[nightly] 30-Oct-2006 build of HEAD on x86_64-unknown-linux (cam-03-unx.europe.corp.microsoft.com)

[Simon Marlow]

Andy Gill wrote:
> There are a lot of failures in here of the form:
> 
> NOTE: Simplifier still going after 4 iterations; bailing out.
> NOTE: Simplifier still going after 4 iterations; bailing out.
> NOTE: Simplifier still going after 4 iterations; bailing out.
> NOTE: Simplifier still going after 3 iterations; bailing out.
> NOTE: Simplifier still going after 4 iterations; bailing out.
> 
> Who looks after re-normalizing these? Or should we codify ignoring  these?

These happen because the stage 3 build failed, so the test suite was run using 
the stage 2 compiler, which is build with DEBUG on and hence emits lots of trace 
messages that are not expected by the test suite.  So the test failures are 
bogus.  It would be nice if this didn't happen, somehow.  Perhaps there should 
be a way to disable trace messages for the purposes of the test suite.

However, there is a real issue with the amount of iterating that the simplifier 
is doing, which Simon PJ knows about.

Cheers,
	Simon