Hi Eric, * Eric Blake wrote on Wed, Jan 14, 2009 at 02:56:27AM CET: > According to Ralf Wildenhues on 1/13/2009 3:37 PM: > > I agree; I'll file a report once I'm through checking that it's not > > already reported (there are several 'set -e'-related bugs there). > > > > As to how to skip the test, how about this patch? > > Looks good; please apply.
Thanks, done (with s/4/3/g; that was a testing leftover; sorry). > By the way, does something like: > > trap 's=$? && set +e; echo $s; as_fn_exit' 0 > > behave any better? Maybe the problem is that when set -e is in effect, > and the first command of the EXIT trap is a shell builtin, that the shell > remembers the non-zero status inherited prior to entrance into the trap. Nope, that doesn't change things. Sorry, I should have said that I had tested && and if...fi as first lists of the trap. Cheers, Ralf
