Re: basebackup: do not verify checksums on pages written before enabling checksums - Mailing list pgsql-hackers

Hi,

On Tue, Aug 18, 2026 at 11:57:39AM +0200, Daniel Gustafsson wrote:
> > On 18 Aug 2026, at 06:18, Bertrand Drouvot <bertranddrouvot.pg@gmail.com> wrote:
> 
> > The test does not check that verification resumes while lastChecksumChangeRecPtr
> > is set. Just to be sure, did you skip that on purpose, or did you miss [1]?
> 
> Sorry, missed to mention that I skipped it on purpose since I don't really
> foresee it providing much value.  That being said, I have been wrong before so
> I ended up adding one in the attached v7 anyway.

Thanks for the update!

I've 2 comments:

=== 1

+  if (hit++ < 5)
+  {
+          if (IS_INJECTION_POINT_ATTACHED("basebackup-fail-checksum-verification"))
+          {


Shouldn't hit be incremented only when the injection point is attached?

=== 2

+  checksum ^= checksum;

That's same as "checksum =  0". That works because 0 is not a valid value but
was the intend to "checksum ^= 1"? to express that we want a different checksum?

Regards,

-- 
Bertrand Drouvot
PostgreSQL Contributors Team
RDS Open Source Databases
Amazon Web Services: https://aws.amazon.com



Attachment

pgsql-hackers by date:

Previous
From: Andrei Lepikhov
Date:
Subject: Re: Allow a prosupport function to be attached to an aggregate
Next
From: Daniel Gustafsson
Date:
Subject: Re: basebackup: do not verify checksums on pages written before enabling checksums