[postgis-devel] Postgis 2.3.1 regression test failure on NetBSD-7 amd64

[Sandro Santilli]

On Mon, Dec 05, 2016 at 01:33:40PM -0500, Stephen Woodbridge wrote:
> On 12/5/2016 12:21 PM, Sandro Santilli wrote:
> >On Mon, Dec 05, 2016 at 10:06:42AM -0500, Greg Troxel wrote:
> >
> >>-contains interrupted on time
> >>+contains interrupted late: 00:00:00.274858
> >
> >Nothing to worry about, unfortunately we do have some time-sensitive
> >tests so if the tester does not receive enough timeslots it may end
> >up interrupting later than expected.
> >
> >I'd love to fix these but hadn't found a way to obtain CPU time
> >of a postgresql session.
> 
> Can't you do something like, get the pid of the backend process and then use
> a system call to get the cpu time maybe using via /proc/[PID]/stat on linux
> 
> and maybe something like this on Windows:
> https://msdn.microsoft.com/en-us/library/windows/desktop/ms683223(v=vs.85).aspx

It would mean using a completely different framework for those tests,
right now we use the .sql and _expect pair as all the rest...

Or, we would need to expose an internal SQL function for this specific
purpose.

--strk;