CC: l...@lists.linux.dev CC: kbuild-...@lists.01.org BCC: l...@intel.com In-Reply-To: <20220304081145.2037182-2-wangyu...@huawei.com> References: <20220304081145.2037182-2-wangyu...@huawei.com> TO: Wang Yufen <wangyu...@huawei.com> TO: john.fastab...@gmail.com TO: dan...@iogearbox.net TO: ja...@cloudflare.com TO: l...@cloudflare.com TO: da...@davemloft.net TO: b...@vger.kernel.org CC: eduma...@google.com CC: yoshf...@linux-ipv6.org CC: dsah...@kernel.org CC: k...@kernel.org CC: a...@kernel.org CC: and...@kernel.org CC: ka...@fb.com CC: songliubrav...@fb.com CC: y...@fb.com CC: kpsi...@kernel.org CC: net...@vger.kernel.org CC: Wang Yufen <wangyu...@huawei.com>
Hi Wang, Thank you for the patch! Perhaps something to improve: [auto build test WARNING on bpf-next/master] url: https://github.com/0day-ci/linux/commits/Wang-Yufen/bpf-sockmap-Fix-memleaks-and-issues-of-mem-charge-uncharge/20220304-155608 base: https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git master :::::: branch date: 7 days ago :::::: commit date: 7 days ago config: riscv-randconfig-c006-20220307 (https://download.01.org/0day-ci/archive/20220312/202203120008.5mallsvp-...@intel.com/config) compiler: clang version 15.0.0 (https://github.com/llvm/llvm-project d271fc04d5b97b12e6b797c6067d3c96a8d7470e) 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 # install riscv cross compiling tool for clang build # apt-get install binutils-riscv64-linux-gnu # https://github.com/0day-ci/linux/commit/686b7d3583903f187fd3e88e1649b07e80a0cfa7 git remote add linux-review https://github.com/0day-ci/linux git fetch --no-tags linux-review Wang-Yufen/bpf-sockmap-Fix-memleaks-and-issues-of-mem-charge-uncharge/20220304-155608 git checkout 686b7d3583903f187fd3e88e1649b07e80a0cfa7 # save the config file to linux build tree COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=riscv clang-analyzer If you fix the issue, kindly add following tag as appropriate Reported-by: kernel test robot <l...@intel.com> clang-analyzer warnings: (new ones prefixed by >>) ^ drivers/scsi/st.c:3121:6: note: Assuming 'SRpnt' is null if (!SRpnt) ^~~~~~ drivers/scsi/st.c:3121:2: note: Taking true branch if (!SRpnt) ^ drivers/scsi/st.c:3122:3: note: Returning without writing to '*partition' return (STp->buffer)->syscall_result; ^ drivers/scsi/st.c:3266:11: note: Returning from 'get_location' if ((i = get_location(STp, &block, &partition, 1)) < 0) ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/scsi/st.c:3266:6: note: Assuming the condition is false if ((i = get_location(STp, &block, &partition, 1)) < 0) ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/scsi/st.c:3266:2: note: Taking false branch if ((i = get_location(STp, &block, &partition, 1)) < 0) ^ drivers/scsi/st.c:3268:16: note: The left operand of '>=' is a garbage value if (partition >= ST_NBR_PARTITIONS) ~~~~~~~~~ ^ drivers/scsi/st.c:4252:7: warning: Although the value stored to 'stp' is used in the enclosing expression, the value is never actually read from 'stp' [clang-analyzer-deadcode.DeadStores] if ((stp = st_incompatible(SDp))) { ^ ~~~~~~~~~~~~~~~~~~~~ drivers/scsi/st.c:4252:7: note: Although the value stored to 'stp' is used in the enclosing expression, the value is never actually read from 'stp' if ((stp = st_incompatible(SDp))) { ^ ~~~~~~~~~~~~~~~~~~~~ Suppressed 7 warnings (7 in non-user code). Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 7 warnings generated. drivers/media/dvb-frontends/si2165.c:528:34: warning: Value stored to 'c' during its initialization is never read [clang-analyzer-deadcode.DeadStores] struct dtv_frontend_properties *c = &fe->dtv_property_cache; ^ ~~~~~~~~~~~~~~~~~~~~~~~ drivers/media/dvb-frontends/si2165.c:528:34: note: Value stored to 'c' during its initialization is never read struct dtv_frontend_properties *c = &fe->dtv_property_cache; ^ ~~~~~~~~~~~~~~~~~~~~~~~ Suppressed 6 warnings (6 in non-user code). Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 7 warnings generated. drivers/media/dvb-frontends/sp2.c:208:14: warning: Value stored to 's' during its initialization is never read [clang-analyzer-deadcode.DeadStores] struct sp2 *s = en50221->data; ^ ~~~~~~~~~~~~~ drivers/media/dvb-frontends/sp2.c:208:14: note: Value stored to 's' during its initialization is never read struct sp2 *s = en50221->data; ^ ~~~~~~~~~~~~~ Suppressed 6 warnings (6 in non-user code). Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 8 warnings generated. drivers/media/dvb-frontends/horus3a.c:141:23: warning: Value stored to 'priv' during its initialization is never read [clang-analyzer-deadcode.DeadStores] struct horus3a_priv *priv = fe->tuner_priv; ^~~~ ~~~~~~~~~~~~~~ drivers/media/dvb-frontends/horus3a.c:141:23: note: Value stored to 'priv' during its initialization is never read struct horus3a_priv *priv = fe->tuner_priv; ^~~~ ~~~~~~~~~~~~~~ drivers/media/dvb-frontends/horus3a.c:149:23: warning: Value stored to 'priv' during its initialization is never read [clang-analyzer-deadcode.DeadStores] struct horus3a_priv *priv = fe->tuner_priv; ^~~~ ~~~~~~~~~~~~~~ drivers/media/dvb-frontends/horus3a.c:149:23: note: Value stored to 'priv' during its initialization is never read struct horus3a_priv *priv = fe->tuner_priv; ^~~~ ~~~~~~~~~~~~~~ Suppressed 6 warnings (6 in non-user code). Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 8 warnings generated. drivers/media/dvb-frontends/ascot2e.c:252:23: warning: Value stored to 'priv' during its initialization is never read [clang-analyzer-deadcode.DeadStores] struct ascot2e_priv *priv = fe->tuner_priv; ^~~~ ~~~~~~~~~~~~~~ drivers/media/dvb-frontends/ascot2e.c:252:23: note: Value stored to 'priv' during its initialization is never read struct ascot2e_priv *priv = fe->tuner_priv; ^~~~ ~~~~~~~~~~~~~~ drivers/media/dvb-frontends/ascot2e.c:272:23: warning: Value stored to 'priv' during its initialization is never read [clang-analyzer-deadcode.DeadStores] struct ascot2e_priv *priv = fe->tuner_priv; ^~~~ ~~~~~~~~~~~~~~ drivers/media/dvb-frontends/ascot2e.c:272:23: note: Value stored to 'priv' during its initialization is never read struct ascot2e_priv *priv = fe->tuner_priv; ^~~~ ~~~~~~~~~~~~~~ Suppressed 6 warnings (6 in non-user code). Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 15 warnings generated. net/core/net-sysfs.c:1769:2: warning: Value stored to 'txq' is never read [clang-analyzer-deadcode.DeadStores] txq = real_tx; ^ ~~~~~~~ net/core/net-sysfs.c:1769:2: note: Value stored to 'txq' is never read txq = real_tx; ^ ~~~~~~~ Suppressed 14 warnings (14 in non-user code). Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 13 warnings generated. Suppressed 13 warnings (13 in non-user code). Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 65 warnings generated. Suppressed 65 warnings (63 in non-user code, 2 with check filters). Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 14 warnings generated. Suppressed 14 warnings (14 in non-user code). Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 13 warnings generated. Suppressed 13 warnings (13 in non-user code). Use -header-filter=.* to display errors from all non-system headers. Use -system-headers to display errors from system headers as well. 15 warnings generated. >> net/core/skmsg.c:569:3: warning: Attempt to free released memory >> [clang-analyzer-unix.Malloc] kfree(msg); ^ net/core/skmsg.c:631:6: note: Assuming field 'skb' is null if (unlikely(state->skb)) { ^ include/linux/compiler.h:78:40: note: expanded from macro 'unlikely' # define unlikely(x) __builtin_expect(!!(x), 0) ^~~~ net/core/skmsg.c:631:2: note: Taking false branch if (unlikely(state->skb)) { ^ net/core/skmsg.c:639:6: note: 'skb' is null if (skb) ^~~ net/core/skmsg.c:639:2: note: Taking false branch if (skb) ^ net/core/skmsg.c:642:2: note: Loop condition is true. Entering loop body while ((skb = skb_dequeue(&psock->ingress_skb))) { ^ net/core/skmsg.c:645:7: note: Assuming the condition is true if (skb_bpf_strparser(skb)) { ^~~~~~~~~~~~~~~~~~~~~~ net/core/skmsg.c:645:3: note: Taking true branch if (skb_bpf_strparser(skb)) { ^ net/core/skmsg.c:656:8: note: Assuming the condition is true if (!sock_flag(psock->sk, SOCK_DEAD)) ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ net/core/skmsg.c:656:4: note: Taking true branch if (!sock_flag(psock->sk, SOCK_DEAD)) ^ net/core/skmsg.c:657:11: note: Calling 'sk_psock_handle_skb' ret = sk_psock_handle_skb(psock, skb, off, ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ net/core/skmsg.c:597:6: note: Assuming 'ingress' is true if (!ingress) { ^~~~~~~~ net/core/skmsg.c:597:2: note: Taking false branch if (!ingress) { ^ net/core/skmsg.c:602:9: note: Calling 'sk_psock_skb_ingress' return sk_psock_skb_ingress(psock, skb, off, len); ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ net/core/skmsg.c:554:15: note: Assuming 'sk' is not equal to field 'sk' if (unlikely(skb->sk == sk)) ^ include/linux/compiler.h:78:42: note: expanded from macro 'unlikely' # define unlikely(x) __builtin_expect(!!(x), 0) ^ net/core/skmsg.c:554:2: note: Taking false branch if (unlikely(skb->sk == sk)) ^ net/core/skmsg.c:556:8: note: Calling 'sk_psock_create_ingress_msg' msg = sk_psock_create_ingress_msg(sk, skb); ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ net/core/skmsg.c:496:6: note: Assuming the condition is false if (atomic_read(&sk->sk_rmem_alloc) > sk->sk_rcvbuf) ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ net/core/skmsg.c:496:2: note: Taking false branch if (atomic_read(&sk->sk_rmem_alloc) > sk->sk_rcvbuf) ^ net/core/skmsg.c:499:2: note: Taking false branch if (!sk_rmem_schedule(sk, skb, skb->truesize)) ^ net/core/skmsg.c:503:15: note: Assuming 'msg' is non-null if (unlikely(!msg)) ^ include/linux/compiler.h:78:42: note: expanded from macro 'unlikely' # define unlikely(x) __builtin_expect(!!(x), 0) ^ net/core/skmsg.c:503:2: note: Taking false branch if (unlikely(!msg)) ^ net/core/skmsg.c:556:8: note: Returning from 'sk_psock_create_ingress_msg' msg = sk_psock_create_ingress_msg(sk, skb); ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ net/core/skmsg.c:557:7: note: 'msg' is non-null if (!msg) ^~~ net/core/skmsg.c:557:2: note: Taking false branch if (!msg) ^ net/core/skmsg.c:567:8: note: Calling 'sk_psock_skb_ingress_enqueue' err = sk_psock_skb_ingress_enqueue(skb, off, len, psock, sk, msg); ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ net/core/skmsg.c:523:2: note: Taking false branch if (skb_linearize(skb)) ^ net/core/skmsg.c:526:15: note: Assuming 'num_sge' is >= 0 if (unlikely(num_sge < 0)) ^ include/linux/compiler.h:78:42: note: expanded from macro 'unlikely' # define unlikely(x) __builtin_expect(!!(x), 0) ^ net/core/skmsg.c:526:2: note: Taking false branch if (unlikely(num_sge < 0)) ^ net/core/skmsg.c:535:2: note: Calling 'sk_psock_queue_msg' sk_psock_queue_msg(psock, msg); vim +569 net/core/skmsg.c 604326b41a6fb9 Daniel Borkmann 2018-10-13 539 7303524e04af49 Liu Jian 2021-10-29 540 static int sk_psock_skb_ingress_self(struct sk_psock *psock, struct sk_buff *skb, 7303524e04af49 Liu Jian 2021-10-29 541 u32 off, u32 len); 2443ca66676d50 John Fastabend 2020-11-16 542 7303524e04af49 Liu Jian 2021-10-29 543 static int sk_psock_skb_ingress(struct sk_psock *psock, struct sk_buff *skb, 7303524e04af49 Liu Jian 2021-10-29 544 u32 off, u32 len) 6fa9201a898983 John Fastabend 2020-11-16 545 { 6fa9201a898983 John Fastabend 2020-11-16 546 struct sock *sk = psock->sk; 6fa9201a898983 John Fastabend 2020-11-16 547 struct sk_msg *msg; 7e6b27a69167f9 John Fastabend 2021-07-12 548 int err; 6fa9201a898983 John Fastabend 2020-11-16 549 2443ca66676d50 John Fastabend 2020-11-16 550 /* If we are receiving on the same sock skb->sk is already assigned, 2443ca66676d50 John Fastabend 2020-11-16 551 * skip memory accounting and owner transition seeing it already set 2443ca66676d50 John Fastabend 2020-11-16 552 * correctly. 2443ca66676d50 John Fastabend 2020-11-16 553 */ 2443ca66676d50 John Fastabend 2020-11-16 554 if (unlikely(skb->sk == sk)) 7303524e04af49 Liu Jian 2021-10-29 555 return sk_psock_skb_ingress_self(psock, skb, off, len); 6fa9201a898983 John Fastabend 2020-11-16 556 msg = sk_psock_create_ingress_msg(sk, skb); 6fa9201a898983 John Fastabend 2020-11-16 557 if (!msg) 6fa9201a898983 John Fastabend 2020-11-16 558 return -EAGAIN; 6fa9201a898983 John Fastabend 2020-11-16 559 6fa9201a898983 John Fastabend 2020-11-16 560 /* This will transition ownership of the data from the socket where 6fa9201a898983 John Fastabend 2020-11-16 561 * the BPF program was run initiating the redirect to the socket 6fa9201a898983 John Fastabend 2020-11-16 562 * we will eventually receive this data on. The data will be released 6fa9201a898983 John Fastabend 2020-11-16 563 * from skb_consume found in __tcp_bpf_recvmsg() after its been copied 6fa9201a898983 John Fastabend 2020-11-16 564 * into user buffers. 6fa9201a898983 John Fastabend 2020-11-16 565 */ 6fa9201a898983 John Fastabend 2020-11-16 566 skb_set_owner_r(skb, sk); 7303524e04af49 Liu Jian 2021-10-29 567 err = sk_psock_skb_ingress_enqueue(skb, off, len, psock, sk, msg); 7e6b27a69167f9 John Fastabend 2021-07-12 568 if (err < 0) 7e6b27a69167f9 John Fastabend 2021-07-12 @569 kfree(msg); 7e6b27a69167f9 John Fastabend 2021-07-12 570 return err; 6fa9201a898983 John Fastabend 2020-11-16 571 } 6fa9201a898983 John Fastabend 2020-11-16 572 --- 0-DAY CI Kernel Test Service https://lists.01.org/hyperkitty/list/kbuild-...@lists.01.org _______________________________________________ kbuild mailing list -- kbuild@lists.01.org To unsubscribe send an email to kbuild-le...@lists.01.org
