Irix6 long doubles implemented wrong? (27_io/ostream_inserter_arith)

[Richard Henderson]

On Tue, Jan 28, 2003 at 08:32:19PM -0500, Kaveh R. Ghazi wrote:
> long double __q_floti( int16_t ); 3
> long double __q_flotju( uint32_t ); 3
> long double __q_flotku( uint64_t ); 3

	floathitf2
	floatunssitf2
	floatunsditf2

> long double __q_neg( double, double ); 3

	negtf2

> int16_t __ii_qint( double, double ); 3

	fix_trunctfhi2

> 2.  Conversely, there were several optabs TF entries for which I couldn't
>     determine something in the irix libc list.  There were:
> 
> unordtf2_libfunc

This one may be a problem.  We don't know what the actual
return values are for the comparison functions.  By rights
it ought to be a tri-state value off of EQ or NE.

> floattitf_libfunc
> fixtfti_libfunc
> fixunstfti_libfunc

There were no TImode functions listed.

> 3.  I notice neither irix nor optabs seems to have functions to
>     convert *unsigned* ints to floating points.

For gcc, see "floatuns<MI><MF>2".
For Irix, see __q_flotju as opposed to __q_flotj.


r~