[Bug math/13886] floorl(-0x1.ffffffffffffffffffffffffff8p+105) returns -0x1.000000000000000000000000000p+106

[azanella at linux dot vnet.ibm.com]

http://sourceware.org/bugzilla/show_bug.cgi?id=13886

Adhemerval Zanella Netto <azanella at linux dot vnet.ibm.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |azanella at linux dot
                   |                            |vnet.ibm.com

--- Comment #1 from Adhemerval Zanella Netto <azanella at linux dot vnet.ibm.com> 2012-04-22 14:34:27 UTC ---
The current 'floorl' implementation for PPC64
(sysdeps/powerpc/powerpc64/fpu/s_floorl.S) produces wrong result in some cases.
It uses a wrong optimization for values lower than 2**52 and also for values
higher.

I proposed a patch in libc-alpha list
(http://sourceware.org/ml/libc-alpha/2012-04/msg00814.html) to just remove this
implementation and use the C version at
'sysdeps/ieee754/ldbl-128ibm/s_floorl.c' (which is used for PPC32).

-- 
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.