Re: Speed up clean meson builds by ~25%

Hi,

On 2024-04-09 19:00:41 -0400, Tom Lane wrote:
> Andres Freund <andres@anarazel.de> writes:
> > I think we need to do something about the compile time of this file, even with
> > gcc. Our main grammar already is an issue and stacking all the ecpg stuff on
> > top makes it considerably worse.
>
> Seems reasonable, if we can.
>
> > Why are strduping all of these?
>
> IIRC, the issue is that the mechanism for concatenating the tokens
> back together frees the input strings

Ah, that explains it - but also seems somewhat unnecessary.


> So that ought to dump core if you don't make all the productions
> return malloc'd strings.  How did you work around that?

I just tried to get to the point of understanding the reasons for slow
compilation, not to actually keep it working :). I.e. I didn't.


> (Maybe it'd be okay to just leak all the strings?)

Hm. The input to ecpg can be fairly large, I guess. And we have fun code like
cat_str(), which afaict is O(arguments^2) in its memory usage if we wouldn't
free?

Not immediately sure what the right path is.

Greetings,

Andres Freund