On 03/09/2013 03:15 PM, Marc Girod wrote: Hi Marc,
thanks for reporting the test failures. > FAIL: tests/tail-2/F-vs-missing.sh (exit: 1) This one is a failure during the cleanup after the actual test ... > + cleanup_ > + : > + cd /proj/vobadm100/tmp/coreutils-8.21 > + chmod -R u+rwx /proj/vobadm100/tmp/coreutils-8.21/gt-F-vs-missing.sh.wlBi > + rm -rf /proj/vobadm100/tmp/coreutils-8.21/gt-F-vs-missing.sh.wlBi > rm: cannot remove > '/proj/vobadm100/tmp/coreutils-8.21/gt-F-vs-missing.sh.wlBi': File exists > + test 0 = 0 > + __st=1 > + exit 1 ... and it reminds me to a similar issue we recently fixed for another test: http://lists.gnu.org/archive/html/coreutils/2013-02/msg00039.html I suspect you're on NFS, and for some reason, the cleanup fails. The reason could be that the tail process didn't exit yet. Does the following additional 'wait' help? diff --git a/tests/tail-2/F-vs-missing.sh b/tests/tail-2/F-vs-missing.sh index 0f74404..22982b6 100755 --- a/tests/tail-2/F-vs-missing.sh +++ b/tests/tail-2/F-vs-missing.sh @@ -45,6 +45,7 @@ tail_re='has appeared' retry_delay_ check_tail_output .1 7 || { echo "$0: file: unexpected delay?"; cat out; fail=1; } kill -HUP $pid +wait cleanup() { > FAIL: tests/tail-2/F-vs-rename.sh (exit: 1) same as above. > FAIL: tests/cp/backup-dir.sh (exit: 1) > FAIL: tests/cp/preserve-link.sh (exit: 1) > FAIL: tests/cp/reflink-perm.sh (exit: 1) > FAIL: tests/cp/src-base-dot.sh (exit: 1) Somehow, preserving fails on that file system with "Operation not applicable". What file system type is it? You're not on MVFS, i.e. in a ClearCase dynamic view, are you? (BTW, I've never seen so many ClearCase-VOBs being cross-mounted, ~2000, wow!) > FAIL: tests/ls/stat-vs-dirent.sh (exit: 1) This one is also strange, and it indicates again that you're maybe using a strange file system type: > ++ stat --format=%i /proj/2gsim > + st_ino=17529 > + test -n 16609 > + test 16609 '!=' 17529 > + echo './tests/ls/stat-vs-dirent.sh: test failed: /proj/2gsim: d_ino(16609) > != st_ino(17529) > This may indicate a flaw in your kernel or file system implementation. > The flaw isn'\''t serious for coreutils, but it might break other > tools, > so you should report it to your operating system vendor.' Have a nice day, Berny
