[prev in list] [next in list] [prev in thread] [next in thread] 

List:       glibc-help
Subject:    tst-cancel4 fails at pselect
From:       ib () wupperonline ! de (Ingo Brueckl)
Date:       2010-03-29 12:47:00
Message-ID: 4bb0af0c () wupperonline ! de
[Download RAW message or body]

I get an error with tst-cancel4 during 'make check', because pselect returns
with zero:

  in-time cancel test of 'read' successful
  in-time cancel test of 'readv' successful
  in-time cancel test of 'select' successful
  tf_pselect: pselect returns with 0 (Success)

Why might this be? What can I do to get 'make check' run successfully?

Most interesting about this is that tst-cancel4 only fails most of the time.
Very rarely it runs successfully:

  in-time cancel test of 'read' successful
  in-time cancel test of 'readv' successful
  in-time cancel test of 'select' successful
  in-time cancel test of 'pselect' successful
  in-time cancel test of 'poll' successful
  in-time cancel test of 'ppoll' successful
  in-time cancel test of 'write' successful
  in-time cancel test of 'writev' successful
  in-time cancel test of 'sleep' successful
  in-time cancel test of 'usleep' successful
  in-time cancel test of 'nanosleep' successful
  in-time cancel test of 'wait' successful
  in-time cancel test of 'waitid' successful
  in-time cancel test of 'waitpid' successful
  in-time cancel test of 'sigpause' successful
  in-time cancel test of 'sigsuspend' successful
  in-time cancel test of 'sigwait' successful
  in-time cancel test of 'sigwaitinfo' successful
  in-time cancel test of 'sigtimedwait' successful
  in-time cancel test of 'pause' successful
  in-time cancel test of 'accept' successful
  in-time cancel test of 'send' successful
  in-time cancel test of 'recv' successful
  in-time cancel test of 'recvfrom' successful
  in-time cancel test of 'recvmsg' successful
  in-time cancel test of 'msgrcv' successful
  early cancel test of 'read' successful
  early cancel test of 'readv' successful
  early cancel test of 'select' successful
  early cancel test of 'pselect' successful
  early cancel test of 'poll' successful
  early cancel test of 'ppoll' successful
  early cancel test of 'write' successful
  early cancel test of 'writev' successful
  early cancel test of 'sleep' successful
  early cancel test of 'usleep' successful
  early cancel test of 'nanosleep' successful
  early cancel test of 'wait' successful
  early cancel test of 'waitid' successful
  early cancel test of 'waitpid' successful
  early cancel test of 'sigpause' successful
  early cancel test of 'sigsuspend' successful
  early cancel test of 'sigwait' successful
  early cancel test of 'sigwaitinfo' successful
  early cancel test of 'sigtimedwait' successful
  early cancel test of 'pause' successful
  early cancel test of 'accept' successful
  early cancel test of 'send' successful
  early cancel test of 'recv' successful
  early cancel test of 'recvfrom' successful
  early cancel test of 'recvmsg' successful
  early cancel test of 'open' successful
  early cancel test of 'close' successful
  early cancel test of 'pread' successful
  early cancel test of 'pwrite' successful
  early cancel test of 'fsync' successful
  early cancel test of 'fdatasync' successful
  early cancel test of 'msync' successful
  early cancel test of 'sendto' successful
  early cancel test of 'sendmsg' successful
  early cancel test of 'creat' successful
  early cancel test of 'connect' successful
  early cancel test of 'tcdrain' successful
  early cancel test of 'msgrcv' successful
  early cancel test of 'msgsnd' successful

Strange. Why does pselect behave this way?

Ingo


[prev in list] [next in list] [prev in thread] [next in thread]