf_dump.c | 8 ++++---- tools/lib/bpf/libbpf.c | 4 ++-- tools/lib/bpf/libbpf_internal.h | 6 +++--- 4 files changed, 21 insertions(+), 21 deletions(-) diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index e137781f9bc6..c98d39710515 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -142,8 +142,8 @@ static inline __u64 ptr_to_u64(const void *ptr) * On success, memory pointer to the beginning of unused memory is returned. * On error, NULL is returned. */ -void *btf_add_mem(void **data, size_t *cap_cnt, size_t elem_sz, - size_t cur_cnt, size_t max_cnt, size_t add_cnt) +void *libbpf_add_mem(void **data, size_t *cap_cnt, size_t elem_sz, + size_t cur_cnt, size_t max_cnt, size_t add_cnt) { size_t new_cnt; void *new_data; @@ -179,14 +179,14 @@ void *btf_add_mem(void **data, size_t *cap_cnt, size_t elem_sz, /* Ensure given dynamically allocated memory region has enough allocated space * to accommodate *need_cnt* elements of size *elem_sz* bytes each */ -int btf_ensure_mem(void **data, size_t *cap_cnt, size_t elem_sz, size_t need_cnt) +int libbpf_ensure_mem(void **data, size_t *cap_cnt, size_t elem_sz, size_t need_cnt) { void *p; if (need_cnt <= *cap_cnt) return 0; - p = btf_add_mem(data, cap_cnt, elem_sz, *cap_cnt, SIZE_MAX, need_cnt - *cap_cnt); + p = libbpf_add_mem(data, cap_cnt, elem_sz, *cap_cnt, SIZE_MAX, need_cnt - *cap_cnt); if (!p) return -ENOMEM; @@ -197,8 +197,8 @@ static int btf_add_type_idx_entry(struct btf *btf, __u32 type_off) { __u32 *p; - p = btf_add_mem((void **)&btf->type_offs, &btf->type_offs_cap, sizeof(__u32), - btf->nr_types, BTF_MAX_NR_TYPES, 1); + p = libbpf_add_mem((void **)&btf->type_offs, &btf->type_offs_cap, sizeof(__u32), + btf->nr_types, BTF_MAX_NR_TYPES, 1); if (!p) return -ENOMEM; @@ -1586,8 +1586,8 @@ static int btf_ensure_modifiable(struct btf *btf) static void *btf_add_str_mem(struct btf *btf, size_t add_sz) { - return btf_add_mem(&btf->strs_data, &btf->strs_data_cap, 1, - btf->hdr->str_len, BTF_MAX_STR_OFFSET, add_sz); + return libbpf_add_mem(&btf->strs_data, &btf->strs_data_cap, 1, + btf->hdr->str_len, BTF_MAX_STR_OFFSET, add_sz); } /* Find an offset in BTF string section that corresponds to a given string *s*. @@ -1683,8 +1683,8 @@ int btf__add_str(struct btf *btf, const char *s) static void *btf_add_type_mem(struct btf *btf, size_t add_sz) { - return btf_add_mem(&btf->types_data, &btf->types_data_cap, 1, - btf->hdr->type_len, UINT_MAX, add_sz); + return libbpf_add_mem(&btf->types_data, &btf->types_data_cap, 1, + btf->hdr->type_len, UINT_MAX, add_sz); } static __u32 btf_type_info(int kind, int vlen, int kflag) @@ -3208,7 +3208,7 @@ static int strs_dedup_remap_str_off(__u32 *str_off_ptr, void *ctx) len = strlen(s) + 1; new_off = d->strs_len; - p = btf_add_mem(&d->strs_data, &d->strs_cap, 1, new_off, BTF_MAX_STR_OFFSET, len); + p = libbpf_add_mem(&d->strs_data, &d->strs_cap, 1, new_off, BTF_MAX_STR_OFFSET, len); if (!p) return -ENOMEM; @@ -3264,7 +3264,7 @@ static int btf_dedup_strings(struct btf_dedup *d) } if (!d->btf->base_btf) { - s = btf_add_mem(&d->strs_data, &d->strs_cap, 1, d->strs_len, BTF_MAX_STR_OFFSET, 1); + s = libbpf_add_mem(&d->strs_data, &d->strs_cap, 1, d->strs_len, BTF_MAX_STR_OFFSET, 1); if (!s) return -ENOMEM; /* initial empty string */ diff --git a/tools/lib/bpf/btf_dump.c b/tools/lib/bpf/btf_dump.c index 5e957fcceee6..b5dbd5adc0e8 100644 --- a/tools/lib/bpf/btf_dump.c +++ b/tools/lib/bpf/btf_dump.c @@ -166,11 +166,11 @@ static int btf_dump_resize(struct btf_dump *d) if (last_id <= d->last_id) return 0; - if (btf_ensure_mem((void **)&d->type_states, &d->type_states_cap, - sizeof(*d->type_states), last_id + 1)) + if (libbpf_ensure_mem((void **)&d->type_states, &d->type_states_cap, + sizeof(*d->type_states), last_id + 1)) return -ENOMEM; - if (btf_ensure_mem((void **)&d->cached_names, &d->cached_names_cap, - sizeof(*d->cached_names), last_id + 1)) + if (libbpf_ensure_mem((void **)&d->cached_names, &d->cached_names_cap, + sizeof(*d->cached_names), last_id + 1)) return -ENOMEM; if (d->last_id == 0) { diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 2f351d3ad3e7..18ba37164e17 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -4867,8 +4867,8 @@ static int load_module_btfs(struct bpf_object *obj) goto err_out; } - err = btf_ensure_mem((void **)&obj->btf_modules, &obj->btf_module_cap, - sizeof(*obj->btf_modules), obj->btf_module_cnt + 1); + err = libbpf_ensure_mem((void **)&obj->btf_modules, &obj->btf_module_cap, + sizeof(*obj->btf_modules), obj->btf_module_cnt + 1); if (err) goto err_out; diff --git a/tools/lib/bpf/libbpf_internal.h b/tools/lib/bpf/libbpf_internal.h index 97b6b9cc9839..e0787900eeca 100644 --- a/tools/lib/bpf/libbpf_internal.h +++ b/tools/lib/bpf/libbpf_internal.h @@ -112,9 +112,9 @@ struct btf_type; struct btf_type *btf_type_by_id(struct btf *btf, __u32 type_id); -void *btf_add_mem(void **data, size_t *cap_cnt, size_t elem_sz, - size_t cur_cnt, size_t max_cnt, size_t add_cnt); -int btf_ensure_mem(void **data, size_t *cap_cnt, size_t elem_sz, size_t need_cnt); +void *libbpf_add_mem(void **data, size_t *cap_cnt, size_t elem_sz, + size_t cur_cnt, size_t max_cnt, size_t add_cnt); +int libbpf_ensure_mem(void **data, size_t *cap_cnt, size_t elem_sz, size_t need_cnt); static inline bool libbpf_validate_opts(const char *opts, size_t opts_sz, size_t user_sz, From patchwork Thu Mar 18 06:31:07 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 404621 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 71986C43331 for ; Thu, 18 Mar 2021 06:32:22 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 45BB864F24 for ; Thu, 18 Mar 2021 06:32:22 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229844AbhCRGbt convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:31:49 -0400 Received: from mx0b-00082601.pphosted.com ([67.231.153.30]:11452 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S229734AbhCRGbf (ORCPT ); Thu, 18 Mar 2021 02:31:35 -0400 Received: from pps.filterd (m0089730.ppops.net [127.0.0.1]) by m0089730.ppops.net (8.16.0.43/8.16.0.43) with SMTP id 12I6QRcf032153 for ; Wed, 17 Mar 2021 23:31:34 -0700 Received: from mail.thefacebook.com ([163.114.132.120]) by m0089730.ppops.net with ESMTP id 37bs1h29fv-4 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:34 -0700 Received: from intmgw001.37.frc1.facebook.com (2620:10d:c085:108::8) by mail.thefacebook.com (2620:10d:c085:21d::7) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:32 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id 6A0742ED24FA; Wed, 17 Mar 2021 23:31:25 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , , Alexei Starovoitov Subject: [PATCH v3 bpf-next 04/12] libbpf: extract internal set-of-strings datastructure APIs Date: Wed, 17 Mar 2021 23:31:07 -0700 Message-ID: <20210318063115.49613-5-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 lowpriorityscore=0 mlxscore=0 mlxlogscore=999 bulkscore=0 priorityscore=1501 clxscore=1015 phishscore=0 impostorscore=0 adultscore=0 suspectscore=0 spamscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Extract BTF logic for maintaining a set of strings data structure, used for BTF strings section construction in writable mode, into separate re-usable API. This data structure is going to be used by bpf_linker to maintains ELF STRTAB section, which has the same layout as BTF strings section. Suggested-by: Alexei Starovoitov Signed-off-by: Andrii Nakryiko --- tools/lib/bpf/Build | 2 +- tools/lib/bpf/btf.c | 255 ++++++++++------------------------------- tools/lib/bpf/strset.c | 176 ++++++++++++++++++++++++++++ tools/lib/bpf/strset.h | 21 ++++ 4 files changed, 259 insertions(+), 195 deletions(-) create mode 100644 tools/lib/bpf/strset.c create mode 100644 tools/lib/bpf/strset.h diff --git a/tools/lib/bpf/Build b/tools/lib/bpf/Build index 190366d05588..8136186a453f 100644 --- a/tools/lib/bpf/Build +++ b/tools/lib/bpf/Build @@ -1,3 +1,3 @@ libbpf-y := libbpf.o bpf.o nlattr.o btf.o libbpf_errno.o str_error.o \ netlink.o bpf_prog_linfo.o libbpf_probes.o xsk.o hashmap.o \ - btf_dump.o ringbuf.o + btf_dump.o ringbuf.o strset.o diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index c98d39710515..fe087592ad35 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -21,6 +21,7 @@ #include "libbpf.h" #include "libbpf_internal.h" #include "hashmap.h" +#include "strset.h" #define BTF_MAX_NR_TYPES 0x7fffffffU #define BTF_MAX_STR_OFFSET 0x7fffffffU @@ -67,7 +68,7 @@ struct btf { * | | | * hdr | | * types_data----+ | - * strs_data------------------+ + * strset__data(strs_set)-----+ * * +----------+---------+-----------+ * | Header | Types | Strings | @@ -105,20 +106,15 @@ struct btf { */ int start_str_off; + /* only one of strs_data or strs_set can be non-NULL, depending on + * whether BTF is in a modifiable state (strs_set is used) or not + * (strs_data points inside raw_data) + */ void *strs_data; - size_t strs_data_cap; /* used size stored in hdr->str_len */ - - /* lookup index for each unique string in strings section */ - struct hashmap *strs_hash; + /* a set of unique strings */ + struct strset *strs_set; /* whether strings are already deduplicated */ bool strs_deduped; - /* extra indirection layer to make strings hashmap work with stable - * string offsets and ability to transparently choose between - * btf->strs_data or btf_dedup->strs_data as a source of strings. - * This is used for BTF strings dedup to transfer deduplicated strings - * data back to struct btf without re-building strings index. - */ - void **strs_data_ptr; /* BTF object FD, if loaded into kernel */ int fd; @@ -738,7 +734,7 @@ void btf__free(struct btf *btf) */ free(btf->hdr); free(btf->types_data); - free(btf->strs_data); + strset__free(btf->strs_set); } free(btf->raw_data); free(btf->raw_data_swapped); @@ -1246,6 +1242,11 @@ void btf__set_fd(struct btf *btf, int fd) btf->fd = fd; } +static const void *btf_strs_data(const struct btf *btf) +{ + return btf->strs_data ? btf->strs_data : strset__data(btf->strs_set); +} + static void *btf_get_raw_data(const struct btf *btf, __u32 *size, bool swap_endian) { struct btf_header *hdr = btf->hdr; @@ -1286,7 +1287,7 @@ static void *btf_get_raw_data(const struct btf *btf, __u32 *size, bool swap_endi } p += hdr->type_len; - memcpy(p, btf->strs_data, hdr->str_len); + memcpy(p, btf_strs_data(btf), hdr->str_len); p += hdr->str_len; *size = data_sz; @@ -1320,7 +1321,7 @@ const char *btf__str_by_offset(const struct btf *btf, __u32 offset) if (offset < btf->start_str_off) return btf__str_by_offset(btf->base_btf, offset); else if (offset - btf->start_str_off < btf->hdr->str_len) - return btf->strs_data + (offset - btf->start_str_off); + return btf_strs_data(btf) + (offset - btf->start_str_off); else return NULL; } @@ -1474,25 +1475,6 @@ int btf__get_map_kv_tids(const struct btf *btf, const char *map_name, return 0; } -static size_t strs_hash_fn(const void *key, void *ctx) -{ - const struct btf *btf = ctx; - const char *strs = *btf->strs_data_ptr; - const char *str = strs + (long)key; - - return str_hash(str); -} - -static bool strs_hash_equal_fn(const void *key1, const void *key2, void *ctx) -{ - const struct btf *btf = ctx; - const char *strs = *btf->strs_data_ptr; - const char *str1 = strs + (long)key1; - const char *str2 = strs + (long)key2; - - return strcmp(str1, str2) == 0; -} - static void btf_invalidate_raw_data(struct btf *btf) { if (btf->raw_data) { @@ -1511,10 +1493,9 @@ static void btf_invalidate_raw_data(struct btf *btf) */ static int btf_ensure_modifiable(struct btf *btf) { - void *hdr, *types, *strs, *strs_end, *s; - struct hashmap *hash = NULL; - long off; - int err; + void *hdr, *types; + struct strset *set = NULL; + int err = -ENOMEM; if (btf_is_modifiable(btf)) { /* any BTF modification invalidates raw_data */ @@ -1525,44 +1506,25 @@ static int btf_ensure_modifiable(struct btf *btf) /* split raw data into three memory regions */ hdr = malloc(btf->hdr->hdr_len); types = malloc(btf->hdr->type_len); - strs = malloc(btf->hdr->str_len); - if (!hdr || !types || !strs) + if (!hdr || !types) goto err_out; memcpy(hdr, btf->hdr, btf->hdr->hdr_len); memcpy(types, btf->types_data, btf->hdr->type_len); - memcpy(strs, btf->strs_data, btf->hdr->str_len); - - /* make hashmap below use btf->strs_data as a source of strings */ - btf->strs_data_ptr = &btf->strs_data; /* build lookup index for all strings */ - hash = hashmap__new(strs_hash_fn, strs_hash_equal_fn, btf); - if (IS_ERR(hash)) { - err = PTR_ERR(hash); - hash = NULL; + set = strset__new(BTF_MAX_STR_OFFSET, btf->strs_data, btf->hdr->str_len); + if (IS_ERR(set)) { + err = PTR_ERR(set); goto err_out; } - strs_end = strs + btf->hdr->str_len; - for (off = 0, s = strs; s < strs_end; off += strlen(s) + 1, s = strs + off) { - /* hashmap__add() returns EEXIST if string with the same - * content already is in the hash map - */ - err = hashmap__add(hash, (void *)off, (void *)off); - if (err == -EEXIST) - continue; /* duplicate */ - if (err) - goto err_out; - } - /* only when everything was successful, update internal state */ btf->hdr = hdr; btf->types_data = types; btf->types_data_cap = btf->hdr->type_len; - btf->strs_data = strs; - btf->strs_data_cap = btf->hdr->str_len; - btf->strs_hash = hash; + btf->strs_data = NULL; + btf->strs_set = set; /* if BTF was created from scratch, all strings are guaranteed to be * unique and deduplicated */ @@ -1577,17 +1539,10 @@ static int btf_ensure_modifiable(struct btf *btf) return 0; err_out: - hashmap__free(hash); + strset__free(set); free(hdr); free(types); - free(strs); - return -ENOMEM; -} - -static void *btf_add_str_mem(struct btf *btf, size_t add_sz) -{ - return libbpf_add_mem(&btf->strs_data, &btf->strs_data_cap, 1, - btf->hdr->str_len, BTF_MAX_STR_OFFSET, add_sz); + return err; } /* Find an offset in BTF string section that corresponds to a given string *s*. @@ -1598,34 +1553,23 @@ static void *btf_add_str_mem(struct btf *btf, size_t add_sz) */ int btf__find_str(struct btf *btf, const char *s) { - long old_off, new_off, len; - void *p; + int off; if (btf->base_btf) { - int ret; - - ret = btf__find_str(btf->base_btf, s); - if (ret != -ENOENT) - return ret; + off = btf__find_str(btf->base_btf, s); + if (off != -ENOENT) + return off; } /* BTF needs to be in a modifiable state to build string lookup index */ if (btf_ensure_modifiable(btf)) return -ENOMEM; - /* see btf__add_str() for why we do this */ - len = strlen(s) + 1; - p = btf_add_str_mem(btf, len); - if (!p) - return -ENOMEM; - - new_off = btf->hdr->str_len; - memcpy(p, s, len); + off = strset__find_str(btf->strs_set, s); + if (off < 0) + return off; - if (hashmap__find(btf->strs_hash, (void *)new_off, (void **)&old_off)) - return btf->start_str_off + old_off; - - return -ENOENT; + return btf->start_str_off + off; } /* Add a string s to the BTF string section. @@ -1635,50 +1579,24 @@ int btf__find_str(struct btf *btf, const char *s) */ int btf__add_str(struct btf *btf, const char *s) { - long old_off, new_off, len; - void *p; - int err; + int off; if (btf->base_btf) { - int ret; - - ret = btf__find_str(btf->base_btf, s); - if (ret != -ENOENT) - return ret; + off = btf__find_str(btf->base_btf, s); + if (off != -ENOENT) + return off; } if (btf_ensure_modifiable(btf)) return -ENOMEM; - /* Hashmap keys are always offsets within btf->strs_data, so to even - * look up some string from the "outside", we need to first append it - * at the end, so that it can be addressed with an offset. Luckily, - * until btf->hdr->str_len is incremented, that string is just a piece - * of garbage for the rest of BTF code, so no harm, no foul. On the - * other hand, if the string is unique, it's already appended and - * ready to be used, only a simple btf->hdr->str_len increment away. - */ - len = strlen(s) + 1; - p = btf_add_str_mem(btf, len); - if (!p) - return -ENOMEM; + off = strset__add_str(btf->strs_set, s); + if (off < 0) + return off; - new_off = btf->hdr->str_len; - memcpy(p, s, len); + btf->hdr->str_len = strset__data_size(btf->strs_set); - /* Now attempt to add the string, but only if the string with the same - * contents doesn't exist already (HASHMAP_ADD strategy). If such - * string exists, we'll get its offset in old_off (that's old_key). - */ - err = hashmap__insert(btf->strs_hash, (void *)new_off, (void *)new_off, - HASHMAP_ADD, (const void **)&old_off, NULL); - if (err == -EEXIST) - return btf->start_str_off + old_off; /* duplicated string, return existing offset */ - if (err) - return err; - - btf->hdr->str_len += len; /* new unique string, adjust data length */ - return btf->start_str_off + new_off; + return btf->start_str_off + off; } static void *btf_add_type_mem(struct btf *btf, size_t add_sz) @@ -3016,10 +2934,7 @@ struct btf_dedup { /* Various option modifying behavior of algorithm */ struct btf_dedup_opts opts; /* temporary strings deduplication state */ - void *strs_data; - size_t strs_cap; - size_t strs_len; - struct hashmap* strs_hash; + struct strset *strs_set; }; static long hash_combine(long h, long value) @@ -3185,10 +3100,8 @@ static int strs_dedup_remap_str_off(__u32 *str_off_ptr, void *ctx) { struct btf_dedup *d = ctx; __u32 str_off = *str_off_ptr; - long old_off, new_off, len; const char *s; - void *p; - int err; + int off, err; /* don't touch empty string or string in main BTF */ if (str_off == 0 || str_off < d->btf->start_str_off) @@ -3205,29 +3118,11 @@ static int strs_dedup_remap_str_off(__u32 *str_off_ptr, void *ctx) return err; } - len = strlen(s) + 1; + off = strset__add_str(d->strs_set, s); + if (off < 0) + return off; - new_off = d->strs_len; - p = libbpf_add_mem(&d->strs_data, &d->strs_cap, 1, new_off, BTF_MAX_STR_OFFSET, len); - if (!p) - return -ENOMEM; - - memcpy(p, s, len); - - /* Now attempt to add the string, but only if the string with the same - * contents doesn't exist already (HASHMAP_ADD strategy). If such - * string exists, we'll get its offset in old_off (that's old_key). - */ - err = hashmap__insert(d->strs_hash, (void *)new_off, (void *)new_off, - HASHMAP_ADD, (const void **)&old_off, NULL); - if (err == -EEXIST) { - *str_off_ptr = d->btf->start_str_off + old_off; - } else if (err) { - return err; - } else { - *str_off_ptr = d->btf->start_str_off + new_off; - d->strs_len += len; - } + *str_off_ptr = d->btf->start_str_off + off; return 0; } @@ -3244,39 +3139,23 @@ static int strs_dedup_remap_str_off(__u32 *str_off_ptr, void *ctx) */ static int btf_dedup_strings(struct btf_dedup *d) { - char *s; int err; if (d->btf->strs_deduped) return 0; - /* temporarily switch to use btf_dedup's strs_data for strings for hash - * functions; later we'll just transfer hashmap to struct btf as is, - * along the strs_data - */ - d->btf->strs_data_ptr = &d->strs_data; - - d->strs_hash = hashmap__new(strs_hash_fn, strs_hash_equal_fn, d->btf); - if (IS_ERR(d->strs_hash)) { - err = PTR_ERR(d->strs_hash); - d->strs_hash = NULL; + d->strs_set = strset__new(BTF_MAX_STR_OFFSET, NULL, 0); + if (IS_ERR(d->strs_set)) { + err = PTR_ERR(d->strs_set); goto err_out; } if (!d->btf->base_btf) { - s = libbpf_add_mem(&d->strs_data, &d->strs_cap, 1, d->strs_len, BTF_MAX_STR_OFFSET, 1); - if (!s) - return -ENOMEM; - /* initial empty string */ - s[0] = 0; - d->strs_len = 1; - /* insert empty string; we won't be looking it up during strings * dedup, but it's good to have it for generic BTF string lookups */ - err = hashmap__insert(d->strs_hash, (void *)0, (void *)0, - HASHMAP_ADD, NULL, NULL); - if (err) + err = strset__add_str(d->strs_set, ""); + if (err < 0) goto err_out; } @@ -3286,28 +3165,16 @@ static int btf_dedup_strings(struct btf_dedup *d) goto err_out; /* replace BTF string data and hash with deduped ones */ - free(d->btf->strs_data); - hashmap__free(d->btf->strs_hash); - d->btf->strs_data = d->strs_data; - d->btf->strs_data_cap = d->strs_cap; - d->btf->hdr->str_len = d->strs_len; - d->btf->strs_hash = d->strs_hash; - /* now point strs_data_ptr back to btf->strs_data */ - d->btf->strs_data_ptr = &d->btf->strs_data; - - d->strs_data = d->strs_hash = NULL; - d->strs_len = d->strs_cap = 0; + strset__free(d->btf->strs_set); + d->btf->hdr->str_len = strset__data_size(d->strs_set); + d->btf->strs_set = d->strs_set; + d->strs_set = NULL; d->btf->strs_deduped = true; return 0; err_out: - free(d->strs_data); - hashmap__free(d->strs_hash); - d->strs_data = d->strs_hash = NULL; - d->strs_len = d->strs_cap = 0; - - /* restore strings pointer for existing d->btf->strs_hash back */ - d->btf->strs_data_ptr = &d->strs_data; + strset__free(d->strs_set); + d->strs_set = NULL; return err; } diff --git a/tools/lib/bpf/strset.c b/tools/lib/bpf/strset.c new file mode 100644 index 000000000000..1fb8b49de1d6 --- /dev/null +++ b/tools/lib/bpf/strset.c @@ -0,0 +1,176 @@ +// SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) +/* Copyright (c) 2021 Facebook */ +#include +#include +#include +#include +#include +#include "hashmap.h" +#include "libbpf_internal.h" +#include "strset.h" + +struct strset { + void *strs_data; + size_t strs_data_len; + size_t strs_data_cap; + size_t strs_data_max_len; + + /* lookup index for each unique string in strings set */ + struct hashmap *strs_hash; +}; + +static size_t strset_hash_fn(const void *key, void *ctx) +{ + const struct strset *s = ctx; + const char *str = s->strs_data + (long)key; + + return str_hash(str); +} + +static bool strset_equal_fn(const void *key1, const void *key2, void *ctx) +{ + const struct strset *s = ctx; + const char *str1 = s->strs_data + (long)key1; + const char *str2 = s->strs_data + (long)key2; + + return strcmp(str1, str2) == 0; +} + +struct strset *strset__new(size_t max_data_sz, const char *init_data, size_t init_data_sz) +{ + struct strset *set = calloc(1, sizeof(*set)); + struct hashmap *hash; + int err = -ENOMEM; + + if (!set) + return ERR_PTR(-ENOMEM); + + hash = hashmap__new(strset_hash_fn, strset_equal_fn, set); + if (IS_ERR(hash)) + goto err_out; + + set->strs_data_max_len = max_data_sz; + set->strs_hash = hash; + + if (init_data) { + long off; + + set->strs_data = malloc(init_data_sz); + if (!set->strs_data) + goto err_out; + + memcpy(set->strs_data, init_data, init_data_sz); + set->strs_data_len = init_data_sz; + set->strs_data_cap = init_data_sz; + + for (off = 0; off < set->strs_data_len; off += strlen(set->strs_data + off) + 1) { + /* hashmap__add() returns EEXIST if string with the same + * content already is in the hash map + */ + err = hashmap__add(hash, (void *)off, (void *)off); + if (err == -EEXIST) + continue; /* duplicate */ + if (err) + goto err_out; + } + } + + return set; +err_out: + strset__free(set); + return ERR_PTR(err); +} + +void strset__free(struct strset *set) +{ + if (IS_ERR_OR_NULL(set)) + return; + + hashmap__free(set->strs_hash); + free(set->strs_data); +} + +size_t strset__data_size(const struct strset *set) +{ + return set->strs_data_len; +} + +const char *strset__data(const struct strset *set) +{ + return set->strs_data; +} + +static void *strset_add_str_mem(struct strset *set, size_t add_sz) +{ + return libbpf_add_mem(&set->strs_data, &set->strs_data_cap, 1, + set->strs_data_len, set->strs_data_max_len, add_sz); +} + +/* Find string offset that corresponds to a given string *s*. + * Returns: + * - >0 offset into string data, if string is found; + * - -ENOENT, if string is not in the string data; + * - <0, on any other error. + */ +int strset__find_str(struct strset *set, const char *s) +{ + long old_off, new_off, len; + void *p; + + /* see strset__add_str() for why we do this */ + len = strlen(s) + 1; + p = strset_add_str_mem(set, len); + if (!p) + return -ENOMEM; + + new_off = set->strs_data_len; + memcpy(p, s, len); + + if (hashmap__find(set->strs_hash, (void *)new_off, (void **)&old_off)) + return old_off; + + return -ENOENT; +} + +/* Add a string s to the string data. If the string already exists, return its + * offset within string data. + * Returns: + * - > 0 offset into string data, on success; + * - < 0, on error. + */ +int strset__add_str(struct strset *set, const char *s) +{ + long old_off, new_off, len; + void *p; + int err; + + /* Hashmap keys are always offsets within set->strs_data, so to even + * look up some string from the "outside", we need to first append it + * at the end, so that it can be addressed with an offset. Luckily, + * until set->strs_data_len is incremented, that string is just a piece + * of garbage for the rest of the code, so no harm, no foul. On the + * other hand, if the string is unique, it's already appended and + * ready to be used, only a simple set->strs_data_len increment away. + */ + len = strlen(s) + 1; + p = strset_add_str_mem(set, len); + if (!p) + return -ENOMEM; + + new_off = set->strs_data_len; + memcpy(p, s, len); + + /* Now attempt to add the string, but only if the string with the same + * contents doesn't exist already (HASHMAP_ADD strategy). If such + * string exists, we'll get its offset in old_off (that's old_key). + */ + err = hashmap__insert(set->strs_hash, (void *)new_off, (void *)new_off, + HASHMAP_ADD, (const void **)&old_off, NULL); + if (err == -EEXIST) + return old_off; /* duplicated string, return existing offset */ + if (err) + return err; + + set->strs_data_len += len; /* new unique string, adjust data length */ + return new_off; +} diff --git a/tools/lib/bpf/strset.h b/tools/lib/bpf/strset.h new file mode 100644 index 000000000000..b6ddf77a83c2 --- /dev/null +++ b/tools/lib/bpf/strset.h @@ -0,0 +1,21 @@ +/* SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) */ + +/* Copyright (c) 2021 Facebook */ +#ifndef __LIBBPF_STRSET_H +#define __LIBBPF_STRSET_H + +#include +#include + +struct strset; + +struct strset *strset__new(size_t max_data_sz, const char *init_data, size_t init_data_sz); +void strset__free(struct strset *set); + +const char *strset__data(const struct strset *set); +size_t strset__data_size(const struct strset *set); + +int strset__find_str(struct strset *set, const char *s); +int strset__add_str(struct strset *set, const char *s); + +#endif /* __LIBBPF_STRSET_H */ From patchwork Thu Mar 18 06:31:08 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 405378 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id D0125C43333 for ; Thu, 18 Mar 2021 06:32:22 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id A125264F04 for ; Thu, 18 Mar 2021 06:32:22 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229867AbhCRGbu convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:31:50 -0400 Received: from mx0b-00082601.pphosted.com ([67.231.153.30]:20744 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S229766AbhCRGbk (ORCPT ); Thu, 18 Mar 2021 02:31:40 -0400 Received: from pps.filterd (m0089730.ppops.net [127.0.0.1]) by m0089730.ppops.net (8.16.0.43/8.16.0.43) with SMTP id 12I6QbSS001920 for ; Wed, 17 Mar 2021 23:31:39 -0700 Received: from maileast.thefacebook.com ([163.114.130.16]) by m0089730.ppops.net with ESMTP id 37bs1h29cb-20 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:39 -0700 Received: from intmgw001.38.frc1.facebook.com (2620:10d:c0a8:1b::d) by mail.thefacebook.com (2620:10d:c0a8:83::7) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:35 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id 8D2DF2ED24FA; Wed, 17 Mar 2021 23:31:27 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , Subject: [PATCH v3 bpf-next 05/12] libbpf: add generic BTF type shallow copy API Date: Wed, 17 Mar 2021 23:31:08 -0700 Message-ID: <20210318063115.49613-6-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 lowpriorityscore=0 mlxscore=0 mlxlogscore=999 bulkscore=0 priorityscore=1501 clxscore=1015 phishscore=0 impostorscore=0 adultscore=0 suspectscore=0 spamscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Add btf__add_type() API that performs shallow copy of a given BTF type from the source BTF into the destination BTF. All the information and type IDs are preserved, but all the strings encountered are added into the destination BTF and corresponding offsets are rewritten. BTF type IDs are assumed to be correct or such that will be (somehow) modified afterwards. Signed-off-by: Andrii Nakryiko --- tools/lib/bpf/btf.c | 48 ++++++++++++++++++++++++++++++++++++++++ tools/lib/bpf/btf.h | 2 ++ tools/lib/bpf/libbpf.map | 1 + 3 files changed, 51 insertions(+) diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index fe087592ad35..d30e67e7e1e5 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -1629,6 +1629,54 @@ static int btf_commit_type(struct btf *btf, int data_sz) return btf->start_id + btf->nr_types - 1; } +struct btf_pipe { + const struct btf *src; + struct btf *dst; +}; + +static int btf_rewrite_str(__u32 *str_off, void *ctx) +{ + struct btf_pipe *p = ctx; + int off; + + if (!*str_off) /* nothing to do for empty strings */ + return 0; + + off = btf__add_str(p->dst, btf__str_by_offset(p->src, *str_off)); + if (off < 0) + return off; + + *str_off = off; + return 0; +} + +int btf__add_type(struct btf *btf, const struct btf *src_btf, const struct btf_type *src_type) +{ + struct btf_pipe p = { .src = src_btf, .dst = btf }; + struct btf_type *t; + int sz, err; + + sz = btf_type_size(src_type); + if (sz < 0) + return sz; + + /* deconstruct BTF, if necessary, and invalidate raw_data */ + if (btf_ensure_modifiable(btf)) + return -ENOMEM; + + t = btf_add_type_mem(btf, sz); + if (!t) + return -ENOMEM; + + memcpy(t, src_type, sz); + + err = btf_type_visit_str_offs(t, btf_rewrite_str, &p); + if (err) + return err; + + return btf_commit_type(btf, sz); +} + /* * Append new BTF_KIND_INT type with: * - *name* - non-empty, non-NULL type name; diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h index 029a9cfc8c2d..3b0b17ba94a1 100644 --- a/tools/lib/bpf/btf.h +++ b/tools/lib/bpf/btf.h @@ -93,6 +93,8 @@ LIBBPF_API struct btf *libbpf_find_kernel_btf(void); LIBBPF_API int btf__find_str(struct btf *btf, const char *s); LIBBPF_API int btf__add_str(struct btf *btf, const char *s); +LIBBPF_API int btf__add_type(struct btf *btf, const struct btf *src_btf, + const struct btf_type *src_type); LIBBPF_API int btf__add_int(struct btf *btf, const char *name, size_t byte_sz, int encoding); LIBBPF_API int btf__add_float(struct btf *btf, const char *name, size_t byte_sz); diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index ec898f464ab9..d31d8c968097 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -354,4 +354,5 @@ LIBBPF_0.3.0 { LIBBPF_0.4.0 { global: btf__add_float; + btf__add_type; } LIBBPF_0.3.0; From patchwork Thu Mar 18 06:31:09 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 404619 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 81D92C4361A for ; Thu, 18 Mar 2021 06:32:23 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 4FA4264F10 for ; Thu, 18 Mar 2021 06:32:23 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229878AbhCRGbw convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:31:52 -0400 Received: from mx0a-00082601.pphosted.com ([67.231.145.42]:61900 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229785AbhCRGbn (ORCPT ); Thu, 18 Mar 2021 02:31:43 -0400 Received: from pps.filterd (m0044012.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.43/8.16.0.43) with SMTP id 12I6PA6K004513 for ; Wed, 17 Mar 2021 23:31:43 -0700 Received: from maileast.thefacebook.com ([163.114.130.16]) by mx0a-00082601.pphosted.com with ESMTP id 37b3bs194h-5 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:43 -0700 Received: from intmgw001.38.frc1.facebook.com (2620:10d:c0a8:1b::d) by mail.thefacebook.com (2620:10d:c0a8:82::d) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:35 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id BEF342ED24FA; Wed, 17 Mar 2021 23:31:29 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , Subject: [PATCH v3 bpf-next 06/12] libbpf: add BPF static linker APIs Date: Wed, 17 Mar 2021 23:31:09 -0700 Message-ID: <20210318063115.49613-7-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 phishscore=0 mlxlogscore=999 suspectscore=0 malwarescore=0 impostorscore=0 mlxscore=0 spamscore=0 clxscore=1015 bulkscore=0 adultscore=0 priorityscore=1501 lowpriorityscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Introduce BPF static linker APIs to libbpf. BPF static linker allows to perform static linking of multiple BPF object files into a single combined resulting object file, preserving all the BPF programs, maps, global variables, etc. Data sections (.bss, .data, .rodata, .maps, maps, etc) with the same name are concatenated together. Similarly, code sections are also concatenated. All the symbols and ELF relocations are also concatenated in their respective ELF sections and are adjusted accordingly to the new object file layout. Static variables and functions are handled correctly as well, adjusting BPF instructions offsets to reflect new variable/function offset within the combined ELF section. Such relocations are referencing STT_SECTION symbols and that stays intact. Data sections in different files can have different alignment requirements, so that is taken care of as well, adjusting sizes and offsets as necessary to satisfy both old and new alignment requirements. DWARF data sections are stripped out, currently. As well as LLLVM_ADDRSIG section, which is ignored by libbpf in bpf_object__open() anyways. So, in a way, BPF static linker is an analogue to `llvm-strip -g`, which is a pretty nice property, especially if resulting .o file is then used to generate BPF skeleton. Original string sections are ignored and instead we construct our own set of unique strings using libbpf-internal `struct strset` API. To reduce the size of the patch, all the .BTF and .BTF.ext processing was moved into a separate patch. The high-level API consists of just 4 functions: - bpf_linker__new() creates an instance of BPF static linker. It accepts output filename and (currently empty) options struct; - bpf_linker__add_file() takes input filename and appends it to the already processed ELF data; it can be called multiple times, one for each BPF ELF object file that needs to be linked in; - bpf_linker__finalize() needs to be called to dump final ELF contents into the output file, specified when bpf_linker was created; after bpf_linker__finalize() is called, no more bpf_linker__add_file() and bpf_linker__finalize() calls are allowed, they will return error; - regardless of whether bpf_linker__finalize() was called or not, bpf_linker__free() will free up all the used resources. Currently, BPF static linker doesn't resolve cross-object file references (extern variables and/or functions). This will be added in the follow up patch set. Signed-off-by: Andrii Nakryiko --- tools/lib/bpf/Build | 2 +- tools/lib/bpf/libbpf.c | 11 +- tools/lib/bpf/libbpf.h | 13 + tools/lib/bpf/libbpf.map | 4 + tools/lib/bpf/libbpf_internal.h | 20 + tools/lib/bpf/linker.c | 1176 +++++++++++++++++++++++++++++++ 6 files changed, 1215 insertions(+), 11 deletions(-) create mode 100644 tools/lib/bpf/linker.c diff --git a/tools/lib/bpf/Build b/tools/lib/bpf/Build index 8136186a453f..9b057cc7650a 100644 --- a/tools/lib/bpf/Build +++ b/tools/lib/bpf/Build @@ -1,3 +1,3 @@ libbpf-y := libbpf.o bpf.o nlattr.o btf.o libbpf_errno.o str_error.o \ netlink.o bpf_prog_linfo.o libbpf_probes.o xsk.o hashmap.o \ - btf_dump.o ringbuf.o strset.o + btf_dump.o ringbuf.o strset.o linker.o diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 18ba37164e17..058b643cbcb1 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -55,10 +55,6 @@ #include "libbpf_internal.h" #include "hashmap.h" -#ifndef EM_BPF -#define EM_BPF 247 -#endif - #ifndef BPF_FS_MAGIC #define BPF_FS_MAGIC 0xcafe4a11 #endif @@ -1134,11 +1130,6 @@ static void bpf_object__elf_finish(struct bpf_object *obj) obj->efile.obj_buf_sz = 0; } -/* if libelf is old and doesn't support mmap(), fall back to read() */ -#ifndef ELF_C_READ_MMAP -#define ELF_C_READ_MMAP ELF_C_READ -#endif - static int bpf_object__elf_init(struct bpf_object *obj) { int err = 0; @@ -2807,7 +2798,7 @@ static bool ignore_elf_section(GElf_Shdr *hdr, const char *name) return true; /* ignore .llvm_addrsig section as well */ - if (hdr->sh_type == 0x6FFF4C03 /* SHT_LLVM_ADDRSIG */) + if (hdr->sh_type == SHT_LLVM_ADDRSIG) return true; /* no subprograms will lead to an empty .text section, ignore it */ diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index 3d690d4e785c..a1a424b9b8ff 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h @@ -760,6 +760,19 @@ enum libbpf_tristate { TRI_MODULE = 2, }; +struct bpf_linker_opts { + /* size of this struct, for forward/backward compatiblity */ + size_t sz; +}; +#define bpf_linker_opts__last_field sz + +struct bpf_linker; + +LIBBPF_API struct bpf_linker *bpf_linker__new(const char *filename, struct bpf_linker_opts *opts); +LIBBPF_API int bpf_linker__add_file(struct bpf_linker *linker, const char *filename); +LIBBPF_API int bpf_linker__finalize(struct bpf_linker *linker); +LIBBPF_API void bpf_linker__free(struct bpf_linker *linker); + #ifdef __cplusplus } /* extern "C" */ #endif diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map index d31d8c968097..279ae861f568 100644 --- a/tools/lib/bpf/libbpf.map +++ b/tools/lib/bpf/libbpf.map @@ -355,4 +355,8 @@ LIBBPF_0.4.0 { global: btf__add_float; btf__add_type; + bpf_linker__add_file; + bpf_linker__finalize; + bpf_linker__free; + bpf_linker__new; } LIBBPF_0.3.0; diff --git a/tools/lib/bpf/libbpf_internal.h b/tools/lib/bpf/libbpf_internal.h index e0787900eeca..6017902c687e 100644 --- a/tools/lib/bpf/libbpf_internal.h +++ b/tools/lib/bpf/libbpf_internal.h @@ -20,6 +20,26 @@ #include "libbpf.h" +#ifndef EM_BPF +#define EM_BPF 247 +#endif + +#ifndef R_BPF_64_64 +#define R_BPF_64_64 1 +#endif +#ifndef R_BPF_64_32 +#define R_BPF_64_32 10 +#endif + +#ifndef SHT_LLVM_ADDRSIG +#define SHT_LLVM_ADDRSIG 0x6FFF4C03 +#endif + +/* if libelf is old and doesn't support mmap(), fall back to read() */ +#ifndef ELF_C_READ_MMAP +#define ELF_C_READ_MMAP ELF_C_READ +#endif + #define BTF_INFO_ENC(kind, kind_flag, vlen) \ ((!!(kind_flag) << 31) | ((kind) << 24) | ((vlen) & BTF_MAX_VLEN)) #define BTF_TYPE_ENC(name, info, size_or_type) (name), (info), (size_or_type) diff --git a/tools/lib/bpf/linker.c b/tools/lib/bpf/linker.c new file mode 100644 index 000000000000..01b56939a835 --- /dev/null +++ b/tools/lib/bpf/linker.c @@ -0,0 +1,1176 @@ +// SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) +/* + * BPF static linker + * + * Copyright (c) 2021 Facebook + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "libbpf.h" +#include "btf.h" +#include "libbpf_internal.h" +#include "strset.h" + +struct src_sec { + const char *sec_name; + /* positional (not necessarily ELF) index in an array of sections */ + int id; + /* positional (not necessarily ELF) index of a matching section in a final object file */ + int dst_id; + /* section data offset in a matching output section */ + int dst_off; + /* whether section is omitted from the final ELF file */ + bool skipped; + + /* ELF info */ + size_t sec_idx; + Elf_Scn *scn; + Elf64_Shdr *shdr; + Elf_Data *data; +}; + +struct src_obj { + const char *filename; + int fd; + Elf *elf; + /* Section header strings section index */ + size_t shstrs_sec_idx; + /* SYMTAB section index */ + size_t symtab_sec_idx; + + /* List of sections. Slot zero is unused. */ + struct src_sec *secs; + int sec_cnt; + + /* mapping of symbol indices from src to dst ELF */ + int *sym_map; +}; + +struct dst_sec { + char *sec_name; + /* positional (not necessarily ELF) index in an array of sections */ + int id; + + /* ELF info */ + size_t sec_idx; + Elf_Scn *scn; + Elf64_Shdr *shdr; + Elf_Data *data; + + /* final output section size */ + int sec_sz; + /* final output contents of the section */ + void *raw_data; + + /* corresponding STT_SECTION symbol index in SYMTAB */ + int sec_sym_idx; +}; + +struct bpf_linker { + char *filename; + int fd; + Elf *elf; + Elf64_Ehdr *elf_hdr; + + /* Output sections metadata */ + struct dst_sec *secs; + int sec_cnt; + + struct strset *strtab_strs; /* STRTAB unique strings */ + size_t strtab_sec_idx; /* STRTAB section index */ + size_t symtab_sec_idx; /* SYMTAB section index */ +}; + +#define pr_warn_elf(fmt, ...) \ +do { \ + libbpf_print(LIBBPF_WARN, "libbpf: " fmt ": %s\n", ##__VA_ARGS__, elf_errmsg(-1)); \ +} while (0) + +static int init_output_elf(struct bpf_linker *linker, const char *file); + +static int linker_load_obj_file(struct bpf_linker *linker, const char *filename, struct src_obj *obj); +static int linker_sanity_check_elf(struct src_obj *obj); +static int linker_append_sec_data(struct bpf_linker *linker, struct src_obj *obj); +static int linker_append_elf_syms(struct bpf_linker *linker, struct src_obj *obj); +static int linker_append_elf_relos(struct bpf_linker *linker, struct src_obj *obj); + +void bpf_linker__free(struct bpf_linker *linker) +{ + int i; + + if (!linker) + return; + + free(linker->filename); + + if (linker->elf) + elf_end(linker->elf); + + if (linker->fd >= 0) + close(linker->fd); + + strset__free(linker->strtab_strs); + + for (i = 1; i < linker->sec_cnt; i++) { + struct dst_sec *sec = &linker->secs[i]; + + free(sec->sec_name); + free(sec->raw_data); + } + free(linker->secs); + + free(linker); +} + +struct bpf_linker *bpf_linker__new(const char *filename, struct bpf_linker_opts *opts) +{ + struct bpf_linker *linker; + int err; + + if (!OPTS_VALID(opts, bpf_linker_opts)) + return NULL; + + if (elf_version(EV_CURRENT) == EV_NONE) { + pr_warn_elf("libelf initialization failed"); + return NULL; + } + + linker = calloc(1, sizeof(*linker)); + if (!linker) + return NULL; + + linker->fd = -1; + + err = init_output_elf(linker, filename); + if (err) + goto err_out; + + return linker; + +err_out: + bpf_linker__free(linker); + return NULL; +} + +static struct dst_sec *add_dst_sec(struct bpf_linker *linker, const char *sec_name) +{ + struct dst_sec *secs = linker->secs, *sec; + size_t new_cnt = linker->sec_cnt ? linker->sec_cnt + 1 : 2; + + secs = libbpf_reallocarray(secs, new_cnt, sizeof(*secs)); + if (!secs) + return NULL; + + /* zero out newly allocated memory */ + memset(secs + linker->sec_cnt, 0, (new_cnt - linker->sec_cnt) * sizeof(*secs)); + + linker->secs = secs; + linker->sec_cnt = new_cnt; + + sec = &linker->secs[new_cnt - 1]; + sec->id = new_cnt - 1; + sec->sec_name = strdup(sec_name); + if (!sec->sec_name) + return NULL; + + return sec; +} + +static Elf64_Sym *add_new_sym(struct bpf_linker *linker, size_t *sym_idx) +{ + struct dst_sec *symtab = &linker->secs[linker->symtab_sec_idx]; + Elf64_Sym *syms, *sym; + size_t sym_cnt = symtab->sec_sz / sizeof(*sym); + + syms = libbpf_reallocarray(symtab->raw_data, sym_cnt + 1, sizeof(*sym)); + if (!syms) + return NULL; + + sym = &syms[sym_cnt]; + memset(sym, 0, sizeof(*sym)); + + symtab->raw_data = syms; + symtab->sec_sz += sizeof(*sym); + symtab->shdr->sh_size += sizeof(*sym); + symtab->data->d_size += sizeof(*sym); + + if (sym_idx) + *sym_idx = sym_cnt; + + return sym; +} + +static int init_output_elf(struct bpf_linker *linker, const char *file) +{ + int err, str_off; + Elf64_Sym *init_sym; + struct dst_sec *sec; + + linker->filename = strdup(file); + if (!linker->filename) + return -ENOMEM; + + linker->fd = open(file, O_WRONLY | O_CREAT | O_TRUNC, 0644); + if (linker->fd < 0) { + err = -errno; + pr_warn("failed to create '%s': %d\n", file, err); + return err; + } + + linker->elf = elf_begin(linker->fd, ELF_C_WRITE, NULL); + if (!linker->elf) { + pr_warn_elf("failed to create ELF object"); + return -EINVAL; + } + + /* ELF header */ + linker->elf_hdr = elf64_newehdr(linker->elf); + if (!linker->elf_hdr){ + pr_warn_elf("failed to create ELF header"); + return -EINVAL; + } + + linker->elf_hdr->e_machine = EM_BPF; + linker->elf_hdr->e_type = ET_REL; +#if __BYTE_ORDER == __LITTLE_ENDIAN + linker->elf_hdr->e_ident[EI_DATA] = ELFDATA2LSB; +#elif __BYTE_ORDER == __BIG_ENDIAN + linker->elf_hdr->e_ident[EI_DATA] = ELFDATA2MSB; +#else +#error "Unknown __BYTE_ORDER" +#endif + + /* STRTAB */ + /* initialize strset with an empty string to conform to ELF */ + linker->strtab_strs = strset__new(INT_MAX, "", sizeof("")); + if (libbpf_get_error(linker->strtab_strs)) + return libbpf_get_error(linker->strtab_strs); + + sec = add_dst_sec(linker, ".strtab"); + if (!sec) + return -ENOMEM; + + sec->scn = elf_newscn(linker->elf); + if (!sec->scn) { + pr_warn_elf("failed to create STRTAB section"); + return -EINVAL; + } + + sec->shdr = elf64_getshdr(sec->scn); + if (!sec->shdr) + return -EINVAL; + + sec->data = elf_newdata(sec->scn); + if (!sec->data) { + pr_warn_elf("failed to create STRTAB data"); + return -EINVAL; + } + + str_off = strset__add_str(linker->strtab_strs, sec->sec_name); + if (str_off < 0) + return str_off; + + sec->sec_idx = elf_ndxscn(sec->scn); + linker->elf_hdr->e_shstrndx = sec->sec_idx; + linker->strtab_sec_idx = sec->sec_idx; + + sec->shdr->sh_name = str_off; + sec->shdr->sh_type = SHT_STRTAB; + sec->shdr->sh_flags = SHF_STRINGS; + sec->shdr->sh_offset = 0; + sec->shdr->sh_link = 0; + sec->shdr->sh_info = 0; + sec->shdr->sh_addralign = 1; + sec->shdr->sh_size = sec->sec_sz = 0; + sec->shdr->sh_entsize = 0; + + /* SYMTAB */ + sec = add_dst_sec(linker, ".symtab"); + if (!sec) + return -ENOMEM; + + sec->scn = elf_newscn(linker->elf); + if (!sec->scn) { + pr_warn_elf("failed to create SYMTAB section"); + return -EINVAL; + } + + sec->shdr = elf64_getshdr(sec->scn); + if (!sec->shdr) + return -EINVAL; + + sec->data = elf_newdata(sec->scn); + if (!sec->data) { + pr_warn_elf("failed to create SYMTAB data"); + return -EINVAL; + } + + str_off = strset__add_str(linker->strtab_strs, sec->sec_name); + if (str_off < 0) + return str_off; + + sec->sec_idx = elf_ndxscn(sec->scn); + linker->symtab_sec_idx = sec->sec_idx; + + sec->shdr->sh_name = str_off; + sec->shdr->sh_type = SHT_SYMTAB; + sec->shdr->sh_flags = 0; + sec->shdr->sh_offset = 0; + sec->shdr->sh_link = linker->strtab_sec_idx; + /* sh_info should be one greater than the index of the last local + * symbol (i.e., binding is STB_LOCAL). But why and who cares? + */ + sec->shdr->sh_info = 0; + sec->shdr->sh_addralign = 8; + sec->shdr->sh_entsize = sizeof(Elf64_Sym); + + /* add the special all-zero symbol */ + init_sym = add_new_sym(linker, NULL); + if (!init_sym) + return -EINVAL; + + init_sym->st_name = 0; + init_sym->st_info = 0; + init_sym->st_other = 0; + init_sym->st_shndx = SHN_UNDEF; + init_sym->st_value = 0; + init_sym->st_size = 0; + + return 0; +} + +int bpf_linker__add_file(struct bpf_linker *linker, const char *filename) +{ + struct src_obj obj = {}; + int err = 0; + + if (!linker->elf) + return -EINVAL; + + err = err ?: linker_load_obj_file(linker, filename, &obj); + err = err ?: linker_append_sec_data(linker, &obj); + err = err ?: linker_append_elf_syms(linker, &obj); + err = err ?: linker_append_elf_relos(linker, &obj); + + /* free up src_obj resources */ + free(obj.secs); + free(obj.sym_map); + if (obj.elf) + elf_end(obj.elf); + if (obj.fd >= 0) + close(obj.fd); + + return err; +} + +static bool is_dwarf_sec_name(const char *name) +{ + /* approximation, but the actual list is too long */ + return strncmp(name, ".debug_", sizeof(".debug_") - 1) == 0; +} + +static bool is_ignored_sec(struct src_sec *sec) +{ + Elf64_Shdr *shdr = sec->shdr; + const char *name = sec->sec_name; + + /* no special handling of .strtab */ + if (shdr->sh_type == SHT_STRTAB) + return true; + + /* ignore .llvm_addrsig section as well */ + if (shdr->sh_type == SHT_LLVM_ADDRSIG) + return true; + + /* no subprograms will lead to an empty .text section, ignore it */ + if (shdr->sh_type == SHT_PROGBITS && shdr->sh_size == 0 && + strcmp(sec->sec_name, ".text") == 0) + return true; + + /* DWARF sections */ + if (is_dwarf_sec_name(sec->sec_name)) + return true; + + if (strncmp(name, ".rel", sizeof(".rel") - 1) == 0) { + name += sizeof(".rel") - 1; + /* DWARF section relocations */ + if (is_dwarf_sec_name(name)) + return true; + + /* .BTF and .BTF.ext don't need relocations */ + if (strcmp(name, BTF_ELF_SEC) == 0 || + strcmp(name, BTF_EXT_ELF_SEC) == 0) + return true; + } + + return false; +} + +static struct src_sec *add_src_sec(struct src_obj *obj, const char *sec_name) +{ + struct src_sec *secs = obj->secs, *sec; + size_t new_cnt = obj->sec_cnt ? obj->sec_cnt + 1 : 2; + + secs = libbpf_reallocarray(secs, new_cnt, sizeof(*secs)); + if (!secs) + return NULL; + + /* zero out newly allocated memory */ + memset(secs + obj->sec_cnt, 0, (new_cnt - obj->sec_cnt) * sizeof(*secs)); + + obj->secs = secs; + obj->sec_cnt = new_cnt; + + sec = &obj->secs[new_cnt - 1]; + sec->id = new_cnt - 1; + sec->sec_name = sec_name; + + return sec; +} + +static int linker_load_obj_file(struct bpf_linker *linker, const char *filename, struct src_obj *obj) +{ +#if __BYTE_ORDER == __LITTLE_ENDIAN + const int host_endianness = ELFDATA2LSB; +#elif __BYTE_ORDER == __BIG_ENDIAN + const int host_endianness = ELFDATA2MSB; +#else +#error "Unknown __BYTE_ORDER" +#endif + int err = 0; + Elf_Scn *scn; + Elf_Data *data; + Elf64_Ehdr *ehdr; + Elf64_Shdr *shdr; + struct src_sec *sec; + + pr_debug("linker: adding object file '%s'...\n", filename); + + obj->filename = filename; + + obj->fd = open(filename, O_RDONLY); + if (obj->fd < 0) { + err = -errno; + pr_warn("failed to open file '%s': %d\n", filename, err); + return err; + } + obj->elf = elf_begin(obj->fd, ELF_C_READ_MMAP, NULL); + if (!obj->elf) { + err = -errno; + pr_warn_elf("failed to parse ELF file '%s'", filename); + return err; + } + + /* Sanity check ELF file high-level properties */ + ehdr = elf64_getehdr(obj->elf); + if (!ehdr) { + err = -errno; + pr_warn_elf("failed to get ELF header for %s", filename); + return err; + } + if (ehdr->e_ident[EI_DATA] != host_endianness) { + err = -EOPNOTSUPP; + pr_warn_elf("unsupported byte order of ELF file %s", filename); + return err; + } + if (ehdr->e_type != ET_REL + || ehdr->e_machine != EM_BPF + || ehdr->e_ident[EI_CLASS] != ELFCLASS64) { + err = -EOPNOTSUPP; + pr_warn_elf("unsupported kind of ELF file %s", filename); + return err; + } + + if (elf_getshdrstrndx(obj->elf, &obj->shstrs_sec_idx)) { + err = -errno; + pr_warn_elf("failed to get SHSTRTAB section index for %s", filename); + return err; + } + + scn = NULL; + while ((scn = elf_nextscn(obj->elf, scn)) != NULL) { + size_t sec_idx = elf_ndxscn(scn); + const char *sec_name; + + shdr = elf64_getshdr(scn); + if (!shdr) { + err = -errno; + pr_warn_elf("failed to get section #%zu header for %s", + sec_idx, filename); + return err; + } + + sec_name = elf_strptr(obj->elf, obj->shstrs_sec_idx, shdr->sh_name); + if (!sec_name) { + err = -errno; + pr_warn_elf("failed to get section #%zu name for %s", + sec_idx, filename); + return err; + } + + data = elf_getdata(scn, 0); + if (!data) { + err = -errno; + pr_warn_elf("failed to get section #%zu (%s) data from %s", + sec_idx, sec_name, filename); + return err; + } + + sec = add_src_sec(obj, sec_name); + if (!sec) + return -ENOMEM; + + sec->scn = scn; + sec->shdr = shdr; + sec->data = data; + sec->sec_idx = elf_ndxscn(scn); + + if (is_ignored_sec(sec)) { + sec->skipped = true; + continue; + } + + switch (shdr->sh_type) { + case SHT_SYMTAB: + if (obj->symtab_sec_idx) { + err = -EOPNOTSUPP; + pr_warn("multiple SYMTAB sections found, not supported\n"); + return err; + } + obj->symtab_sec_idx = sec_idx; + break; + case SHT_STRTAB: + /* we'll construct our own string table */ + break; + case SHT_PROGBITS: + if (strcmp(sec_name, BTF_ELF_SEC) == 0) { + sec->skipped = true; + continue; + } + if (strcmp(sec_name, BTF_EXT_ELF_SEC) == 0) { + sec->skipped = true; + continue; + } + + /* data & code */ + break; + case SHT_NOBITS: + /* BSS */ + break; + case SHT_REL: + /* relocations */ + break; + default: + pr_warn("unrecognized section #%zu (%s) in %s\n", + sec_idx, sec_name, filename); + err = -EINVAL; + return err; + } + } + + err = err ?: linker_sanity_check_elf(obj); + + return err; +} + +static bool is_pow_of_2(size_t x) +{ + return x && (x & (x - 1)) == 0; +} + +static int linker_sanity_check_elf(struct src_obj *obj) +{ + struct src_sec *sec, *link_sec; + int i, j, n; + + if (!obj->symtab_sec_idx) { + pr_warn("ELF is missing SYMTAB section in %s\n", obj->filename); + return -EINVAL; + } + if (!obj->shstrs_sec_idx) { + pr_warn("ELF is missing section headers STRTAB section in %s\n", obj->filename); + return -EINVAL; + } + + for (i = 1; i < obj->sec_cnt; i++) { + sec = &obj->secs[i]; + + if (sec->sec_name[0] == '\0') { + pr_warn("ELF section #%zu has empty name in %s\n", sec->sec_idx, obj->filename); + return -EINVAL; + } + + if (sec->shdr->sh_addralign && !is_pow_of_2(sec->shdr->sh_addralign)) + return -EINVAL; + if (sec->shdr->sh_addralign != sec->data->d_align) + return -EINVAL; + + if (sec->shdr->sh_size != sec->data->d_size) + return -EINVAL; + + switch (sec->shdr->sh_type) { + case SHT_SYMTAB: { + Elf64_Sym *sym; + + if (sec->shdr->sh_entsize != sizeof(Elf64_Sym)) + return -EINVAL; + if (sec->shdr->sh_size % sec->shdr->sh_entsize != 0) + return -EINVAL; + + if (!sec->shdr->sh_link || sec->shdr->sh_link >= obj->sec_cnt) { + pr_warn("ELF SYMTAB section #%zu points to missing STRTAB section #%zu in %s\n", + sec->sec_idx, (size_t)sec->shdr->sh_link, obj->filename); + return -EINVAL; + } + link_sec = &obj->secs[sec->shdr->sh_link]; + if (link_sec->shdr->sh_type != SHT_STRTAB) { + pr_warn("ELF SYMTAB section #%zu points to invalid STRTAB section #%zu in %s\n", + sec->sec_idx, (size_t)sec->shdr->sh_link, obj->filename); + return -EINVAL; + } + + n = sec->shdr->sh_size / sec->shdr->sh_entsize; + sym = sec->data->d_buf; + for (j = 0; j < n; j++, sym++) { + if (sym->st_shndx + && sym->st_shndx < SHN_LORESERVE + && sym->st_shndx >= obj->sec_cnt) { + pr_warn("ELF sym #%d in section #%zu points to missing section #%zu in %s\n", + j, sec->sec_idx, (size_t)sym->st_shndx, obj->filename); + return -EINVAL; + } + if (ELF64_ST_TYPE(sym->st_info) == STT_SECTION) { + if (sym->st_value != 0) + return -EINVAL; + } + } + break; + } + case SHT_STRTAB: + break; + case SHT_PROGBITS: + if (sec->shdr->sh_flags & SHF_EXECINSTR) { + if (sec->shdr->sh_size % sizeof(struct bpf_insn) != 0) + return -EINVAL; + } + break; + case SHT_NOBITS: + break; + case SHT_REL: { + Elf64_Rel *relo; + struct src_sec *sym_sec; + + if (sec->shdr->sh_entsize != sizeof(Elf64_Rel)) + return -EINVAL; + if (sec->shdr->sh_size % sec->shdr->sh_entsize != 0) + return -EINVAL; + + /* SHT_REL's sh_link should point to SYMTAB */ + if (sec->shdr->sh_link != obj->symtab_sec_idx) { + pr_warn("ELF relo section #%zu points to invalid SYMTAB section #%zu in %s\n", + sec->sec_idx, (size_t)sec->shdr->sh_link, obj->filename); + return -EINVAL; + } + + /* SHT_REL's sh_info points to relocated section */ + if (!sec->shdr->sh_info || sec->shdr->sh_info >= obj->sec_cnt) { + pr_warn("ELF relo section #%zu points to missing section #%zu in %s\n", + sec->sec_idx, (size_t)sec->shdr->sh_info, obj->filename); + return -EINVAL; + } + link_sec = &obj->secs[sec->shdr->sh_info]; + + /* .rel -> pattern is followed */ + if (strncmp(sec->sec_name, ".rel", sizeof(".rel") - 1) != 0 + || strcmp(sec->sec_name + sizeof(".rel") - 1, link_sec->sec_name) != 0) { + pr_warn("ELF relo section #%zu name has invalid name in %s\n", + sec->sec_idx, obj->filename); + return -EINVAL; + } + + /* don't further validate relocations for ignored sections */ + if (link_sec->skipped) + break; + + /* relocatable section is data or instructions */ + if (link_sec->shdr->sh_type != SHT_PROGBITS + && link_sec->shdr->sh_type != SHT_NOBITS) { + pr_warn("ELF relo section #%zu points to invalid section #%zu in %s\n", + sec->sec_idx, (size_t)sec->shdr->sh_info, obj->filename); + return -EINVAL; + } + + /* check sanity of each relocation */ + n = sec->shdr->sh_size / sec->shdr->sh_entsize; + relo = sec->data->d_buf; + sym_sec = &obj->secs[obj->symtab_sec_idx]; + for (j = 0; j < n; j++, relo++) { + size_t sym_idx = ELF64_R_SYM(relo->r_info); + size_t sym_type = ELF64_R_TYPE(relo->r_info); + + if (sym_type != R_BPF_64_64 && sym_type != R_BPF_64_32) { + pr_warn("ELF relo #%d in section #%zu has unexpected type %zu in %s\n", + j, sec->sec_idx, sym_type, obj->filename); + return -EINVAL; + } + + if (!sym_idx || sym_idx * sizeof(Elf64_Sym) >= sym_sec->shdr->sh_size) { + pr_warn("ELF relo #%d in section #%zu points to invalid symbol #%zu in %s\n", + j, sec->sec_idx, sym_idx, obj->filename); + return -EINVAL; + } + + if (link_sec->shdr->sh_flags & SHF_EXECINSTR) { + if (relo->r_offset % sizeof(struct bpf_insn) != 0) { + pr_warn("ELF relo #%d in section #%zu points to missing symbol #%zu in %s\n", + j, sec->sec_idx, sym_idx, obj->filename); + return -EINVAL; + } + } + } + break; + } + case SHT_LLVM_ADDRSIG: + break; + default: + pr_warn("ELF section #%zu (%s) has unrecognized type %zu in %s\n", + sec->sec_idx, sec->sec_name, (size_t)sec->shdr->sh_type, obj->filename); + return -EINVAL; + } + } + + return 0; +} + +static int init_sec(struct bpf_linker *linker, struct dst_sec *dst_sec, struct src_sec *src_sec) +{ + Elf_Scn *scn; + Elf_Data *data; + Elf64_Shdr *shdr; + int name_off; + + dst_sec->sec_sz = 0; + dst_sec->sec_idx = 0; + + scn = elf_newscn(linker->elf); + if (!scn) + return -1; + data = elf_newdata(scn); + if (!data) + return -1; + shdr = elf64_getshdr(scn); + if (!shdr) + return -1; + + dst_sec->scn = scn; + dst_sec->shdr = shdr; + dst_sec->data = data; + dst_sec->sec_idx = elf_ndxscn(scn); + + name_off = strset__add_str(linker->strtab_strs, src_sec->sec_name); + if (name_off < 0) + return name_off; + + shdr->sh_name = name_off; + shdr->sh_type = src_sec->shdr->sh_type; + shdr->sh_flags = src_sec->shdr->sh_flags; + shdr->sh_size = 0; + /* sh_link and sh_info have different meaning for different types of + * sections, so we leave it up to the caller code to fill them in, if + * necessary + */ + shdr->sh_link = 0; + shdr->sh_info = 0; + shdr->sh_addralign = src_sec->shdr->sh_addralign; + shdr->sh_entsize = src_sec->shdr->sh_entsize; + + data->d_type = src_sec->data->d_type; + data->d_size = 0; + data->d_buf = NULL; + data->d_align = src_sec->data->d_align; + data->d_off = 0; + + return 0; +} + +static struct dst_sec *find_dst_sec_by_name(struct bpf_linker *linker, const char *sec_name) +{ + struct dst_sec *sec; + int i; + + for (i = 1; i < linker->sec_cnt; i++) { + sec = &linker->secs[i]; + + if (strcmp(sec->sec_name, sec_name) == 0) + return sec; + } + + return NULL; +} + +static bool secs_match(struct dst_sec *dst, struct src_sec *src) +{ + if (dst->shdr->sh_type != src->shdr->sh_type) { + pr_warn("sec %s types mismatch\n", dst->sec_name); + return false; + } + if (dst->shdr->sh_flags != src->shdr->sh_flags) { + pr_warn("sec %s flags mismatch\n", dst->sec_name); + return false; + } + if (dst->shdr->sh_entsize != src->shdr->sh_entsize) { + pr_warn("sec %s entsize mismatch\n", dst->sec_name); + return false; + } + + return true; +} + +static bool sec_content_is_same(struct dst_sec *dst_sec, struct src_sec *src_sec) +{ + if (dst_sec->sec_sz != src_sec->shdr->sh_size) + return false; + if (memcmp(dst_sec->raw_data, src_sec->data->d_buf, dst_sec->sec_sz) != 0) + return false; + return true; +} + +static int extend_sec(struct dst_sec *dst, struct src_sec *src) +{ + void *tmp; + size_t dst_align = dst->shdr->sh_addralign; + size_t src_align = src->shdr->sh_addralign; + size_t dst_align_sz, dst_final_sz; + + if (dst_align == 0) + dst_align = 1; + if (dst_align < src_align) + dst_align = src_align; + + dst_align_sz = (dst->sec_sz + dst_align - 1) / dst_align * dst_align; + + /* no need to re-align final size */ + dst_final_sz = dst_align_sz + src->shdr->sh_size; + + if (src->shdr->sh_type != SHT_NOBITS) { + tmp = realloc(dst->raw_data, dst_final_sz); + if (!tmp) + return -ENOMEM; + dst->raw_data = tmp; + + /* pad dst section, if it's alignment forced size increase */ + memset(dst->raw_data + dst->sec_sz, 0, dst_align_sz - dst->sec_sz); + /* now copy src data at a properly aligned offset */ + memcpy(dst->raw_data + dst_align_sz, src->data->d_buf, src->shdr->sh_size); + } + + dst->sec_sz = dst_final_sz; + dst->shdr->sh_size = dst_final_sz; + dst->data->d_size = dst_final_sz; + + dst->shdr->sh_addralign = dst_align; + dst->data->d_align = dst_align; + + src->dst_off = dst_align_sz; + + return 0; +} + +static bool is_data_sec(struct src_sec *sec) +{ + if (!sec || sec->skipped) + return false; + return sec->shdr->sh_type == SHT_PROGBITS || sec->shdr->sh_type == SHT_NOBITS; +} + +static bool is_relo_sec(struct src_sec *sec) +{ + if (!sec || sec->skipped) + return false; + return sec->shdr->sh_type == SHT_REL; +} + +static int linker_append_sec_data(struct bpf_linker *linker, struct src_obj *obj) +{ + int i, err; + + for (i = 1; i < obj->sec_cnt; i++) { + struct src_sec *src_sec; + struct dst_sec *dst_sec; + + src_sec = &obj->secs[i]; + if (!is_data_sec(src_sec)) + continue; + + dst_sec = find_dst_sec_by_name(linker, src_sec->sec_name); + if (!dst_sec) { + dst_sec = add_dst_sec(linker, src_sec->sec_name); + if (!dst_sec) + return -ENOMEM; + err = init_sec(linker, dst_sec, src_sec); + if (err) { + pr_warn("failed to init section '%s'\n", src_sec->sec_name); + return err; + } + } else { + if (!secs_match(dst_sec, src_sec)) { + pr_warn("ELF sections %s are incompatible\n", src_sec->sec_name); + return -1; + } + + /* "license" and "version" sections are deduped */ + if (strcmp(src_sec->sec_name, "license") == 0 + || strcmp(src_sec->sec_name, "version") == 0) { + if (!sec_content_is_same(dst_sec, src_sec)) { + pr_warn("non-identical contents of section '%s' are not supported\n", src_sec->sec_name); + return -EINVAL; + } + src_sec->skipped = true; + src_sec->dst_id = dst_sec->id; + continue; + } + } + + /* record mapped section index */ + src_sec->dst_id = dst_sec->id; + + err = extend_sec(dst_sec, src_sec); + if (err) + return err; + } + + return 0; +} + +static int linker_append_elf_syms(struct bpf_linker *linker, struct src_obj *obj) +{ + struct src_sec *symtab = &obj->secs[obj->symtab_sec_idx]; + Elf64_Sym *sym = symtab->data->d_buf, *dst_sym; + int i, n = symtab->shdr->sh_size / symtab->shdr->sh_entsize; + int str_sec_idx = symtab->shdr->sh_link; + + obj->sym_map = calloc(n + 1, sizeof(*obj->sym_map)); + if (!obj->sym_map) + return -ENOMEM; + + for (i = 0; i < n; i++, sym++) { + struct src_sec *src_sec = NULL; + struct dst_sec *dst_sec = NULL; + const char *sym_name; + size_t dst_sym_idx; + int name_off; + + /* we already have all-zero initial symbol */ + if (sym->st_name == 0 && sym->st_info == 0 && + sym->st_other == 0 && sym->st_shndx == SHN_UNDEF && + sym->st_value == 0 && sym->st_size ==0) + continue; + + sym_name = elf_strptr(obj->elf, str_sec_idx, sym->st_name); + if (!sym_name) { + pr_warn("can't fetch symbol name for symbol #%d in '%s'\n", i, obj->filename); + return -1; + } + + if (sym->st_shndx && sym->st_shndx < SHN_LORESERVE) { + src_sec = &obj->secs[sym->st_shndx]; + if (src_sec->skipped) + continue; + dst_sec = &linker->secs[src_sec->dst_id]; + + /* allow only one STT_SECTION symbol per section */ + if (ELF64_ST_TYPE(sym->st_info) == STT_SECTION && dst_sec->sec_sym_idx) { + obj->sym_map[i] = dst_sec->sec_sym_idx; + continue; + } + } + + name_off = strset__add_str(linker->strtab_strs, sym_name); + if (name_off < 0) + return name_off; + + dst_sym = add_new_sym(linker, &dst_sym_idx); + if (!dst_sym) + return -ENOMEM; + + dst_sym->st_name = name_off; + dst_sym->st_info = sym->st_info; + dst_sym->st_other = sym->st_other; + dst_sym->st_shndx = src_sec ? dst_sec->sec_idx : sym->st_shndx; + dst_sym->st_value = (src_sec ? src_sec->dst_off : 0) + sym->st_value; + dst_sym->st_size = sym->st_size; + + obj->sym_map[i] = dst_sym_idx; + + if (ELF64_ST_TYPE(sym->st_info) == STT_SECTION && dst_sym) { + dst_sec->sec_sym_idx = dst_sym_idx; + dst_sym->st_value = 0; + } + + } + + return 0; +} + +static int linker_append_elf_relos(struct bpf_linker *linker, struct src_obj *obj) +{ + struct src_sec *src_symtab = &obj->secs[obj->symtab_sec_idx]; + struct dst_sec *dst_symtab = &linker->secs[linker->symtab_sec_idx]; + int i, err; + + for (i = 1; i < obj->sec_cnt; i++) { + struct src_sec *src_sec, *src_linked_sec; + struct dst_sec *dst_sec, *dst_linked_sec; + Elf64_Rel *src_rel, *dst_rel; + int j, n; + + src_sec = &obj->secs[i]; + if (!is_relo_sec(src_sec)) + continue; + + /* shdr->sh_info points to relocatable section */ + src_linked_sec = &obj->secs[src_sec->shdr->sh_info]; + if (src_linked_sec->skipped) + continue; + + dst_sec = find_dst_sec_by_name(linker, src_sec->sec_name); + if (!dst_sec) { + dst_sec = add_dst_sec(linker, src_sec->sec_name); + if (!dst_sec) + return -ENOMEM; + err = init_sec(linker, dst_sec, src_sec); + if (err) { + pr_warn("failed to init section '%s'\n", src_sec->sec_name); + return err; + } + } else if (!secs_match(dst_sec, src_sec)) { + pr_warn("Secs %s are not compatible\n", src_sec->sec_name); + return -1; + } + + /* shdr->sh_link points to SYMTAB */ + dst_sec->shdr->sh_link = linker->symtab_sec_idx; + + /* shdr->sh_info points to relocated section */ + dst_linked_sec = &linker->secs[src_linked_sec->dst_id]; + dst_sec->shdr->sh_info = dst_linked_sec->sec_idx; + + src_sec->dst_id = dst_sec->id; + err = extend_sec(dst_sec, src_sec); + if (err) + return err; + + src_rel = src_sec->data->d_buf; + dst_rel = dst_sec->raw_data + src_sec->dst_off; + n = src_sec->shdr->sh_size / src_sec->shdr->sh_entsize; + for (j = 0; j < n; j++, src_rel++, dst_rel++) { + size_t src_sym_idx = ELF64_R_SYM(src_rel->r_info); + size_t sym_type = ELF64_R_TYPE(src_rel->r_info); + Elf64_Sym *src_sym, *dst_sym; + size_t dst_sym_idx; + + src_sym_idx = ELF64_R_SYM(src_rel->r_info); + src_sym = src_symtab->data->d_buf + sizeof(*src_sym) * src_sym_idx; + + dst_sym_idx = obj->sym_map[src_sym_idx]; + dst_sym = dst_symtab->raw_data + sizeof(*dst_sym) * dst_sym_idx; + dst_rel->r_offset += src_linked_sec->dst_off; + sym_type = ELF64_R_TYPE(src_rel->r_info); + dst_rel->r_info = ELF64_R_INFO(dst_sym_idx, sym_type); + + if (ELF64_ST_TYPE(src_sym->st_info) == STT_SECTION) { + struct src_sec *sec = &obj->secs[src_sym->st_shndx]; + struct bpf_insn *insn; + + if (src_linked_sec->shdr->sh_flags & SHF_EXECINSTR) { + /* calls to the very first static function inside + * .text section at offset 0 will + * reference section symbol, not the + * function symbol. Fix that up, + * otherwise it won't be possible to + * relocate calls to two different + * static functions with the same name + * (rom two different object files) + */ + insn = dst_linked_sec->raw_data + dst_rel->r_offset; + if (insn->code == (BPF_JMP | BPF_CALL)) + insn->imm += sec->dst_off / sizeof(struct bpf_insn); + else + insn->imm += sec->dst_off; + } else { + pr_warn("relocation against STT_SECTION in non-exec section is not supported!\n"); + return -EINVAL; + } + } + + } + } + + return 0; +} + +int bpf_linker__finalize(struct bpf_linker *linker) +{ + struct dst_sec *sec; + size_t strs_sz; + const void *strs; + int err, i; + + if (!linker->elf) + return -EINVAL; + + /* Finalize strings */ + strs_sz = strset__data_size(linker->strtab_strs); + strs = strset__data(linker->strtab_strs); + + sec = &linker->secs[linker->strtab_sec_idx]; + sec->data->d_align = 1; + sec->data->d_off = 0LL; + sec->data->d_buf = (void *)strs; + sec->data->d_type = ELF_T_BYTE; + sec->data->d_size = strs_sz; + sec->shdr->sh_size = strs_sz; + + for (i = 1; i < linker->sec_cnt; i++) { + sec = &linker->secs[i]; + + /* STRTAB is handled specially above */ + if (sec->sec_idx == linker->strtab_sec_idx) + continue; + + sec->data->d_buf = sec->raw_data; + } + + /* Finalize ELF layout */ + if (elf_update(linker->elf, ELF_C_NULL) < 0) { + err = -errno; + pr_warn_elf("failed to finalize ELF layout"); + return err; + } + + /* Write out final ELF contents */ + if (elf_update(linker->elf, ELF_C_WRITE) < 0) { + err = -errno; + pr_warn_elf("failed to write ELF contents"); + return err; + } + + elf_end(linker->elf); + close(linker->fd); + + linker->elf = NULL; + linker->fd = -1; + + return 0; +} From patchwork Thu Mar 18 06:31:10 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 404617 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 1871CC43381 for ; Thu, 18 Mar 2021 06:32:52 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id D943D64F2A for ; Thu, 18 Mar 2021 06:32:51 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229913AbhCRGcX convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:32:23 -0400 Received: from mx0b-00082601.pphosted.com ([67.231.153.30]:52248 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S229584AbhCRGbq (ORCPT ); Thu, 18 Mar 2021 02:31:46 -0400 Received: from pps.filterd (m0089730.ppops.net [127.0.0.1]) by m0089730.ppops.net (8.16.0.43/8.16.0.43) with SMTP id 12I6QRcm032153 for ; Wed, 17 Mar 2021 23:31:46 -0700 Received: from mail.thefacebook.com ([163.114.132.120]) by m0089730.ppops.net with ESMTP id 37bs1h29fv-8 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:45 -0700 Received: from intmgw001.05.ash7.facebook.com (2620:10d:c085:108::8) by mail.thefacebook.com (2620:10d:c085:21d::7) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:34 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id F0B422ED24FA; Wed, 17 Mar 2021 23:31:31 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , Subject: [PATCH v3 bpf-next 07/12] libbpf: add BPF static linker BTF and BTF.ext support Date: Wed, 17 Mar 2021 23:31:10 -0700 Message-ID: <20210318063115.49613-8-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 lowpriorityscore=0 mlxscore=0 mlxlogscore=999 bulkscore=0 priorityscore=1501 clxscore=1015 phishscore=0 impostorscore=0 adultscore=0 suspectscore=0 spamscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Add .BTF and .BTF.ext static linking logic. When multiple BPF object files are linked together, their respective .BTF and .BTF.ext sections are merged together. BTF types are not just concatenated, but also deduplicated. .BTF.ext data is grouped by type (func info, line info, core_relos) and target section names, and then all the records are concatenated together, preserving their relative order. All the BTF type ID references and string offsets are updated as necessary, to take into account possibly deduplicated strings and types. BTF DATASEC types are handled specially. Their respective var_secinfos are accumulated separately in special per-section data and then final DATASEC types are emitted at the very end during bpf_linker__finalize() operation, just before emitting final ELF output file. BTF data can also provide "section annotations" for some extern variables. Such concept is missing in ELF, but BTF will have DATASEC types for such special extern datasections (e.g., .kconfig, .ksyms). Such sections are called "ephemeral" internally. Internally linker will keep metadata for each such section, collecting variables information, but those sections won't be emitted into the final ELF file. Also, given LLVM/Clang during compilation emits BTF DATASECS that are incomplete, missing section size and variable offsets for static variables, BPF static linker will initially fix up such DATASECs, using ELF symbols data. The final DATASECs will preserve section sizes and all variable offsets. This is handled correctly by libbpf already, so won't cause any new issues. On the other hand, it's actually a nice property to have a complete BTF data without runtime adjustments done during bpf_object__open() by libbpf. In that sense, BPF static linker is also a BTF normalizer. Signed-off-by: Andrii Nakryiko --- tools/lib/bpf/linker.c | 769 ++++++++++++++++++++++++++++++++++++++++- 1 file changed, 767 insertions(+), 2 deletions(-) diff --git a/tools/lib/bpf/linker.c b/tools/lib/bpf/linker.c index 01b56939a835..b4fff912dce2 100644 --- a/tools/lib/bpf/linker.c +++ b/tools/lib/bpf/linker.c @@ -32,12 +32,17 @@ struct src_sec { int dst_off; /* whether section is omitted from the final ELF file */ bool skipped; + /* whether section is an ephemeral section, not mapped to an ELF section */ + bool ephemeral; /* ELF info */ size_t sec_idx; Elf_Scn *scn; Elf64_Shdr *shdr; Elf_Data *data; + + /* corresponding BTF DATASEC type ID */ + int sec_type_id; }; struct src_obj { @@ -49,12 +54,24 @@ struct src_obj { /* SYMTAB section index */ size_t symtab_sec_idx; - /* List of sections. Slot zero is unused. */ + struct btf *btf; + struct btf_ext *btf_ext; + + /* List of sections (including ephemeral). Slot zero is unused. */ struct src_sec *secs; int sec_cnt; /* mapping of symbol indices from src to dst ELF */ int *sym_map; + /* mapping from the src BTF type IDs to dst ones */ + int *btf_type_map; +}; + +/* single .BTF.ext data section */ +struct btf_ext_sec_data { + size_t rec_cnt; + __u32 rec_sz; + void *recs; }; struct dst_sec { @@ -75,6 +92,15 @@ struct dst_sec { /* corresponding STT_SECTION symbol index in SYMTAB */ int sec_sym_idx; + + /* section's DATASEC variable info, emitted on BTF finalization */ + int sec_var_cnt; + struct btf_var_secinfo *sec_vars; + + /* section's .BTF.ext data */ + struct btf_ext_sec_data func_info; + struct btf_ext_sec_data line_info; + struct btf_ext_sec_data core_relo_info; }; struct bpf_linker { @@ -90,6 +116,9 @@ struct bpf_linker { struct strset *strtab_strs; /* STRTAB unique strings */ size_t strtab_sec_idx; /* STRTAB section index */ size_t symtab_sec_idx; /* SYMTAB section index */ + + struct btf *btf; + struct btf_ext *btf_ext; }; #define pr_warn_elf(fmt, ...) \ @@ -101,9 +130,17 @@ static int init_output_elf(struct bpf_linker *linker, const char *file); static int linker_load_obj_file(struct bpf_linker *linker, const char *filename, struct src_obj *obj); static int linker_sanity_check_elf(struct src_obj *obj); +static int linker_sanity_check_btf(struct src_obj *obj); +static int linker_sanity_check_btf_ext(struct src_obj *obj); +static int linker_fixup_btf(struct src_obj *obj); static int linker_append_sec_data(struct bpf_linker *linker, struct src_obj *obj); static int linker_append_elf_syms(struct bpf_linker *linker, struct src_obj *obj); static int linker_append_elf_relos(struct bpf_linker *linker, struct src_obj *obj); +static int linker_append_btf(struct bpf_linker *linker, struct src_obj *obj); +static int linker_append_btf_ext(struct bpf_linker *linker, struct src_obj *obj); + +static int finalize_btf(struct bpf_linker *linker); +static int finalize_btf_ext(struct bpf_linker *linker); void bpf_linker__free(struct bpf_linker *linker) { @@ -122,11 +159,19 @@ void bpf_linker__free(struct bpf_linker *linker) strset__free(linker->strtab_strs); + btf__free(linker->btf); + btf_ext__free(linker->btf_ext); + for (i = 1; i < linker->sec_cnt; i++) { struct dst_sec *sec = &linker->secs[i]; free(sec->sec_name); free(sec->raw_data); + free(sec->sec_vars); + + free(sec->func_info.recs); + free(sec->line_info.recs); + free(sec->core_relo_info.recs); } free(linker->secs); @@ -335,6 +380,12 @@ static int init_output_elf(struct bpf_linker *linker, const char *file) sec->shdr->sh_addralign = 8; sec->shdr->sh_entsize = sizeof(Elf64_Sym); + /* .BTF */ + linker->btf = btf__new_empty(); + err = libbpf_get_error(linker->btf); + if (err) + return err; + /* add the special all-zero symbol */ init_sym = add_new_sym(linker, NULL); if (!init_sym) @@ -362,8 +413,13 @@ int bpf_linker__add_file(struct bpf_linker *linker, const char *filename) err = err ?: linker_append_sec_data(linker, &obj); err = err ?: linker_append_elf_syms(linker, &obj); err = err ?: linker_append_elf_relos(linker, &obj); + err = err ?: linker_append_btf(linker, &obj); + err = err ?: linker_append_btf_ext(linker, &obj); /* free up src_obj resources */ + free(obj.btf_type_map); + btf__free(obj.btf); + btf_ext__free(obj.btf_ext); free(obj.secs); free(obj.sym_map); if (obj.elf) @@ -555,10 +611,22 @@ static int linker_load_obj_file(struct bpf_linker *linker, const char *filename, break; case SHT_PROGBITS: if (strcmp(sec_name, BTF_ELF_SEC) == 0) { + obj->btf = btf__new(data->d_buf, shdr->sh_size); + err = libbpf_get_error(obj->btf); + if (err) { + pr_warn("failed to parse .BTF from %s: %d\n", filename, err); + return err; + } sec->skipped = true; continue; } if (strcmp(sec_name, BTF_EXT_ELF_SEC) == 0) { + obj->btf_ext = btf_ext__new(data->d_buf, shdr->sh_size); + err = libbpf_get_error(obj->btf_ext); + if (err) { + pr_warn("failed to parse .BTF.ext from '%s': %d\n", filename, err); + return err; + } sec->skipped = true; continue; } @@ -580,6 +648,9 @@ static int linker_load_obj_file(struct bpf_linker *linker, const char *filename, } err = err ?: linker_sanity_check_elf(obj); + err = err ?: linker_sanity_check_btf(obj); + err = err ?: linker_sanity_check_btf_ext(obj); + err = err ?: linker_fixup_btf(obj); return err; } @@ -753,6 +824,69 @@ static int linker_sanity_check_elf(struct src_obj *obj) return 0; } +static int check_btf_type_id(__u32 *type_id, void *ctx) +{ + struct btf *btf = ctx; + + if (*type_id > btf__get_nr_types(btf)) + return -EINVAL; + + return 0; +} + +static int check_btf_str_off(__u32 *str_off, void *ctx) +{ + struct btf *btf = ctx; + const char *s; + + s = btf__str_by_offset(btf, *str_off); + + if (!s) + return -EINVAL; + + return 0; +} + +static int linker_sanity_check_btf(struct src_obj *obj) +{ + struct btf_type *t; + int i, n, err = 0; + + if (!obj->btf) + return 0; + + n = btf__get_nr_types(obj->btf); + for (i = 1; i <= n; i++) { + t = btf_type_by_id(obj->btf, i); + + err = err ?: btf_type_visit_type_ids(t, check_btf_type_id, obj->btf); + err = err ?: btf_type_visit_str_offs(t, check_btf_str_off, obj->btf); + if (err) + return err; + } + + return 0; +} + +static int linker_sanity_check_btf_ext(struct src_obj *obj) +{ + int err = 0; + + if (!obj->btf_ext) + return 0; + + /* can't use .BTF.ext without .BTF */ + if (!obj->btf) + return -EINVAL; + + err = err ?: btf_ext_visit_type_ids(obj->btf_ext, check_btf_type_id, obj->btf); + err = err ?: btf_ext_visit_str_offs(obj->btf_ext, check_btf_str_off, obj->btf); + if (err) + return err; + + return 0; +} + static int init_sec(struct bpf_linker *linker, struct dst_sec *dst_sec, struct src_sec *src_sec) { Elf_Scn *scn; @@ -763,6 +897,10 @@ static int init_sec(struct bpf_linker *linker, struct dst_sec *dst_sec, struct s dst_sec->sec_sz = 0; dst_sec->sec_idx = 0; + /* ephemeral sections are just thin section shells lacking most parts */ + if (src_sec->ephemeral) + return 0; + scn = elf_newscn(linker->elf); if (!scn) return -1; @@ -891,12 +1029,15 @@ static bool is_data_sec(struct src_sec *sec) { if (!sec || sec->skipped) return false; + /* ephemeral sections are data sections, e.g., .kconfig, .ksyms */ + if (sec->ephemeral) + return true; return sec->shdr->sh_type == SHT_PROGBITS || sec->shdr->sh_type == SHT_NOBITS; } static bool is_relo_sec(struct src_sec *sec) { - if (!sec || sec->skipped) + if (!sec || sec->skipped || sec->ephemeral) return false; return sec->shdr->sh_type == SHT_REL; } @@ -945,6 +1086,9 @@ static int linker_append_sec_data(struct bpf_linker *linker, struct src_obj *obj /* record mapped section index */ src_sec->dst_id = dst_sec->id; + if (src_sec->ephemeral) + continue; + err = extend_sec(dst_sec, src_sec); if (err) return err; @@ -1120,6 +1264,339 @@ static int linker_append_elf_relos(struct bpf_linker *linker, struct src_obj *ob return 0; } +static struct src_sec *find_src_sec_by_name(struct src_obj *obj, const char *sec_name) +{ + struct src_sec *sec; + int i; + + for (i = 1; i < obj->sec_cnt; i++) { + sec = &obj->secs[i]; + + if (strcmp(sec->sec_name, sec_name) == 0) + return sec; + } + + return NULL; +} + +static Elf64_Sym *find_sym_by_name(struct src_obj *obj, size_t sec_idx, + int sym_type, const char *sym_name) +{ + struct src_sec *symtab = &obj->secs[obj->symtab_sec_idx]; + Elf64_Sym *sym = symtab->data->d_buf; + int i, n = symtab->shdr->sh_size / symtab->shdr->sh_entsize; + int str_sec_idx = symtab->shdr->sh_link; + const char *name; + + for (i = 0; i < n; i++, sym++) { + if (sym->st_shndx != sec_idx) + continue; + if (ELF64_ST_TYPE(sym->st_info) != sym_type) + continue; + + name = elf_strptr(obj->elf, str_sec_idx, sym->st_name); + if (!name) + return NULL; + + if (strcmp(sym_name, name) != 0) + continue; + + return sym; + } + + return NULL; +} + +static int linker_fixup_btf(struct src_obj *obj) +{ + const char *sec_name; + struct src_sec *sec; + int i, j, n, m; + + n = btf__get_nr_types(obj->btf); + for (i = 1; i <= n; i++) { + struct btf_var_secinfo *vi; + struct btf_type *t; + + t = btf_type_by_id(obj->btf, i); + if (btf_kind(t) != BTF_KIND_DATASEC) + continue; + + sec_name = btf__str_by_offset(obj->btf, t->name_off); + sec = find_src_sec_by_name(obj, sec_name); + if (sec) { + /* record actual section size, unless ephemeral */ + if (sec->shdr) + t->size = sec->shdr->sh_size; + } else { + /* BTF can have some sections that are not represented + * in ELF, e.g., .kconfig and .ksyms, which are used + * for special extern variables. Here we'll + * pre-create "section shells" for them to be able to + * keep track of extra per-section metadata later + * (e.g., BTF variables). + */ + sec = add_src_sec(obj, sec_name); + if (!sec) + return -ENOMEM; + + sec->ephemeral = true; + sec->sec_idx = 0; /* will match UNDEF shndx in ELF */ + } + + /* remember ELF section and its BTF type ID match */ + sec->sec_type_id = i; + + /* fix up variable offsets */ + vi = btf_var_secinfos(t); + for (j = 0, m = btf_vlen(t); j < m; j++, vi++) { + const struct btf_type *vt = btf__type_by_id(obj->btf, vi->type); + const char *var_name = btf__str_by_offset(obj->btf, vt->name_off); + int var_linkage = btf_var(vt)->linkage; + Elf64_Sym *sym; + + /* no need to patch up static or extern vars */ + if (var_linkage != BTF_VAR_GLOBAL_ALLOCATED) + continue; + + sym = find_sym_by_name(obj, sec->sec_idx, STT_OBJECT, var_name); + if (!sym) { + pr_warn("failed to find symbol for variable '%s' in section '%s'\n", var_name, sec_name); + return -ENOENT; + } + + vi->offset = sym->st_value; + } + } + + return 0; +} + +static int remap_type_id(__u32 *type_id, void *ctx) +{ + int *id_map = ctx; + + *type_id = id_map[*type_id]; + + return 0; +} + +static int linker_append_btf(struct bpf_linker *linker, struct src_obj *obj) +{ + const struct btf_type *t; + int i, j, n, start_id, id; + + if (!obj->btf) + return 0; + + start_id = btf__get_nr_types(linker->btf) + 1; + n = btf__get_nr_types(obj->btf); + + obj->btf_type_map = calloc(n + 1, sizeof(int)); + if (!obj->btf_type_map) + return -ENOMEM; + + for (i = 1; i <= n; i++) { + t = btf__type_by_id(obj->btf, i); + + /* DATASECs are handled specially below */ + if (btf_kind(t) == BTF_KIND_DATASEC) + continue; + + id = btf__add_type(linker->btf, obj->btf, t); + if (id < 0) { + pr_warn("failed to append BTF type #%d from file '%s'\n", i, obj->filename); + return id; + } + + obj->btf_type_map[i] = id; + } + + /* remap all the types except DATASECs */ + n = btf__get_nr_types(linker->btf); + for (i = start_id; i <= n; i++) { + struct btf_type *dst_t = btf_type_by_id(linker->btf, i); + + if (btf_type_visit_type_ids(dst_t, remap_type_id, obj->btf_type_map)) + return -EINVAL; + } + + /* append DATASEC info */ + for (i = 1; i < obj->sec_cnt; i++) { + struct src_sec *src_sec; + struct dst_sec *dst_sec; + const struct btf_var_secinfo *src_var; + struct btf_var_secinfo *dst_var; + + src_sec = &obj->secs[i]; + if (!src_sec->sec_type_id || src_sec->skipped) + continue; + dst_sec = &linker->secs[src_sec->dst_id]; + + t = btf__type_by_id(obj->btf, src_sec->sec_type_id); + src_var = btf_var_secinfos(t); + n = btf_vlen(t); + for (j = 0; j < n; j++, src_var++) { + void *sec_vars = dst_sec->sec_vars; + + sec_vars = libbpf_reallocarray(sec_vars, + dst_sec->sec_var_cnt + 1, + sizeof(*dst_sec->sec_vars)); + if (!sec_vars) + return -ENOMEM; + + dst_sec->sec_vars = sec_vars; + dst_sec->sec_var_cnt++; + + dst_var = &dst_sec->sec_vars[dst_sec->sec_var_cnt - 1]; + dst_var->type = obj->btf_type_map[src_var->type]; + dst_var->size = src_var->size; + dst_var->offset = src_sec->dst_off + src_var->offset; + } + } + + return 0; +} + +static void *add_btf_ext_rec(struct btf_ext_sec_data *ext_data, const void *src_rec) +{ + void *tmp; + + tmp = libbpf_reallocarray(ext_data->recs, ext_data->rec_cnt + 1, ext_data->rec_sz); + if (!tmp) + return NULL; + ext_data->recs = tmp; + + tmp += ext_data->rec_cnt * ext_data->rec_sz; + memcpy(tmp, src_rec, ext_data->rec_sz); + + ext_data->rec_cnt++; + + return tmp; +} + +static int linker_append_btf_ext(struct bpf_linker *linker, struct src_obj *obj) +{ + const struct btf_ext_info_sec *ext_sec; + const char *sec_name, *s; + struct src_sec *src_sec; + struct dst_sec *dst_sec; + int rec_sz, str_off, i; + + if (!obj->btf_ext) + return 0; + + rec_sz = obj->btf_ext->func_info.rec_size; + for_each_btf_ext_sec(&obj->btf_ext->func_info, ext_sec) { + struct bpf_func_info_min *src_rec, *dst_rec; + + sec_name = btf__name_by_offset(obj->btf, ext_sec->sec_name_off); + src_sec = find_src_sec_by_name(obj, sec_name); + if (!src_sec) { + pr_warn("can't find section '%s' referenced from .BTF.ext\n", sec_name); + return -EINVAL; + } + dst_sec = &linker->secs[src_sec->dst_id]; + + if (dst_sec->func_info.rec_sz == 0) + dst_sec->func_info.rec_sz = rec_sz; + if (dst_sec->func_info.rec_sz != rec_sz) { + pr_warn("incompatible .BTF.ext record sizes for section '%s'\n", sec_name); + return -EINVAL; + } + + for_each_btf_ext_rec(&obj->btf_ext->func_info, ext_sec, i, src_rec) { + dst_rec = add_btf_ext_rec(&dst_sec->func_info, src_rec); + if (!dst_rec) + return -ENOMEM; + + dst_rec->insn_off += src_sec->dst_off; + dst_rec->type_id = obj->btf_type_map[dst_rec->type_id]; + } + } + + rec_sz = obj->btf_ext->line_info.rec_size; + for_each_btf_ext_sec(&obj->btf_ext->line_info, ext_sec) { + struct bpf_line_info_min *src_rec, *dst_rec; + + sec_name = btf__name_by_offset(obj->btf, ext_sec->sec_name_off); + src_sec = find_src_sec_by_name(obj, sec_name); + if (!src_sec) { + pr_warn("can't find section '%s' referenced from .BTF.ext\n", sec_name); + return -EINVAL; + } + dst_sec = &linker->secs[src_sec->dst_id]; + + if (dst_sec->line_info.rec_sz == 0) + dst_sec->line_info.rec_sz = rec_sz; + if (dst_sec->line_info.rec_sz != rec_sz) { + pr_warn("incompatible .BTF.ext record sizes for section '%s'\n", sec_name); + return -EINVAL; + } + + for_each_btf_ext_rec(&obj->btf_ext->line_info, ext_sec, i, src_rec) { + dst_rec = add_btf_ext_rec(&dst_sec->line_info, src_rec); + if (!dst_rec) + return -ENOMEM; + + dst_rec->insn_off += src_sec->dst_off; + + s = btf__str_by_offset(obj->btf, src_rec->file_name_off); + str_off = btf__add_str(linker->btf, s); + if (str_off < 0) + return -ENOMEM; + dst_rec->file_name_off = str_off; + + s = btf__str_by_offset(obj->btf, src_rec->line_off); + str_off = btf__add_str(linker->btf, s); + if (str_off < 0) + return -ENOMEM; + dst_rec->line_off = str_off; + + /* dst_rec->line_col is fine */ + } + } + + rec_sz = obj->btf_ext->core_relo_info.rec_size; + for_each_btf_ext_sec(&obj->btf_ext->core_relo_info, ext_sec) { + struct bpf_core_relo *src_rec, *dst_rec; + + sec_name = btf__name_by_offset(obj->btf, ext_sec->sec_name_off); + src_sec = find_src_sec_by_name(obj, sec_name); + if (!src_sec) { + pr_warn("can't find section '%s' referenced from .BTF.ext\n", sec_name); + return -EINVAL; + } + dst_sec = &linker->secs[src_sec->dst_id]; + + if (dst_sec->core_relo_info.rec_sz == 0) + dst_sec->core_relo_info.rec_sz = rec_sz; + if (dst_sec->core_relo_info.rec_sz != rec_sz) { + pr_warn("incompatible .BTF.ext record sizes for section '%s'\n", sec_name); + return -EINVAL; + } + + for_each_btf_ext_rec(&obj->btf_ext->core_relo_info, ext_sec, i, src_rec) { + dst_rec = add_btf_ext_rec(&dst_sec->core_relo_info, src_rec); + if (!dst_rec) + return -ENOMEM; + + dst_rec->insn_off += src_sec->dst_off; + dst_rec->type_id = obj->btf_type_map[dst_rec->type_id]; + + s = btf__str_by_offset(obj->btf, src_rec->access_str_off); + str_off = btf__add_str(linker->btf, s); + if (str_off < 0) + return -ENOMEM; + dst_rec->access_str_off = str_off; + + /* dst_rec->kind is fine */ + } + } + + return 0; +} + int bpf_linker__finalize(struct bpf_linker *linker) { struct dst_sec *sec; @@ -1130,6 +1607,10 @@ int bpf_linker__finalize(struct bpf_linker *linker) if (!linker->elf) return -EINVAL; + err = finalize_btf(linker); + if (err) + return err; + /* Finalize strings */ strs_sz = strset__data_size(linker->strtab_strs); strs = strset__data(linker->strtab_strs); @@ -1149,6 +1630,10 @@ int bpf_linker__finalize(struct bpf_linker *linker) if (sec->sec_idx == linker->strtab_sec_idx) continue; + /* special ephemeral sections (.ksyms, .kconfig, etc) */ + if (!sec->scn) + continue; + sec->data->d_buf = sec->raw_data; } @@ -1174,3 +1659,283 @@ int bpf_linker__finalize(struct bpf_linker *linker) return 0; } + +static int emit_elf_data_sec(struct bpf_linker *linker, const char *sec_name, + size_t align, const void *raw_data, size_t raw_sz) +{ + Elf_Scn *scn; + Elf_Data *data; + Elf64_Shdr *shdr; + int name_off; + + name_off = strset__add_str(linker->strtab_strs, sec_name); + if (name_off < 0) + return name_off; + + scn = elf_newscn(linker->elf); + if (!scn) + return -ENOMEM; + data = elf_newdata(scn); + if (!data) + return -ENOMEM; + shdr = elf64_getshdr(scn); + if (!shdr) + return -EINVAL; + + shdr->sh_name = name_off; + shdr->sh_type = SHT_PROGBITS; + shdr->sh_flags = 0; + shdr->sh_size = raw_sz; + shdr->sh_link = 0; + shdr->sh_info = 0; + shdr->sh_addralign = align; + shdr->sh_entsize = 0; + + data->d_type = ELF_T_BYTE; + data->d_size = raw_sz; + data->d_buf = (void *)raw_data; + data->d_align = align; + data->d_off = 0; + + return 0; +} + +static int finalize_btf(struct bpf_linker *linker) +{ + struct btf *btf = linker->btf; + const void *raw_data; + int i, j, id, err; + __u32 raw_sz; + + /* bail out if no BTF data was produced */ + if (btf__get_nr_types(linker->btf) == 0) + return 0; + + for (i = 1; i < linker->sec_cnt; i++) { + struct dst_sec *sec = &linker->secs[i]; + + if (!sec->sec_var_cnt) + continue; + + id = btf__add_datasec(btf, sec->sec_name, sec->sec_sz); + if (id < 0) { + pr_warn("failed to add consolidated BTF type for datasec '%s': %d\n", + sec->sec_name, id); + return id; + } + + for (j = 0; j < sec->sec_var_cnt; j++) { + struct btf_var_secinfo *vi = &sec->sec_vars[j]; + + if (btf__add_datasec_var_info(btf, vi->type, vi->offset, vi->size)) + return -EINVAL; + } + } + + err = finalize_btf_ext(linker); + if (err) { + pr_warn(".BTF.ext generation failed: %d\n", err); + return err; + } + + err = btf__dedup(linker->btf, linker->btf_ext, NULL); + if (err) { + pr_warn("BTF dedup failed: %d\n", err); + return err; + } + + /* Emit .BTF section */ + raw_data = btf__get_raw_data(linker->btf, &raw_sz); + if (!raw_data) + return -ENOMEM; + + err = emit_elf_data_sec(linker, BTF_ELF_SEC, 8, raw_data, raw_sz); + if (err) { + pr_warn("failed to write out .BTF ELF section: %d\n", err); + return err; + } + + /* Emit .BTF.ext section */ + if (linker->btf_ext) { + raw_data = btf_ext__get_raw_data(linker->btf_ext, &raw_sz); + if (!raw_data) + return -ENOMEM; + + err = emit_elf_data_sec(linker, BTF_EXT_ELF_SEC, 8, raw_data, raw_sz); + if (err) { + pr_warn("failed to write out .BTF.ext ELF section: %d\n", err); + return err; + } + } + + return 0; +} + +static int emit_btf_ext_data(struct bpf_linker *linker, void *output, + const char *sec_name, struct btf_ext_sec_data *sec_data) +{ + struct btf_ext_info_sec *sec_info; + void *cur = output; + int str_off; + size_t sz; + + if (!sec_data->rec_cnt) + return 0; + + str_off = btf__add_str(linker->btf, sec_name); + if (str_off < 0) + return -ENOMEM; + + sec_info = cur; + sec_info->sec_name_off = str_off; + sec_info->num_info = sec_data->rec_cnt; + cur += sizeof(struct btf_ext_info_sec); + + sz = sec_data->rec_cnt * sec_data->rec_sz; + memcpy(cur, sec_data->recs, sz); + cur += sz; + + return cur - output; +} + +static int finalize_btf_ext(struct bpf_linker *linker) +{ + size_t funcs_sz = 0, lines_sz = 0, core_relos_sz = 0, total_sz = 0; + size_t func_rec_sz = 0, line_rec_sz = 0, core_relo_rec_sz = 0; + struct btf_ext_header *hdr; + void *data, *cur; + int i, err, sz; + + /* validate that all sections have the same .BTF.ext record sizes + * and calculate total data size for each type of data (func info, + * line info, core relos) + */ + for (i = 1; i < linker->sec_cnt; i++) { + struct dst_sec *sec = &linker->secs[i]; + + if (sec->func_info.rec_cnt) { + if (func_rec_sz == 0) + func_rec_sz = sec->func_info.rec_sz; + if (func_rec_sz != sec->func_info.rec_sz) { + pr_warn("mismatch in func_info record size %zu != %u\n", + func_rec_sz, sec->func_info.rec_sz); + return -EINVAL; + } + + funcs_sz += sizeof(struct btf_ext_info_sec) + func_rec_sz * sec->func_info.rec_cnt; + } + if (sec->line_info.rec_cnt) { + if (line_rec_sz == 0) + line_rec_sz = sec->line_info.rec_sz; + if (line_rec_sz != sec->line_info.rec_sz) { + pr_warn("mismatch in line_info record size %zu != %u\n", + line_rec_sz, sec->line_info.rec_sz); + return -EINVAL; + } + + lines_sz += sizeof(struct btf_ext_info_sec) + line_rec_sz * sec->line_info.rec_cnt; + } + if (sec->core_relo_info.rec_cnt) { + if (core_relo_rec_sz == 0) + core_relo_rec_sz = sec->core_relo_info.rec_sz; + if (core_relo_rec_sz != sec->core_relo_info.rec_sz) { + pr_warn("mismatch in core_relo_info record size %zu != %u\n", + core_relo_rec_sz, sec->core_relo_info.rec_sz); + return -EINVAL; + } + + core_relos_sz += sizeof(struct btf_ext_info_sec) + core_relo_rec_sz * sec->core_relo_info.rec_cnt; + } + } + + if (!funcs_sz && !lines_sz && !core_relos_sz) + return 0; + + total_sz += sizeof(struct btf_ext_header); + if (funcs_sz) { + funcs_sz += sizeof(__u32); /* record size prefix */ + total_sz += funcs_sz; + } + if (lines_sz) { + lines_sz += sizeof(__u32); /* record size prefix */ + total_sz += lines_sz; + } + if (core_relos_sz) { + core_relos_sz += sizeof(__u32); /* record size prefix */ + total_sz += core_relos_sz; + } + + cur = data = calloc(1, total_sz); + if (!data) + return -ENOMEM; + + hdr = cur; + hdr->magic = BTF_MAGIC; + hdr->version = BTF_VERSION; + hdr->flags = 0; + hdr->hdr_len = sizeof(struct btf_ext_header); + cur += sizeof(struct btf_ext_header); + + /* All offsets are in bytes relative to the end of this header */ + hdr->func_info_off = 0; + hdr->func_info_len = funcs_sz; + hdr->line_info_off = funcs_sz; + hdr->line_info_len = lines_sz; + hdr->core_relo_off = funcs_sz + lines_sz;; + hdr->core_relo_len = core_relos_sz; + + if (funcs_sz) { + *(__u32 *)cur = func_rec_sz; + cur += sizeof(__u32); + + for (i = 1; i < linker->sec_cnt; i++) { + struct dst_sec *sec = &linker->secs[i]; + + sz = emit_btf_ext_data(linker, cur, sec->sec_name, &sec->func_info); + if (sz < 0) + return sz; + + cur += sz; + } + } + + if (lines_sz) { + *(__u32 *)cur = line_rec_sz; + cur += sizeof(__u32); + + for (i = 1; i < linker->sec_cnt; i++) { + struct dst_sec *sec = &linker->secs[i]; + + sz = emit_btf_ext_data(linker, cur, sec->sec_name, &sec->line_info); + if (sz < 0) + return sz; + + cur += sz; + } + } + + if (core_relos_sz) { + *(__u32 *)cur = core_relo_rec_sz; + cur += sizeof(__u32); + + for (i = 1; i < linker->sec_cnt; i++) { + struct dst_sec *sec = &linker->secs[i]; + + sz = emit_btf_ext_data(linker, cur, sec->sec_name, &sec->core_relo_info); + if (sz < 0) + return sz; + + cur += sz; + } + } + + linker->btf_ext = btf_ext__new(data, total_sz); + err = libbpf_get_error(linker->btf_ext); + if (err) { + linker->btf_ext = NULL; + pr_warn("failed to parse final .BTF.ext data: %d\n", err); + return err; + } + + return 0; +} From patchwork Thu Mar 18 06:31:11 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 404618 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 018DFC43603 for ; Thu, 18 Mar 2021 06:32:23 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id CAC0B64F53 for ; Thu, 18 Mar 2021 06:32:22 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229854AbhCRGbt convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:31:49 -0400 Received: from mx0a-00082601.pphosted.com ([67.231.145.42]:20206 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229745AbhCRGbg (ORCPT ); Thu, 18 Mar 2021 02:31:36 -0400 Received: from pps.filterd (m0044010.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.43/8.16.0.43) with SMTP id 12I6QC0Z028109 for ; Wed, 17 Mar 2021 23:31:36 -0700 Received: from mail.thefacebook.com ([163.114.132.120]) by mx0a-00082601.pphosted.com with ESMTP id 37bgdhdvhn-2 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:36 -0700 Received: from intmgw002.06.ash9.facebook.com (2620:10d:c085:208::11) by mail.thefacebook.com (2620:10d:c085:11d::7) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:35 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id 3A8C52ED24FA; Wed, 17 Mar 2021 23:31:34 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , Subject: [PATCH v3 bpf-next 08/12] bpftool: add ability to specify custom skeleton object name Date: Wed, 17 Mar 2021 23:31:11 -0700 Message-ID: <20210318063115.49613-9-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 mlxlogscore=999 clxscore=1015 phishscore=0 priorityscore=1501 malwarescore=0 mlxscore=0 bulkscore=0 spamscore=0 impostorscore=0 lowpriorityscore=0 adultscore=0 suspectscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Add optional name OBJECT_NAME parameter to `gen skeleton` command to override default object name, normally derived from input file name. This allows much more flexibility during build time. Signed-off-by: Andrii Nakryiko --- .../bpf/bpftool/Documentation/bpftool-gen.rst | 13 +++++---- tools/bpf/bpftool/bash-completion/bpftool | 11 +++++++- tools/bpf/bpftool/gen.c | 27 +++++++++++++++++-- 3 files changed, 43 insertions(+), 8 deletions(-) diff --git a/tools/bpf/bpftool/Documentation/bpftool-gen.rst b/tools/bpf/bpftool/Documentation/bpftool-gen.rst index 84cf0639696f..d4e7338e22e7 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-gen.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-gen.rst @@ -19,7 +19,7 @@ SYNOPSIS GEN COMMANDS ============= -| **bpftool** **gen skeleton** *FILE* +| **bpftool** **gen skeleton** *FILE* [**name** *OBJECT_NAME*] | **bpftool** **gen help** DESCRIPTION @@ -75,10 +75,13 @@ DESCRIPTION specific maps, programs, etc. As part of skeleton, few custom functions are generated. - Each of them is prefixed with object name, derived from - object file name. I.e., if BPF object file name is - **example.o**, BPF object name will be **example**. The - following custom functions are provided in such case: + Each of them is prefixed with object name. Object name can + either be derived from object file name, i.e., if BPF object + file name is **example.o**, BPF object name will be + **example**. Object name can be also specified explicitly + through **name** *OBJECT_NAME* parameter. The following + custom functions are provided (assuming **example** as + the object name): - **example__open** and **example__open_opts**. These functions are used to instantiate skeleton. It diff --git a/tools/bpf/bpftool/bash-completion/bpftool b/tools/bpf/bpftool/bash-completion/bpftool index fdffbc64c65c..bf7b4bdbb23a 100644 --- a/tools/bpf/bpftool/bash-completion/bpftool +++ b/tools/bpf/bpftool/bash-completion/bpftool @@ -982,7 +982,16 @@ _bpftool() gen) case $command in skeleton) - _filedir + case $prev in + $command) + _filedir + return 0 + ;; + *) + _bpftool_once_attr 'name' + return 0 + ;; + esac ;; *) [[ $prev == $object ]] && \ diff --git a/tools/bpf/bpftool/gen.c b/tools/bpf/bpftool/gen.c index 4033c46d83e7..cd35163f2d08 100644 --- a/tools/bpf/bpftool/gen.c +++ b/tools/bpf/bpftool/gen.c @@ -288,6 +288,28 @@ static int do_skeleton(int argc, char **argv) } file = GET_ARG(); + while (argc) { + if (!REQ_ARGS(2)) + return -1; + + if (is_prefix(*argv, "name")) { + NEXT_ARG(); + + if (obj_name[0] != '\0') { + p_err("object name already specified"); + return -1; + } + + strncpy(obj_name, *argv, MAX_OBJ_NAME_LEN - 1); + obj_name[MAX_OBJ_NAME_LEN - 1] = '\0'; + } else { + p_err("unknown arg %s", *argv); + return -1; + } + + NEXT_ARG(); + } + if (argc) { p_err("extra unknown arguments"); return -1; @@ -310,7 +332,8 @@ static int do_skeleton(int argc, char **argv) p_err("failed to mmap() %s: %s", file, strerror(errno)); goto out; } - get_obj_name(obj_name, file); + if (obj_name[0] == '\0') + get_obj_name(obj_name, file); opts.object_name = obj_name; obj = bpf_object__open_mem(obj_data, file_sz, &opts); if (IS_ERR(obj)) { @@ -599,7 +622,7 @@ static int do_help(int argc, char **argv) } fprintf(stderr, - "Usage: %1$s %2$s skeleton FILE\n" + "Usage: %1$s %2$s skeleton FILE [name OBJECT_NAME]\n" " %1$s %2$s help\n" "\n" " " HELP_SPEC_OPTIONS "\n" From patchwork Thu Mar 18 06:31:12 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 405377 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, URIBL_BLOCKED, USER_AGENT_GIT autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 274E2C4360C for ; Thu, 18 Mar 2021 06:32:23 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id F1E8A64F04 for ; Thu, 18 Mar 2021 06:32:22 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229874AbhCRGbv convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:31:51 -0400 Received: from mx0a-00082601.pphosted.com ([67.231.145.42]:43558 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229770AbhCRGbl (ORCPT ); Thu, 18 Mar 2021 02:31:41 -0400 Received: from pps.filterd (m0044010.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.43/8.16.0.43) with SMTP id 12I6Qk4Q028589 for ; Wed, 17 Mar 2021 23:31:41 -0700 Received: from mail.thefacebook.com ([163.114.132.120]) by mx0a-00082601.pphosted.com with ESMTP id 37bgdhdvj7-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:41 -0700 Received: from intmgw001.06.ash9.facebook.com (2620:10d:c085:208::11) by mail.thefacebook.com (2620:10d:c085:21d::4) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:40 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id 62E002ED24FA; Wed, 17 Mar 2021 23:31:36 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , , Quentin Monnet Subject: [PATCH v3 bpf-next 09/12] bpftool: add `gen object` command to perform BPF static linking Date: Wed, 17 Mar 2021 23:31:12 -0700 Message-ID: <20210318063115.49613-10-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 mlxlogscore=999 clxscore=1034 phishscore=0 priorityscore=1501 malwarescore=0 mlxscore=0 bulkscore=0 spamscore=0 impostorscore=0 lowpriorityscore=0 adultscore=0 suspectscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Add `bpftool gen object ...` command to statically link multiple BPF ELF object files into a single output BPF ELF object file. This patch also updates bash completions and man page. Man page gets a short section on `gen object` command, but also updates the skeleton example to show off workflow for BPF application with two .bpf.c files, compiled individually with Clang, then resulting object files are linked together with `gen object`, and then final object file is used to generate usable BPF skeleton. This should help new users understand realistic workflow w.r.t. compiling mutli-file BPF application. Reviewed-by: Quentin Monnet Signed-off-by: Andrii Nakryiko --- .../bpf/bpftool/Documentation/bpftool-gen.rst | 65 +++++++++++++++---- tools/bpf/bpftool/bash-completion/bpftool | 6 +- tools/bpf/bpftool/gen.c | 45 ++++++++++++- 3 files changed, 100 insertions(+), 16 deletions(-) diff --git a/tools/bpf/bpftool/Documentation/bpftool-gen.rst b/tools/bpf/bpftool/Documentation/bpftool-gen.rst index d4e7338e22e7..7cd6681137f3 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-gen.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-gen.rst @@ -14,16 +14,37 @@ SYNOPSIS *OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] } - *COMMAND* := { **skeleton** | **help** } + *COMMAND* := { **object** | **skeleton** | **help** } GEN COMMANDS ============= +| **bpftool** **gen object** *OUTPUT_FILE* *INPUT_FILE* [*INPUT_FILE*...] | **bpftool** **gen skeleton** *FILE* [**name** *OBJECT_NAME*] | **bpftool** **gen help** DESCRIPTION =========== + **bpftool gen object** *OUTPUT_FILE* *INPUT_FILE* [*INPUT_FILE*...] + Statically link (combine) together one or more *INPUT_FILE*'s + into a single resulting *OUTPUT_FILE*. All the files involved + are BPF ELF object files. + + The rules of BPF static linking are mostly the same as for + user-space object files, but in addition to combining data + and instruction sections, .BTF and .BTF.ext (if present in + any of the input files) data are combined together. .BTF + data is deduplicated, so all the common types across + *INPUT_FILE*'s will only be represented once in the resulting + BTF information. + + BPF static linking allows to partition BPF source code into + individually compiled files that are then linked into + a single resulting BPF object file, which can be used to + generated BPF skeleton (with **gen skeleton** command) or + passed directly into **libbpf** (using **bpf_object__open()** + family of APIs). + **bpftool gen skeleton** *FILE* Generate BPF skeleton C header file for a given *FILE*. @@ -133,26 +154,19 @@ OPTIONS EXAMPLES ======== -**$ cat example.c** +**$ cat example1.bpf.c** :: #include #include #include - #include "bpf_helpers.h" + #include const volatile int param1 = 42; bool global_flag = true; struct { int x; } data = {}; - struct { - __uint(type, BPF_MAP_TYPE_HASH); - __uint(max_entries, 128); - __type(key, int); - __type(value, long); - } my_map SEC(".maps"); - SEC("raw_tp/sys_enter") int handle_sys_enter(struct pt_regs *ctx) { @@ -164,6 +178,21 @@ EXAMPLES return 0; } +**$ cat example2.bpf.c** + +:: + + #include + #include + #include + + struct { + __uint(type, BPF_MAP_TYPE_HASH); + __uint(max_entries, 128); + __type(key, int); + __type(value, long); + } my_map SEC(".maps"); + SEC("raw_tp/sys_exit") int handle_sys_exit(struct pt_regs *ctx) { @@ -173,9 +202,17 @@ EXAMPLES } This is example BPF application with two BPF programs and a mix of BPF maps -and global variables. +and global variables. Source code is split across two source code files. -**$ bpftool gen skeleton example.o** +**$ clang -target bpf -g example1.bpf.c -o example1.bpf.o** +**$ clang -target bpf -g example2.bpf.c -o example2.bpf.o** +**$ bpftool gen object example.bpf.o example1.bpf.o example2.bpf.o** + +This set of commands compiles *example1.bpf.c* and *example2.bpf.c* +individually and then statically links respective object files into the final +BPF ELF object file *example.bpf.o*. + +**$ bpftool gen skeleton example.bpf.o name example | tee example.skel.h** :: @@ -230,7 +267,7 @@ and global variables. #endif /* __EXAMPLE_SKEL_H__ */ -**$ cat example_user.c** +**$ cat example.c** :: @@ -273,7 +310,7 @@ and global variables. return err; } -**# ./example_user** +**# ./example** :: diff --git a/tools/bpf/bpftool/bash-completion/bpftool b/tools/bpf/bpftool/bash-completion/bpftool index bf7b4bdbb23a..d67518bcbd44 100644 --- a/tools/bpf/bpftool/bash-completion/bpftool +++ b/tools/bpf/bpftool/bash-completion/bpftool @@ -981,6 +981,10 @@ _bpftool() ;; gen) case $command in + object) + _filedir + return 0 + ;; skeleton) case $prev in $command) @@ -995,7 +999,7 @@ _bpftool() ;; *) [[ $prev == $object ]] && \ - COMPREPLY=( $( compgen -W 'skeleton help' -- "$cur" ) ) + COMPREPLY=( $( compgen -W 'object skeleton help' -- "$cur" ) ) ;; esac ;; diff --git a/tools/bpf/bpftool/gen.c b/tools/bpf/bpftool/gen.c index cd35163f2d08..8c993b597028 100644 --- a/tools/bpf/bpftool/gen.c +++ b/tools/bpf/bpftool/gen.c @@ -614,6 +614,47 @@ static int do_skeleton(int argc, char **argv) return err; } +static int do_object(int argc, char **argv) +{ + struct bpf_linker *linker; + const char *output_file, *file; + int err = 0; + + if (!REQ_ARGS(2)) { + usage(); + return -1; + } + + output_file = GET_ARG(); + + linker = bpf_linker__new(output_file, NULL); + if (!linker) { + p_err("failed to create BPF linker instance"); + return -1; + } + + while (argc) { + file = GET_ARG(); + + err = bpf_linker__add_file(linker, file); + if (err) { + p_err("failed to link '%s': %s (%d)", file, strerror(err), err); + goto out; + } + } + + err = bpf_linker__finalize(linker); + if (err) { + p_err("failed to finalize ELF file: %s (%d)", strerror(err), err); + goto out; + } + + err = 0; +out: + bpf_linker__free(linker); + return err; +} + static int do_help(int argc, char **argv) { if (json_output) { @@ -622,7 +663,8 @@ static int do_help(int argc, char **argv) } fprintf(stderr, - "Usage: %1$s %2$s skeleton FILE [name OBJECT_NAME]\n" + "Usage: %1$s %2$s object OUTPUT_FILE INPUT_FILE [INPUT_FILE...]\n" + " %1$s %2$s skeleton FILE [name OBJECT_NAME]\n" " %1$s %2$s help\n" "\n" " " HELP_SPEC_OPTIONS "\n" @@ -633,6 +675,7 @@ static int do_help(int argc, char **argv) } static const struct cmd cmds[] = { + { "object", do_object }, { "skeleton", do_skeleton }, { "help", do_help }, { 0 } From patchwork Thu Mar 18 06:31:13 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 405373 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 51E62C4332E for ; Thu, 18 Mar 2021 06:32:52 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 251FA64F53 for ; Thu, 18 Mar 2021 06:32:52 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229927AbhCRGcY convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:32:24 -0400 Received: from mx0a-00082601.pphosted.com ([67.231.145.42]:44352 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229775AbhCRGbs (ORCPT ); Thu, 18 Mar 2021 02:31:48 -0400 Received: from pps.filterd (m0044010.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.43/8.16.0.43) with SMTP id 12I6QC0h028109 for ; Wed, 17 Mar 2021 23:31:48 -0700 Received: from mail.thefacebook.com ([163.114.132.120]) by mx0a-00082601.pphosted.com with ESMTP id 37bgdhdvhn-10 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:48 -0700 Received: from intmgw001.05.ash9.facebook.com (2620:10d:c085:108::4) by mail.thefacebook.com (2620:10d:c085:11d::7) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:42 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id 8E2432ED24FA; Wed, 17 Mar 2021 23:31:38 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , Subject: [PATCH v3 bpf-next 10/12] selftests/bpf: re-generate vmlinux.h and BPF skeletons if bpftool changed Date: Wed, 17 Mar 2021 23:31:13 -0700 Message-ID: <20210318063115.49613-11-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 mlxlogscore=999 clxscore=1015 phishscore=0 priorityscore=1501 malwarescore=0 mlxscore=0 bulkscore=0 spamscore=0 impostorscore=0 lowpriorityscore=0 adultscore=0 suspectscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Trigger vmlinux.h and BPF skeletons re-generation if detected that bpftool was re-compiled. Otherwise full `make clean` is required to get updated skeletons, if bpftool is modified. Fixes: acbd06206bbb ("selftests/bpf: Add vmlinux.h selftest exercising tracing of syscalls") Signed-off-by: Andrii Nakryiko --- tools/testing/selftests/bpf/Makefile | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index d0db2b673c6f..dbca39f45382 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -232,7 +232,7 @@ $(HOST_BPFOBJ): $(wildcard $(BPFDIR)/*.[ch] $(BPFDIR)/Makefile) \ DESTDIR=$(HOST_SCRATCH_DIR)/ prefix= all install_headers endif -$(INCLUDE_DIR)/vmlinux.h: $(VMLINUX_BTF) | $(BPFTOOL) $(INCLUDE_DIR) +$(INCLUDE_DIR)/vmlinux.h: $(VMLINUX_BTF) $(BPFTOOL) | $(INCLUDE_DIR) ifeq ($(VMLINUX_H),) $(call msg,GEN,,$@) $(Q)$(BPFTOOL) btf dump file $(VMLINUX_BTF) format c > $@ @@ -357,7 +357,8 @@ $(TRUNNER_BPF_OBJS): $(TRUNNER_OUTPUT)/%.o: \ $(TRUNNER_BPF_SKELS): $(TRUNNER_OUTPUT)/%.skel.h: \ $(TRUNNER_OUTPUT)/%.o \ - | $(BPFTOOL) $(TRUNNER_OUTPUT) + $(BPFTOOL) \ + | $(TRUNNER_OUTPUT) $$(call msg,GEN-SKEL,$(TRUNNER_BINARY),$$@) $(Q)$$(BPFTOOL) gen skeleton $$< > $$@ endif From patchwork Thu Mar 18 06:31:14 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 405375 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 55F12C4321A for ; Thu, 18 Mar 2021 06:32:23 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 2F7A064F30 for ; Thu, 18 Mar 2021 06:32:23 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229884AbhCRGbx convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:31:53 -0400 Received: from mx0a-00082601.pphosted.com ([67.231.145.42]:63538 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229787AbhCRGbp (ORCPT ); Thu, 18 Mar 2021 02:31:45 -0400 Received: from pps.filterd (m0109334.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.43/8.16.0.43) with SMTP id 12I6QS5i015959 for ; Wed, 17 Mar 2021 23:31:45 -0700 Received: from mail.thefacebook.com ([163.114.132.120]) by mx0a-00082601.pphosted.com with ESMTP id 37bdyy6har-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:45 -0700 Received: from intmgw001.25.frc3.facebook.com (2620:10d:c085:208::11) by mail.thefacebook.com (2620:10d:c085:11d::5) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:44 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id B6B582ED24FA; Wed, 17 Mar 2021 23:31:40 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , Subject: [PATCH v3 bpf-next 11/12] selftests/bpf: pass all BPF .o's through BPF static linker Date: Wed, 17 Mar 2021 23:31:14 -0700 Message-ID: <20210318063115.49613-12-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 priorityscore=1501 bulkscore=0 lowpriorityscore=0 clxscore=1015 malwarescore=0 mlxlogscore=999 spamscore=0 phishscore=0 mlxscore=0 impostorscore=0 suspectscore=0 adultscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Pass all individual BPF object files (generated from progs/*.c) through `bpftool gen object` command to validate that BPF static linker doesn't corrupt them. As an additional sanity checks, validate that passing resulting object files through linker again results in identical ELF files. Exact same ELF contents can be guaranteed only after two passes, as after the first pass ELF sections order changes, and thus .BTF.ext data sections order changes. That, in turn, means that strings are added into the final BTF string sections in different order, so .BTF strings data might not be exactly the same. But doing another round of linking afterwards should result in the identical ELF file, which is checked with additional `diff` command. Signed-off-by: Andrii Nakryiko --- tools/testing/selftests/bpf/Makefile | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index dbca39f45382..3a7c02add70c 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -355,12 +355,13 @@ $(TRUNNER_BPF_OBJS): $(TRUNNER_OUTPUT)/%.o: \ $$(call $(TRUNNER_BPF_BUILD_RULE),$$<,$$@, \ $(TRUNNER_BPF_CFLAGS)) -$(TRUNNER_BPF_SKELS): $(TRUNNER_OUTPUT)/%.skel.h: \ - $(TRUNNER_OUTPUT)/%.o \ - $(BPFTOOL) \ - | $(TRUNNER_OUTPUT) +$(TRUNNER_BPF_SKELS): %.skel.h: %.o $(BPFTOOL) | $(TRUNNER_OUTPUT) $$(call msg,GEN-SKEL,$(TRUNNER_BINARY),$$@) - $(Q)$$(BPFTOOL) gen skeleton $$< > $$@ + $(Q)$$(BPFTOOL) gen object $$(<:.o=.linked.o) $$< + $(Q)$$(BPFTOOL) gen object $$(<:.o=.linked2.o) $$(<:.o=.linked.o) + $(Q)$$(BPFTOOL) gen object $$(<:.o=.linked3.o) $$(<:.o=.linked2.o) + $(Q)diff $$(<:.o=.linked2.o) $$(<:.o=.linked3.o) + $(Q)$$(BPFTOOL) gen skeleton $$(<:.o=.linked.o) name $$(notdir $$(<:.o=)) > $$@ endif # ensure we set up tests.h header generation rule just once From patchwork Thu Mar 18 06:31:15 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrii Nakryiko X-Patchwork-Id: 404616 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-17.0 required=3.0 tests=BAYES_00, INCLUDES_CR_TRAILER, INCLUDES_PATCH, MAILING_LIST_MULTI, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 707CBC4332B for ; Thu, 18 Mar 2021 06:32:52 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 4B57664F40 for ; Thu, 18 Mar 2021 06:32:52 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229942AbhCRGc0 convert rfc822-to-8bit (ORCPT ); Thu, 18 Mar 2021 02:32:26 -0400 Received: from mx0a-00082601.pphosted.com ([67.231.145.42]:19078 "EHLO mx0a-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229840AbhCRGbt (ORCPT ); Thu, 18 Mar 2021 02:31:49 -0400 Received: from pps.filterd (m0044010.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.43/8.16.0.43) with SMTP id 12I6QC0k028109 for ; Wed, 17 Mar 2021 23:31:48 -0700 Received: from mail.thefacebook.com ([163.114.132.120]) by mx0a-00082601.pphosted.com with ESMTP id 37bgdhdvhn-13 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Mar 2021 23:31:48 -0700 Received: from intmgw001.06.ash9.facebook.com (2620:10d:c085:208::11) by mail.thefacebook.com (2620:10d:c085:11d::7) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2176.2; Wed, 17 Mar 2021 23:31:45 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id EE4322ED24FA; Wed, 17 Mar 2021 23:31:42 -0700 (PDT) From: Andrii Nakryiko To: , , , CC: , Subject: [PATCH v3 bpf-next 12/12] selftests/bpf: add multi-file statically linked BPF object file test Date: Wed, 17 Mar 2021 23:31:15 -0700 Message-ID: <20210318063115.49613-13-andrii@kernel.org> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210318063115.49613-1-andrii@kernel.org> References: <20210318063115.49613-1-andrii@kernel.org> MIME-Version: 1.0 X-FB-Internal: Safe X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.369, 18.0.761 definitions=2021-03-18_02:2021-03-17,2021-03-18 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 mlxlogscore=999 clxscore=1034 phishscore=0 priorityscore=1501 malwarescore=0 mlxscore=0 bulkscore=0 spamscore=0 impostorscore=0 lowpriorityscore=0 adultscore=0 suspectscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2009150000 definitions=main-2103180050 X-FB-Internal: deliver Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Add Makefile infra to specify multi-file BPF object files (and derivative skeletons). Add first selftest validating BPF static linker can merge together successfully two independent BPF object files and resulting object and skeleton are correct and usable. Use the same F(F(F(X))) = F(F(X)) identity test on linked object files as for the case of single BPF object files. Signed-off-by: Andrii Nakryiko --- tools/testing/selftests/bpf/Makefile | 21 ++++++++-- .../selftests/bpf/prog_tests/static_linked.c | 40 +++++++++++++++++++ .../selftests/bpf/progs/test_static_linked1.c | 30 ++++++++++++++ .../selftests/bpf/progs/test_static_linked2.c | 31 ++++++++++++++ 4 files changed, 119 insertions(+), 3 deletions(-) create mode 100644 tools/testing/selftests/bpf/prog_tests/static_linked.c create mode 100644 tools/testing/selftests/bpf/progs/test_static_linked1.c create mode 100644 tools/testing/selftests/bpf/progs/test_static_linked2.c diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index 3a7c02add70c..c740e76b3b54 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -303,6 +303,10 @@ endef SKEL_BLACKLIST := btf__% test_pinning_invalid.c test_sk_assign.c +LINKED_SKELS := test_static_linked.skel.h + +test_static_linked.skel.h-deps := test_static_linked1.o test_static_linked2.o + # Set up extra TRUNNER_XXX "temporary" variables in the environment (relies on # $eval()) and pass control to DEFINE_TEST_RUNNER_RULES. # Parameters: @@ -323,6 +327,7 @@ TRUNNER_BPF_OBJS := $$(patsubst %.c,$$(TRUNNER_OUTPUT)/%.o, $$(TRUNNER_BPF_SRCS) TRUNNER_BPF_SKELS := $$(patsubst %.c,$$(TRUNNER_OUTPUT)/%.skel.h, \ $$(filter-out $(SKEL_BLACKLIST), \ $$(TRUNNER_BPF_SRCS))) +TRUNNER_BPF_SKELS_LINKED := $$(addprefix $$(TRUNNER_OUTPUT)/,$(LINKED_SKELS)) TEST_GEN_FILES += $$(TRUNNER_BPF_OBJS) # Evaluate rules now with extra TRUNNER_XXX variables above already defined @@ -357,11 +362,20 @@ $(TRUNNER_BPF_OBJS): $(TRUNNER_OUTPUT)/%.o: \ $(TRUNNER_BPF_SKELS): %.skel.h: %.o $(BPFTOOL) | $(TRUNNER_OUTPUT) $$(call msg,GEN-SKEL,$(TRUNNER_BINARY),$$@) - $(Q)$$(BPFTOOL) gen object $$(<:.o=.linked.o) $$< - $(Q)$$(BPFTOOL) gen object $$(<:.o=.linked2.o) $$(<:.o=.linked.o) + $(Q)$$(BPFTOOL) gen object $$(<:.o=.linked1.o) $$< + $(Q)$$(BPFTOOL) gen object $$(<:.o=.linked2.o) $$(<:.o=.linked1.o) $(Q)$$(BPFTOOL) gen object $$(<:.o=.linked3.o) $$(<:.o=.linked2.o) $(Q)diff $$(<:.o=.linked2.o) $$(<:.o=.linked3.o) - $(Q)$$(BPFTOOL) gen skeleton $$(<:.o=.linked.o) name $$(notdir $$(<:.o=)) > $$@ + $(Q)$$(BPFTOOL) gen skeleton $$(<:.o=.linked3.o) name $$(notdir $$(<:.o=)) > $$@ + +$(TRUNNER_BPF_SKELS_LINKED): $(TRUNNER_BPF_OBJS) $(BPFTOOL) | $(TRUNNER_OUTPUT) + $$(call msg,LINK-BPF,$(TRUNNER_BINARY),$$(@:.skel.h=.o)) + $(Q)$$(BPFTOOL) gen object $$(@:.skel.h=.linked1.o) $$(addprefix $(TRUNNER_OUTPUT)/,$$($$(@F)-deps)) + $(Q)$$(BPFTOOL) gen object $$(@:.skel.h=.linked2.o) $$(@:.skel.h=.linked1.o) + $(Q)$$(BPFTOOL) gen object $$(@:.skel.h=.linked3.o) $$(@:.skel.h=.linked2.o) + $(Q)diff $$(<:.o=.linked2.o) $$(<:.o=.linked3.o) + $$(call msg,GEN-SKEL,$(TRUNNER_BINARY),$$@) + $(Q)$$(BPFTOOL) gen skeleton $$(@:.skel.h=.linked3.o) name $$(notdir $$(@:.skel.h=)) > $$@ endif # ensure we set up tests.h header generation rule just once @@ -383,6 +397,7 @@ $(TRUNNER_TEST_OBJS): $(TRUNNER_OUTPUT)/%.test.o: \ $(TRUNNER_EXTRA_HDRS) \ $(TRUNNER_BPF_OBJS) \ $(TRUNNER_BPF_SKELS) \ + $(TRUNNER_BPF_SKELS_LINKED) \ $$(BPFOBJ) | $(TRUNNER_OUTPUT) $$(call msg,TEST-OBJ,$(TRUNNER_BINARY),$$@) $(Q)cd $$(@D) && $$(CC) -I. $$(CFLAGS) -c $(CURDIR)/$$< $$(LDLIBS) -o $$(@F) diff --git a/tools/testing/selftests/bpf/prog_tests/static_linked.c b/tools/testing/selftests/bpf/prog_tests/static_linked.c new file mode 100644 index 000000000000..46556976dccc --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/static_linked.c @@ -0,0 +1,40 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2019 Facebook */ + +#include +#include "test_static_linked.skel.h" + +void test_static_linked(void) +{ + int err; + struct test_static_linked* skel; + + skel = test_static_linked__open(); + if (!ASSERT_OK_PTR(skel, "skel_open")) + return; + + skel->rodata->rovar1 = 1; + skel->bss->static_var1 = 2; + skel->bss->static_var11 = 3; + + skel->rodata->rovar2 = 4; + skel->bss->static_var2 = 5; + skel->bss->static_var22 = 6; + + err = test_static_linked__load(skel); + if (!ASSERT_OK(err, "skel_load")) + goto cleanup; + + err = test_static_linked__attach(skel); + if (!ASSERT_OK(err, "skel_attach")) + goto cleanup; + + /* trigger */ + usleep(1); + + ASSERT_EQ(skel->bss->var1, 1 * 2 + 2 + 3, "var1"); + ASSERT_EQ(skel->bss->var2, 4 * 3 + 5 + 6, "var2"); + +cleanup: + test_static_linked__destroy(skel); +} diff --git a/tools/testing/selftests/bpf/progs/test_static_linked1.c b/tools/testing/selftests/bpf/progs/test_static_linked1.c new file mode 100644 index 000000000000..ea1a6c4c7172 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/test_static_linked1.c @@ -0,0 +1,30 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2021 Facebook */ + +#include +#include + +/* 8-byte aligned .bss */ +static volatile long static_var1; +static volatile int static_var11; +int var1 = 0; +/* 4-byte aligned .rodata */ +const volatile int rovar1; + +/* same "subprog" name in both files */ +static __noinline int subprog(int x) +{ + /* but different formula */ + return x * 2; +} + +SEC("raw_tp/sys_enter") +int handler1(const void *ctx) +{ + var1 = subprog(rovar1) + static_var1 + static_var11; + + return 0; +} + +char LICENSE[] SEC("license") = "GPL"; +int VERSION SEC("version") = 1; diff --git a/tools/testing/selftests/bpf/progs/test_static_linked2.c b/tools/testing/selftests/bpf/progs/test_static_linked2.c new file mode 100644 index 000000000000..54d8d1ab577c --- /dev/null +++ b/tools/testing/selftests/bpf/progs/test_static_linked2.c @@ -0,0 +1,31 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2021 Facebook */ + +#include +#include + +/* 4-byte aligned .bss */ +static volatile int static_var2; +static volatile int static_var22; +int var2 = 0; +/* 8-byte aligned .rodata */ +const volatile long rovar2; + +/* same "subprog" name in both files */ +static __noinline int subprog(int x) +{ + /* but different formula */ + return x * 3; +} + +SEC("raw_tp/sys_enter") +int handler2(const void *ctx) +{ + var2 = subprog(rovar2) + static_var2 + static_var22; + + return 0; +} + +/* different name and/or type of the variable doesn't matter */ +char _license[] SEC("license") = "GPL"; +int _version SEC("version") = 1;