OpenBLAS/lapack-netlib/SRC/cungtsqr_row.c

#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;
#ifdef _MSC_VER
static inline _Fcomplex Cf(complex *z) {_Fcomplex zz={z->r , z->i}; return zz;}
static inline _Dcomplex Cd(doublecomplex *z) {_Dcomplex zz={z->r , z->i};return zz;}
static inline _Fcomplex * _pCf(complex *z) {return (_Fcomplex*)z;}
static inline _Dcomplex * _pCd(doublecomplex *z) {return (_Dcomplex*)z;}
#else
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;}
#endif
#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)); }
#ifdef _MSC_VER
#define c_div(c, a, b) {Cf(c)._Val[0] = (Cf(a)._Val[0]/Cf(b)._Val[0]); Cf(c)._Val[1]=(Cf(a)._Val[1]/Cf(b)._Val[1]);}
#define z_div(c, a, b) {Cd(c)._Val[0] = (Cd(a)._Val[0]/Cd(b)._Val[0]); Cd(c)._Val[1]=(Cd(a)._Val[1]/df(b)._Val[1]);}
#else
#define c_div(c, a, b) {pCf(c) = Cf(a)/Cf(b);}
#define z_div(c, a, b) {pCd(c) = Cd(a)/Cd(b);}
#endif
#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) = conjf(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) (cimagf(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;
}
#ifdef _MSC_VER
static _Fcomplex cpow_ui(complex x, integer n) {
	complex pow={1.0,0.0}; unsigned long int u;
		if(n != 0) {
		if(n < 0) n = -n, x.r = 1/x.r, x.i=1/x.i;
		for(u = n; ; ) {
			if(u & 01) pow.r *= x.r, pow.i *= x.i;
			if(u >>= 1) x.r *= x.r, x.i *= x.i;
			else break;
		}
	}
	_Fcomplex p={pow.r, pow.i};
	return p;
}
#else
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;
}
#endif
#ifdef _MSC_VER
static _Dcomplex zpow_ui(_Dcomplex x, integer n) {
	_Dcomplex pow={1.0,0.0}; unsigned long int u;
	if(n != 0) {
		if(n < 0) n = -n, x._Val[0] = 1/x._Val[0], x._Val[1] =1/x._Val[1];
		for(u = n; ; ) {
			if(u & 01) pow._Val[0] *= x._Val[0], pow._Val[1] *= x._Val[1];
			if(u >>= 1) x._Val[0] *= x._Val[0], x._Val[1] *= x._Val[1];
			else break;
		}
	}
	_Dcomplex p = {pow._Val[0], pow._Val[1]};
	return p;
}
#else
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;
}
#endif
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;
#ifdef _MSC_VER
	_Fcomplex zdotc = {0.0, 0.0};
	if (incx == 1 && incy == 1) {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc._Val[0] += conjf(Cf(&x[i]))._Val[0] * Cf(&y[i])._Val[0];
			zdotc._Val[1] += conjf(Cf(&x[i]))._Val[1] * Cf(&y[i])._Val[1];
		}
	} else {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc._Val[0] += conjf(Cf(&x[i*incx]))._Val[0] * Cf(&y[i*incy])._Val[0];
			zdotc._Val[1] += conjf(Cf(&x[i*incx]))._Val[1] * Cf(&y[i*incy])._Val[1];
		}
	}
	pCf(z) = zdotc;
}
#else
	_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;
}
#endif
static inline void zdotc_(doublecomplex *z, integer *n_, doublecomplex *x, integer *incx_, doublecomplex *y, integer *incy_) {
	integer n = *n_, incx = *incx_, incy = *incy_, i;
#ifdef _MSC_VER
	_Dcomplex zdotc = {0.0, 0.0};
	if (incx == 1 && incy == 1) {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc._Val[0] += conj(Cd(&x[i]))._Val[0] * Cd(&y[i])._Val[0];
			zdotc._Val[1] += conj(Cd(&x[i]))._Val[1] * Cd(&y[i])._Val[1];
		}
	} else {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc._Val[0] += conj(Cd(&x[i*incx]))._Val[0] * Cd(&y[i*incy])._Val[0];
			zdotc._Val[1] += conj(Cd(&x[i*incx]))._Val[1] * Cd(&y[i*incy])._Val[1];
		}
	}
	pCd(z) = zdotc;
}
#else
	_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;
}
#endif	
static inline void cdotu_(complex *z, integer *n_, complex *x, integer *incx_, complex *y, integer *incy_) {
	integer n = *n_, incx = *incx_, incy = *incy_, i;
#ifdef _MSC_VER
	_Fcomplex zdotc = {0.0, 0.0};
	if (incx == 1 && incy == 1) {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc._Val[0] += Cf(&x[i])._Val[0] * Cf(&y[i])._Val[0];
			zdotc._Val[1] += Cf(&x[i])._Val[1] * Cf(&y[i])._Val[1];
		}
	} else {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc._Val[0] += Cf(&x[i*incx])._Val[0] * Cf(&y[i*incy])._Val[0];
			zdotc._Val[1] += Cf(&x[i*incx])._Val[1] * Cf(&y[i*incy])._Val[1];
		}
	}
	pCf(z) = zdotc;
}
#else
	_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;
}
#endif
static inline void zdotu_(doublecomplex *z, integer *n_, doublecomplex *x, integer *incx_, doublecomplex *y, integer *incy_) {
	integer n = *n_, incx = *incx_, incy = *incy_, i;
#ifdef _MSC_VER
	_Dcomplex zdotc = {0.0, 0.0};
	if (incx == 1 && incy == 1) {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc._Val[0] += Cd(&x[i])._Val[0] * Cd(&y[i])._Val[0];
			zdotc._Val[1] += Cd(&x[i])._Val[1] * Cd(&y[i])._Val[1];
		}
	} else {
		for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
			zdotc._Val[0] += Cd(&x[i*incx])._Val[0] * Cd(&y[i*incy])._Val[0];
			zdotc._Val[1] += Cd(&x[i*incx])._Val[1] * Cd(&y[i*incy])._Val[1];
		}
	}
	pCd(z) = zdotc;
}
#else
	_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 complex c_b1 = {1.f,0.f};
static complex c_b2 = {0.f,0.f};
static integer c__0 = 0;
static integer c__1 = 1;

/* > \brief \b CUNGTSQR_ROW */

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

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

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

/*       SUBROUTINE CUNGTSQR_ROW( M, N, MB, NB, A, LDA, T, LDT, WORK, */
/*      $                         LWORK, INFO ) */
/*       IMPLICIT NONE */

/*       INTEGER           INFO, LDA, LDT, LWORK, M, N, MB, NB */
/*       COMPLEX           A( LDA, * ), T( LDT, * ), WORK( * ) */

/* > \par Purpose: */
/*  ============= */
/* > */
/* > \verbatim */
/* > */
/* > CUNGTSQR_ROW generates an M-by-N complex matrix Q_out with */
/* > orthonormal columns from the output of CLATSQR. These N orthonormal */
/* > columns are the first N columns of a product of complex unitary */
/* > matrices Q(k)_in of order M, which are returned by CLATSQR in */
/* > a special format. */
/* > */
/* >      Q_out = first_N_columns_of( Q(1)_in * Q(2)_in * ... * Q(k)_in ). */
/* > */
/* > The input matrices Q(k)_in are stored in row and column blocks in A. */
/* > See the documentation of CLATSQR for more details on the format of */
/* > Q(k)_in, where each Q(k)_in is represented by block Householder */
/* > transformations. This routine calls an auxiliary routine CLARFB_GETT, */
/* > where the computation is performed on each individual block. The */
/* > algorithm first sweeps NB-sized column blocks from the right to left */
/* > starting in the bottom row block and continues to the top row block */
/* > (hence _ROW in the routine name). This sweep is in reverse order of */
/* > the order in which CLATSQR generates the output blocks. */
/* > \endverbatim */

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

/* > \param[in] M */
/* > \verbatim */
/* >          M is INTEGER */
/* >          The number of rows of the matrix A.  M >= 0. */
/* > \endverbatim */
/* > */
/* > \param[in] N */
/* > \verbatim */
/* >          N is INTEGER */
/* >          The number of columns of the matrix A. M >= N >= 0. */
/* > \endverbatim */
/* > */
/* > \param[in] MB */
/* > \verbatim */
/* >          MB is INTEGER */
/* >          The row block size used by CLATSQR to return */
/* >          arrays A and T. MB > N. */
/* >          (Note that if MB > M, then M is used instead of MB */
/* >          as the row block size). */
/* > \endverbatim */
/* > */
/* > \param[in] NB */
/* > \verbatim */
/* >          NB is INTEGER */
/* >          The column block size used by CLATSQR to return */
/* >          arrays A and T. NB >= 1. */
/* >          (Note that if NB > N, then N is used instead of NB */
/* >          as the column block size). */
/* > \endverbatim */
/* > */
/* > \param[in,out] A */
/* > \verbatim */
/* >          A is COMPLEX array, dimension (LDA,N) */
/* > */
/* >          On entry: */
/* > */
/* >             The elements on and above the diagonal are not used as */
/* >             input. The elements below the diagonal represent the unit */
/* >             lower-trapezoidal blocked matrix V computed by CLATSQR */
/* >             that defines the input matrices Q_in(k) (ones on the */
/* >             diagonal are not stored). See CLATSQR for more details. */
/* > */
/* >          On exit: */
/* > */
/* >             The array A contains an M-by-N orthonormal matrix Q_out, */
/* >             i.e the columns of A are orthogonal unit vectors. */
/* > \endverbatim */
/* > */
/* > \param[in] LDA */
/* > \verbatim */
/* >          LDA is INTEGER */
/* >          The leading dimension of the array A.  LDA >= f2cmax(1,M). */
/* > \endverbatim */
/* > */
/* > \param[in] T */
/* > \verbatim */
/* >          T is COMPLEX array, */
/* >          dimension (LDT, N * NIRB) */
/* >          where NIRB = Number_of_input_row_blocks */
/* >                     = MAX( 1, CEIL((M-N)/(MB-N)) ) */
/* >          Let NICB = Number_of_input_col_blocks */
/* >                   = CEIL(N/NB) */
/* > */
/* >          The upper-triangular block reflectors used to define the */
/* >          input matrices Q_in(k), k=(1:NIRB*NICB). The block */
/* >          reflectors are stored in compact form in NIRB block */
/* >          reflector sequences. Each of the NIRB block reflector */
/* >          sequences is stored in a larger NB-by-N column block of T */
/* >          and consists of NICB smaller NB-by-NB upper-triangular */
/* >          column blocks. See CLATSQR for more details on the format */
/* >          of T. */
/* > \endverbatim */
/* > */
/* > \param[in] LDT */
/* > \verbatim */
/* >          LDT is INTEGER */
/* >          The leading dimension of the array T. */
/* >          LDT >= f2cmax(1,f2cmin(NB,N)). */
/* > \endverbatim */
/* > */
/* > \param[out] WORK */
/* > \verbatim */
/* >          (workspace) COMPLEX array, dimension (MAX(1,LWORK)) */
/* >          On exit, if INFO = 0, WORK(1) returns the optimal LWORK. */
/* > \endverbatim */
/* > */
/* > \param[in] LWORK */
/* > \verbatim */
/* >          The dimension of the array WORK. */
/* >          LWORK >= NBLOCAL * MAX(NBLOCAL,(N-NBLOCAL)), */
/* >          where NBLOCAL=MIN(NB,N). */
/* >          If LWORK = -1, then a workspace query is assumed. */
/* >          The routine only calculates the optimal size of the WORK */
/* >          array, returns this value as the first entry of the WORK */
/* >          array, and no error message related to LWORK is issued */
/* >          by XERBLA. */
/* > \endverbatim */
/* > */
/* > \param[out] INFO */
/* > \verbatim */
/* >          INFO is INTEGER */
/* >          = 0:  successful exit */
/* >          < 0:  if INFO = -i, the i-th argument had an illegal value */
/* > \endverbatim */
/* > */
/*  Authors: */
/*  ======== */

/* > \author Univ. of Tennessee */
/* > \author Univ. of California Berkeley */
/* > \author Univ. of Colorado Denver */
/* > \author NAG Ltd. */

/* > \ingroup complexOTHERcomputational */

/* > \par Contributors: */
/*  ================== */
/* > */
/* > \verbatim */
/* > */
/* > November 2020, Igor Kozachenko, */
/* >                Computer Science Division, */
/* >                University of California, Berkeley */
/* > */
/* > \endverbatim */
/* > */
/*  ===================================================================== */
/* Subroutine */ int cungtsqr_row_(integer *m, integer *n, integer *mb, 
	integer *nb, complex *a, integer *lda, complex *t, integer *ldt, 
	complex *work, integer *lwork, integer *info)
{
    /* System generated locals */
    integer a_dim1, a_offset, t_dim1, t_offset, i__1, i__2, i__3, i__4, i__5;
    complex q__1;

    /* Local variables */
    integer jb_t__, itmp, lworkopt;
    complex dummy[1]	/* was [1][1] */;
    integer ib_bottom__, ib, kb;
    extern /* Subroutine */ int claset_(char *, integer *, integer *, complex 
	    *, complex *, complex *, integer *), xerbla_(char *, 
	    integer *, ftnlen);
    integer mb1, mb2, m_plus_one__;
    logical lquery;
    integer num_all_row_blocks__, imb, knb, nblocal, kb_last__;
    extern /* Subroutine */ int clarfb_gett_(char *, integer *, integer *, 
	    integer *, complex *, integer *, complex *, integer *, complex *, 
	    integer *, complex *, integer *);


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


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


/*     Test the input parameters */

    /* Parameter adjustments */
    a_dim1 = *lda;
    a_offset = 1 + a_dim1 * 1;
    a -= a_offset;
    t_dim1 = *ldt;
    t_offset = 1 + t_dim1 * 1;
    t -= t_offset;
    --work;

    /* Function Body */
    *info = 0;
    lquery = *lwork == -1;
    if (*m < 0) {
	*info = -1;
    } else if (*n < 0 || *m < *n) {
	*info = -2;
    } else if (*mb <= *n) {
	*info = -3;
    } else if (*nb < 1) {
	*info = -4;
    } else if (*lda < f2cmax(1,*m)) {
	*info = -6;
    } else /* if(complicated condition) */ {
/* Computing MAX */
	i__1 = 1, i__2 = f2cmin(*nb,*n);
	if (*ldt < f2cmax(i__1,i__2)) {
	    *info = -8;
	} else if (*lwork < 1 && ! lquery) {
	    *info = -10;
	}
    }

    nblocal = f2cmin(*nb,*n);

/*     Determine the workspace size. */

    if (*info == 0) {
/* Computing MAX */
	i__1 = nblocal, i__2 = *n - nblocal;
	lworkopt = nblocal * f2cmax(i__1,i__2);
    }

/*     Handle error in the input parameters and handle the workspace query. */

    if (*info != 0) {
	i__1 = -(*info);
	xerbla_("CUNGTSQR_ROW", &i__1, (ftnlen)12);
	return 0;
    } else if (lquery) {
	q__1.r = (real) lworkopt, q__1.i = 0.f;
	work[1].r = q__1.r, work[1].i = q__1.i;
	return 0;
    }

/*     Quick return if possible */

    if (f2cmin(*m,*n) == 0) {
	q__1.r = (real) lworkopt, q__1.i = 0.f;
	work[1].r = q__1.r, work[1].i = q__1.i;
	return 0;
    }

/*     (0) Set the upper-triangular part of the matrix A to zero and */
/*     its diagonal elements to one. */

    claset_("U", m, n, &c_b2, &c_b1, &a[a_offset], lda);

/*     KB_LAST is the column index of the last column block reflector */
/*     in the matrices T and V. */

    kb_last__ = (*n - 1) / nblocal * nblocal + 1;


/*     (1) Bottom-up loop over row blocks of A, except the top row block. */
/*     NOTE: If MB>=M, then the loop is never executed. */

    if (*mb < *m) {

/*        MB2 is the row blocking size for the row blocks before the */
/*        first top row block in the matrix A. IB is the row index for */
/*        the row blocks in the matrix A before the first top row block. */
/*        IB_BOTTOM is the row index for the last bottom row block */
/*        in the matrix A. JB_T is the column index of the corresponding */
/*        column block in the matrix T. */

/*        Initialize variables. */

/*        NUM_ALL_ROW_BLOCKS is the number of row blocks in the matrix A */
/*        including the first row block. */

	mb2 = *mb - *n;
	m_plus_one__ = *m + 1;
	itmp = (*m - *mb - 1) / mb2;
	ib_bottom__ = itmp * mb2 + *mb + 1;
	num_all_row_blocks__ = itmp + 2;
	jb_t__ = num_all_row_blocks__ * *n + 1;

	i__1 = *mb + 1;
	i__2 = -mb2;
	for (ib = ib_bottom__; i__2 < 0 ? ib >= i__1 : ib <= i__1; ib += i__2)
		 {

/*           Determine the block size IMB for the current row block */
/*           in the matrix A. */

/* Computing MIN */
	    i__3 = m_plus_one__ - ib;
	    imb = f2cmin(i__3,mb2);

/*           Determine the column index JB_T for the current column block */
/*           in the matrix T. */

	    jb_t__ -= *n;

/*           Apply column blocks of H in the row block from right to left. */

/*           KB is the column index of the current column block reflector */
/*           in the matrices T and V. */

	    i__3 = -nblocal;
	    for (kb = kb_last__; i__3 < 0 ? kb >= 1 : kb <= 1; kb += i__3) {

/*              Determine the size of the current column block KNB in */
/*              the matrices T and V. */

/* Computing MIN */
		i__4 = nblocal, i__5 = *n - kb + 1;
		knb = f2cmin(i__4,i__5);

		i__4 = *n - kb + 1;
		clarfb_gett_("I", &imb, &i__4, &knb, &t[(jb_t__ + kb - 1) * 
			t_dim1 + 1], ldt, &a[kb + kb * a_dim1], lda, &a[ib + 
			kb * a_dim1], lda, &work[1], &knb);

	    }

	}

    }

/*     (2) Top row block of A. */
/*     NOTE: If MB>=M, then we have only one row block of A of size M */
/*     and we work on the entire matrix A. */

    mb1 = f2cmin(*mb,*m);

/*     Apply column blocks of H in the top row block from right to left. */

/*     KB is the column index of the current block reflector in */
/*     the matrices T and V. */

    i__2 = -nblocal;
    for (kb = kb_last__; i__2 < 0 ? kb >= 1 : kb <= 1; kb += i__2) {

/*        Determine the size of the current column block KNB in */
/*        the matrices T and V. */

/* Computing MIN */
	i__1 = nblocal, i__3 = *n - kb + 1;
	knb = f2cmin(i__1,i__3);

	if (mb1 - kb - knb + 1 == 0) {

/*           In SLARFB_GETT parameters, when M=0, then the matrix B */
/*           does not exist, hence we need to pass a dummy array */
/*           reference DUMMY(1,1) to B with LDDUMMY=1. */

	    i__1 = *n - kb + 1;
	    clarfb_gett_("N", &c__0, &i__1, &knb, &t[kb * t_dim1 + 1], ldt, &
		    a[kb + kb * a_dim1], lda, dummy, &c__1, &work[1], &knb);
	} else {
	    i__1 = mb1 - kb - knb + 1;
	    i__3 = *n - kb + 1;
	    clarfb_gett_("N", &i__1, &i__3, &knb, &t[kb * t_dim1 + 1], ldt, &
		    a[kb + kb * a_dim1], lda, &a[kb + knb + kb * a_dim1], lda,
		     &work[1], &knb);
	}

    }

    q__1.r = (real) lworkopt, q__1.i = 0.f;
    work[1].r = q__1.r, work[1].i = q__1.i;
    return 0;

/*     End of CUNGTSQR_ROW */

} /* cungtsqr_row__ */