From: Geliang Tang tanggeliang@kylinos.cn
Run the following BPF selftests on Loongarch:
./test_sockmap
A Kernel panic occurs:
''' Oops[#1]: CPU: 20 PID: 23245 Comm: test_sockmap Tainted: G OE 6.10.0-rc2+ #32 Hardware name: LOONGSON Dabieshan/Loongson-TC542F0, BIOS Loongson-UDK2018-V4.0.11 pc 900000000426cd1c ra 90000000043a315c tp 900010008bfbc000 sp 900010008bfbf8a0 a0 ffffffffffffffe4 a1 900010008bfbfe20 a2 9000100089cd9400 a3 0000000000000003 a4 900010008bfbfb80 a5 900010008bfbfe20 a6 0000000000000000 a7 00000000000000d3 t0 0000000000000000 t1 0000000000000000 t2 0000000000008000 t3 0000000000000000 t4 0000000000000000 t5 0000000000000000 t6 0000000000000006 t7 fffffef1fea12c80 t8 fffffffffffffffc u0 0000000400000005 s9 0000000000000003 s0 0000000000000000 s1 0000000000000012 s2 900010008b9bbc00 s3 0000000000000018 s4 0000020000000000 s5 fffffffffffffffc s6 000000007fffffff s7 0000000000000002 s8 9000100089cd9400 ra: 90000000043a315c tcp_bpf_sendmsg+0x23c/0x420 ERA: 900000000426cd1c sk_msg_memcopy_from_iter+0xbc/0x220 CRMD: 000000b0 (PLV0 -IE -DA +PG DACF=CC DACM=CC -WE) PRMD: 0000000c (PPLV0 +PIE +PWE) EUEN: 00000007 (+FPE +SXE +ASXE -BTE) ECFG: 00071c1d (LIE=0,2-4,10-12 VS=7) ESTAT: 00010000 [PIL] (IS= ECode=1 EsubCode=0) BADV: 0000000000000040 PRID: 0014c011 (Loongson-64bit, Loongson-3C5000) Modules linked in: tls xt_CHECKSUM xt_MASQUERADE xt_conntrack ipt_REJECT Process test_sockmap (pid: 23245, threadinfo=00000000aeb68043, task=00000000781bb2f1) Stack : 0000000000000000 900010008bfbfe20 0000000000000000 0000000000000003 0000000000000000 900010008bfbf94c 900010008bfbf950 0000000000000000 0000000000000003 0000000000000003 900010008bfbfe10 900010008beeb400 9000100089cd9400 0000000000000003 900010008b9bbc00 90000000043a315c 0000000000084000 900010008bfbfe20 900010008bfbf958 900010008beeb5ac 900010087fffd500 0000000000000000 7fffffffffffffff 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 0000000000000000 ... Call Trace: [<900000000426cd1c>] sk_msg_memcopy_from_iter+0xbc/0x220 [<90000000043a315c>] tcp_bpf_sendmsg+0x23c/0x420 [<90000000041cafc8>] __sock_sendmsg+0x68/0xe0 [<90000000041cc4bc>] ____sys_sendmsg+0x2bc/0x360 [<90000000041cea18>] ___sys_sendmsg+0xb8/0x120 [<90000000041cf1f8>] __sys_sendmsg+0x98/0x100 [<90000000045b76ec>] do_syscall+0x8c/0xc0 [<90000000030e1da4>] handle_syscall+0xc4/0x160
Code: 001532f7 0014f210 001036ed <28c10204> 298043ed 28c8632c 0010c5ef 0010bc84 0014ed8c
---[ end trace 0000000000000000 ]--- '''
This is because "sg_page(sge)" is NULL in that case. This patch adds null check for it in sk_msg_memcopy_from_iter() to fix this error.
Fixes: 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface") Signed-off-by: Geliang Tang tanggeliang@kylinos.cn --- net/core/skmsg.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/net/core/skmsg.c b/net/core/skmsg.c index bafcc1e2eadf..495b18b5dce5 100644 --- a/net/core/skmsg.c +++ b/net/core/skmsg.c @@ -375,6 +375,8 @@ int sk_msg_memcopy_from_iter(struct sock *sk, struct iov_iter *from,
do { sge = sk_msg_elem(msg, i); + if (!sg_page(sge)) + goto out; /* This is possible if a trim operation shrunk the buffer */ if (msg->sg.copybreak >= sge->length) { msg->sg.copybreak = 0;