Hi Dmitry, Thank you for the patch! Perhaps something to improve: [auto build test WARNING on 058affafc65a74cf54499fb578b66ad0b18f939b] url: https://github.com/intel-lab-lkp/linux/commits/Dmitry-Safonov/net-crypto-Introduce-crypto_pool/20220727-041830 base: 058affafc65a74cf54499fb578b66ad0b18f939b config: i386-randconfig-a005 (https://download.01.org/0day-ci/archive/20220814/202208142357.bDGLpecB-lkp@xxxxxxxxx/config) compiler: gcc-11 (Debian 11.3.0-3) 11.3.0 reproduce (this is a W=1 build): # https://github.com/intel-lab-lkp/linux/commit/a4ee3ecdaada036ed6747ed86eaf7270d3f27bab git remote add linux-review https://github.com/intel-lab-lkp/linux git fetch --no-tags linux-review Dmitry-Safonov/net-crypto-Introduce-crypto_pool/20220727-041830 git checkout a4ee3ecdaada036ed6747ed86eaf7270d3f27bab # save the config file mkdir build_dir && cp config build_dir/.config make W=1 O=build_dir ARCH=i386 SHELL=/bin/bash net/ipv4/ If you fix the issue, kindly add following tag where applicable Reported-by: kernel test robot <lkp@xxxxxxxxx> All warnings (new ones prefixed by >>): >> net/ipv4/tcp_ipv4.c:1174:5: warning: no previous prototype for '__tcp_md5_do_add' [-Wmissing-prototypes] 1174 | int __tcp_md5_do_add(struct sock *sk, const union tcp_md5_addr *addr, | ^~~~~~~~~~~~~~~~ vim +/__tcp_md5_do_add +1174 net/ipv4/tcp_ipv4.c 1172 1173 /* This can be called on a newly created socket, from other files */ > 1174 int __tcp_md5_do_add(struct sock *sk, const union tcp_md5_addr *addr, 1175 int family, u8 prefixlen, int l3index, u8 flags, 1176 const u8 *newkey, u8 newkeylen, gfp_t gfp) 1177 { 1178 /* Add Key to the list */ 1179 struct tcp_md5sig_key *key; 1180 struct tcp_sock *tp = tcp_sk(sk); 1181 struct tcp_md5sig_info *md5sig; 1182 1183 key = tcp_md5_do_lookup_exact(sk, addr, family, prefixlen, l3index, flags); 1184 if (key) { 1185 /* Pre-existing entry - just update that one. 1186 * Note that the key might be used concurrently. 1187 * data_race() is telling kcsan that we do not care of 1188 * key mismatches, since changing MD5 key on live flows 1189 * can lead to packet drops. 1190 */ 1191 data_race(memcpy(key->key, newkey, newkeylen)); 1192 1193 /* Pairs with READ_ONCE() in tcp_md5_hash_key(). 1194 * Also note that a reader could catch new key->keylen value 1195 * but old key->key[], this is the reason we use __GFP_ZERO 1196 * at sock_kmalloc() time below these lines. 1197 */ 1198 WRITE_ONCE(key->keylen, newkeylen); 1199 1200 return 0; 1201 } 1202 1203 md5sig = rcu_dereference_protected(tp->md5sig_info, 1204 lockdep_sock_is_held(sk)); 1205 1206 key = sock_kmalloc(sk, sizeof(*key), gfp | __GFP_ZERO); 1207 if (!key) 1208 return -ENOMEM; 1209 if (!tcp_alloc_md5sig_pool()) { 1210 sock_kfree_s(sk, key, sizeof(*key)); 1211 return -ENOMEM; 1212 } 1213 1214 memcpy(key->key, newkey, newkeylen); 1215 key->keylen = newkeylen; 1216 key->family = family; 1217 key->prefixlen = prefixlen; 1218 key->l3index = l3index; 1219 key->flags = flags; 1220 memcpy(&key->addr, addr, 1221 (IS_ENABLED(CONFIG_IPV6) && family == AF_INET6) ? sizeof(struct in6_addr) : 1222 sizeof(struct in_addr)); 1223 hlist_add_head_rcu(&key->node, &md5sig->head); 1224 return 0; 1225 } 1226 -- 0-DAY CI Kernel Test Service https://01.org/lkp