Re: CVE-2017-7484-induced bugs, or, btree cmp functions are not leakproof? - Mailing list pgsql-hackers

From Amit Langote
Subject Re: CVE-2017-7484-induced bugs, or, btree cmp functions are not leakproof?
Date
Msg-id CA+HiwqEvB-JELhjPgwFNDq2hPEfTFUjUOsvJyf0S50vOLm2mAA@mail.gmail.com
Whole thread Raw
In response to Re: CVE-2017-7484-induced bugs, or, btree cmp functions are not leakproof?  (Dilip Kumar <dilipbalaut@gmail.com>)
Responses Re: CVE-2017-7484-induced bugs, or, btree cmp functions are not leakproof?  (Dilip Kumar <dilipbalaut@gmail.com>)
List pgsql-hackers
Hi Dilip,

On Wed, Jul 10, 2019 at 1:29 PM Dilip Kumar <dilipbalaut@gmail.com> wrote:
> On Wed, Jul 10, 2019 at 9:44 AM Dilip Kumar <dilipbalaut@gmail.com> wrote:
> > On Fri, Nov 2, 2018 at 1:34 PM Amit Langote wrote:
> > > Okay, here are two patches:
> > >
> > > 0001 adds a new RelOptInfo member inh_root_parent that's set for
> > > inheritance child otherrels and contains the RT index of the inheritance
> > > parent table mentioned in the query from which they originated.
> > >
> > > 0002 is your patch that modifies examine_variable, etc. to use the
> > > permissions granted on parent before reading stats on otherrel inheritance
> > > child tables. I've added your name as the author in the 2nd patch.
> > >
> >
> > I have looked into the patches and these look fine to me.  I have also
> > added it to the next commitfest.
> >
> Hi Amit,
>
> I have reviewed your 0001 patch and I think you have already taken a
> look on 0002.  So should I move it to "Ready for Committer" or you
> want to review it further?

Thanks for checking.  There has been a lot of churn in the inheritance
planning code since my last email on this thread, so I'd like to
reconsider.  I'm busy this week with some things, so I'll try posting
something on next Tuesday.

Thanks,
Amit



pgsql-hackers by date:

Previous
From: Michael Paquier
Date:
Subject: Re: PGOPTIONS="-fh" make check gets stuck since Postgres 11
Next
From: Michael Paquier
Date:
Subject: Re: Add parallelism and glibc dependent only options to reindexdb