Re: Mark ItemPointer parameters as const in tuple/table lock functions

From: Peter Eisentraut <peter(at)eisentraut(dot)org>
To: Chao Li <li(dot)evan(dot)chao(at)gmail(dot)com>
Cc: Postgres hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: Re: Mark ItemPointer parameters as const in tuple/table lock functions
Date: 2025-08-29 05:42:41
Message-ID: 812f1ee6-eb2f-4793-92d2-98a800273245@eisentraut.org
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 28.08.25 04:27, Chao Li wrote:
>> On Aug 27, 2025, at 17:24, Peter Eisentraut <peter(at)eisentraut(dot)org> wrote:
>>
>> This style of having Foo be a type alias for pointer-to-FooData is an
>> ancient Postgres coding convention that does not map well to modern C
>> that has an emphasis on judicious use of qualifiers and attributes,
>> and so if this abstraction gets in the way, we sometimes crack it
>> open, like in the case of ItemPointerGetBlockNumber() etc.
>
> You are right, we want to protect the stuff that “tid” points to instead
> of “tid” itself:
>
>      *tid = something; // should hit compile error
>      tid = something; // ok
>
> Also, thanks for telling the history. I have updated the patch to use
> “const ItemPointerData *” in the same way as ItemPointerGetBlockNumber().
>
> Attached is the v2 patch.

This patch still causes a compiler warning:

../src/backend/storage/lmgr/lmgr.c: In function 'XactLockTableWait':
../src/backend/storage/lmgr/lmgr.c:681:27: error: assignment discards
'const' qualifier from pointer target type [-Werror=discarded-qualifiers]

I have fixed that and committed your patch.

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Ajin Cherian 2025-08-29 06:12:39 Re: Improve pg_sync_replication_slots() to wait for primary to advance
Previous Message Peter Eisentraut 2025-08-29 05:23:27 Re: Remove unneeded cast in heap_xlog_lock.