rpcbind: detect support of remote calls
Subsystems (2)
Hi Steve,
Ricardo found that TI-RPC rpc_pmap_rmtcall [1] tirpc_rpcb_rmtcall [2] tests are failing when they use rpcbind *without* --enable-rmtcalls (the default since 2018, see 2e9c289 ("rpcbind: Disable remote calls by default") [3]).
TL;DR: Is there a way to detect missing support from rpcbind? Because we cannot blindly expect that timeout means disabled remote calls (it could be also caused by regression). Other option is just to disable these tests by default (detection is preferred).
# export PATH="/opt/ltp/testcases/bin:$PATH" # rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... tirpc_rpcb_rmtcall 10.0.0.2 536875000 rpc_test 1 TFAIL: tirpc_rpcb_rmtcall 10.0.0.2 536875000 failed unexpectedly
As the name of the test suggests they are using pmap_rmtcall() and rpcb_rmtcall(). A bit debug info.
Modified rpc_test.sh to use strace:
+++ b/testcases/network/rpc/rpc-tirpc/rpc_test.sh
@@ -87,6 +87,8 @@ do_test()
done
fi
+ echo "$CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS" # FIXME: debug
+ EXPECT_RHOST_PASS strace -o /tmp/a $CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS
EXPECT_RHOST_PASS $CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS
}
I see the test timeouts (full strace output below):
# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall
...
sendto(5, "h=\r}\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\5\0\0\0\0\0\0\0\0"..., 60, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 60
poll([{fd=5, events=POLLIN}], 1, 1000) = 0 (Timeout)Using rpcbind 1.2.7-1.2 (from Tumbleweed), output when run with debug mode:
# /usr/sbin/rpcbind -w -f -d rpcbind: PMAPPROC_DUMP
rpcbind: RPCB_UNSET request for (536875000, 1, ) :
rpcbind: RPCB_UNSET: succeeded
rpcbind: RPCB_SET request for (536875000, 1, udp, 0.0.0.0.223.168) :
rpcbind: RPCB_SET: succeeded
rpcbind: RPCB_GETADDR req for (100000, 2, tcp) from 127.0.0.1.3.98:
mergeaddr: contact uaddr = 127.0.0.1.0.111 addrmerge(caller, 0.0.0.0.0.111, 127.0.0.1.0.111, tcp
addrmerge: hint 127.0.0.1.0.111
addrmerge: returning 127.0.0.1.0.111
mergeaddr: uaddr = 0.0.0.0.0.111, merged uaddr = 127.0.0.1.0.111
rpcbind: getaddr: 127.0.0.1.0.111
rpcbind: PMAPPROC_DUMP
rpcbind: RPCB_GETADDR req for (536875000, 1, udp) from 10.0.0.1.3.105:
mergeaddr: contact uaddr = 10.0.0.2.0.111
addrmerge(caller, 0.0.0.0.223.168, 10.0.0.2.0.111, udp
addrmerge: hint 10.0.0.2.0.111
addrmerge: returning 10.0.0.2.223.168
mergeaddr: uaddr = 0.0.0.0.223.168, merged uaddr = 10.0.0.2.223.168
rpcbind: getaddr: 10.0.0.2.223.168
rpcbind: RPCBPROC_BCAST
rpcbind: rpcb_indirect callit req for (536875000, 1, 1, udp) from 10.0.0.1.3.105 :
rpcbind: found at uaddr 0.0.0.0.223.168
addrmerge(caller, 0.0.0.0.223.168, NULL, udp
addrmerge: hint 127.0.0.1.0.111
addrmerge: returning 127.0.0.1.223.168
addrmerge(caller, 0.0.0.0.223.168, NULL, udp
addrmerge: hint 10.0.0.1.3.105
addrmerge: returning 192.168.122.43.223.168
rpcbind: merged uaddr 192.168.122.43.223.168
rpcbind: RPCB_UNSET request for (536875000, 1, ) :
rpcbind: Suppression RPC_UNSET(map_unset)
rpcbind: rbl->rpcb_map.r_owner=superuser
rpcbind: owner=superuser
rpcbind: RPCB_UNSET: succeeded
Obviously, if I compile rpcbind with --enable-rmtcalls and run it, both tests work:
$ ./autogen.sh && ./configure --enable-debug --enable-warmstarts --enable-rmtcalls --with-rpcuser=rpc --with-nss-modules="files usrfiles" $ make -j`nproc` # ./rpcbind -w -d -f
# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... rpc_test 1 TINFO: using libtirpc: yes tirpc_rpcb_rmtcall 10.0.0.2 536875000 rpc_test 1 TPASS: tirpc_rpcb_rmtcall 10.0.0.2 536875000 passed as expected
# rpc_test.sh -s rpc_svc_1 -c rpc_pmap_rmtcall ... rpc_pmap_rmtcall 10.0.0.2 536875000 rpc_test 1 TPASS: rpc_pmap_rmtcall 10.0.0.2 536875000 passed as expected
And the rpcbind outpt contains also:
rpcbind: rpcbproc_callit_com: original XID 683f1705, new XID f68e200
rpcbind: my_svc_run: polled on forwarding fd 7, netid udp - calling handle_reply
Also, wouldn't it be worth mention --enable-rmtcalls in functions' man pages? (Or have I overlooked that in man?)
Thanks for any hint.
Kind regards,
Petr
[1] https://github.com/linux-test-project/ltp/tree/master/testcases/network/rpc/rpc-tirpc/tests_pack/rpc_suite/rpc/rpc_addrmanagmt_pmap_rmtcall/rpc_pmap_rmtcall.c
[2] https://github.com/linux-test-project/ltp/tree/master/testcases/network/rpc/rpc-tirpc/tests_pack/rpc_suite/tirpc/tirpc_expertlevel_rpcb_rmtcall/tirpc_rpcb_rmtcall.c
[3] https://git.linux-nfs.org/?p=steved/rpcbind.git;a=commitdiff;h=2e9c289246c647e25649914bdb0d9400c66f486e
Full strace on rpcbind compiled without --enable-rmtcalls (the default, thus how it's shipped to the new distros):
# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall
execve("/opt/ltp/testcases/bin/tirpc_rpcb_rmtcall", ["tirpc_rpcb_rmtcall", "10.0.0.2", "536875000"], 0x7ffee8701b10 /* 228 vars */) = 0
...openat(AT_FDCWD, "/etc/services", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)openat(AT_FDCWD, "/etc/ld.so.cache", O_RDONLY|O_CLOEXEC) = 5
...openat(AT_FDCWD, "/usr/etc/services", O_RDONLY|O_CLOEXEC) = 5fstat(5, {st_mode=S_IFREG|0644, st_size=868338, ...}) = 0read(5, "#\n# Network services, Internet s"..., 4096) = 4096read(5, "[Jon_Postel]\ndaytime "..., 4096) = 4096read(5, "gs 44/udp # MPM F"..., 4096) = 4096read(5, "emote Job Service \nnetrjs-2 "..., 4096) = 4096read(5, "Jon_Postel]\nhostname 1"..., 4096) = 4096
close(5) = 0socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP) = 5getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16]) = 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4]) = 0openat(AT_FDCWD, "/etc/bindresvport.blacklist", O_RDONLY) = 6fstat(6, {st_mode=S_IFREG|0644, st_size=415, ...}) = 0read(6, "#\n# This file contains a list of"..., 4096) = 415read(6, "", 4096) = 0
close(6) = 0getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16]) = 0getpid() = 28530bind(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, 16) = 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0getsockname(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, [128 => 16]) = 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4]) = 0gettimeofday({tv_sec=1748867467, tv_usec=890549}, NULL) = 0getpid() = 28530setsockopt(5, SOL_IP, IP_RECVERR, [1], 4) = 0ioctl(5, FIONBIO, [1]) = 0
...rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8) = 0
sendto(5, "h0`M\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\3\0\0\0\0\0\0\0\0"..., 88, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 88 poll([{fd=5, events=POLLIN}], 1, 15000) = 1 ([{fd=5, revents=POLLIN}])
recvfrom(5, "h0`M\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\02010.0"..., 8800, 0, NULL, NULL) = 44rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8) = 0
close(5) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP) = 5getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16]) = 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4]) = 0getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16]) = 0getpid() = 28530bind(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, 16) = 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0getsockname(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, [128 => 16]) = 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4]) = 0gettimeofday({tv_sec=1748867467, tv_usec=892984}, NULL) = 0getpid() = 28530setsockopt(5, SOL_IP, IP_RECVERR, [1], 4) = 0ioctl(5, FIONBIO, [1]) = 0
...
sendto(5, "h0V\302\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\5\0\0\0\0\0\0\0\0"..., 60, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 60 poll([{fd=5, events=POLLIN}], 1, 1000) = 0 (Timeout)rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8) = 0
close(5) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0fstat(1, {st_mode=S_IFIFO|0600, st_size=0, ...}) = 0write(1, "1\n", 2) = 2exit_group(1) = ?
+++ exited with 1 +++
Hey!
Quoted 10 lines · Petr Vorel
Hi Steve,
Ricardo found that TI-RPC
rpc_pmap_rmtcall[1]tirpc_rpcb_rmtcall[2] tests are failing when they use rpcbind *without* --enable-rmtcalls (the default since 2018, see 2e9c289 ("rpcbind: Disable remote calls by default") [3]).TL;DR: Is there a way to detect missing support from rpcbind? Because we cannot blindly expect that timeout means disabled remote calls (it could be also caused by regression). Other option is just to disable these tests by default (detection is preferred).
No there is not a way, that I know of, to see if remote calls are or are not enabled since it is a compile time flag.
I really don't know what to say... In Fedora we disabled rmtcalls which broke NIS and in RHEL we left it enable which drives SELinux (and a few customers) nuts.
There is no clear answer... IMHO.
steved.
Quoted 183 lines
# export PATH="/opt/ltp/testcases/bin:$PATH" # rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... tirpc_rpcb_rmtcall 10.0.0.2 536875000 rpc_test 1 TFAIL: tirpc_rpcb_rmtcall 10.0.0.2 536875000 failed unexpectedlyAs the name of the test suggests they are using
pmap_rmtcall()andrpcb_rmtcall(). A bit debug info.Modified
rpc_test.shto use strace:+++
b/testcases/network/rpc/rpc-tirpc/rpc_test.sh(quoted diff) +2
-0@@ -87,6 +87,8 @@ do_test() done fi + echo "$CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS" # FIXME: debug + EXPECT_RHOST_PASS strace -o /tmp/a $CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS EXPECT_RHOST_PASS $CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS }I see the test timeouts (full strace output below):
# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... sendto(5, "h=\r}\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\5\0\0\0\0\0\0\0\0"..., 60, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 60 poll([{fd=5, events=POLLIN}], 1, 1000) = 0 (Timeout)Using rpcbind 1.2.7-1.2 (from Tumbleweed), output when run with debug mode:
# /usr/sbin/rpcbind -w -f -d rpcbind: PMAPPROC_DUMPrpcbind:
RPCB_UNSETrequest for (536875000, 1, ) :
rpcbind:RPCB_UNSET: succeeded
rpcbind:RPCB_SETrequest for (536875000, 1, udp, 0.0.0.0.223.168) :
rpcbind:RPCB_SET: succeeded
rpcbind:RPCB_GETADDRreq for (100000, 2, tcp) from 127.0.0.1.3.98:
mergeaddr: contact uaddr = 127.0.0.1.0.111 addrmerge(caller, 0.0.0.0.0.111, 127.0.0.1.0.111, tcp
addrmerge: hint 127.0.0.1.0.111
addrmerge: returning 127.0.0.1.0.111
mergeaddr: uaddr = 0.0.0.0.0.111, merged uaddr = 127.0.0.1.0.111
rpcbind: getaddr: 127.0.0.1.0.111
rpcbind:PMAPPROC_DUMPrpcbind:
RPCB_GETADDRreq for (536875000, 1, udp) from 10.0.0.1.3.105:
mergeaddr: contact uaddr = 10.0.0.2.0.111
addrmerge(caller, 0.0.0.0.223.168, 10.0.0.2.0.111, udp
addrmerge: hint 10.0.0.2.0.111
addrmerge: returning 10.0.0.2.223.168
mergeaddr: uaddr = 0.0.0.0.223.168, merged uaddr = 10.0.0.2.223.168
rpcbind: getaddr: 10.0.0.2.223.168
rpcbind:RPCBPROC_BCASTrpcbind:
rpcb_indirectcallit req for (536875000, 1, 1, udp) from 10.0.0.1.3.105 :
rpcbind: found at uaddr 0.0.0.0.223.168addrmerge(caller, 0.0.0.0.223.168, NULL, udp
addrmerge: hint 127.0.0.1.0.111
addrmerge: returning 127.0.0.1.223.168
addrmerge(caller, 0.0.0.0.223.168, NULL, udp
addrmerge: hint 10.0.0.1.3.105
addrmerge: returning 192.168.122.43.223.168
rpcbind: merged uaddr 192.168.122.43.223.168rpcbind:
RPCB_UNSETrequest for (536875000, 1, ) :
rpcbind: SuppressionRPC_UNSET(map_unset)
rpcbind: rbl->rpcb_map.r_owner=superuser
rpcbind: owner=superuser
rpcbind:RPCB_UNSET: succeededObviously, if I compile rpcbind with --enable-rmtcalls and run it, both tests work:
$ ./autogen.sh && ./configure --enable-debug --enable-warmstarts --enable-rmtcalls --with-rpcuser=rpc --with-nss-modules="files usrfiles" $ make -j`nproc` # ./rpcbind -w -d -f# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... rpc_test 1 TINFO: using libtirpc: yes tirpc_rpcb_rmtcall 10.0.0.2 536875000 rpc_test 1 TPASS: tirpc_rpcb_rmtcall 10.0.0.2 536875000 passed as expected# rpc_test.sh -s rpc_svc_1 -c rpc_pmap_rmtcall ... rpc_pmap_rmtcall 10.0.0.2 536875000 rpc_test 1 TPASS: rpc_pmap_rmtcall 10.0.0.2 536875000 passed as expectedAnd the rpcbind outpt contains also:
rpcbind:
rpcbproc_callit_com: original XID683f1705, new XID f68e200
rpcbind:my_svc_run: polled on forwarding fd 7, netid udp - callinghandle_replyAlso, wouldn't it be worth mention --enable-rmtcalls in functions' man pages? (Or have I overlooked that in man?)
Thanks for any hint.
Kind regards,
Petr[1] https://github.com/linux-test-project/ltp/tree/master/testcases/network/rpc/rpc-tirpc/tests_pack/rpc_suite/rpc/rpc_addrmanagmt_pmap_rmtcall/rpc_pmap_rmtcall.c
[2] https://github.com/linux-test-project/ltp/tree/master/testcases/network/rpc/rpc-tirpc/tests_pack/rpc_suite/tirpc/tirpc_expertlevel_rpcb_rmtcall/tirpc_rpcb_rmtcall.c
[3] https://git.linux-nfs.org/?p=steved/rpcbind.git;a=commitdiff;h=2e9c289246c647e25649914bdb0d9400c66f486eFull strace on rpcbind compiled without --enable-rmtcalls (the default, thus how it's shipped to the new distros):
# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcallexecve("/opt/ltp/testcases/bin/tirpc_rpcb_rmtcall", ["
tirpc_rpcb_rmtcall", "10.0.0.2", "536875000"],0x7ffee8701b10/* 228 vars */) = 0
...openat(AT_FDCWD, "/etc/services", O_RDONLY|O_CLOEXEC)= -1 ENOENT (No such file or directory)openat(AT_FDCWD, "/etc/ld.so.cache", O_RDONLY|O_CLOEXEC)= 5
...openat(AT_FDCWD, "/usr/etc/services", O_RDONLY|O_CLOEXEC)= 5fstat(5, {st_mode=S_IFREG|0644, st_size=868338, ...})= 0read(5, "#\n# Network services, Internet s"..., 4096)= 4096read(5, "[Jon_Postel]\ndaytime "..., 4096)= 4096read(5, "gs 44/udp # MPM F"..., 4096)= 4096read(5, "emote Job Service \nnetrjs-2 "..., 4096)= 4096read(5, "Jon_Postel]\nhostname 1"..., 4096)= 4096
close(5) = 0socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP)= 5getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0openat(AT_FDCWD, "/etc/bindresvport.blacklist", O_RDONLY)= 6fstat(6, {st_mode=S_IFREG|0644, st_size=415, ...})= 0read(6, "#\n# This file contains a list of"..., 4096)= 415read(6, "", 4096)= 0
close(6) = 0getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getpid()= 28530bind(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, 16)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0getsockname(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0gettimeofday({tv_sec=1748867467, tv_usec=890549}, NULL)= 0getpid()= 28530setsockopt(5, SOL_IP, IP_RECVERR, [1], 4)= 0ioctl(5, FIONBIO, [1])= 0
...rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0
sendto(5, "h0`M\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\3\0\0\0\0\0\0\0\0"..., 88, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 88poll([{fd=5, events=POLLIN}], 1, 15000)= 1 ([{fd=5, revents=POLLIN}])
recvfrom(5, "h0`M\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\02010.0"..., 8800, 0, NULL, NULL) = 44rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0
close(5) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP)= 5getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getpid()= 28530bind(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, 16)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0getsockname(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0gettimeofday({tv_sec=1748867467, tv_usec=892984}, NULL)= 0getpid()= 28530setsockopt(5, SOL_IP, IP_RECVERR, [1], 4)= 0ioctl(5, FIONBIO, [1])= 0
...
sendto(5, "h0V\302\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\5\0\0\0\0\0\0\0\0"..., 60, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 60poll([{fd=5, events=POLLIN}], 1, 1000)= 0 (Timeout)rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0
close(5) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0fstat(1, {st_mode=S_IFIFO|0600, st_size=0, ...})= 0write(1, "1\n", 2)= 2exit_group(1)= ?
+++ exited with 1 +++
Hi Steve,
Quoted 9 lines
Hey!
Quoted 5 lines · Petr Vorel
Hi Steve,
Ricardo found that TI-RPC
rpc_pmap_rmtcall[1]tirpc_rpcb_rmtcall[2] tests are failing when they use rpcbind *without* --enable-rmtcalls (the default since 2018, see 2e9c289 ("rpcbind: Disable remote calls by default") [3]).
Quoted 6 lines
Quoted 4 lines
TL;DR: Is there a way to detect missing support from rpcbind? Because we cannot blindly expect that timeout means disabled remote calls (it could be also caused by regression). Other option is just to disable these tests by default (detection is preferred).
No there is not a way, that I know of, to see if remote calls are or are not enabled since it is a compile time flag.
How about rpcbind to print version via new -v or -V command line option. Then compiled options could be printed as part of the output.
Nobody needed to know rpcbind version, therefore I'm not sure if testing justify it, but that would be an easy way to provide the info.
Quoted 3 lines
I really don't know what to say... In Fedora we disabled rmtcalls which broke NIS and in RHEL we left it enable which drives SELinux (and a few customers) nuts.
Thanks for info, I fully understand the chosen solution.
Kind regards,
Petr
Quoted 185 lines
There is no clear answer... IMHO.
steved.
Quoted 179 lines
# export PATH="/opt/ltp/testcases/bin:$PATH" # rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... tirpc_rpcb_rmtcall 10.0.0.2 536875000 rpc_test 1 TFAIL: tirpc_rpcb_rmtcall 10.0.0.2 536875000 failed unexpectedlyAs the name of the test suggests they are using
pmap_rmtcall()andrpcb_rmtcall(). A bit debug info.Modified
rpc_test.shto use strace:+++
b/testcases/network/rpc/rpc-tirpc/rpc_test.sh(quoted diff) +2
-0@@ -87,6 +87,8 @@ do_test() done fi + echo "$CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS" # FIXME: debug + EXPECT_RHOST_PASS strace -o /tmp/a $CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS EXPECT_RHOST_PASS $CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS }I see the test timeouts (full strace output below):
# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... sendto(5, "h=\r}\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\5\0\0\0\0\0\0\0\0"..., 60, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 60 poll([{fd=5, events=POLLIN}], 1, 1000) = 0 (Timeout)Using rpcbind 1.2.7-1.2 (from Tumbleweed), output when run with debug mode:
# /usr/sbin/rpcbind -w -f -d rpcbind: PMAPPROC_DUMPrpcbind:
RPCB_UNSETrequest for (536875000, 1, ) :
rpcbind:RPCB_UNSET: succeeded
rpcbind:RPCB_SETrequest for (536875000, 1, udp, 0.0.0.0.223.168) :
rpcbind:RPCB_SET: succeeded
rpcbind:RPCB_GETADDRreq for (100000, 2, tcp) from 127.0.0.1.3.98:
mergeaddr: contact uaddr = 127.0.0.1.0.111 addrmerge(caller, 0.0.0.0.0.111, 127.0.0.1.0.111, tcp
addrmerge: hint 127.0.0.1.0.111
addrmerge: returning 127.0.0.1.0.111
mergeaddr: uaddr = 0.0.0.0.0.111, merged uaddr = 127.0.0.1.0.111
rpcbind: getaddr: 127.0.0.1.0.111
rpcbind:PMAPPROC_DUMPrpcbind:
RPCB_GETADDRreq for (536875000, 1, udp) from 10.0.0.1.3.105:
mergeaddr: contact uaddr = 10.0.0.2.0.111
addrmerge(caller, 0.0.0.0.223.168, 10.0.0.2.0.111, udp
addrmerge: hint 10.0.0.2.0.111
addrmerge: returning 10.0.0.2.223.168
mergeaddr: uaddr = 0.0.0.0.223.168, merged uaddr = 10.0.0.2.223.168
rpcbind: getaddr: 10.0.0.2.223.168
rpcbind:RPCBPROC_BCASTrpcbind:
rpcb_indirectcallit req for (536875000, 1, 1, udp) from 10.0.0.1.3.105 :
rpcbind: found at uaddr 0.0.0.0.223.168addrmerge(caller, 0.0.0.0.223.168, NULL, udp
addrmerge: hint 127.0.0.1.0.111
addrmerge: returning 127.0.0.1.223.168
addrmerge(caller, 0.0.0.0.223.168, NULL, udp
addrmerge: hint 10.0.0.1.3.105
addrmerge: returning 192.168.122.43.223.168
rpcbind: merged uaddr 192.168.122.43.223.168rpcbind:
RPCB_UNSETrequest for (536875000, 1, ) :
rpcbind: SuppressionRPC_UNSET(map_unset)
rpcbind: rbl->rpcb_map.r_owner=superuser
rpcbind: owner=superuser
rpcbind:RPCB_UNSET: succeededObviously, if I compile rpcbind with --enable-rmtcalls and run it, both tests work:
$ ./autogen.sh && ./configure --enable-debug --enable-warmstarts --enable-rmtcalls --with-rpcuser=rpc --with-nss-modules="files usrfiles" $ make -j`nproc` # ./rpcbind -w -d -f# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... rpc_test 1 TINFO: using libtirpc: yes tirpc_rpcb_rmtcall 10.0.0.2 536875000 rpc_test 1 TPASS: tirpc_rpcb_rmtcall 10.0.0.2 536875000 passed as expected# rpc_test.sh -s rpc_svc_1 -c rpc_pmap_rmtcall ... rpc_pmap_rmtcall 10.0.0.2 536875000 rpc_test 1 TPASS: rpc_pmap_rmtcall 10.0.0.2 536875000 passed as expectedAnd the rpcbind outpt contains also:
rpcbind:
rpcbproc_callit_com: original XID683f1705, new XID f68e200
rpcbind:my_svc_run: polled on forwarding fd 7, netid udp - callinghandle_replyAlso, wouldn't it be worth mention --enable-rmtcalls in functions' man pages? (Or have I overlooked that in man?)
Thanks for any hint.
Kind regards,
Petr[1] https://github.com/linux-test-project/ltp/tree/master/testcases/network/rpc/rpc-tirpc/tests_pack/rpc_suite/rpc/rpc_addrmanagmt_pmap_rmtcall/rpc_pmap_rmtcall.c
[2] https://github.com/linux-test-project/ltp/tree/master/testcases/network/rpc/rpc-tirpc/tests_pack/rpc_suite/tirpc/tirpc_expertlevel_rpcb_rmtcall/tirpc_rpcb_rmtcall.c
[3] https://git.linux-nfs.org/?p=steved/rpcbind.git;a=commitdiff;h=2e9c289246c647e25649914bdb0d9400c66f486eFull strace on rpcbind compiled without --enable-rmtcalls (the default, thus how it's shipped to the new distros):
# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcallexecve("/opt/ltp/testcases/bin/tirpc_rpcb_rmtcall", ["
tirpc_rpcb_rmtcall", "10.0.0.2", "536875000"],0x7ffee8701b10/* 228 vars */) = 0
...openat(AT_FDCWD, "/etc/services", O_RDONLY|O_CLOEXEC)= -1 ENOENT (No such file or directory)openat(AT_FDCWD, "/etc/ld.so.cache", O_RDONLY|O_CLOEXEC)= 5
...openat(AT_FDCWD, "/usr/etc/services", O_RDONLY|O_CLOEXEC)= 5fstat(5, {st_mode=S_IFREG|0644, st_size=868338, ...})= 0read(5, "#\n# Network services, Internet s"..., 4096)= 4096read(5, "[Jon_Postel]\ndaytime "..., 4096)= 4096read(5, "gs 44/udp # MPM F"..., 4096)= 4096read(5, "emote Job Service \nnetrjs-2 "..., 4096)= 4096read(5, "Jon_Postel]\nhostname 1"..., 4096)= 4096
close(5) = 0socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP)= 5getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0openat(AT_FDCWD, "/etc/bindresvport.blacklist", O_RDONLY)= 6fstat(6, {st_mode=S_IFREG|0644, st_size=415, ...})= 0read(6, "#\n# This file contains a list of"..., 4096)= 415read(6, "", 4096)= 0
close(6) = 0getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getpid()= 28530bind(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, 16)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0getsockname(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0gettimeofday({tv_sec=1748867467, tv_usec=890549}, NULL)= 0getpid()= 28530setsockopt(5, SOL_IP, IP_RECVERR, [1], 4)= 0ioctl(5, FIONBIO, [1])= 0
...rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0
sendto(5, "h0`M\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\3\0\0\0\0\0\0\0\0"..., 88, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 88poll([{fd=5, events=POLLIN}], 1, 15000)= 1 ([{fd=5, revents=POLLIN}])
recvfrom(5, "h0`M\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\02010.0"..., 8800, 0, NULL, NULL) = 44rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0
close(5) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP)= 5getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getpid()= 28530bind(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, 16)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0getsockname(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0gettimeofday({tv_sec=1748867467, tv_usec=892984}, NULL)= 0getpid()= 28530setsockopt(5, SOL_IP, IP_RECVERR, [1], 4)= 0ioctl(5, FIONBIO, [1])= 0
...
sendto(5, "h0V\302\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\5\0\0\0\0\0\0\0\0"..., 60, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 60poll([{fd=5, events=POLLIN}], 1, 1000)= 0 (Timeout)rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0
close(5) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0fstat(1, {st_mode=S_IFIFO|0600, st_size=0, ...})= 0write(1, "1\n", 2)= 2exit_group(1)= ?
+++ exited with 1 +++
Quoted 15 lines · Petr Vorel
Hi Steve,
Quoted 13 lines
Hey!
Quoted 10 lines · Petr Vorel
Hi Steve,
Ricardo found that TI-RPC
rpc_pmap_rmtcall[1]tirpc_rpcb_rmtcall[2] tests are failing when they use rpcbind *without* --enable-rmtcalls (the default since 2018, see 2e9c289 ("rpcbind: Disable remote calls by default") [3]).TL;DR: Is there a way to detect missing support from rpcbind? Because we cannot blindly expect that timeout means disabled remote calls (it could be also caused by regression). Other option is just to disable these tests by default (detection is preferred).
Quoted 5 lines
Quoted 2 lines
No there is not a way, that I know of, to see if remote calls are or are not enabled since it is a compile time flag.
How about rpcbind to print version via new -v or -V command line option. Then compiled options could be printed as part of the output.
That is not a bad idea... It could be a bit messy since they're close to 20 different ifdefs...
Quoted 3 lines
Nobody needed to know rpcbind version, therefore I'm not sure if testing justify it, but that would be an easy way to provide the info.
No testing would be needed... as long as rpcbind does not drop core with the -v option :-) also there would be a manpage up date... but it would not be very invasive.
Quoted 6 lines
Quoted 4 lines
I really don't know what to say... In Fedora we disabled rmtcalls which broke NIS and in RHEL we left it enable which drives SELinux (and a few customers) nuts.
Thanks for info, I fully understand the chosen solution.
I wish other people would!!! 8-)
steved.
Quoted 191 lines
Kind regards,
PetrQuoted 185 lines
There is no clear answer... IMHO.
steved.
Quoted 179 lines
# export PATH="/opt/ltp/testcases/bin:$PATH" # rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... tirpc_rpcb_rmtcall 10.0.0.2 536875000 rpc_test 1 TFAIL: tirpc_rpcb_rmtcall 10.0.0.2 536875000 failed unexpectedlyAs the name of the test suggests they are using
pmap_rmtcall()andrpcb_rmtcall(). A bit debug info.Modified
rpc_test.shto use strace:+++
b/testcases/network/rpc/rpc-tirpc/rpc_test.sh(quoted diff) +2
-0@@ -87,6 +87,8 @@ do_test() done fi + echo "$CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS" # FIXME: debug + EXPECT_RHOST_PASS strace -o /tmp/a $CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS EXPECT_RHOST_PASS $CLIENT $(tst_ipaddr) $PROGNUMNOSVC $CLIENT_EXTRA_OPTS }I see the test timeouts (full strace output below):
# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... sendto(5, "h=\r}\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\5\0\0\0\0\0\0\0\0"..., 60, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 60 poll([{fd=5, events=POLLIN}], 1, 1000) = 0 (Timeout)Using rpcbind 1.2.7-1.2 (from Tumbleweed), output when run with debug mode:
# /usr/sbin/rpcbind -w -f -d rpcbind: PMAPPROC_DUMPrpcbind:
RPCB_UNSETrequest for (536875000, 1, ) :
rpcbind:RPCB_UNSET: succeeded
rpcbind:RPCB_SETrequest for (536875000, 1, udp, 0.0.0.0.223.168) :
rpcbind:RPCB_SET: succeeded
rpcbind:RPCB_GETADDRreq for (100000, 2, tcp) from 127.0.0.1.3.98:
mergeaddr: contact uaddr = 127.0.0.1.0.111 addrmerge(caller, 0.0.0.0.0.111, 127.0.0.1.0.111, tcp
addrmerge: hint 127.0.0.1.0.111
addrmerge: returning 127.0.0.1.0.111
mergeaddr: uaddr = 0.0.0.0.0.111, merged uaddr = 127.0.0.1.0.111
rpcbind: getaddr: 127.0.0.1.0.111
rpcbind:PMAPPROC_DUMPrpcbind:
RPCB_GETADDRreq for (536875000, 1, udp) from 10.0.0.1.3.105:
mergeaddr: contact uaddr = 10.0.0.2.0.111
addrmerge(caller, 0.0.0.0.223.168, 10.0.0.2.0.111, udp
addrmerge: hint 10.0.0.2.0.111
addrmerge: returning 10.0.0.2.223.168
mergeaddr: uaddr = 0.0.0.0.223.168, merged uaddr = 10.0.0.2.223.168
rpcbind: getaddr: 10.0.0.2.223.168
rpcbind:RPCBPROC_BCASTrpcbind:
rpcb_indirectcallit req for (536875000, 1, 1, udp) from 10.0.0.1.3.105 :
rpcbind: found at uaddr 0.0.0.0.223.168addrmerge(caller, 0.0.0.0.223.168, NULL, udp
addrmerge: hint 127.0.0.1.0.111
addrmerge: returning 127.0.0.1.223.168
addrmerge(caller, 0.0.0.0.223.168, NULL, udp
addrmerge: hint 10.0.0.1.3.105
addrmerge: returning 192.168.122.43.223.168
rpcbind: merged uaddr 192.168.122.43.223.168rpcbind:
RPCB_UNSETrequest for (536875000, 1, ) :
rpcbind: SuppressionRPC_UNSET(map_unset)
rpcbind: rbl->rpcb_map.r_owner=superuser
rpcbind: owner=superuser
rpcbind:RPCB_UNSET: succeededObviously, if I compile rpcbind with --enable-rmtcalls and run it, both tests work:
$ ./autogen.sh && ./configure --enable-debug --enable-warmstarts --enable-rmtcalls --with-rpcuser=rpc --with-nss-modules="files usrfiles" $ make -j`nproc` # ./rpcbind -w -d -f# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcall ... rpc_test 1 TINFO: using libtirpc: yes tirpc_rpcb_rmtcall 10.0.0.2 536875000 rpc_test 1 TPASS: tirpc_rpcb_rmtcall 10.0.0.2 536875000 passed as expected# rpc_test.sh -s rpc_svc_1 -c rpc_pmap_rmtcall ... rpc_pmap_rmtcall 10.0.0.2 536875000 rpc_test 1 TPASS: rpc_pmap_rmtcall 10.0.0.2 536875000 passed as expectedAnd the rpcbind outpt contains also:
rpcbind:
rpcbproc_callit_com: original XID683f1705, new XID f68e200
rpcbind:my_svc_run: polled on forwarding fd 7, netid udp - callinghandle_replyAlso, wouldn't it be worth mention --enable-rmtcalls in functions' man pages? (Or have I overlooked that in man?)
Thanks for any hint.
Kind regards,
Petr[1] https://github.com/linux-test-project/ltp/tree/master/testcases/network/rpc/rpc-tirpc/tests_pack/rpc_suite/rpc/rpc_addrmanagmt_pmap_rmtcall/rpc_pmap_rmtcall.c
[2] https://github.com/linux-test-project/ltp/tree/master/testcases/network/rpc/rpc-tirpc/tests_pack/rpc_suite/tirpc/tirpc_expertlevel_rpcb_rmtcall/tirpc_rpcb_rmtcall.c
[3] https://git.linux-nfs.org/?p=steved/rpcbind.git;a=commitdiff;h=2e9c289246c647e25649914bdb0d9400c66f486eFull strace on rpcbind compiled without --enable-rmtcalls (the default, thus how it's shipped to the new distros):
# rpc_test.sh -s tirpc_svc_4 -c tirpc_rpcb_rmtcallexecve("/opt/ltp/testcases/bin/tirpc_rpcb_rmtcall", ["
tirpc_rpcb_rmtcall", "10.0.0.2", "536875000"],0x7ffee8701b10/* 228 vars */) = 0
...openat(AT_FDCWD, "/etc/services", O_RDONLY|O_CLOEXEC)= -1 ENOENT (No such file or directory)openat(AT_FDCWD, "/etc/ld.so.cache", O_RDONLY|O_CLOEXEC)= 5
...openat(AT_FDCWD, "/usr/etc/services", O_RDONLY|O_CLOEXEC)= 5fstat(5, {st_mode=S_IFREG|0644, st_size=868338, ...})= 0read(5, "#\n# Network services, Internet s"..., 4096)= 4096read(5, "[Jon_Postel]\ndaytime "..., 4096)= 4096read(5, "gs 44/udp # MPM F"..., 4096)= 4096read(5, "emote Job Service \nnetrjs-2 "..., 4096)= 4096read(5, "Jon_Postel]\nhostname 1"..., 4096)= 4096
close(5) = 0socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP)= 5getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0openat(AT_FDCWD, "/etc/bindresvport.blacklist", O_RDONLY)= 6fstat(6, {st_mode=S_IFREG|0644, st_size=415, ...})= 0read(6, "#\n# This file contains a list of"..., 4096)= 415read(6, "", 4096)= 0
close(6) = 0getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getpid()= 28530bind(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, 16)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0getsockname(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0gettimeofday({tv_sec=1748867467, tv_usec=890549}, NULL)= 0getpid()= 28530setsockopt(5, SOL_IP, IP_RECVERR, [1], 4)= 0ioctl(5, FIONBIO, [1])= 0
...rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0
sendto(5, "h0`M\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\3\0\0\0\0\0\0\0\0"..., 88, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 88poll([{fd=5, events=POLLIN}], 1, 15000)= 1 ([{fd=5, revents=POLLIN}])
recvfrom(5, "h0`M\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\02010.0"..., 8800, 0, NULL, NULL) = 44rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0
close(5) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP)= 5getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0getsockname(5, {sa_family=AF_INET, sin_port=htons(0), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getpid()= 28530bind(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, 16)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0getsockname(5, {sa_family=AF_INET, sin_port=htons(722), sin_addr=inet_addr("0.0.0.0")}, [128 => 16])= 0getsockopt(5, SOL_SOCKET, SO_TYPE, [2], [4])= 0gettimeofday({tv_sec=1748867467, tv_usec=892984}, NULL)= 0getpid()= 28530setsockopt(5, SOL_IP, IP_RECVERR, [1], 4)= 0ioctl(5, FIONBIO, [1])= 0
...
sendto(5, "h0V\302\0\0\0\0\0\0\0\2\0\1\206\240\0\0\0\4\0\0\0\5\0\0\0\0\0\0\0\0"..., 60, 0, {sa_family=AF_INET, sin_port=htons(111), sin_addr=inet_addr("10.0.0.2")}, 16) = 60poll([{fd=5, events=POLLIN}], 1, 1000)= 0 (Timeout)rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0rt_sigprocmask(SIG_SETMASK, ~[RTMIN RT_1], [], 8)= 0
close(5) = 0rt_sigprocmask(SIG_SETMASK, [], NULL, 8)= 0fstat(1, {st_mode=S_IFIFO|0600, st_size=0, ...})= 0write(1, "1\n", 2)= 2exit_group(1)= ?
+++ exited with 1 +++