src/rt/rc3.c

#ifndef lint
static const char RCSid[] = "$Id: rc3.c,v 2.17 2012/11/15 15:26:52 greg Exp $";
#endif
/*
 * Accumulate ray contributions for a set of materials
 * Controlling process for multiple children
 */

#include <signal.h>
#include "rcontrib.h"
#include "selcall.h"

#define MAXIQ           (int)(PIPE_BUF/(sizeof(FVECT)*2))

/* Modifier contribution queue (results waiting to be output) */
typedef struct s_binq {
        RNUMBER         ndx;            /* index for this entry */
        RNUMBER         nadded;         /* accumulated so far */
        struct s_binq   *next;          /* next in queue */
        MODCONT         *mca[1];        /* contrib. array (extends struct) */
} BINQ;

static BINQ     *out_bq = NULL;         /* output bin queue */
static BINQ     *free_bq = NULL;        /* free queue entries */

static struct {
        RNUMBER r1;                     /* assigned ray starting index */
        SUBPROC pr;                     /* PID, i/o descriptors */
        FILE    *infp;                  /* file pointer to read from process */
        int     nr;                     /* number of rays to sum (0 if free) */
} kida[MAXPROCESS];             /* our child processes */


/* Get new bin queue entry */
static BINQ *
new_binq()
{
        BINQ    *bp;
        int     i;

        if (free_bq != NULL) {          /* something already available? */
                bp = free_bq;
                free_bq = bp->next;
                bp->next = NULL;
                bp->nadded = 0;
                return(bp);
        }
                                        /* else allocate fresh */
        bp = (BINQ *)malloc(sizeof(BINQ) + sizeof(MODCONT *)*(nmods-1));
        if (bp == NULL)
                goto memerr;
        for (i = nmods; i--; ) {
                MODCONT *mp = (MODCONT *)lu_find(&modconttab,modname[i])->data;
                bp->mca[i] = (MODCONT *)malloc(sizeof(MODCONT) +
                                                sizeof(DCOLOR)*(mp->nbins-1));
                if (bp->mca[i] == NULL)
                        goto memerr;
                memcpy(bp->mca[i], mp, sizeof(MODCONT)-sizeof(DCOLOR));
                /* memset(bp->mca[i]->cbin, 0, sizeof(DCOLOR)*mp->nbins); */
        }
        bp->ndx = 0;
        bp->nadded = 0;
        bp->next = NULL;
        return(bp);
memerr:
        error(SYSTEM, "out of memory in new_binq()");
        return(NULL);
}


/* Free a bin queue entry */
static void
free_binq(BINQ *bp)
{
        int     i;

        if (bp == NULL) {               /* signal to release our free list */
                while ((bp = free_bq) != NULL) {
                        free_bq = bp->next;
                        for (i = nmods; i--; )
                                free(bp->mca[i]);
                                /* Note: we don't own bp->mca[i]->binv */
                        free(bp);
                }
                return;
        }
                                        /* clear sums for next use */
/*      for (i = nmods; i--; )
                memset(bp->mca[i]->cbin, 0, sizeof(DCOLOR)*bp->mca[i]->nbins);
*/
        bp->ndx = 0;
        bp->next = free_bq;             /* push onto free list */
        free_bq = bp;
}


/* Add modifier values to accumulation record in queue and clear */
static void
queue_modifiers()
{
        MODCONT *mpin, *mpout;
        int     i, j;

        if ((accumulate > 0) | (out_bq == NULL))
                error(CONSISTENCY, "bad call to queue_modifiers()");

        for (i = nmods; i--; ) {
                mpin = (MODCONT *)lu_find(&modconttab,modname[i])->data;
                mpout = out_bq->mca[i];
                for (j = mpout->nbins; j--; )
                        addcolor(mpout->cbin[j], mpin->cbin[j]);
                memset(mpin->cbin, 0, sizeof(DCOLOR)*mpin->nbins);
        }
        out_bq->nadded++;
}


/* Sum one modifier record into another (updates nadded) */
static void
add_modbin(BINQ *dst, BINQ *src)
{
        int     i, j;
        
        for (i = nmods; i--; ) {
                MODCONT *mpin = src->mca[i];
                MODCONT *mpout = dst->mca[i];
                for (j = mpout->nbins; j--; )
                        addcolor(mpout->cbin[j], mpin->cbin[j]);
        }
        dst->nadded += src->nadded;
}


/* Queue values for later output */
static void
queue_output(BINQ *bp)
{
        BINQ    *b_last, *b_cur;

        if (accumulate <= 0) {          /* just accumulating? */
                if (out_bq == NULL) {
                        bp->next = NULL;
                        out_bq = bp;
                } else {
                        add_modbin(out_bq, bp);
                        free_binq(bp);
                }
                return;
        }
        b_last = NULL;                  /* insert in output queue */
        for (b_cur = out_bq; b_cur != NULL && b_cur->ndx < bp->ndx;
                                b_cur = b_cur->next)
                b_last = b_cur;

        if (b_last != NULL) {
                bp->next = b_cur;
                b_last->next = bp;
        } else {
                bp->next = out_bq;
                out_bq = bp;
        }
        if (accumulate == 1)            /* no accumulation? */
                return;
        b_cur = out_bq;                 /* else merge accumulation entries */
        while (b_cur->next != NULL) {
                if (b_cur->nadded >= accumulate ||
                                (b_cur->ndx-1)/accumulate !=
                                (b_cur->next->ndx-1)/accumulate) {
                        b_cur = b_cur->next;
                        continue;
                }
                add_modbin(b_cur, b_cur->next);
                b_last = b_cur->next;
                b_cur->next = b_last->next;
                b_last->next = NULL;
                free_binq(b_last);
        }
}


/* Count number of records ready for output */
static int
queue_ready()
{
        int     nready = 0;
        BINQ    *bp;

        for (bp = out_bq; bp != NULL && bp->nadded >= accumulate &&
                                bp->ndx == lastdone+nready*accumulate+1;
                                bp = bp->next)
                ++nready;

        return(nready);
}


/* Catch up with output queue by producing ready results */
static int
output_catchup(int nmax)
{
        int     nout = 0;
        BINQ    *bp;
        int     i;
                                        /* output ready results */
        while (out_bq != NULL && out_bq->nadded >= accumulate
                                && out_bq->ndx == lastdone+1) {
                if ((nmax > 0) & (nout >= nmax))
                        break;
                bp = out_bq;                    /* pop off first entry */
                out_bq = bp->next;
                bp->next = NULL;
                for (i = 0; i < nmods; i++)     /* output record */
                        mod_output(bp->mca[i]);
                end_record();
                free_binq(bp);                  /* free this entry */
                lastdone += accumulate;
                ++nout;
        }
        return(nout);
}


/* Put a zero record in results queue & output */
void
put_zero_record(int ndx)
{
        BINQ    *bp = new_binq();
        int     i;

        for (i = nmods; i--; )
                memset(bp->mca[i]->cbin, 0, sizeof(DCOLOR)*bp->mca[i]->nbins);
        bp->ndx = ndx;
        bp->nadded = 1;
        queue_output(bp);
        output_catchup(0);
}


/* Get results from child process and add to queue */
static void
queue_results(int k)
{
        BINQ    *bq = new_binq();       /* get results holder */
        int     j;

        bq->ndx = kida[k].r1;
        bq->nadded = kida[k].nr;
                                        /* read from child */
        for (j = 0; j < nmods; j++)
                if (fread(bq->mca[j]->cbin, sizeof(DCOLOR), bq->mca[j]->nbins,
                                        kida[k].infp) != bq->mca[j]->nbins)
                        error(SYSTEM, "read error from render process");
                        
        queue_output(bq);               /* put results in output queue */
        kida[k].nr = 0;                 /* mark child as available */
}


/* callback to set output spec to NULL (stdout) */
static int
set_stdout(const LUENT *le, void *p)
{
        (*(MODCONT *)le->data).outspec = NULL;
        return(0);
}


/* Start child processes if we can */
int
in_rchild()
{
        int     rval;

        while (nchild < nproc) {        /* fork until target reached */
                errno = 0;
                rval = open_process(&kida[nchild].pr, NULL);
                if (rval < 0)
                        error(SYSTEM, "open_process() call failed");
                if (rval == 0) {        /* if in child, set up & return true */
                        lu_doall(&modconttab, &set_stdout, NULL);
                        lu_done(&ofiletab);
                        while (nchild--) {      /* don't share other pipes */
                                close(kida[nchild].pr.w);
                                fclose(kida[nchild].infp);
                        }
                        inpfmt = (sizeof(RREAL)==sizeof(double)) ? 'd' : 'f';
                        outfmt = 'd';
                        header = 0;
                        yres = 0;
                        raysleft = 0;
                        if (accumulate == 1) {
                                waitflush = xres = 1;
                                account = accumulate = 1;
                        } else {        /* parent controls accumulation */
                                waitflush = xres = 0;
                                account = accumulate = 0;
                        }
                        return(1);      /* return "true" in child */
                }
                if (rval != PIPE_BUF)
                        error(CONSISTENCY, "bad value from open_process()");
                                        /* connect to child's output */
                kida[nchild].infp = fdopen(kida[nchild].pr.r, "rb");
                if (kida[nchild].infp == NULL)
                        error(SYSTEM, "out of memory in in_rchild()");
#ifdef getc_unlocked
                flockfile(kida[nchild].infp);   /* avoid mutex overhead */
#endif
                kida[nchild++].nr = 0;  /* mark as available */
        }
        return(0);                      /* return "false" in parent */
}


/* Close child processes */
void
end_children(int immed)
{
        int     status;
        
        while (nchild > 0) {
                nchild--;
#ifdef SIGKILL
                if (immed)              /* error mode -- quick exit */
                        kill(kida[nchild].pr.pid, SIGKILL);
#endif
                if ((status = close_process(&kida[nchild].pr)) > 0 && !immed) {
                        sprintf(errmsg,
                                "rendering process returned bad status (%d)",
                                        status);
                        error(WARNING, errmsg);
                }
                fclose(kida[nchild].infp);
        }
}


/* Wait for the next available child, managing output queue simultaneously */
static int
next_child_nq(int flushing)
{
        static struct timeval   polling;
        struct timeval          *pmode;
        fd_set                  readset, errset;
        int                     i, n, nr, nqr;

        if (!flushing)                  /* see if there's one free */
                for (i = nchild; i--; )
                        if (!kida[i].nr)
                                return(i);

        nqr = queue_ready();            /* choose blocking mode or polling */
        if ((nqr > 0) & !flushing)
                pmode = &polling;
        else
                pmode = NULL;
tryagain:                               /* catch up with output? */
        if (pmode == &polling) {
                if (nqr > nchild)       /* don't get too far behind */
                        nqr -= output_catchup(nqr-nchild);
        } else if (nqr > 0)             /* clear output before blocking */
                nqr -= output_catchup(0);
                                        /* prepare select() call */
        FD_ZERO(&readset); FD_ZERO(&errset);
        n = nr = 0;
        for (i = nchild; i--; ) {
                if (kida[i].nr) {
                        FD_SET(kida[i].pr.r, &readset);
                        ++nr;
                }
                FD_SET(kida[i].pr.r, &errset);
                if (kida[i].pr.r >= n)
                        n = kida[i].pr.r + 1;
        }
        if (!nr)                        /* nothing to wait for? */
                return(-1);
        if ((nr > 1) | (pmode == &polling)) {
                errno = 0;
                nr = select(n, &readset, NULL, &errset, pmode);
                if (!nr) {
                        pmode = NULL;   /* try again, blocking this time */
                        goto tryagain;
                }
                if (nr < 0)
                        error(SYSTEM, "select() error in next_child_nq()");
        } else
                FD_ZERO(&errset);
        n = -1;                         /* read results from child(ren) */
        for (i = nchild; i--; ) {
                if (FD_ISSET(kida[i].pr.r, &errset))
                        error(USER, "rendering process died");
                if (FD_ISSET(kida[i].pr.r, &readset))
                        queue_results(n = i);
        }
        return(n);                      /* first available child */
}


/* Run parental oversight loop */
void
parental_loop()
{
        const int       qlimit = (accumulate == 1) ? 1 : MAXIQ-1;
        int             ninq = 0;
        FVECT           orgdir[2*MAXIQ];
        int             i, n;
                                        /* load rays from stdin & process */
#ifdef getc_unlocked
        flockfile(stdin);               /* avoid lock/unlock overhead */
#endif
        while (getvec(orgdir[2*ninq]) == 0 && getvec(orgdir[2*ninq+1]) == 0) {
                const int       zero_ray = orgdir[2*ninq+1][0] == 0.0 &&
                                                (orgdir[2*ninq+1][1] == 0.0) &
                                                (orgdir[2*ninq+1][2] == 0.0);
                ninq += !zero_ray;
                                /* Zero ray cannot go in input queue */
                if (zero_ray ? ninq : ninq >= qlimit ||
                            lastray/accumulate != (lastray+ninq)/accumulate) {
                        i = next_child_nq(0);           /* manages output */
                        n = ninq;
                        if (accumulate > 1)             /* need terminator? */
                                memset(orgdir[2*n++], 0, sizeof(FVECT)*2);
                        n *= sizeof(FVECT)*2;           /* send assignment */
                        if (writebuf(kida[i].pr.w, (char *)orgdir, n) != n)
                                error(SYSTEM, "pipe write error");
                        kida[i].r1 = lastray+1;
                        lastray += kida[i].nr = ninq;   /* mark as busy */
                        if (lastray < lastdone) {       /* RNUMBER wrapped? */
                                while (next_child_nq(1) >= 0)
                                        ;
                                lastray -= ninq;
                                lastdone = lastray %= accumulate;
                        }
                        ninq = 0;
                }
                if (zero_ray) {                         /* put bogus record? */
                        if ((yres <= 0) | (xres <= 1) &&
                                        (lastray+1) % accumulate == 0) {
                                while (next_child_nq(1) >= 0)
                                        ;               /* clear the queue */
                                lastdone = lastray = accumulate-1;
                                waitflush = 1;          /* flush next */
                        }
                        put_zero_record(++lastray);
                }
                if (raysleft && !--raysleft)
                        break;                          /* preemptive EOI */
        }
        while (next_child_nq(1) >= 0)           /* empty results queue */
                ;
        if (account < accumulate) {
                error(WARNING, "partial accumulation in final record");
                free_binq(out_bq);              /* XXX just ignore it */
                out_bq = NULL;
        }
        free_binq(NULL);                        /* clean up */
        lu_done(&ofiletab);
        if (raysleft)
                error(USER, "unexpected EOF on input");
}


/* Wait for the next available child by monitoring "to" pipes */
static int
next_child_ready()
{
        fd_set                  writeset, errset;
        int                     i, n;

        for (i = nchild; i--; )         /* see if there's one free first */
                if (!kida[i].nr)
                        return(i);
                                        /* prepare select() call */
        FD_ZERO(&writeset); FD_ZERO(&errset);
        n = 0;
        for (i = nchild; i--; ) {
                FD_SET(kida[i].pr.w, &writeset);
                FD_SET(kida[i].pr.r, &errset);
                if (kida[i].pr.w >= n)
                        n = kida[i].pr.w + 1;
                if (kida[i].pr.r >= n)
                        n = kida[i].pr.r + 1;
        }
        errno = 0;
        n = select(n, NULL, &writeset, &errset, NULL);
        if (n < 0)
                error(SYSTEM, "select() error in next_child_ready()");
        n = -1;                         /* identify waiting child */
        for (i = nchild; i--; ) {
                if (FD_ISSET(kida[i].pr.r, &errset))
                        error(USER, "rendering process died");
                if (FD_ISSET(kida[i].pr.w, &writeset))
                        kida[n = i].nr = 0;
        }
        return(n);                      /* first available child */
}


/* Modified parental loop for full accumulation mode (-c 0) */
void
feeder_loop()
{
        static int      ignore_warning_given = 0;
        int             ninq = 0;
        FVECT           orgdir[2*MAXIQ];
        int             i, n;
                                        /* load rays from stdin & process */
#ifdef getc_unlocked
        flockfile(stdin);               /* avoid lock/unlock overhead */
#endif
        while (getvec(orgdir[2*ninq]) == 0 && getvec(orgdir[2*ninq+1]) == 0) {
                if (orgdir[2*ninq+1][0] == 0.0 &&       /* asking for flush? */
                                (orgdir[2*ninq+1][1] == 0.0) &
                                (orgdir[2*ninq+1][2] == 0.0)) {
                        if (!ignore_warning_given++)
                                error(WARNING,
                                "dummy ray(s) ignored during accumulation\n");
                        continue;
                } 
                if (++ninq >= MAXIQ) {
                        i = next_child_ready();         /* get eager child */
                        n = sizeof(FVECT)*2 * ninq;     /* give assignment */
                        if (writebuf(kida[i].pr.w, (char *)orgdir, n) != n)
                                error(SYSTEM, "pipe write error");
                        kida[i].r1 = lastray+1;
                        lastray += kida[i].nr = ninq;
                        if (lastray < lastdone)         /* RNUMBER wrapped? */
                                lastdone = lastray = 0;
                        ninq = 0;
                }
                if (raysleft && !--raysleft)
                        break;                          /* preemptive EOI */
        }
        if (ninq) {                             /* polish off input */
                i = next_child_ready();
                n = sizeof(FVECT)*2 * ninq;
                if (writebuf(kida[i].pr.w, (char *)orgdir, n) != n)
                        error(SYSTEM, "pipe write error");
                kida[i].r1 = lastray+1;
                lastray += kida[i].nr = ninq;
                ninq = 0;
        }
        memset(orgdir, 0, sizeof(FVECT)*2);     /* get results */
        for (i = nchild; i--; ) {
                writebuf(kida[i].pr.w, (char *)orgdir, sizeof(FVECT)*2);
                queue_results(i);
        }
        if (recover)                            /* and from before? */
                queue_modifiers();
        end_children(0);                        /* free up file descriptors */
        for (i = 0; i < nmods; i++)
                mod_output(out_bq->mca[i]);     /* output accumulated record */
        end_record();
        free_binq(out_bq);                      /* clean up */
        out_bq = NULL;
        free_binq(NULL);
        lu_done(&ofiletab);
        if (raysleft)
                error(USER, "unexpected EOF on input");
}
Revision:	2.18
Committed:	Thu Nov 15 19:41:03 2012 UTC (11 years, 5 months ago) by greg
Content type:	text/plain
Branch:	MAIN
CVS Tags:	rad4R2P2, rad4R2, rad4R2P1
Changes since 2.17:	+20 -27 lines
Log Message:	Tweaks and fixes related to flushing with -c >1
#	User	Rev	Content
1	greg	2.1	#ifndef lint
2	greg	2.18	static const char RCSid[] = "$Id: rc3.c,v 2.17 2012/11/15 15:26:52 greg Exp $";
3	greg	2.1	#endif
4			/*
5			* Accumulate ray contributions for a set of materials
6			* Controlling process for multiple children
7			*/
8
9	greg	2.14	#include <signal.h>
10	greg	2.1	#include "rcontrib.h"
11			#include "selcall.h"
12
13	greg	2.10	#define MAXIQ (int)(PIPE_BUF/(sizeof(FVECT)*2))
14
15	greg	2.1	/* Modifier contribution queue (results waiting to be output) */
16			typedef struct s_binq {
17	greg	2.6	RNUMBER ndx; /* index for this entry */
18			RNUMBER nadded; /* accumulated so far */
19	greg	2.1	struct s_binq next; / next in queue */
20			MODCONT mca[1]; / contrib. array (extends struct) */
21			} BINQ;
22
23			static BINQ out_bq = NULL; / output bin queue */
24			static BINQ free_bq = NULL; / free queue entries */
25
26	greg	2.6	static struct {
27			RNUMBER r1; /* assigned ray starting index */
28			SUBPROC pr; /* PID, i/o descriptors */
29			FILE infp; / file pointer to read from process */
30	greg	2.10	int nr; /* number of rays to sum (0 if free) */
31	greg	2.6	} kida[MAXPROCESS]; /* our child processes */
32	greg	2.1
33
34	greg	2.4	/* Get new bin queue entry */
35	greg	2.1	static BINQ *
36			new_binq()
37			{
38	greg	2.4	BINQ *bp;
39	greg	2.1	int i;
40
41	greg	2.4	if (free_bq != NULL) { /* something already available? */
42			bp = free_bq;
43	greg	2.1	free_bq = bp->next;
44			bp->next = NULL;
45	greg	2.6	bp->nadded = 0;
46	greg	2.1	return(bp);
47			}
48			/* else allocate fresh */
49	greg	2.4	bp = (BINQ )malloc(sizeof(BINQ) + sizeof(MODCONT )*(nmods-1));
50	greg	2.1	if (bp == NULL)
51			goto memerr;
52			for (i = nmods; i--; ) {
53			MODCONT mp = (MODCONT )lu_find(&modconttab,modname[i])->data;
54			bp->mca[i] = (MODCONT *)malloc(sizeof(MODCONT) +
55			sizeof(DCOLOR)*(mp->nbins-1));
56			if (bp->mca[i] == NULL)
57			goto memerr;
58			memcpy(bp->mca[i], mp, sizeof(MODCONT)-sizeof(DCOLOR));
59			/* memset(bp->mca[i]->cbin, 0, sizeof(DCOLOR)mp->nbins); /
60			}
61			bp->ndx = 0;
62	greg	2.6	bp->nadded = 0;
63	greg	2.1	bp->next = NULL;
64			return(bp);
65			memerr:
66			error(SYSTEM, "out of memory in new_binq()");
67			return(NULL);
68			}
69
70
71			/* Free a bin queue entry */
72			static void
73			free_binq(BINQ *bp)
74			{
75			int i;
76
77			if (bp == NULL) { /* signal to release our free list */
78			while ((bp = free_bq) != NULL) {
79			free_bq = bp->next;
80			for (i = nmods; i--; )
81			free(bp->mca[i]);
82			/* Note: we don't own bp->mca[i]->binv */
83			free(bp);
84			}
85			return;
86			}
87			/* clear sums for next use */
88			/* for (i = nmods; i--; )
89			memset(bp->mca[i]->cbin, 0, sizeof(DCOLOR)*bp->mca[i]->nbins);
90			*/
91			bp->ndx = 0;
92			bp->next = free_bq; /* push onto free list */
93			free_bq = bp;
94			}
95
96
97			/* Add modifier values to accumulation record in queue and clear */
98	greg	2.10	static void
99	greg	2.1	queue_modifiers()
100			{
101			MODCONT mpin, mpout;
102			int i, j;
103
104			if ((accumulate > 0) \| (out_bq == NULL))
105			error(CONSISTENCY, "bad call to queue_modifiers()");
106
107			for (i = nmods; i--; ) {
108			mpin = (MODCONT *)lu_find(&modconttab,modname[i])->data;
109			mpout = out_bq->mca[i];
110			for (j = mpout->nbins; j--; )
111			addcolor(mpout->cbin[j], mpin->cbin[j]);
112			memset(mpin->cbin, 0, sizeof(DCOLOR)*mpin->nbins);
113			}
114	greg	2.5	out_bq->nadded++;
115	greg	2.1	}
116
117
118	greg	2.4	/* Sum one modifier record into another (updates nadded) */
119	greg	2.1	static void
120			add_modbin(BINQ dst, BINQ src)
121			{
122			int i, j;
123
124			for (i = nmods; i--; ) {
125			MODCONT *mpin = src->mca[i];
126			MODCONT *mpout = dst->mca[i];
127			for (j = mpout->nbins; j--; )
128			addcolor(mpout->cbin[j], mpin->cbin[j]);
129			}
130	greg	2.4	dst->nadded += src->nadded;
131	greg	2.1	}
132
133
134	greg	2.2	/* Queue values for later output */
135			static void
136	greg	2.1	queue_output(BINQ *bp)
137			{
138			BINQ b_last, b_cur;
139
140			if (accumulate <= 0) { /* just accumulating? */
141			if (out_bq == NULL) {
142			bp->next = NULL;
143			out_bq = bp;
144			} else {
145			add_modbin(out_bq, bp);
146			free_binq(bp);
147			}
148	greg	2.2	return;
149	greg	2.1	}
150	greg	2.10	b_last = NULL; /* insert in output queue */
151	greg	2.1	for (b_cur = out_bq; b_cur != NULL && b_cur->ndx < bp->ndx;
152			b_cur = b_cur->next)
153			b_last = b_cur;
154	greg	2.4
155	greg	2.1	if (b_last != NULL) {
156			bp->next = b_cur;
157			b_last->next = bp;
158			} else {
159			bp->next = out_bq;
160			out_bq = bp;
161			}
162	greg	2.3	if (accumulate == 1) /* no accumulation? */
163	greg	2.2	return;
164			b_cur = out_bq; /* else merge accumulation entries */
165			while (b_cur->next != NULL) {
166	greg	2.4	if (b_cur->nadded >= accumulate \|\|
167	greg	2.2	(b_cur->ndx-1)/accumulate !=
168			(b_cur->next->ndx-1)/accumulate) {
169			b_cur = b_cur->next;
170			continue;
171	greg	2.1	}
172	greg	2.2	add_modbin(b_cur, b_cur->next);
173			b_last = b_cur->next;
174			b_cur->next = b_last->next;
175			b_last->next = NULL;
176			free_binq(b_last);
177	greg	2.1	}
178	greg	2.2	}
179
180
181	greg	2.4	/* Count number of records ready for output */
182			static int
183			queue_ready()
184			{
185			int nready = 0;
186			BINQ *bp;
187
188			for (bp = out_bq; bp != NULL && bp->nadded >= accumulate &&
189			bp->ndx == lastdone+nready*accumulate+1;
190			bp = bp->next)
191			++nready;
192
193			return(nready);
194			}
195
196
197			/* Catch up with output queue by producing ready results */
198	greg	2.2	static int
199	greg	2.4	output_catchup(int nmax)
200	greg	2.2	{
201			int nout = 0;
202			BINQ *bp;
203			int i;
204	greg	2.10	/* output ready results */
205	greg	2.4	while (out_bq != NULL && out_bq->nadded >= accumulate
206			&& out_bq->ndx == lastdone+1) {
207			if ((nmax > 0) & (nout >= nmax))
208			break;
209	greg	2.2	bp = out_bq; /* pop off first entry */
210			out_bq = bp->next;
211			bp->next = NULL;
212	greg	2.1	for (i = 0; i < nmods; i++) /* output record */
213	greg	2.2	mod_output(bp->mca[i]);
214	greg	2.1	end_record();
215	greg	2.2	free_binq(bp); /* free this entry */
216	greg	2.1	lastdone += accumulate;
217			++nout;
218			}
219			return(nout);
220			}
221
222
223	greg	2.3	/* Put a zero record in results queue & output */
224	greg	2.1	void
225	greg	2.3	put_zero_record(int ndx)
226	greg	2.1	{
227			BINQ *bp = new_binq();
228			int i;
229
230			for (i = nmods; i--; )
231			memset(bp->mca[i]->cbin, 0, sizeof(DCOLOR)*bp->mca[i]->nbins);
232			bp->ndx = ndx;
233	greg	2.6	bp->nadded = 1;
234	greg	2.1	queue_output(bp);
235	greg	2.4	output_catchup(0);
236	greg	2.1	}
237
238
239	greg	2.6	/* Get results from child process and add to queue */
240			static void
241			queue_results(int k)
242			{
243			BINQ bq = new_binq(); / get results holder */
244			int j;
245
246			bq->ndx = kida[k].r1;
247			bq->nadded = kida[k].nr;
248			/* read from child */
249			for (j = 0; j < nmods; j++)
250			if (fread(bq->mca[j]->cbin, sizeof(DCOLOR), bq->mca[j]->nbins,
251			kida[k].infp) != bq->mca[j]->nbins)
252			error(SYSTEM, "read error from render process");
253
254			queue_output(bq); /* put results in output queue */
255			kida[k].nr = 0; /* mark child as available */
256			}
257
258
259	greg	2.1	/* callback to set output spec to NULL (stdout) */
260			static int
261			set_stdout(const LUENT le, void p)
262			{
263			((MODCONT )le->data).outspec = NULL;
264			return(0);
265			}
266
267
268			/* Start child processes if we can */
269			int
270			in_rchild()
271			{
272	greg	2.9	int rval;
273
274			while (nchild < nproc) { /* fork until target reached */
275	greg	2.3	errno = 0;
276	greg	2.9	rval = open_process(&kida[nchild].pr, NULL);
277			if (rval < 0)
278			error(SYSTEM, "open_process() call failed");
279			if (rval == 0) { /* if in child, set up & return true */
280	greg	2.18	lu_doall(&modconttab, &set_stdout, NULL);
281	greg	2.4	lu_done(&ofiletab);
282	greg	2.5	while (nchild--) { /* don't share other pipes */
283	greg	2.6	close(kida[nchild].pr.w);
284			fclose(kida[nchild].infp);
285	greg	2.4	}
286	greg	2.1	inpfmt = (sizeof(RREAL)==sizeof(double)) ? 'd' : 'f';
287			outfmt = 'd';
288			header = 0;
289			yres = 0;
290			raysleft = 0;
291	greg	2.6	if (accumulate == 1) {
292			waitflush = xres = 1;
293			account = accumulate = 1;
294			} else { /* parent controls accumulation */
295			waitflush = xres = 0;
296			account = accumulate = 0;
297			}
298	greg	2.9	return(1); /* return "true" in child */
299	greg	2.1	}
300	greg	2.9	if (rval != PIPE_BUF)
301			error(CONSISTENCY, "bad value from open_process()");
302			/* connect to child's output */
303			kida[nchild].infp = fdopen(kida[nchild].pr.r, "rb");
304	greg	2.6	if (kida[nchild].infp == NULL)
305	greg	2.2	error(SYSTEM, "out of memory in in_rchild()");
306	greg	2.3	#ifdef getc_unlocked
307	greg	2.6	flockfile(kida[nchild].infp); /* avoid mutex overhead */
308	greg	2.3	#endif
309	greg	2.6	kida[nchild++].nr = 0; /* mark as available */
310	greg	2.1	}
311	greg	2.9	return(0); /* return "false" in parent */
312	greg	2.1	}
313
314
315			/* Close child processes */
316			void
317	greg	2.12	end_children(int immed)
318	greg	2.1	{
319			int status;
320
321	greg	2.5	while (nchild > 0) {
322			nchild--;
323	greg	2.15	#ifdef SIGKILL
324	greg	2.12	if (immed) /* error mode -- quick exit */
325			kill(kida[nchild].pr.pid, SIGKILL);
326	greg	2.15	#endif
327	greg	2.12	if ((status = close_process(&kida[nchild].pr)) > 0 && !immed) {
328	greg	2.1	sprintf(errmsg,
329			"rendering process returned bad status (%d)",
330			status);
331			error(WARNING, errmsg);
332			}
333	greg	2.9	fclose(kida[nchild].infp);
334	greg	2.2	}
335	greg	2.1	}
336
337
338	greg	2.5	/* Wait for the next available child, managing output queue simultaneously */
339	greg	2.1	static int
340	greg	2.5	next_child_nq(int flushing)
341	greg	2.1	{
342	greg	2.2	static struct timeval polling;
343	greg	2.4	struct timeval *pmode;
344	greg	2.2	fd_set readset, errset;
345	greg	2.6	int i, n, nr, nqr;
346	greg	2.1
347	greg	2.5	if (!flushing) /* see if there's one free */
348	greg	2.1	for (i = nchild; i--; )
349	greg	2.6	if (!kida[i].nr)
350	greg	2.1	return(i);
351	greg	2.4
352	greg	2.5	nqr = queue_ready(); /* choose blocking mode or polling */
353			if ((nqr > 0) & !flushing)
354			pmode = &polling;
355			else
356	greg	2.4	pmode = NULL;
357	greg	2.5	tryagain: /* catch up with output? */
358			if (pmode == &polling) {
359			if (nqr > nchild) /* don't get too far behind */
360			nqr -= output_catchup(nqr-nchild);
361			} else if (nqr > 0) /* clear output before blocking */
362			nqr -= output_catchup(0);
363	greg	2.1	/* prepare select() call */
364			FD_ZERO(&readset); FD_ZERO(&errset);
365			n = nr = 0;
366			for (i = nchild; i--; ) {
367	greg	2.6	if (kida[i].nr) {
368			FD_SET(kida[i].pr.r, &readset);
369	greg	2.1	++nr;
370			}
371	greg	2.6	FD_SET(kida[i].pr.r, &errset);
372			if (kida[i].pr.r >= n)
373			n = kida[i].pr.r + 1;
374	greg	2.1	}
375	greg	2.3	if (!nr) /* nothing to wait for? */
376			return(-1);
377	greg	2.2	if ((nr > 1) \| (pmode == &polling)) {
378	greg	2.1	errno = 0;
379	greg	2.5	nr = select(n, &readset, NULL, &errset, pmode);
380			if (!nr) {
381	greg	2.2	pmode = NULL; /* try again, blocking this time */
382			goto tryagain;
383			}
384	greg	2.5	if (nr < 0)
385	greg	2.3	error(SYSTEM, "select() error in next_child_nq()");
386	greg	2.1	} else
387			FD_ZERO(&errset);
388			n = -1; /* read results from child(ren) */
389			for (i = nchild; i--; ) {
390	greg	2.6	if (FD_ISSET(kida[i].pr.r, &errset))
391	greg	2.1	error(USER, "rendering process died");
392	greg	2.6	if (FD_ISSET(kida[i].pr.r, &readset))
393			queue_results(n = i);
394	greg	2.1	}
395	greg	2.3	return(n); /* first available child */
396	greg	2.1	}
397
398
399			/* Run parental oversight loop */
400			void
401			parental_loop()
402			{
403	greg	2.17	const int qlimit = (accumulate == 1) ? 1 : MAXIQ-1;
404	greg	2.6	int ninq = 0;
405			FVECT orgdir[2*MAXIQ];
406			int i, n;
407	greg	2.1	/* load rays from stdin & process */
408			#ifdef getc_unlocked
409			flockfile(stdin); /* avoid lock/unlock overhead */
410			#endif
411	greg	2.6	while (getvec(orgdir[2ninq]) == 0 && getvec(orgdir[2ninq+1]) == 0) {
412	greg	2.18	const int zero_ray = orgdir[2*ninq+1][0] == 0.0 &&
413			(orgdir[2*ninq+1][1] == 0.0) &
414			(orgdir[2*ninq+1][2] == 0.0);
415			ninq += !zero_ray;
416			/* Zero ray cannot go in input queue */
417			if (zero_ray ? ninq : ninq >= qlimit \|\|
418	greg	2.6	lastray/accumulate != (lastray+ninq)/accumulate) {
419	greg	2.5	i = next_child_nq(0); /* manages output */
420	greg	2.6	n = ninq;
421	greg	2.18	if (accumulate > 1) /* need terminator? */
422	greg	2.6	memset(orgdir[2n++], 0, sizeof(FVECT)2);
423			n = sizeof(FVECT)2; /* send assignment */
424			if (writebuf(kida[i].pr.w, (char *)orgdir, n) != n)
425	greg	2.1	error(SYSTEM, "pipe write error");
426	greg	2.6	kida[i].r1 = lastray+1;
427			lastray += kida[i].nr = ninq; /* mark as busy */
428			if (lastray < lastdone) { /* RNUMBER wrapped? */
429			while (next_child_nq(1) >= 0)
430			;
431	greg	2.17	lastray -= ninq;
432			lastdone = lastray %= accumulate;
433	greg	2.6	}
434	greg	2.17	ninq = 0;
435	greg	2.1	}
436	greg	2.18	if (zero_ray) { /* put bogus record? */
437			if ((yres <= 0) \| (xres <= 1) &&
438			(lastray+1) % accumulate == 0) {
439			while (next_child_nq(1) >= 0)
440			; /* clear the queue */
441			lastdone = lastray = accumulate-1;
442			waitflush = 1; /* flush next */
443			}
444			put_zero_record(++lastray);
445			}
446	greg	2.1	if (raysleft && !--raysleft)
447			break; /* preemptive EOI */
448			}
449			while (next_child_nq(1) >= 0) /* empty results queue */
450			;
451	greg	2.10	if (account < accumulate) {
452			error(WARNING, "partial accumulation in final record");
453			free_binq(out_bq); /* XXX just ignore it */
454	greg	2.3	out_bq = NULL;
455	greg	2.1	}
456	greg	2.10	free_binq(NULL); /* clean up */
457			lu_done(&ofiletab);
458	greg	2.1	if (raysleft)
459			error(USER, "unexpected EOF on input");
460	greg	2.10	}
461
462
463			/* Wait for the next available child by monitoring "to" pipes */
464			static int
465			next_child_ready()
466			{
467			fd_set writeset, errset;
468	greg	2.11	int i, n;
469	greg	2.10
470			for (i = nchild; i--; ) /* see if there's one free first */
471			if (!kida[i].nr)
472			return(i);
473			/* prepare select() call */
474			FD_ZERO(&writeset); FD_ZERO(&errset);
475			n = 0;
476			for (i = nchild; i--; ) {
477			FD_SET(kida[i].pr.w, &writeset);
478			FD_SET(kida[i].pr.r, &errset);
479			if (kida[i].pr.w >= n)
480			n = kida[i].pr.w + 1;
481			if (kida[i].pr.r >= n)
482			n = kida[i].pr.r + 1;
483			}
484			errno = 0;
485			n = select(n, NULL, &writeset, &errset, NULL);
486			if (n < 0)
487			error(SYSTEM, "select() error in next_child_ready()");
488			n = -1; /* identify waiting child */
489			for (i = nchild; i--; ) {
490			if (FD_ISSET(kida[i].pr.r, &errset))
491			error(USER, "rendering process died");
492			if (FD_ISSET(kida[i].pr.w, &writeset))
493			kida[n = i].nr = 0;
494			}
495			return(n); /* first available child */
496			}
497
498
499			/* Modified parental loop for full accumulation mode (-c 0) */
500			void
501			feeder_loop()
502			{
503			static int ignore_warning_given = 0;
504			int ninq = 0;
505			FVECT orgdir[2*MAXIQ];
506			int i, n;
507			/* load rays from stdin & process */
508			#ifdef getc_unlocked
509			flockfile(stdin); /* avoid lock/unlock overhead */
510			#endif
511			while (getvec(orgdir[2ninq]) == 0 && getvec(orgdir[2ninq+1]) == 0) {
512			if (orgdir[2ninq+1][0] == 0.0 && / asking for flush? */
513			(orgdir[2*ninq+1][1] == 0.0) &
514			(orgdir[2*ninq+1][2] == 0.0)) {
515			if (!ignore_warning_given++)
516			error(WARNING,
517			"dummy ray(s) ignored during accumulation\n");
518			continue;
519			}
520			if (++ninq >= MAXIQ) {
521			i = next_child_ready(); /* get eager child */
522			n = sizeof(FVECT)2 ninq; /* give assignment */
523			if (writebuf(kida[i].pr.w, (char *)orgdir, n) != n)
524			error(SYSTEM, "pipe write error");
525			kida[i].r1 = lastray+1;
526			lastray += kida[i].nr = ninq;
527			if (lastray < lastdone) /* RNUMBER wrapped? */
528			lastdone = lastray = 0;
529	greg	2.18	ninq = 0;
530	greg	2.10	}
531			if (raysleft && !--raysleft)
532			break; /* preemptive EOI */
533			}
534			if (ninq) { /* polish off input */
535			i = next_child_ready();
536			n = sizeof(FVECT)2 ninq;
537			if (writebuf(kida[i].pr.w, (char *)orgdir, n) != n)
538			error(SYSTEM, "pipe write error");
539			kida[i].r1 = lastray+1;
540			lastray += kida[i].nr = ninq;
541			ninq = 0;
542			}
543	greg	2.13	memset(orgdir, 0, sizeof(FVECT)2); / get results */
544			for (i = nchild; i--; ) {
545			writebuf(kida[i].pr.w, (char )orgdir, sizeof(FVECT)2);
546	greg	2.10	queue_results(i);
547			}
548			if (recover) /* and from before? */
549			queue_modifiers();
550	greg	2.12	end_children(0); /* free up file descriptors */
551	greg	2.10	for (i = 0; i < nmods; i++)
552			mod_output(out_bq->mca[i]); /* output accumulated record */
553			end_record();
554			free_binq(out_bq); /* clean up */
555			out_bq = NULL;
556			free_binq(NULL);
557	greg	2.1	lu_done(&ofiletab);
558	greg	2.10	if (raysleft)
559			error(USER, "unexpected EOF on input");
560	greg	2.1	}