OpenBLAS/lapack-netlib/SRC/ssb2st_kernels.c

/* f2c.h  --  Standard Fortran to C header file */

/**  barf  [ba:rf]  2.  "He suggested using FORTRAN, and everybody barfed."

	- From The Shogakukan DICTIONARY OF NEW ENGLISH (Second edition) */

#ifndef F2C_INCLUDE
#define F2C_INCLUDE

#include <math.h>
#include <stdlib.h>
#include <string.h>
#include <stdio.h>
#include <complex.h>
#ifdef complex
#undef complex
#endif
#ifdef I
#undef I
#endif

#if defined(_WIN64)
typedef long long BLASLONG;
typedef unsigned long long BLASULONG;
#else
typedef long BLASLONG;
typedef unsigned long BLASULONG;
#endif

#ifdef LAPACK_ILP64
typedef BLASLONG blasint;
#if defined(_WIN64)
#define blasabs(x) llabs(x)
#else
#define blasabs(x) labs(x)
#endif
#else
typedef int blasint;
#define blasabs(x) abs(x)
#endif

typedef blasint integer;

typedef unsigned int uinteger;
typedef char *address;
typedef short int shortint;
typedef float real;
typedef double doublereal;
typedef struct { real r, i; } complex;
typedef struct { doublereal r, i; } doublecomplex;
static inline _Complex float Cf(complex *z) {return z->r + z->i*_Complex_I;}
static inline _Complex double Cd(doublecomplex *z) {return z->r + z->i*_Complex_I;}
static inline _Complex float * _pCf(complex *z) {return (_Complex float*)z;}
static inline _Complex double * _pCd(doublecomplex *z) {return (_Complex double*)z;}
#define pCf(z) (*_pCf(z))
#define pCd(z) (*_pCd(z))
typedef int logical;
typedef short int shortlogical;
typedef char logical1;
typedef char integer1;

#define TRUE_ (1)
#define FALSE_ (0)

/* Extern is for use with -E */
#ifndef Extern
#define Extern extern
#endif

/* I/O stuff */

typedef int flag;
typedef int ftnlen;
typedef int ftnint;

/*external read, write*/
typedef struct
{	flag cierr;
	ftnint ciunit;
	flag ciend;
	char *cifmt;
	ftnint cirec;
} cilist;

/*internal read, write*/
typedef struct
{	flag icierr;
	char *iciunit;
	flag iciend;
	char *icifmt;
	ftnint icirlen;
	ftnint icirnum;
} icilist;

/*open*/
typedef struct
{	flag oerr;
	ftnint ounit;
	char *ofnm;
	ftnlen ofnmlen;
	char *osta;
	char *oacc;
	char *ofm;
	ftnint orl;
	char *oblnk;
} olist;

/*close*/
typedef struct
{	flag cerr;
	ftnint cunit;
	char *csta;
} cllist;

/*rewind, backspace, endfile*/
typedef struct
{	flag aerr;
	ftnint aunit;
} alist;

/* inquire */
typedef struct
{	flag inerr;
	ftnint inunit;
	char *infile;
	ftnlen infilen;
	ftnint	*inex;	/*parameters in standard's order*/
	ftnint	*inopen;
	ftnint	*innum;
	ftnint	*innamed;
	char	*inname;
	ftnlen	innamlen;
	char	*inacc;
	ftnlen	inacclen;
	char	*inseq;
	ftnlen	inseqlen;
	char 	*indir;
	ftnlen	indirlen;
	char	*infmt;
	ftnlen	infmtlen;
	char	*inform;
	ftnint	informlen;
	char	*inunf;
	ftnlen	inunflen;
	ftnint	*inrecl;
	ftnint	*innrec;
	char	*inblank;
	ftnlen	inblanklen;
} inlist;

#define VOID void

union Multitype {	/* for multiple entry points */
	integer1 g;
	shortint h;
	integer i;
	/* longint j; */
	real r;
	doublereal d;
	complex c;
	doublecomplex z;
	};

typedef union Multitype Multitype;

struct Vardesc {	/* for Namelist */
	char *name;
	char *addr;
	ftnlen *dims;
	int  type;
	};
typedef struct Vardesc Vardesc;

struct Namelist {
	char *name;
	Vardesc **vars;
	int nvars;
	};
typedef struct Namelist Namelist;

#define abs(x) ((x) >= 0 ? (x) : -(x))
#define dabs(x) (fabs(x))
#define f2cmin(a,b) ((a) <= (b) ? (a) : (b))
#define f2cmax(a,b) ((a) >= (b) ? (a) : (b))
#define dmin(a,b) (f2cmin(a,b))
#define dmax(a,b) (f2cmax(a,b))
#define bit_test(a,b)	((a) >> (b) & 1)
#define bit_clear(a,b)	((a) & ~((uinteger)1 << (b)))
#define bit_set(a,b)	((a) |  ((uinteger)1 << (b)))

#define abort_() { sig_die("Fortran abort routine called", 1); }
#define c_abs(z) (cabsf(Cf(z)))
#define c_cos(R,Z) { pCf(R)=ccos(Cf(Z)); }
#define c_div(c, a, b) {pCf(c) = Cf(a)/Cf(b);}
#define z_div(c, a, b) {pCd(c) = Cd(a)/Cd(b);}
#define c_exp(R, Z) {pCf(R) = cexpf(Cf(Z));}
#define c_log(R, Z) {pCf(R) = clogf(Cf(Z));}
#define c_sin(R, Z) {pCf(R) = csinf(Cf(Z));}
//#define c_sqrt(R, Z) {*(R) = csqrtf(Cf(Z));}
#define c_sqrt(R, Z) {pCf(R) = csqrtf(Cf(Z));}
#define d_abs(x) (fabs(*(x)))
#define d_acos(x) (acos(*(x)))
#define d_asin(x) (asin(*(x)))
#define d_atan(x) (atan(*(x)))
#define d_atn2(x, y) (atan2(*(x),*(y)))
#define d_cnjg(R, Z) { pCd(R) = conj(Cd(Z)); }
#define r_cnjg(R, Z) { pCf(R) = conj(Cf(Z)); }
#define d_cos(x) (cos(*(x)))
#define d_cosh(x) (cosh(*(x)))
#define d_dim(__a, __b) ( *(__a) > *(__b) ? *(__a) - *(__b) : 0.0 )
#define d_exp(x) (exp(*(x)))
#define d_imag(z) (cimag(Cd(z)))
#define r_imag(z) (cimag(Cf(z)))
#define d_int(__x) (*(__x)>0 ? floor(*(__x)) : -floor(- *(__x)))
#define r_int(__x) (*(__x)>0 ? floor(*(__x)) : -floor(- *(__x)))
#define d_lg10(x) ( 0.43429448190325182765 * log(*(x)) )
#define r_lg10(x) ( 0.43429448190325182765 * log(*(x)) )
#define d_log(x) (log(*(x)))
#define d_mod(x, y) (fmod(*(x), *(y)))
#define u_nint(__x) ((__x)>=0 ? floor((__x) + .5) : -floor(.5 - (__x)))
#define d_nint(x) u_nint(*(x))
#define u_sign(__a,__b) ((__b) >= 0 ? ((__a) >= 0 ? (__a) : -(__a)) : -((__a) >= 0 ? (__a) : -(__a)))
#define d_sign(a,b) u_sign(*(a),*(b))
#define r_sign(a,b) u_sign(*(a),*(b))
#define d_sin(x) (sin(*(x)))
#define d_sinh(x) (sinh(*(x)))
#define d_sqrt(x) (sqrt(*(x)))
#define d_tan(x) (tan(*(x)))
#define d_tanh(x) (tanh(*(x)))
#define i_abs(x) abs(*(x))
#define i_dnnt(x) ((integer)u_nint(*(x)))
#define i_len(s, n) (n)
#define i_nint(x) ((integer)u_nint(*(x)))
#define i_sign(a,b) ((integer)u_sign((integer)*(a),(integer)*(b)))
#define pow_dd(ap, bp) ( pow(*(ap), *(bp)))
#define pow_si(B,E) spow_ui(*(B),*(E))
#define pow_ri(B,E) spow_ui(*(B),*(E))
#define pow_di(B,E) dpow_ui(*(B),*(E))
#define pow_zi(p, a, b) {pCd(p) = zpow_ui(Cd(a), *(b));}
#define pow_ci(p, a, b) {pCf(p) = cpow_ui(Cf(a), *(b));}
#define pow_zz(R,A,B) {pCd(R) = cpow(Cd(A),*(B));}
#define s_cat(lpp, rpp, rnp, np, llp) { 	ftnlen i, nc, ll; char *f__rp, *lp; 	ll = (llp); lp = (lpp); 	for(i=0; i < (int)*(np); ++i) {         	nc = ll; 	        if((rnp)[i] < nc) nc = (rnp)[i]; 	        ll -= nc;         	f__rp = (rpp)[i]; 	        while(--nc >= 0) *lp++ = *(f__rp)++;         } 	while(--ll >= 0) *lp++ = ' '; }
#define s_cmp(a,b,c,d) ((integer)strncmp((a),(b),f2cmin((c),(d))))
#define s_copy(A,B,C,D) { int __i,__m; for (__i=0, __m=f2cmin((C),(D)); __i<__m && (B)[__i] != 0; ++__i) (A)[__i] = (B)[__i]; }
#define sig_die(s, kill) { exit(1); }
#define s_stop(s, n) {exit(0);}
static char junk[] = "\n@(#)LIBF77 VERSION 19990503\n";
#define z_abs(z) (cabs(Cd(z)))
#define z_exp(R, Z) {pCd(R) = cexp(Cd(Z));}
#define z_sqrt(R, Z) {pCd(R) = csqrt(Cd(Z));}
#define myexit_() break;
#define mycycle() continue;
#define myceiling(w) {ceil(w)}
#define myhuge(w) {HUGE_VAL}
//#define mymaxloc_(w,s,e,n) {if (sizeof(*(w)) == sizeof(double)) dmaxloc_((w),*(s),*(e),n); else dmaxloc_((w),*(s),*(e),n);}
#define mymaxloc(w,s,e,n) {dmaxloc_(w,*(s),*(e),n)}

/* procedure parameter types for -A and -C++ */

#define F2C_proc_par_types 1
#ifdef __cplusplus
typedef logical (*L_fp)(...);
#else
typedef logical (*L_fp)();
#endif

static float spow_ui(float x, integer n) {
	float pow=1.0; unsigned long int u;
	if(n != 0) {
		if(n < 0) n = -n, x = 1/x;
		for(u = n; ; ) {
			if(u & 01) pow *= x;
			if(u >>= 1) x *= x;
			else break;
		}
	}
	return pow;
}
static double dpow_ui(double x, integer n) {
	double pow=1.0; unsigned long int u;
	if(n != 0) {
		if(n < 0) n = -n, x = 1/x;
		for(u = n; ; ) {
			if(u & 01) pow *= x;
			if(u >>= 1) x *= x;
			else break;
		}
	}
	return pow;
}
static _Complex float cpow_ui(_Complex float x, integer n) {
	_Complex float pow=1.0; unsigned long int u;
	if(n != 0) {
		if(n < 0) n = -n, x = 1/x;
		for(u = n; ; ) {
			if(u & 01) pow *= x;
			if(u >>= 1) x *= x;
			else break;
		}
	}
	return pow;
}
static _Complex double zpow_ui(_Complex double x, integer n) {
	_Complex double pow=1.0; unsigned long int u;
	if(n != 0) {
		if(n < 0) n = -n, x = 1/x;
		for(u = n; ; ) {
			if(u & 01) pow *= x;
			if(u >>= 1) x *= x;
			else break;
		}
	}
	return pow;
}
static integer pow_ii(integer x, integer n) {
	integer pow; unsigned long int u;
	if (n <= 0) {
		if (n == 0 || x == 1) pow = 1;
		else if (x != -1) pow = x == 0 ? 1/x : 0;
		else n = -n;
	}
	if ((n > 0) || !(n == 0 || x == 1 || x != -1)) {
		u = n;
		for(pow = 1; ; ) {
			if(u & 01) pow *= x;
			if(u >>= 1) x *= x;
			else break;
		}
	}
	return pow;
}
static integer dmaxloc_(double *w, integer s, integer e, integer *n)
{
	double m; integer i, mi;
	for(m=w[s-1], mi=s, i=s+1; i<=e; i++)
		if (w[i-1]>m) mi=i ,m=w[i-1];
	return mi-s+1;
}
static integer smaxloc_(float *w, integer s, integer e, integer *n)
{
	float m; integer i, mi;
	for(m=w[s-1], mi=s, i=s+1; i<=e; i++)
		if (w[i-1]>m) mi=i ,m=w[i-1];
	return mi-s+1;
}
static inline void cdotc_(complex *z, integer *n_, complex *x, integer *incx_, complex *y, integer *incy_) {
	integer n = *n_, incx = *incx_, incy = *incy_, i;
	_Complex float zdotc = 0.0;
	if (incx == 1 && incy == 1) {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc += conjf(Cf(&x[i])) * Cf(&y[i]);
		}
	} else {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc += conjf(Cf(&x[i*incx])) * Cf(&y[i*incy]);
		}
	}
	pCf(z) = zdotc;
}
static inline void zdotc_(doublecomplex *z, integer *n_, doublecomplex *x, integer *incx_, doublecomplex *y, integer *incy_) {
	integer n = *n_, incx = *incx_, incy = *incy_, i;
	_Complex double zdotc = 0.0;
	if (incx == 1 && incy == 1) {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc += conj(Cd(&x[i])) * Cd(&y[i]);
		}
	} else {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc += conj(Cd(&x[i*incx])) * Cd(&y[i*incy]);
		}
	}
	pCd(z) = zdotc;
}	
static inline void cdotu_(complex *z, integer *n_, complex *x, integer *incx_, complex *y, integer *incy_) {
	integer n = *n_, incx = *incx_, incy = *incy_, i;
	_Complex float zdotc = 0.0;
	if (incx == 1 && incy == 1) {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc += Cf(&x[i]) * Cf(&y[i]);
		}
	} else {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc += Cf(&x[i*incx]) * Cf(&y[i*incy]);
		}
	}
	pCf(z) = zdotc;
}
static inline void zdotu_(doublecomplex *z, integer *n_, doublecomplex *x, integer *incx_, doublecomplex *y, integer *incy_) {
	integer n = *n_, incx = *incx_, incy = *incy_, i;
	_Complex double zdotc = 0.0;
	if (incx == 1 && incy == 1) {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc += Cd(&x[i]) * Cd(&y[i]);
		}
	} else {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc += Cd(&x[i*incx]) * Cd(&y[i*incy]);
		}
	}
	pCd(z) = zdotc;
}
#endif
/*  -- translated by f2c (version 20000121).
   You must link the resulting object file with the libraries:
	-lf2c -lm   (in that order)
*/



/* Table of constant values */

static integer c__1 = 1;

/* > \brief \b SSB2ST_KERNELS */

/*  @generated from zhb2st_kernels.f, fortran z -> s, Wed Dec  7 08:22:40 2016 */

/*  =========== DOCUMENTATION =========== */

/* Online html documentation available at */
/*            http://www.netlib.org/lapack/explore-html/ */

/* > \htmlonly */
/* > Download SSB2ST_KERNELS + dependencies */
/* > <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/ssb2st_
kernels.f"> */
/* > [TGZ]</a> */
/* > <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/ssb2st_
kernels.f"> */
/* > [ZIP]</a> */
/* > <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/ssb2st_
kernels.f"> */
/* > [TXT]</a> */
/* > \endhtmlonly */

/*  Definition: */
/*  =========== */

/*       SUBROUTINE  SSB2ST_KERNELS( UPLO, WANTZ, TTYPE, */
/*                                   ST, ED, SWEEP, N, NB, IB, */
/*                                   A, LDA, V, TAU, LDVT, WORK) */

/*       IMPLICIT NONE */

/*       CHARACTER          UPLO */
/*       LOGICAL            WANTZ */
/*       INTEGER            TTYPE, ST, ED, SWEEP, N, NB, IB, LDA, LDVT */
/*       REAL               A( LDA, * ), V( * ), */
/*                          TAU( * ), WORK( * ) */

/* > \par Purpose: */
/*  ============= */
/* > */
/* > \verbatim */
/* > */
/* > SSB2ST_KERNELS is an internal routine used by the SSYTRD_SB2ST */
/* > subroutine. */
/* > \endverbatim */

/*  Arguments: */
/*  ========== */

/* > \param[in] UPLO */
/* > \verbatim */
/* >          UPLO is CHARACTER*1 */
/* > \endverbatim */
/* > */
/* > \param[in] WANTZ */
/* > \verbatim */
/* >          WANTZ is LOGICAL which indicate if Eigenvalue are requested or both */
/* >          Eigenvalue/Eigenvectors. */
/* > \endverbatim */
/* > */
/* > \param[in] TTYPE */
/* > \verbatim */
/* >          TTYPE is INTEGER */
/* > \endverbatim */
/* > */
/* > \param[in] ST */
/* > \verbatim */
/* >          ST is INTEGER */
/* >          internal parameter for indices. */
/* > \endverbatim */
/* > */
/* > \param[in] ED */
/* > \verbatim */
/* >          ED is INTEGER */
/* >          internal parameter for indices. */
/* > \endverbatim */
/* > */
/* > \param[in] SWEEP */
/* > \verbatim */
/* >          SWEEP is INTEGER */
/* >          internal parameter for indices. */
/* > \endverbatim */
/* > */
/* > \param[in] N */
/* > \verbatim */
/* >          N is INTEGER. The order of the matrix A. */
/* > \endverbatim */
/* > */
/* > \param[in] NB */
/* > \verbatim */
/* >          NB is INTEGER. The size of the band. */
/* > \endverbatim */
/* > */
/* > \param[in] IB */
/* > \verbatim */
/* >          IB is INTEGER. */
/* > \endverbatim */
/* > */
/* > \param[in, out] A */
/* > \verbatim */
/* >          A is REAL array. A pointer to the matrix A. */
/* > \endverbatim */
/* > */
/* > \param[in] LDA */
/* > \verbatim */
/* >          LDA is INTEGER. The leading dimension of the matrix A. */
/* > \endverbatim */
/* > */
/* > \param[out] V */
/* > \verbatim */
/* >          V is REAL array, dimension 2*n if eigenvalues only are */
/* >          requested or to be queried for vectors. */
/* > \endverbatim */
/* > */
/* > \param[out] TAU */
/* > \verbatim */
/* >          TAU is REAL array, dimension (2*n). */
/* >          The scalar factors of the Householder reflectors are stored */
/* >          in this array. */
/* > \endverbatim */
/* > */
/* > \param[in] LDVT */
/* > \verbatim */
/* >          LDVT is INTEGER. */
/* > \endverbatim */
/* > */
/* > \param[out] WORK */
/* > \verbatim */
/* >          WORK is REAL array. Workspace of size nb. */
/* > \endverbatim */
/* > @param[in] n */
/* >          The order of the matrix A. */
/* > */
/* > */
/* > \par Further Details: */
/*  ===================== */
/* > */
/* > \verbatim */
/* > */
/* >  Implemented by Azzam Haidar. */
/* > */
/* >  All details are available on technical report, SC11, SC13 papers. */
/* > */
/* >  Azzam Haidar, Hatem Ltaief, and Jack Dongarra. */
/* >  Parallel reduction to condensed forms for symmetric eigenvalue problems */
/* >  using aggregated fine-grained and memory-aware kernels. In Proceedings */
/* >  of 2011 International Conference for High Performance Computing, */
/* >  Networking, Storage and Analysis (SC '11), New York, NY, USA, */
/* >  Article 8 , 11 pages. */
/* >  http://doi.acm.org/10.1145/2063384.2063394 */
/* > */
/* >  A. Haidar, J. Kurzak, P. Luszczek, 2013. */
/* >  An improved parallel singular value algorithm and its implementation */
/* >  for multicore hardware, In Proceedings of 2013 International Conference */
/* >  for High Performance Computing, Networking, Storage and Analysis (SC '13). */
/* >  Denver, Colorado, USA, 2013. */
/* >  Article 90, 12 pages. */
/* >  http://doi.acm.org/10.1145/2503210.2503292 */
/* > */
/* >  A. Haidar, R. Solca, S. Tomov, T. Schulthess and J. Dongarra. */
/* >  A novel hybrid CPU-GPU generalized eigensolver for electronic structure */
/* >  calculations based on fine-grained memory aware tasks. */
/* >  International Journal of High Performance Computing Applications. */
/* >  Volume 28 Issue 2, Pages 196-209, May 2014. */
/* >  http://hpc.sagepub.com/content/28/2/196 */
/* > */
/* > \endverbatim */
/* > */
/*  ===================================================================== */
/* Subroutine */ int ssb2st_kernels_(char *uplo, logical *wantz, integer *
	ttype, integer *st, integer *ed, integer *sweep, integer *n, integer *
	nb, integer *ib, real *a, integer *lda, real *v, real *tau, integer *
	ldvt, real *work)
{
    /* System generated locals */
    integer a_dim1, a_offset, i__1, i__2;
    real r__1;

    /* Local variables */
    real ctmp;
    integer dpos, vpos, i__;
    extern logical lsame_(char *, char *);
    logical upper;
    integer j1, j2, lm, ln, ajeter;
    extern /* Subroutine */ int slarfg_(integer *, real *, real *, integer *, 
	    real *);
    integer ofdpos;
    extern /* Subroutine */ int slarfx_(char *, integer *, integer *, real *, 
	    real *, real *, integer *, real *), slarfy_(char *, 
	    integer *, real *, integer *, real *, real *, integer *, real *);
    integer taupos;



/*  -- LAPACK computational routine (version 3.7.1) -- */
/*  -- LAPACK is a software package provided by Univ. of Tennessee,    -- */
/*  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..-- */
/*     June 2017 */


/*  ===================================================================== */


    /* Parameter adjustments */
    a_dim1 = *lda;
    a_offset = 1 + a_dim1 * 1;
    a -= a_offset;
    --v;
    --tau;
    --work;

    /* Function Body */
    ajeter = *ib + *ldvt;
    upper = lsame_(uplo, "U");
    if (upper) {
	dpos = (*nb << 1) + 1;
	ofdpos = *nb << 1;
    } else {
	dpos = 1;
	ofdpos = 2;
    }

/*     Upper case */

    if (upper) {

	if (*wantz) {
	    vpos = (*sweep - 1) % 2 * *n + *st;
	    taupos = (*sweep - 1) % 2 * *n + *st;
	} else {
	    vpos = (*sweep - 1) % 2 * *n + *st;
	    taupos = (*sweep - 1) % 2 * *n + *st;
	}

	if (*ttype == 1) {
	    lm = *ed - *st + 1;

	    v[vpos] = 1.f;
	    i__1 = lm - 1;
	    for (i__ = 1; i__ <= i__1; ++i__) {
		v[vpos + i__] = a[ofdpos - i__ + (*st + i__) * a_dim1];
		a[ofdpos - i__ + (*st + i__) * a_dim1] = 0.f;
/* L10: */
	    }
	    ctmp = a[ofdpos + *st * a_dim1];
	    slarfg_(&lm, &ctmp, &v[vpos + 1], &c__1, &tau[taupos]);
	    a[ofdpos + *st * a_dim1] = ctmp;

	    lm = *ed - *st + 1;
	    r__1 = tau[taupos];
	    i__1 = *lda - 1;
	    slarfy_(uplo, &lm, &v[vpos], &c__1, &r__1, &a[dpos + *st * a_dim1]
		    , &i__1, &work[1]);
	}

	if (*ttype == 3) {

	    lm = *ed - *st + 1;
	    r__1 = tau[taupos];
	    i__1 = *lda - 1;
	    slarfy_(uplo, &lm, &v[vpos], &c__1, &r__1, &a[dpos + *st * a_dim1]
		    , &i__1, &work[1]);
	}

	if (*ttype == 2) {
	    j1 = *ed + 1;
/* Computing MIN */
	    i__1 = *ed + *nb;
	    j2 = f2cmin(i__1,*n);
	    ln = *ed - *st + 1;
	    lm = j2 - j1 + 1;
	    if (lm > 0) {
		r__1 = tau[taupos];
		i__1 = *lda - 1;
		slarfx_("Left", &ln, &lm, &v[vpos], &r__1, &a[dpos - *nb + j1 
			* a_dim1], &i__1, &work[1]);

		if (*wantz) {
		    vpos = (*sweep - 1) % 2 * *n + j1;
		    taupos = (*sweep - 1) % 2 * *n + j1;
		} else {
		    vpos = (*sweep - 1) % 2 * *n + j1;
		    taupos = (*sweep - 1) % 2 * *n + j1;
		}

		v[vpos] = 1.f;
		i__1 = lm - 1;
		for (i__ = 1; i__ <= i__1; ++i__) {
		    v[vpos + i__] = a[dpos - *nb - i__ + (j1 + i__) * a_dim1];
		    a[dpos - *nb - i__ + (j1 + i__) * a_dim1] = 0.f;
/* L30: */
		}
		ctmp = a[dpos - *nb + j1 * a_dim1];
		slarfg_(&lm, &ctmp, &v[vpos + 1], &c__1, &tau[taupos]);
		a[dpos - *nb + j1 * a_dim1] = ctmp;

		i__1 = ln - 1;
		i__2 = *lda - 1;
		slarfx_("Right", &i__1, &lm, &v[vpos], &tau[taupos], &a[dpos 
			- *nb + 1 + j1 * a_dim1], &i__2, &work[1]);
	    }
	}

/*     Lower case */

    } else {

	if (*wantz) {
	    vpos = (*sweep - 1) % 2 * *n + *st;
	    taupos = (*sweep - 1) % 2 * *n + *st;
	} else {
	    vpos = (*sweep - 1) % 2 * *n + *st;
	    taupos = (*sweep - 1) % 2 * *n + *st;
	}

	if (*ttype == 1) {
	    lm = *ed - *st + 1;

	    v[vpos] = 1.f;
	    i__1 = lm - 1;
	    for (i__ = 1; i__ <= i__1; ++i__) {
		v[vpos + i__] = a[ofdpos + i__ + (*st - 1) * a_dim1];
		a[ofdpos + i__ + (*st - 1) * a_dim1] = 0.f;
/* L20: */
	    }
	    slarfg_(&lm, &a[ofdpos + (*st - 1) * a_dim1], &v[vpos + 1], &c__1,
		     &tau[taupos]);

	    lm = *ed - *st + 1;

	    r__1 = tau[taupos];
	    i__1 = *lda - 1;
	    slarfy_(uplo, &lm, &v[vpos], &c__1, &r__1, &a[dpos + *st * a_dim1]
		    , &i__1, &work[1]);
	}

	if (*ttype == 3) {
	    lm = *ed - *st + 1;

	    r__1 = tau[taupos];
	    i__1 = *lda - 1;
	    slarfy_(uplo, &lm, &v[vpos], &c__1, &r__1, &a[dpos + *st * a_dim1]
		    , &i__1, &work[1]);
	}

	if (*ttype == 2) {
	    j1 = *ed + 1;
/* Computing MIN */
	    i__1 = *ed + *nb;
	    j2 = f2cmin(i__1,*n);
	    ln = *ed - *st + 1;
	    lm = j2 - j1 + 1;

	    if (lm > 0) {
		i__1 = *lda - 1;
		slarfx_("Right", &lm, &ln, &v[vpos], &tau[taupos], &a[dpos + *
			nb + *st * a_dim1], &i__1, &work[1]);

		if (*wantz) {
		    vpos = (*sweep - 1) % 2 * *n + j1;
		    taupos = (*sweep - 1) % 2 * *n + j1;
		} else {
		    vpos = (*sweep - 1) % 2 * *n + j1;
		    taupos = (*sweep - 1) % 2 * *n + j1;
		}

		v[vpos] = 1.f;
		i__1 = lm - 1;
		for (i__ = 1; i__ <= i__1; ++i__) {
		    v[vpos + i__] = a[dpos + *nb + i__ + *st * a_dim1];
		    a[dpos + *nb + i__ + *st * a_dim1] = 0.f;
/* L40: */
		}
		slarfg_(&lm, &a[dpos + *nb + *st * a_dim1], &v[vpos + 1], &
			c__1, &tau[taupos]);

		i__1 = ln - 1;
		r__1 = tau[taupos];
		i__2 = *lda - 1;
		slarfx_("Left", &lm, &i__1, &v[vpos], &r__1, &a[dpos + *nb - 
			1 + (*st + 1) * a_dim1], &i__2, &work[1]);
	    }
	}
    }

    return 0;

/*     END OF SSB2ST_KERNELS */

} /* ssb2st_kernels__ */