| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Michael Paquier <michael(at)paquier(dot)xyz> |
| Cc: | "Hsu, John" <hsuchen(at)amazon(dot)com>, "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>, Peter Eisentraut <peter(dot)eisentraut(at)2ndquadrant(dot)com> |
| Subject: | Re: Include RELKIND_TOASTVALUE in get_relkind_objtype |
| Date: | 2019-11-04 20:31:27 |
| Message-ID: | 14056.1572899487@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Michael Paquier <michael(at)paquier(dot)xyz> writes:
> Okay. Attached is what I was thinking about, with extra regression
> tests to cover the ground for toast tables and indexes that are able
> to reproduce the original failure, and more comments for the routines
> as they should be used only for ACL error messages.
I'd rather do something like the attached, which makes it more of an
explicit goal that we won't fail on bad input. (As written, we'd only
fail on bad classId, which is a case that really shouldn't happen.)
Tests are the same as yours, but I revised the commentary and got
rid of the elog-for-bad-relkind. I also made some cosmetic changes
in commands/alter.c, so as to (1) make it clear by inspection that
those calls are only used to feed aclcheck_error, and (2) avoid
uselessly computing a value that we won't need in normal non-error
cases.
regards, tom lane
| Attachment | Content-Type | Size |
|---|---|---|
| toast-acl-errors-2.patch | text/x-diff | 5.8 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Thomas Munro | 2019-11-04 20:34:52 | Re: 64 bit transaction id |
| Previous Message | Stephen Frost | 2019-11-04 20:12:05 | Re: cost based vacuum (parallel) |