Re: Noah Misch 2014-06-08 <>
> Here's an update that places the socket in a temporary subdirectory of /tmp.
> The first attached patch adds NetBSD mkdtemp() to libpgport.  The second,
> principal, patch uses mkdtemp() to implement this design in pg_regress.  The
> corresponding change to contrib/pg_upgrade/ is based on the "configure"
> script's arrangements for its temporary directory.


I believe pg_upgrade itself still needs a fix. While it's not a
security problem to put the socket in $CWD while upgrading (it is
using -c unix_socket_permissions=0700), this behavior is pretty
unexpected, and does fail if your $CWD is > 107 bytes.

In f545d233ebce6971b6f9847680e48b679e707d22 Peter fixed the pg_ctl
perl tests to avoid that problem, so imho it would make even more
sense to fix pg_upgrade which could also fail in production.

This has been discussed here and elsewhere [1] before, but was
rejected as not being in line what the other utilities do, but now
pg_upgrade is the lone outlier. Noah's changes let Debian drop 4 out
of 5 pg_regress-sockdir patches, having this fixed would also let us
get rid of the last one [2].


-- |

Sent via pgsql-hackers mailing list (
To make changes to your subscription:

Reply via email to