SSIA Reproducer: $ sudo strace -o /tmp/xxyz xrpm -qvv --requires --specfile time.spec ptrace: umoven: Input/output error ptrace: umoven: Input/output error ptrace: umoven: Input/output error /sbin/install-info /bin/sh /bin/sh (athough I've seen with every strace invocation in the last week or so) SELinux is disabled if that matters. $ rpm -qf /usr/bin/strace strace-4.5.14-2.i386.rpm
$ rpm -q glibc glibc-2.4.90-17.i386.rpm $ uname -a Linux wellfleet.jbj.org 2.6.17-1.2307_FC6PAE #1 SMP Fri Jun 23 17:10:16 EDT 2006 i686 i686 i386 GNU/ Linux
Please check the output. Instead of -o, use > file 2>&1 and then look at the part of the output near the "umoven" error message. Does it refer to an inappropriate syscall like sendto or another socket-related call that is not what the program is really doing? If so, this is the same as bug 202620, for which I already have a fix.
Here's the requested output: set_thread_area({entry_number:-1 -> 6, base_addr:0x9c47850, limit:1048575, seg_32bit:1, contents:0, read_exec_only:0, limit_in_pages:1, seg_not_present:0, useable:1}) = 0 set_tid_address(0x9c47898) = 31449 sendto(163870880, ptrace: umoven: Input/output error 0xc, 163870800, MSG_OOB|MSG_PROXY|MSG_FIN|MSG_SYN|0x20000, ptrace: umoven: Input/output error {...}, 3220895864) = 0 rt_sigaction(SIGRTMIN, {0x8450134, [], SA_SIGINFO}, NULL, 8) = 0 rt_sigaction(SIGRT_1, {0x8450080, [], SA_RESTART|SA_SIGINFO}, NULL, 8) = 0 rt_sigprocmask(SIG_UNBLOCK, [RTMIN RT_1], NULL, 8) = 0
This should be fixed in 4.5.14-3.