Обсуждение: Re: [committers] pgsql: RLS fixes, new hooks, and new test module
* Stephen Frost wrote: > RLS fixes, new hooks, and new test module The buildfarm says that with -DCLOBBER_CACHE_ALWAYS, the RLS violations get blamed on the wrong tables. Mostly, they are catalogs (I have seen pg_opclass, pg_am, and pg_amproc), but some also come up with binary garbage instead, e.g. - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=markhor&dt=2015-04-23%2000%3A00%3A12 - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=friarbird&dt=2015-04-23%2004%3A20%3A00 - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=tick&dt=2015-04-22%2019%3A56%3A53 -- Christian
Christian, * Christian Ullrich (chris@chrullrich.net) wrote: > * Stephen Frost wrote: > > >RLS fixes, new hooks, and new test module > > The buildfarm says that with -DCLOBBER_CACHE_ALWAYS, the RLS > violations get blamed on the wrong tables. Mostly, they are catalogs > (I have seen pg_opclass, pg_am, and pg_amproc), but some also come > up with binary garbage instead, e.g. > > - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=markhor&dt=2015-04-23%2000%3A00%3A12 > - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=friarbird&dt=2015-04-23%2004%3A20%3A00 > - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=tick&dt=2015-04-22%2019%3A56%3A53 Yup, thanks, Robert pointed this out on another thread and I'm looking into it. Thanks again! Stephen
On 24 April 2015 at 04:31, Stephen Frost <sfrost@snowman.net> wrote: > Christian, > > * Christian Ullrich (chris@chrullrich.net) wrote: >> * Stephen Frost wrote: >> >> >RLS fixes, new hooks, and new test module >> >> The buildfarm says that with -DCLOBBER_CACHE_ALWAYS, the RLS >> violations get blamed on the wrong tables. Mostly, they are catalogs >> (I have seen pg_opclass, pg_am, and pg_amproc), but some also come >> up with binary garbage instead, e.g. >> >> - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=markhor&dt=2015-04-23%2000%3A00%3A12 >> - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=friarbird&dt=2015-04-23%2004%3A20%3A00 >> - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=tick&dt=2015-04-22%2019%3A56%3A53 > > Yup, thanks, Robert pointed this out on another thread and I'm looking > into it. > Ah, yes it looks like a couple of places in get_row_security_policies() ought to be making a copy of the relation name when setting it on the WCO. Regards, Dean
Dean, * Dean Rasheed (dean.a.rasheed@gmail.com) wrote: > On 24 April 2015 at 04:31, Stephen Frost <sfrost@snowman.net> wrote: > > * Christian Ullrich (chris@chrullrich.net) wrote: > >> * Stephen Frost wrote: > >> >RLS fixes, new hooks, and new test module > >> > >> The buildfarm says that with -DCLOBBER_CACHE_ALWAYS, the RLS > >> violations get blamed on the wrong tables. Mostly, they are catalogs > >> (I have seen pg_opclass, pg_am, and pg_amproc), but some also come > >> up with binary garbage instead, e.g. > >> > >> - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=markhor&dt=2015-04-23%2000%3A00%3A12 > >> - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=friarbird&dt=2015-04-23%2004%3A20%3A00 > >> - http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=tick&dt=2015-04-22%2019%3A56%3A53 > > > > Yup, thanks, Robert pointed this out on another thread and I'm looking > > into it. > > > > Ah, yes it looks like a couple of places in > get_row_security_policies() ought to be making a copy of the relation > name when setting it on the WCO. Yup, that was the same conclusion that I came to last night, just hadn't finished testing before needing sleep. :) Will push a fix shortly for it. Thanks! Stephen