[COMMITTED] hppa: Regenerate libm-testulps.

[Joseph S. Myers]

On Mon, 3 Feb 2014, Joseph S. Myers wrote:

> except for m68k where probably some long double functions need their 
> accuracy improved, all ulps are 7 or less.)

Specifically, since those ulps of 8 and 9 for m68k are for tgamma, I 
suspect they may arise out of known issues with powl on m68k (which could 
be fixed by adapting sysdeps/x86/fpu/powl_helper.c for m68k - the 
adaptation should actually be simpler than the x86 version since there 
should be no need to convert to base 2 logarithms).

-- 
Joseph S. Myers
joseph@codesourcery.com