From: Jim Bell (Jim_at_[hidden])
Date: 2008-08-20 11:42:06
Was: Re: [Boost-testing] bjam hangs in object.py/list.py (trunk, FreeBSD-7)
Thanks for the responses.
Here's the current state of the matter: bjam consumes CPU as one would
expect until it hits the python part. Then it appears to make very little
progress for hours, but eventually completes. The total trunk regression
test takes five hours on an AMD Athlon(tm) 64 X2 Dual Core 5200+ 2.61 GHz.
Another wrinkle: FreeBSD is running as a VMWare guest (1GB RAM) under a
Windows XP host (4GB RAM).
To answer the questions posed:
Markus Sch?pflin wrote:
> Has this just started recently, or did it never work for you?
I thought it worked differently until recently, but now I think it may have
always worked this way.
> What python version do you use?
Python 2.5.1, FreeBSD-7's stock release.
> When you look at the bjam.log, do you see any progress? What are the tests
that are currently executed?
I do see very slow progress, and, given enough time, the test seems to
David Abrahams wrote:
> Huh; this looks really familiar. ... a build of Boost.Python appeared to
hang on MacOS. I think this must be some BSD/bjam interaction; I suggest
you follow up to the boost-build list...
Thanks. I'm copying both lists on this message, to report that it's working
again on FreeBSD.
I held out for input from the FreeBSD community, but haven't heard back.
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk