Horton

Re: HOT chain validation in verify_heapam()

Robert Haas <robertmhaas@gmail.com>

View thread View on postgr.es
From: Robert Haas <robertmhaas@gmail.com>
To: Himanshu Upadhyaya <upadhyaya.himanshu@gmail.com>
Cc: Aleksander Alekseev <aleksander@timescale.com>, PostgreSQL Hackers <pgsql-hackers@postgresql.org>, Andres Freund <andres@anarazel.de>
Date: 2023-01-24T19:18:02Z
Lists: pgsql-hackers 
On Sun, Jan 22, 2023 at 10:19 AM Himanshu Upadhyaya
<upadhyaya.himanshu@gmail.com> wrote:
> I was trying to use lp_valid as I need to identify the root of the HOT chain and we are doing validation on the root of the HOT chain when we loop over the predecessor array.
> Was resetting lp_valid in the last patch because we don't add data to predecessor[] and while looping over the predecessor array we need to isolate (and identify) all cases of missing data in the predecessor array to exactly identify the root of HOT chain.
> One solution is to always add data to predecessor array while looping over successor array and then while looping over predecessor array we can continue for other validation "if (lp_valid [predecessor[currentoffnum]] && lp_valid[currentoffnum]" is true but in this case also our third loop will also look at lp_valid[].

I don't mind if the third loop looks at lp_valid if it has a reason to
do that, but I don't think we should be resetting values from true to
false. Once we know a line pointer to be valid, it doesn't stop being
valid later because we found out some other thing about something
else.

-- 
Robert Haas
EDB: http://www.enterprisedb.com

Commits

  1. amcheck: Generalize one of the recently-added update chain checks.

  2. amcheck: Tighten up validation of redirect line pointers.

  3. amcheck: Fix verify_heapam for tuples where xmin or xmax is 0.

  4. amcheck: Fix a few bugs in new update chain validation.

  5. Fix new test case to work on (some?) big-endian architectures.

  6. Don't test HEAP_XMAX_INVALID when freezing xmax.