diff options
Diffstat (limited to 'usr/src/libm/src/mvec/vsqrt_.c')
-rw-r--r-- | usr/src/libm/src/mvec/vsqrt_.c | 87 |
1 files changed, 87 insertions, 0 deletions
diff --git a/usr/src/libm/src/mvec/vsqrt_.c b/usr/src/libm/src/mvec/vsqrt_.c new file mode 100644 index 0000000..c0d6bbf --- /dev/null +++ b/usr/src/libm/src/mvec/vsqrt_.c @@ -0,0 +1,87 @@ +/* + * CDDL HEADER START + * + * The contents of this file are subject to the terms of the + * Common Development and Distribution License (the "License"). + * You may not use this file except in compliance with the License. + * + * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE + * or http://www.opensolaris.org/os/licensing. + * See the License for the specific language governing permissions + * and limitations under the License. + * + * When distributing Covered Code, include this CDDL HEADER in each + * file and include the License file at usr/src/OPENSOLARIS.LICENSE. + * If applicable, add the following below this CDDL HEADER, with the + * fields enclosed by brackets "[]" replaced with your own identifying + * information: Portions Copyright [yyyy] [name of copyright owner] + * + * CDDL HEADER END + */ + +/* + * Copyright 2006 Sun Microsystems, Inc. All rights reserved. + * Use is subject to license terms. + */ + +#pragma ident "@(#)vsqrt_.c 1.4 06/01/31 SMI" + +extern void __vsqrt( int, double *, int, double *, int ); + +#pragma weak vsqrt_ = __vsqrt_ + +#ifndef LIBMTSK_BASED + +/* just invoke the serial function */ +void +__vsqrt_( int *n, double *x, int *stridex, double *y, int *stridey ) +{ + __vsqrt( *n, x, *stridex, y, *stridey ); +} + +#else + +#include "mtsk.h" + +static double *xp, *yp; +static int sx, sy; + +/* m-function for parallel vsqrt */ +void +__vsqrt_mfunc( struct MFunctionBlock *MFunctionBlockPtr, int LowerBound, + int UpperBound, int Step ) +{ + __vsqrt( UpperBound - LowerBound + 1, xp + sx * LowerBound, sx, + yp + sy * LowerBound, sy ); +} + +void +__vsqrt_( int *n, double *x, int *stridex, double *y, int *stridey ) +{ + struct MFunctionBlock m; + int i; + + /* if ncpus < 2, we are already in a parallel construct, or there + aren't enough vector elements to bother parallelizing, just + invoke the serial function */ + i = __mt_getncpus_(); + if ( i < 2 || *n < ( i << 3 ) || __mt_inepc_() || __mt_inapc_() ) + { + __vsqrt( *n, x, *stridex, y, *stridey ); + return; + } + + /* should be safe, we already know we're not in a parallel region */ + xp = x; + sx = *stridex; + yp = y; + sy = *stridey; + + m.MFunctionPtr = &__vsqrt_mfunc; + m.LowerBound = 0; + m.UpperBound = *n - 1; + m.Step = 1; + __mt_dopar_vfun_( m.MFunctionPtr, m.LowerBound, m.UpperBound, m.Step ); +} + +#endif |