[Bug target/15383] New: __float128 function name is used for long double

[Bug target/15383] New: __float128 function name is used for long double

[hjl at lucon dot org]

 

-- 
           Summary: __float128 function name is used for long double
           Product: gcc
           Version: 3.4.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: target
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: hjl at lucon dot org
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: ia64-unknown-linux-gnu
  GCC host triplet: ia64-unknown-linux-gnu
GCC target triplet: ia64-unknown-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15383

[Bug target/15383] __float128 function name is used for long double

[hjl at lucon dot org]

------- Additional Comments From hjl at lucon dot org  2004-05-11 19:40 -------
__divtf3 is an alias for __divxf3, which is a 80bit long double
function for backward compatibility. Since gcc 3.4.0 supports
__float128, gcc will use __divtf3 for 128bit floating point number.
__divtf3 should be either removed or implement __float128 operation.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15383

[Bug target/15383] __float128 function name is used for long double

[zack at codesourcery dot com]

------- Additional Comments From zack at codesourcery dot com  2004-05-11 20:37 -------
Subject: Re:  __float128 function name is used for long
 double

> __divtf3 is an alias for __divxf3, which is a 80bit long double
> function for backward compatibility. Since gcc 3.4.0 supports
> __float128, gcc will use __divtf3 for 128bit floating point number.
> __divtf3 should be either removed or implement __float128 operation.

I think __divtf3 must remain as it is now, or we break the C ABI.

When I originally implemented this stuff I proposed using the HPUX
function names for __float128 operations (_U_Qf* -- see
ia64_hpux_init_libfuncs) but it was considered to be moot as long as
we didn't actually have an 128bit floating point library.  Has that
changed?

zw


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15383

[Bug target/15383] __float128 function name is used for long double

[hjl at lucon dot org]

------- Additional Comments From hjl at lucon dot org  2004-05-11 22:21 -------
We can't have the same symbol for 2 different purposes. Yes,
I am working on __float128 support for Linux.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15383

[Bug target/15383] __float128 function name is used for long double

[hjl at lucon dot org]

------- Additional Comments From hjl at lucon dot org  2004-05-12 19:00 -------
Created an attachment (id=6267)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=6267&action=view)
A testcase

[hjl@gnu-2 float128]$ /usr/gcc-3.4/bin/gcc -O bug.c
[hjl@gnu-2 float128]$ ./a.out
Aborted

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15383

[Bug target/15383] __float128 function name is used for long double

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-05-20 02:15 -------
Subject: Bug 15383

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	hjl@gcc.gnu.org	2004-05-20 02:15:46

Modified files:
	gcc            : ChangeLog 
	gcc/config/ia64: ia64.c sysv4.h 

Log message:
	2004-05-19  H.J. Lu  <hongjiu.lu@intel.com>
	
	PR target/15383
	* config/ia64/ia64.c (ia64_expand_compare): Don't check
	TARGET_HPUX for TFmode compare. Abort if op0 is in TFmode and
	cmptf_libfunc isn't set.
	(ia64_init_libfuncs): Rename TFmode libfuncs using the HPUX
	conventions.
	(ia64_sysv4_init_libfuncs): New.
	
	* config/ia64/sysv4.h (TARGET_INIT_LIBFUNCS): New. Defined as
	ia64_sysv4_init_libfuncs.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.3702&r2=2.3703
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/ia64/ia64.c.diff?cvsroot=gcc&r1=1.286&r2=1.287
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/config/ia64/sysv4.h.diff?cvsroot=gcc&r1=1.30&r2=1.31



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15383

[Bug target/15383] __float128 function name is used for long double

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-05-20 02:34 -------
Fixed.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |3.5.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15383