Re: pgsql: Use slots in trigger infrastructure, except for theactual invoc

Hi,

On 2019-02-27 11:35:18 -0500, Tom Lane wrote:
> Andres Freund <andres@anarazel.de> writes:
> > Use slots in trigger infrastructure, except for the actual invocation.
> 
> I believe it's this commit that is resulting in my compiler bleating
> about
> 
> trigger.c: In function 'afterTriggerInvokeEvents':
> trigger.c:4493: warning: 'rInfo' may be used uninitialized in this function

Hm, yea, I can see why a compiler, especially without doing more
expensive control flow analysis, would get this wrong. Easier to
understand if we NULL initialize rInfo, not just rel, too. Pushed.

Greetings,

Andres Freund