Re: new regression test

On Monday, April 21, 2003, at 05:27 AM, David Schultz wrote:

> On Sun, Apr 20, 2003, David Leimbach wrote:
>> test-printfloat.c: In function `main':
>> test-printfloat.c:80: `NAN' undeclared (first use this function)
>> test-printfloat.c:80: (Each undeclared identifier is reported only 
>> once
>> test-printfloat.c:80: for each function it appears in.)
>> test-printfloat.c:88: `HUGE_VALL' undeclared (first use this function)
>> *** Error code 1
>>
>> Am I running these from the proper directory?
>
> You appear to be running these tests with an old world (before
> February 8th), or you are running them on -STABLE, which lacks the
> features many of these tests are designed to exercise.
>

What I have is a mirror of the FreeBSD CVS via a CVSup script which
has gotten me at least a 5.0 kernel.  I keep seeing the most recent
changes come down such as the switch to UFS2 default from yesterday.

I then use a local CVS checkout command with no tag or revision to get
the head.  If this is wrong I don't know what's going on.

I even have the new %g float test from a few days back.

Doesn't make buildworld/installworld update all the system headers?

That could be my problem right there.

Dave