diff options
| author | Jim Meyering <jim@meyering.net> | 2003-03-13 21:05:59 +0000 |
|---|---|---|
| committer | Jim Meyering <jim@meyering.net> | 2003-03-13 21:05:59 +0000 |
| commit | 689922fc0147b013a9ba00b1a93cd6bc0ec9795e (patch) | |
| tree | 4e73bad335bec82c01dda1ce11dee0d58fd67631 /tests/du/8gb | |
| parent | 419e2147999bba7b96a8375aab44e3ecf84fec7e (diff) | |
| download | coreutils-689922fc0147b013a9ba00b1a93cd6bc0ec9795e.tar.xz | |
Work around what appears to be an NFS failure that
would make this test fail on some systems.
Diffstat (limited to 'tests/du/8gb')
| -rwxr-xr-x | tests/du/8gb | 14 |
1 files changed, 14 insertions, 0 deletions
diff --git a/tests/du/8gb b/tests/du/8gb index bd6ffb3b8..140036996 100755 --- a/tests/du/8gb +++ b/tests/du/8gb @@ -30,6 +30,20 @@ if test $? != 0; then (exit 77); exit 77 fi +# FIXME: this should be a test of dd. +# On some systems (at least linux-2.4.18 + NFS to disks on a Solaris system) +# the `dd' command above mistakenly creates a file of length `0', yet +# doesn't fail. The root of that failure is that the ftruncate call +# returns zero but doesn't do its job. Detect this failure. +set x `ls -gG big` +size=$4 +if test "$size" = 0; then + echo "$0: cannot create a file large enough for this test;" 1>&2 + echo "$0: possibly because this system's NFS support is buggy;" 1>&2 + echo "$0: Consider rerunning this test on a different file system." 1>&2 + (exit 77); exit 77 +fi + fail=0 # This would print `0 big' with coreutils-4.5.8. |