> On Fri, Aug 15, 2025 at 10:36:47AM -0500, Sami Imseih wrote:
> > But that could potentially be dangerous if code in the startup hook gets
> > re-executed? I guess the doc below is giving a vague warning that one
> > should be careful what they put in that hook.
>
> The docs seem reasonably clear that these hooks need to be careful to not
> re-initialize shared memory that was already initialized by another backend
> [0].
>
> > Thanks, I missed the doc update. Yes, that is inconsistent between platforms,
> > and if we must live with this behavior, should the doc give a bigger warning
> > about the code that goes in that hook?
>
> The docs should definitely be updated for accuracy, but I'm not following
> what sort of additional warning you think we need. Could you share a
> concrete example of what you have in mind?
I noticed a few things where this behavior becomes very suspicious.
For example, in pgss_startup_hook, every time startup_hook is run
we take an exclusive LW lock. so, all backends now may be competing
for that lock by nature of connection establishment.
test_slru calls LWLockNewTrancheId inside that hook.
So, this tells me that the caller needs to be aware of such caveats.
I am thinking something like this:
"Because this hook is executed by the postmaster and invoked by backends via
EXEC_BACKEND, it is essential to ensure that any code intended to run only
during postmaster startup is properly protected against repeated execution.
This can be enforced by verifying !IsUnderPostmaster before invocation."
--
Sami
