The best testsuite result on Linux/x86-64 in months

[Mike Stump]

On May 23, 2007, at 8:58 AM, H. J. Lu wrote:
> The testsuite result, including both 32bit and 64bit, of gcc 4.3
> revision 124987 on Linux/x86-64:
>
> http://gcc.gnu.org/ml/gcc-testresults/2007-05/msg01148.html
>
> is the best in months.

So close, yet, so far...  :-)  If we pushed to get it down to zero,  
would it stay there...  Do we have enough discipline to keep it  
there?  Does someone want to step forward and propose xfails for all  
of the remaining expected fails that are unexpected currently?