| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <202206231904.F79Q8wfi-lkp@intel.com>
Date: Thu, 23 Jun 2022 20:44:26 +0800
From: kernel test robot <lkp@...el.com>
To: Jiri Olsa <jolsa@...nel.org>
Cc: llvm@...ts.linux.dev, kbuild-all@...ts.01.org,
linux-kernel@...r.kernel.org
Subject: [jolsa-perf:bpf/tramp_11 17/21] kernel/bpf/syscall.c:2994:8: error:
call to undeclared function 'bpf_tramp_id_alloc'; ISO C99 and later do not
support implicit function declarations
tree: https://git.kernel.org/pub/scm/linux/kernel/git/jolsa/perf.git bpf/tramp_11
head: c807110aa605281263da02a78af57847f5c4275e
commit: ac6153f8956b357b52e16ea68b70ea51aff1752c [17/21] bpf: Use bpf_tramp_id object in attach
config: hexagon-randconfig-r041-20220622 (https://download.01.org/0day-ci/archive/20220623/202206231904.F79Q8wfi-lkp@intel.com/config)
compiler: clang version 15.0.0 (https://github.com/llvm/llvm-project 46be5faaf03466c3751f8a2882bef5a217e15926)
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# https://git.kernel.org/pub/scm/linux/kernel/git/jolsa/perf.git/commit/?id=ac6153f8956b357b52e16ea68b70ea51aff1752c
git remote add jolsa-perf https://git.kernel.org/pub/scm/linux/kernel/git/jolsa/perf.git
git fetch --no-tags jolsa-perf bpf/tramp_11
git checkout ac6153f8956b357b52e16ea68b70ea51aff1752c
# save the config file
mkdir build_dir && cp config build_dir/.config
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir ARCH=hexagon SHELL=/bin/bash kernel/bpf/
If you fix the issue, kindly add following tag where applicable
Reported-by: kernel test robot <lkp@...el.com>
All error/warnings (new ones prefixed by >>):
>> kernel/bpf/syscall.c:2994:8: error: call to undeclared function 'bpf_tramp_id_alloc'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
id = bpf_tramp_id_alloc();
^
kernel/bpf/syscall.c:2994:8: note: did you mean 'bpf_map_kzalloc'?
include/linux/bpf.h:1656:1: note: 'bpf_map_kzalloc' declared here
bpf_map_kzalloc(const struct bpf_map *map, size_t size, gfp_t flags)
^
>> kernel/bpf/syscall.c:2994:6: warning: incompatible integer to pointer conversion assigning to 'struct bpf_tramp_id *' from 'int' [-Wint-conversion]
id = bpf_tramp_id_alloc();
^ ~~~~~~~~~~~~~~~~~~~~
>> kernel/bpf/syscall.c:3007:3: error: call to undeclared function 'bpf_tramp_id_init'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
bpf_tramp_id_init(id, tgt_prog, NULL, btf_id);
^
kernel/bpf/syscall.c:3053:8: error: call to undeclared function 'bpf_tramp_id_alloc'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
id = bpf_tramp_id_alloc();
^
kernel/bpf/syscall.c:3053:6: warning: incompatible integer to pointer conversion assigning to 'struct bpf_tramp_id *' from 'int' [-Wint-conversion]
id = bpf_tramp_id_alloc();
^ ~~~~~~~~~~~~~~~~~~~~
kernel/bpf/syscall.c:3059:3: error: call to undeclared function 'bpf_tramp_id_init'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
bpf_tramp_id_init(id, NULL, prog->aux->attach_btf, btf_id);
^
>> kernel/bpf/syscall.c:3062:29: error: call to undeclared function 'bpf_tramp_id_is_equal'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
if (!prog->aux->dst_id || !bpf_tramp_id_is_equal(id, prog->aux->dst_id)) {
^
2 warnings and 5 errors generated.
--
>> kernel/bpf/verifier.c:15073:7: error: call to undeclared function 'bpf_tramp_id_alloc'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
id = bpf_tramp_id_alloc();
^
kernel/bpf/verifier.c:15073:7: note: did you mean 'bpf_map_kzalloc'?
include/linux/bpf.h:1656:1: note: 'bpf_map_kzalloc' declared here
bpf_map_kzalloc(const struct bpf_map *map, size_t size, gfp_t flags)
^
>> kernel/bpf/verifier.c:15073:5: warning: incompatible integer to pointer conversion assigning to 'struct bpf_tramp_id *' from 'int' [-Wint-conversion]
id = bpf_tramp_id_alloc();
^ ~~~~~~~~~~~~~~~~~~~~
>> kernel/bpf/verifier.c:15077:2: error: call to undeclared function 'bpf_tramp_id_init'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
bpf_tramp_id_init(id, tgt_prog, prog->aux->attach_btf, btf_id);
^
1 warning and 2 errors generated.
vim +/bpf_tramp_id_alloc +2994 kernel/bpf/syscall.c
2942
2943 static int bpf_tracing_prog_attach(struct bpf_prog *prog,
2944 int tgt_prog_fd,
2945 u32 btf_id,
2946 u64 bpf_cookie)
2947 {
2948 struct bpf_attach_target_info tgt_info_buf = {};
2949 struct bpf_attach_target_info *tgt_info = &tgt_info_buf;
2950 struct bpf_link_primer link_primer;
2951 struct bpf_prog *tgt_prog = NULL;
2952 struct bpf_tramp_id *id = NULL;
2953 struct bpf_tracing_link *link;
2954 int err;
2955
2956 switch (prog->type) {
2957 case BPF_PROG_TYPE_TRACING:
2958 if (prog->expected_attach_type != BPF_TRACE_FENTRY &&
2959 prog->expected_attach_type != BPF_TRACE_FEXIT &&
2960 prog->expected_attach_type != BPF_MODIFY_RETURN) {
2961 err = -EINVAL;
2962 goto out_put_prog;
2963 }
2964 break;
2965 case BPF_PROG_TYPE_EXT:
2966 if (prog->expected_attach_type != 0) {
2967 err = -EINVAL;
2968 goto out_put_prog;
2969 }
2970 break;
2971 case BPF_PROG_TYPE_LSM:
2972 if (prog->expected_attach_type != BPF_LSM_MAC) {
2973 err = -EINVAL;
2974 goto out_put_prog;
2975 }
2976 break;
2977 default:
2978 err = -EINVAL;
2979 goto out_put_prog;
2980 }
2981
2982 if (!!tgt_prog_fd != !!btf_id) {
2983 err = -EINVAL;
2984 goto out_put_prog;
2985 }
2986
2987 if (tgt_prog_fd) {
2988 /* For now we only allow new targets for BPF_PROG_TYPE_EXT */
2989 if (prog->type != BPF_PROG_TYPE_EXT) {
2990 err = -EINVAL;
2991 goto out_put_prog;
2992 }
2993
> 2994 id = bpf_tramp_id_alloc();
2995 if (!id) {
2996 err = -ENOMEM;
2997 goto out_put_prog;
2998 }
2999
3000 tgt_prog = bpf_prog_get(tgt_prog_fd);
3001 if (IS_ERR(tgt_prog)) {
3002 err = PTR_ERR(tgt_prog);
3003 tgt_prog = NULL;
3004 goto out_put_prog;
3005 }
3006
> 3007 bpf_tramp_id_init(id, tgt_prog, NULL, btf_id);
3008 }
3009
3010 link = kzalloc(sizeof(*link), GFP_USER);
3011 if (!link) {
3012 err = -ENOMEM;
3013 goto out_put_prog;
3014 }
3015 bpf_link_init(&link->link, BPF_LINK_TYPE_TRACING,
3016 &bpf_tracing_link_lops, prog);
3017 link->attach_type = prog->expected_attach_type;
3018 link->attach.tp.cookie = bpf_cookie;
3019 link->attach.tp.prog = prog;
3020
3021 mutex_lock(&prog->aux->dst_mutex);
3022
3023 /* There are a few possible cases here:
3024 *
3025 * - if prog->aux->dst_key is set, the program was just loaded
3026 * and not yet attached to anything, so we can use the values stored
3027 * in prog->aux
3028 *
3029 * - if prog->aux->dst_key is NULL, the program has already been
3030 * attached to a target and its initial target was cleared (below)
3031 *
3032 * - if tgt_prog != NULL, the caller specified tgt_prog_fd +
3033 * target_btf_id using the link_create API.
3034 *
3035 * - if tgt_prog == NULL when this function was called using the old
3036 * raw_tracepoint_open API, and we need a target from prog->aux
3037 *
3038 * - if prog->aux->dst_key and tgt_prog is NULL, the program
3039 * was detached and is going for re-attachment.
3040 */
3041 if (!prog->aux->dst_id && !tgt_prog) {
3042 /*
3043 * Allow re-attach for TRACING and LSM programs. If it's
3044 * currently linked, bpf_trampoline_link_prog will fail.
3045 * EXT programs need to specify tgt_prog_fd, so they
3046 * re-attach in separate code path.
3047 */
3048 if (prog->type != BPF_PROG_TYPE_TRACING &&
3049 prog->type != BPF_PROG_TYPE_LSM) {
3050 err = -EINVAL;
3051 goto out_unlock;
3052 }
> 3053 id = bpf_tramp_id_alloc();
3054 if (!id) {
3055 err = -ENOMEM;
3056 goto out_unlock;
3057 }
3058 btf_id = prog->aux->attach_btf_id;
3059 bpf_tramp_id_init(id, NULL, prog->aux->attach_btf, btf_id);
3060 }
3061
> 3062 if (!prog->aux->dst_id || !bpf_tramp_id_is_equal(id, prog->aux->dst_id)) {
3063 /* If there is no saved target, or the specified target is
3064 * different from the destination specified at load time, we
3065 * need a new trampoline and a check for compatibility
3066 */
3067
3068 err = bpf_check_attach_target(NULL, prog, tgt_prog, btf_id,
3069 tgt_info);
3070 if (err)
3071 goto out_unlock;
3072 } else {
3073 /* The caller didn't specify a target, or the target was the
3074 * same as the destination supplied during program load. This
3075 * means we can reuse the trampoline and reference from program
3076 * load time, and there is no need to allocate a new one. This
3077 * can only happen once for any program, as the saved values in
3078 * prog->aux are cleared below.
3079 */
3080 id = prog->aux->dst_id;
3081 tgt_prog = prog->aux->dst_prog;
3082 tgt_info = &prog->aux->dst_tgt_info;
3083 }
3084
3085 link->attach.id = id;
3086
3087 err = bpf_link_prime(&link->link, &link_primer);
3088 if (err)
3089 goto out_unlock;
3090
3091 err = bpf_trampoline_attach(&link->attach, tgt_info);
3092 if (err) {
3093 bpf_link_cleanup(&link_primer);
3094 link = NULL;
3095 goto out_unlock;
3096 }
3097
3098 link->tgt_prog = tgt_prog;
3099
3100 /* Always clear the trampoline and target prog from prog->aux to make
3101 * sure the original attach destination is not kept alive after a
3102 * program is (re-)attached to another target.
3103 */
3104 if (prog->aux->dst_prog &&
3105 (tgt_prog_fd || !bpf_tramp_id_is_equal(id, prog->aux->dst_id)))
3106 /* got extra prog ref from syscall, or attaching to different prog */
3107 bpf_prog_put(prog->aux->dst_prog);
3108
3109 prog->aux->dst_prog = NULL;
3110 prog->aux->dst_id = NULL;
3111 mutex_unlock(&prog->aux->dst_mutex);
3112
3113 return bpf_link_settle(&link_primer);
3114 out_unlock:
3115 mutex_unlock(&prog->aux->dst_mutex);
3116 kfree(link);
3117 out_put_prog:
3118 if (tgt_prog_fd && tgt_prog)
3119 bpf_prog_put(tgt_prog);
3120 return err;
3121 }
3122
--
0-DAY CI Kernel Test Service
https://01.org/lkp
Powered by blists - more mailing lists