Paul Eggert wrote: > I found this one by tracking down a bug with the latest snapshot > on Solaris. Still haven't got to the real bug yet, but this bug > was getting in the way. > > Subject: [PATCH] tests: set fail=0 by default > > * tests/init.sh (setup_): Set fail=0. This was the intent as per > <http://lists.gnu.org/archive/html/bug-coreutils/2010-12/msg00058.html> > but the assignment in mktempd_ is ineffective, since mktempd_ > is used inside `` and its assignments are in a subshell. > --- > tests/init.sh | 3 +++ > 1 files changed, 3 insertions(+), 0 deletions(-) > > diff --git a/tests/init.sh b/tests/init.sh > index a57de77..7be5e9e 100644 > --- a/tests/init.sh > +++ b/tests/init.sh > @@ -336,6 +336,9 @@ setup_() > for sig_ in 1 2 3 13 15; do > eval "trap 'Exit $(expr $sig_ + 128)' $sig_" > done > + > + # Set up for the "Exit $fail" at the end of many tests. > + fail=0
Thanks for spotting and fixing that. This is a new bug. test-lib.sh had a working fail=0 initialization since Oct 30: tests: factor 350 fail=0 initializations into test-lib.sh http://git.savannah.gnu.org/cgit/coreutils.git/commit/?id=3c88587b2eadee1 but then I converted all bourne shell test scripts to use init.sh rather than the aging test-lib.sh, and init.sh did not have that initialization. This change probably belongs in gnulib's tests/init.sh, too.