Hi,
On 04/02/25 at 16:17 +0100, Christoph Berg wrote:
> Control: tags -1 unreproducible
>
> Re: Lucas Nussbaum
> > > # Failed test 'Action 'autovac_freeze' self-identifies correctly'
> > > # at t/02_autovac_freeze.t line 40.
> > > # 'uid=65534 topuid=10 euid=104 path='/tmp' at
> > > /usr/lib/x86_64-linux-gnu/perl-base/File/Temp.pm line 686.
> > > # File::Temp::_is_safe("/tmp", SCALAR(0x5621b5172be8)) called at
> > > /usr/lib/x86_64-linux-gnu/perl-base/File/Temp.pm line 472
> > > # File::Temp::_gettemp("/tmp/XXXXXXXXXX", "open", 0, "mkdir", 1,
> > > "suffixlen", 0, "ErrStr", ...) called at
> > > /usr/lib/x86_64-linux-gnu/perl-base/File/Temp.pm line 1712
> > > # File::Temp::tempdir("CLEANUP", 1) called at check_postgres.pl
> > > line 3033
> > > # check_postgres::run_command("SELECT freez, txns,
> > > ROUND(100*(txns/freez::float)) AS perc, d"..., HASH(0x5621b57c2458))
> > > called at check_postgres.pl line 4129
> > > # check_postgres::check_autovac_freeze() called at
> > > check_postgres.pl line 2764
> > > # Error in tempdir() using /tmp/XXXXXXXXXX: Parent directory (/tmp) is
> > > not safe (Directory owned neither by root nor the current user) at
> > > check_postgres.pl line 3033.
>
> This doesn't seem like a problem with check-postgres. Is /tmp botched
> in that build chroot?
I don't think so. There's another bug with /tmp strangeness (#1094100),
also raised by File::Temp.
Lucas
