Boost Testing :
Subject: Re: [Boost-testing] Trying to run the regression tests
From: Steven Watanabe (watanabesj_at_[hidden])
Date: 2009-05-25 13:39:57
Marshall Clow wrote:
> At 9:28 AM -0700 5/25/09, Steven Watanabe wrote:
>> Marshall Clow wrote:
>>> Intel-based Mac Mini. Mac OS X 10.5.7. XCode 3.2.1 (gcc 4.0.1)
>>> I executed:
>>> $ python run.py --runner="mclow-darwin-intel" --toolsets=darwin
>>> --bjam-toolset=darwin --pjl-toolset=darwin
>>> as suggested by John Pavel on this list back in 2007
>>> end of page.
>>> But it fails (see attached output). It appears that the tools get
>>> confused trying to build
>>> process_jam_log, and decide that "darwin" is a target, rather than a
>> You need to have a user-config.jam with
>> using darwin ;
> Wouldn't that make it hard to test multiple different toolsets on the
> same machine?
In that case you would need
using darwin : someversion : /path/to/g++ ;
using darwin : anotherversion : /path/to/g++ ;
and pass --toolsets="darwin-someversion darwin-anotherversion" to run.py
>> It should be fairly easy to fix the regression testing framework, by
>> passing toolset=darwin instead of darwin.
> I'm looking at the notes by the folx doing regression testing on darwin.
> One of them (Sandia-leopard-x86) says:
> Command Line
> regression.py invoked as follows (trailing carets are line
> continuation characters)
> python regression.py --runner="Sandia-leopard-x86"
> --bjam-toolset=darwin ^
> --pjl-toolset=darwin --toolsets="darwin-4.0.1 intel-10.1"
> Now I'm even more confused. What's the difference between
> "regression.py" and "run.py"?
The top level script is run.py. regression.py does most of the work.
The references to regression.py are probably just out of date.
> I see that the page at
> talks about "run.py", except that it mentions "regression.py" once, in
> the "patching boost sources" section.
> Is that page the correct place to get instructions?