Re: Query is over 2x slower with jit=on

From: Andres Freund <andres(at)anarazel(dot)de>
To: Pierre Ducroquet <p(dot)psql(at)pinaraf(dot)info>
Cc: pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: Query is over 2x slower with jit=on
Date: 2018-08-22 19:02:51
Views: Raw Message | Whole Thread | Download mbox | Resend email
Lists: pgsql-hackers

On 2018-08-22 20:48:48 +0200, Pierre Ducroquet wrote:
> On Wednesday, August 22, 2018 6:51:55 PM CEST Andres Freund wrote:
> > On 2018-08-22 18:39:18 +0200, Andreas Joseph Krogh wrote:
> > > Just to be clear; The query really runs slower (wall-clock time), it's not
> > > just the timing.
> >
> > I bet it's not actually running slower, it "just" takes longer to start
> > up due to the JITing in each worker. I suspect what we should do is to
> > multiple the cost limits by the number of workers, to model that. But
> > without the fixed instrumentation that's harder to see...
> It depends on the query. It has been shown in other threads that query can
> indeed take longer to run because of JITing : if the cost is too low to fire
> LLVM optimizer, the generated code can be so bad it will be slower than the
> non-JIT executor.

This largely seems to be orthogonal to what I'm talking about.

> Cf for instance a previous discussion here :

I'd wish people stopped using It's *NOT* maintained, in fact I do not know who does. It does shows
ads when downloading links, which I'm personally not ok with.


Andres Freund

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Daniel Verite 2018-08-22 20:05:41 Re: csv format for psql
Previous Message Andreas Joseph Krogh 2018-08-22 18:58:45 Re: Query is over 2x slower with jit=on