Postgres Pro
Downloads
Home   >   mailing lists

Re:Why not set builder->next_phase_at=InvalidTransactionId in SnapBuildRestore function? - Mailing list pgsql-general

From zxwsbg
Subject Re:Why not set builder->next_phase_at=InvalidTransactionId in SnapBuildRestore function?
Date
Msg-id tencent_DD447707A5E37A1D6B1A428FCE4C1057EB06@qq.com
Whole thread Raw
In response to Why not set builder->next_phase_at=InvalidTransactionId in SnapBuildRestore function?  ("zxwsbg" <zxwsbg@qq.com>)
List pgsql-general
Tree view



I don't know what happened, but an image seems to be better.
------------------ Original ------------------
From: "zxwsbg" <zxwsbg@qq.com>;
Date: Thu, Nov 24, 2022 04:01 PM
To: "pgsql-general"<pgsql-general@postgresql.org>;
Subject: Why not set builder->next_phase_at=InvalidTransactionId in SnapBuildRestore function?

Last email has some problem, I resend one.

In PG14 or higher version, I notice that SnapBuildRestore function don't set builder->next_phase_at=InvalidTransactionId .

But in SnapBuildSerialize function, the assert check this condition.

Assert(builder->next_phase_at == InvalidTransactionId);

I'd like to know is this right?

Thanks.

Attachment

pgsql-general by date:

Previous
From: "zxwsbg"
Date:
Subject: Why not set builder->next_phase_at=InvalidTransactionId in SnapBuildRestore function?
Next
From: "Ryo Yamaji (Fujitsu)"
Date:
Subject: About row locking ordering