Re: [PATCH] Fix ProcKill lock-group vs procLatch recycle race - Mailing list pgsql-hackers

From Vlad Lesin
Subject Re: [PATCH] Fix ProcKill lock-group vs procLatch recycle race
Date
Msg-id 600f21bb-d6d1-48a5-85fb-117177b88c97@gmail.com
Whole thread
In response to Re: [PATCH] Fix ProcKill lock-group vs procLatch recycle race  (Michael Paquier <michael@paquier.xyz>)
List pgsql-hackers
On 5/28/26 04:08, Michael Paquier wrote:
> Using two different points, one for the leader and one for the
> follower is indeed the correct way to do things.
I am not sure about that. It is quite confusing to have several waiters 
when only one of them is awakened when calling the "wake up" function. 
If this logic is forbidden, it should be documented in the code 
comments, or an error should be thrown when attempting to add more than 
one waiter to the same injection point.

-- 
Best regards,
Vlad



pgsql-hackers by date:

Previous
From: Vlad Lesin
Date:
Subject: Re: [PATCH] Fix ProcKill lock-group vs procLatch recycle race
Next
From: Thom Brown
Date:
Subject: Re: Fix HAVING-to-WHERE pushdown with mismatched operator families