Re: Failures in constraints regression test, "read only 0 of 8192 bytes"

On 3/8/24 21:29, Thomas Munro wrote:
> On Sat, Mar 9, 2024 at 2:36 AM Tomas Vondra
> <tomas(dot)vondra(at)enterprisedb(dot)com> wrote:
>> On 3/8/24 13:21, Tomas Vondra wrote:
>>> My guess would be 8af25652489, as it's the only storage-related commit.
>>>
>>> I'm currently running tests to verify this.
>>>
>>
>> Yup, the breakage starts with this commit. I haven't looked into the
>> root cause, or whether the commit maybe just made some pre-existing
>> issue easier to hit. Also, I haven't followed the discussion on the
>> pgsql-bugs thread [1], maybe there are some interesting findings.
>
> Adding Heikki.

I spent a bit of time investigating this today, but haven't made much
progress due to (a) my unfamiliarity with the smgr code in general and
the patch in particular, and (b) CLOBBER_CACHE_ALWAYS making it quite
time consuming to iterate and experiment.

However, the smallest schedule that still reproduces the issue is:

-------------------
test: test_setup

test: create_aggregate create_function_sql create_cast constraints
triggers select inherit typed_table vacuum drop_if_exists
updatable_views roleattributes create_am hash_func errors infinite_recurse
-------------------

I tried to reduce the second step to just "constraints", but that does
not fail. Clearly there's some concurrency at play, and having just a
single backend makes that go away.

regards

--
Tomas Vondra
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company