Re: Skip building the test programs in parallel/threadingbuildingblocks

[Charlotte Koch <charlotte%netbsd.org@localhost>]

On Thu, 11 Aug 2022, Taylor R Campbell wrote:

> > From: Charlotte Koch <charlotte%netbsd.org@localhost>
> >
> > So, I think we can speed up the build significantly in the typical use
> > case by building the test programs iff we want to run the tests in the
> > first place. What do we think of my diff (attached) ?
>
> How much time is saved?  What is the fraction of the build time?

In this specific case, the time saved is substantial:

WITHOUT the patch: 
1482.64 real      5349.65 user       339.51 sys

WITH the patch:
53.80 real       106.20 user        18.05 sys

(Built on an unimpressive T440)

It's probably an outlier -- on the average, I doubt we'd see savings
this dramatic.

> With the patch, how would pkgsrc know whether the user is about to do
> `bmake test' manually, while not setting PKGSRC_RUN_TEST?  I do this
> regularly while developing packages.

I'm not sure the best way to go about that... I literally discovered
PKGSRC_RUN_TEST even exists just last night, heh.

But I do agree with what other folks have been saying, that running
'make test' manually should always work.

Charlotte