src/rt/rc3.c

#ifndef lint
static const char RCSid[] = "$Id: rc3.c,v 2.22 2016/08/18 00:52:48 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 SUBPROC  kidpr[MAXPROCESS];      /* our child processes */

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


/* 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);
*/
        if (bp->next != NULL)
                error(CONSISTENCY, "free_binq() handed list");
        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 (getbinary(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 (call only once in parent!) */
int
in_rchild()
{
        int     rval;

        while (nchild < nproc) {        /* fork until target reached */
                errno = 0;
                rval = open_process(&kidpr[nchild], 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(kidpr[nchild].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(kidpr[nchild].r, "rb");
                if (kida[nchild].infp == NULL)
                        error(SYSTEM, "out of memory in in_rchild()");
                kida[nchild++].nr = 0;  /* mark as available */
        }
#ifdef getc_unlocked
        for (rval = nchild; rval--; )   /* avoid mutex overhead */
                flockfile(kida[rval].infp);
#endif
        return(0);                      /* return "false" in parent */
}


/* Close child processes */
void
end_children(int immed)
{
        int     i;

#ifdef SIGKILL                          /* error mode -- quick exit */
        for (i = nchild*immed; i-- > 0; )
                kill(kidpr[nchild].pid, SIGKILL);
#endif
        if ((i = close_processes(kidpr, nchild)) > 0 && !immed) {
                sprintf(errmsg, "rendering process returned bad status (%d)",
                                        i);
                        error(WARNING, errmsg);
        }
        while (nchild-- > 0)
                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(kidpr[i].r, &readset);
                        ++nr;
                }
                FD_SET(kidpr[i].r, &errset);
                if (kidpr[i].r >= n)
                        n = kidpr[i].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(kidpr[i].r, &errset))
                        error(USER, "rendering process died");
                if (FD_ISSET(kidpr[i].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(kidpr[i].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(kidpr[i].w, &writeset);
                FD_SET(kidpr[i].r, &errset);
                if (kidpr[i].w >= n)
                        n = kidpr[i].w + 1;
                if (kidpr[i].r >= n)
                        n = kidpr[i].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(kidpr[i].r, &errset))
                        error(USER, "rendering process died");
                if (FD_ISSET(kidpr[i].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(kidpr[i].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(kidpr[i].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(kidpr[i].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:	2.23
Committed:	Sun May 27 18:35:57 2018 UTC (5 years, 11 months ago) by greg
Content type:	text/plain
Branch:	MAIN
CVS Tags:	rad5R2, rad5R3
Changes since 2.22:	+3 -1 lines
Log Message:	Added internal error check
#	User	Rev	Content
1	greg	2.1	#ifndef lint
2	greg	2.23	static const char RCSid[] = "$Id: rc3.c,v 2.22 2016/08/18 00:52:48 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	2.21	static SUBPROC kidpr[MAXPROCESS]; /* our child processes */
27
28	greg	2.6	static struct {
29			RNUMBER r1; /* assigned ray starting index */
30			FILE infp; / file pointer to read from process */
31	greg	2.10	int nr; /* number of rays to sum (0 if free) */
32	greg	2.21	} kida[MAXPROCESS]; /* our child process i/o */
33	greg	2.1
34
35	greg	2.4	/* Get new bin queue entry */
36	greg	2.1	static BINQ *
37			new_binq()
38			{
39	greg	2.4	BINQ *bp;
40	greg	2.1	int i;
41
42	greg	2.4	if (free_bq != NULL) { /* something already available? */
43			bp = free_bq;
44	greg	2.1	free_bq = bp->next;
45			bp->next = NULL;
46	greg	2.6	bp->nadded = 0;
47	greg	2.1	return(bp);
48			}
49			/* else allocate fresh */
50	greg	2.4	bp = (BINQ )malloc(sizeof(BINQ) + sizeof(MODCONT )*(nmods-1));
51	greg	2.1	if (bp == NULL)
52			goto memerr;
53			for (i = nmods; i--; ) {
54			MODCONT mp = (MODCONT )lu_find(&modconttab,modname[i])->data;
55			bp->mca[i] = (MODCONT *)malloc(sizeof(MODCONT) +
56			sizeof(DCOLOR)*(mp->nbins-1));
57			if (bp->mca[i] == NULL)
58			goto memerr;
59			memcpy(bp->mca[i], mp, sizeof(MODCONT)-sizeof(DCOLOR));
60			/* memset(bp->mca[i]->cbin, 0, sizeof(DCOLOR)mp->nbins); /
61			}
62			bp->ndx = 0;
63	greg	2.6	bp->nadded = 0;
64	greg	2.1	bp->next = NULL;
65			return(bp);
66			memerr:
67			error(SYSTEM, "out of memory in new_binq()");
68			return(NULL);
69			}
70
71
72			/* Free a bin queue entry */
73			static void
74			free_binq(BINQ *bp)
75			{
76			int i;
77
78			if (bp == NULL) { /* signal to release our free list */
79			while ((bp = free_bq) != NULL) {
80			free_bq = bp->next;
81			for (i = nmods; i--; )
82			free(bp->mca[i]);
83			/* Note: we don't own bp->mca[i]->binv */
84			free(bp);
85			}
86			return;
87			}
88			/* clear sums for next use */
89			/* for (i = nmods; i--; )
90			memset(bp->mca[i]->cbin, 0, sizeof(DCOLOR)*bp->mca[i]->nbins);
91			*/
92	greg	2.23	if (bp->next != NULL)
93			error(CONSISTENCY, "free_binq() handed list");
94	greg	2.1	bp->ndx = 0;
95			bp->next = free_bq; /* push onto free list */
96			free_bq = bp;
97			}
98
99
100			/* Add modifier values to accumulation record in queue and clear */
101	greg	2.10	static void
102	greg	2.1	queue_modifiers()
103			{
104			MODCONT mpin, mpout;
105			int i, j;
106
107			if ((accumulate > 0) \| (out_bq == NULL))
108			error(CONSISTENCY, "bad call to queue_modifiers()");
109
110			for (i = nmods; i--; ) {
111			mpin = (MODCONT *)lu_find(&modconttab,modname[i])->data;
112			mpout = out_bq->mca[i];
113			for (j = mpout->nbins; j--; )
114			addcolor(mpout->cbin[j], mpin->cbin[j]);
115			memset(mpin->cbin, 0, sizeof(DCOLOR)*mpin->nbins);
116			}
117	greg	2.5	out_bq->nadded++;
118	greg	2.1	}
119
120
121	greg	2.4	/* Sum one modifier record into another (updates nadded) */
122	greg	2.1	static void
123			add_modbin(BINQ dst, BINQ src)
124			{
125			int i, j;
126
127			for (i = nmods; i--; ) {
128			MODCONT *mpin = src->mca[i];
129			MODCONT *mpout = dst->mca[i];
130			for (j = mpout->nbins; j--; )
131			addcolor(mpout->cbin[j], mpin->cbin[j]);
132			}
133	greg	2.4	dst->nadded += src->nadded;
134	greg	2.1	}
135
136
137	greg	2.2	/* Queue values for later output */
138			static void
139	greg	2.1	queue_output(BINQ *bp)
140			{
141			BINQ b_last, b_cur;
142
143			if (accumulate <= 0) { /* just accumulating? */
144			if (out_bq == NULL) {
145			bp->next = NULL;
146			out_bq = bp;
147			} else {
148			add_modbin(out_bq, bp);
149			free_binq(bp);
150			}
151	greg	2.2	return;
152	greg	2.1	}
153	greg	2.10	b_last = NULL; /* insert in output queue */
154	greg	2.1	for (b_cur = out_bq; b_cur != NULL && b_cur->ndx < bp->ndx;
155			b_cur = b_cur->next)
156			b_last = b_cur;
157	greg	2.4
158	greg	2.1	if (b_last != NULL) {
159			bp->next = b_cur;
160			b_last->next = bp;
161			} else {
162			bp->next = out_bq;
163			out_bq = bp;
164			}
165	greg	2.3	if (accumulate == 1) /* no accumulation? */
166	greg	2.2	return;
167			b_cur = out_bq; /* else merge accumulation entries */
168			while (b_cur->next != NULL) {
169	greg	2.4	if (b_cur->nadded >= accumulate \|\|
170	greg	2.2	(b_cur->ndx-1)/accumulate !=
171			(b_cur->next->ndx-1)/accumulate) {
172			b_cur = b_cur->next;
173			continue;
174	greg	2.1	}
175	greg	2.2	add_modbin(b_cur, b_cur->next);
176			b_last = b_cur->next;
177			b_cur->next = b_last->next;
178			b_last->next = NULL;
179			free_binq(b_last);
180	greg	2.1	}
181	greg	2.2	}
182
183
184	greg	2.4	/* Count number of records ready for output */
185			static int
186			queue_ready()
187			{
188			int nready = 0;
189			BINQ *bp;
190
191			for (bp = out_bq; bp != NULL && bp->nadded >= accumulate &&
192			bp->ndx == lastdone+nready*accumulate+1;
193			bp = bp->next)
194			++nready;
195
196			return(nready);
197			}
198
199
200			/* Catch up with output queue by producing ready results */
201	greg	2.2	static int
202	greg	2.4	output_catchup(int nmax)
203	greg	2.2	{
204			int nout = 0;
205			BINQ *bp;
206			int i;
207	greg	2.10	/* output ready results */
208	greg	2.4	while (out_bq != NULL && out_bq->nadded >= accumulate
209			&& out_bq->ndx == lastdone+1) {
210			if ((nmax > 0) & (nout >= nmax))
211			break;
212	greg	2.2	bp = out_bq; /* pop off first entry */
213			out_bq = bp->next;
214			bp->next = NULL;
215	greg	2.1	for (i = 0; i < nmods; i++) /* output record */
216	greg	2.2	mod_output(bp->mca[i]);
217	greg	2.1	end_record();
218	greg	2.2	free_binq(bp); /* free this entry */
219	greg	2.1	lastdone += accumulate;
220			++nout;
221			}
222			return(nout);
223			}
224
225
226	greg	2.3	/* Put a zero record in results queue & output */
227	greg	2.1	void
228	greg	2.3	put_zero_record(int ndx)
229	greg	2.1	{
230			BINQ *bp = new_binq();
231			int i;
232
233			for (i = nmods; i--; )
234			memset(bp->mca[i]->cbin, 0, sizeof(DCOLOR)*bp->mca[i]->nbins);
235			bp->ndx = ndx;
236	greg	2.6	bp->nadded = 1;
237	greg	2.1	queue_output(bp);
238	greg	2.4	output_catchup(0);
239	greg	2.1	}
240
241
242	greg	2.6	/* Get results from child process and add to queue */
243			static void
244			queue_results(int k)
245			{
246			BINQ bq = new_binq(); / get results holder */
247			int j;
248
249			bq->ndx = kida[k].r1;
250			bq->nadded = kida[k].nr;
251			/* read from child */
252			for (j = 0; j < nmods; j++)
253	greg	2.22	if (getbinary(bq->mca[j]->cbin, sizeof(DCOLOR), bq->mca[j]->nbins,
254	greg	2.6	kida[k].infp) != bq->mca[j]->nbins)
255			error(SYSTEM, "read error from render process");
256
257			queue_output(bq); /* put results in output queue */
258			kida[k].nr = 0; /* mark child as available */
259			}
260
261
262	greg	2.1	/* callback to set output spec to NULL (stdout) */
263			static int
264			set_stdout(const LUENT le, void p)
265			{
266			((MODCONT )le->data).outspec = NULL;
267			return(0);
268			}
269
270
271	greg	2.20	/* Start child processes if we can (call only once in parent!) */
272	greg	2.1	int
273			in_rchild()
274			{
275	greg	2.9	int rval;
276
277			while (nchild < nproc) { /* fork until target reached */
278	greg	2.3	errno = 0;
279	greg	2.21	rval = open_process(&kidpr[nchild], NULL);
280	greg	2.9	if (rval < 0)
281			error(SYSTEM, "open_process() call failed");
282			if (rval == 0) { /* if in child, set up & return true */
283	greg	2.18	lu_doall(&modconttab, &set_stdout, NULL);
284	greg	2.4	lu_done(&ofiletab);
285	greg	2.5	while (nchild--) { /* don't share other pipes */
286	greg	2.21	close(kidpr[nchild].w);
287	greg	2.6	fclose(kida[nchild].infp);
288	greg	2.4	}
289	greg	2.1	inpfmt = (sizeof(RREAL)==sizeof(double)) ? 'd' : 'f';
290			outfmt = 'd';
291			header = 0;
292			yres = 0;
293			raysleft = 0;
294	greg	2.6	if (accumulate == 1) {
295			waitflush = xres = 1;
296			account = accumulate = 1;
297			} else { /* parent controls accumulation */
298			waitflush = xres = 0;
299			account = accumulate = 0;
300			}
301	greg	2.9	return(1); /* return "true" in child */
302	greg	2.1	}
303	greg	2.9	if (rval != PIPE_BUF)
304			error(CONSISTENCY, "bad value from open_process()");
305			/* connect to child's output */
306	greg	2.21	kida[nchild].infp = fdopen(kidpr[nchild].r, "rb");
307	greg	2.6	if (kida[nchild].infp == NULL)
308	greg	2.2	error(SYSTEM, "out of memory in in_rchild()");
309	greg	2.6	kida[nchild++].nr = 0; /* mark as available */
310	greg	2.1	}
311	greg	2.20	#ifdef getc_unlocked
312			for (rval = nchild; rval--; ) /* avoid mutex overhead */
313			flockfile(kida[rval].infp);
314			#endif
315	greg	2.9	return(0); /* return "false" in parent */
316	greg	2.1	}
317
318
319			/* Close child processes */
320			void
321	greg	2.12	end_children(int immed)
322	greg	2.1	{
323	greg	2.21	int i;
324
325			#ifdef SIGKILL /* error mode -- quick exit */
326			for (i = nchild*immed; i-- > 0; )
327			kill(kidpr[nchild].pid, SIGKILL);
328	greg	2.15	#endif
329	greg	2.21	if ((i = close_processes(kidpr, nchild)) > 0 && !immed) {
330			sprintf(errmsg, "rendering process returned bad status (%d)",
331			i);
332	greg	2.1	error(WARNING, errmsg);
333	greg	2.21	}
334			while (nchild-- > 0)
335	greg	2.9	fclose(kida[nchild].infp);
336	greg	2.1	}
337
338
339	greg	2.5	/* Wait for the next available child, managing output queue simultaneously */
340	greg	2.1	static int
341	greg	2.5	next_child_nq(int flushing)
342	greg	2.1	{
343	greg	2.2	static struct timeval polling;
344	greg	2.4	struct timeval *pmode;
345	greg	2.2	fd_set readset, errset;
346	greg	2.6	int i, n, nr, nqr;
347	greg	2.1
348	greg	2.5	if (!flushing) /* see if there's one free */
349	greg	2.1	for (i = nchild; i--; )
350	greg	2.6	if (!kida[i].nr)
351	greg	2.1	return(i);
352	greg	2.4
353	greg	2.5	nqr = queue_ready(); /* choose blocking mode or polling */
354			if ((nqr > 0) & !flushing)
355			pmode = &polling;
356			else
357	greg	2.4	pmode = NULL;
358	greg	2.5	tryagain: /* catch up with output? */
359			if (pmode == &polling) {
360			if (nqr > nchild) /* don't get too far behind */
361			nqr -= output_catchup(nqr-nchild);
362			} else if (nqr > 0) /* clear output before blocking */
363			nqr -= output_catchup(0);
364	greg	2.1	/* prepare select() call */
365			FD_ZERO(&readset); FD_ZERO(&errset);
366			n = nr = 0;
367			for (i = nchild; i--; ) {
368	greg	2.6	if (kida[i].nr) {
369	greg	2.21	FD_SET(kidpr[i].r, &readset);
370	greg	2.1	++nr;
371			}
372	greg	2.21	FD_SET(kidpr[i].r, &errset);
373			if (kidpr[i].r >= n)
374			n = kidpr[i].r + 1;
375	greg	2.1	}
376	greg	2.3	if (!nr) /* nothing to wait for? */
377			return(-1);
378	greg	2.2	if ((nr > 1) \| (pmode == &polling)) {
379	greg	2.1	errno = 0;
380	greg	2.5	nr = select(n, &readset, NULL, &errset, pmode);
381			if (!nr) {
382	greg	2.2	pmode = NULL; /* try again, blocking this time */
383			goto tryagain;
384			}
385	greg	2.5	if (nr < 0)
386	greg	2.3	error(SYSTEM, "select() error in next_child_nq()");
387	greg	2.1	} else
388			FD_ZERO(&errset);
389			n = -1; /* read results from child(ren) */
390			for (i = nchild; i--; ) {
391	greg	2.21	if (FD_ISSET(kidpr[i].r, &errset))
392	greg	2.1	error(USER, "rendering process died");
393	greg	2.21	if (FD_ISSET(kidpr[i].r, &readset))
394	greg	2.6	queue_results(n = i);
395	greg	2.1	}
396	greg	2.3	return(n); /* first available child */
397	greg	2.1	}
398
399
400			/* Run parental oversight loop */
401			void
402			parental_loop()
403			{
404	greg	2.17	const int qlimit = (accumulate == 1) ? 1 : MAXIQ-1;
405	greg	2.6	int ninq = 0;
406			FVECT orgdir[2*MAXIQ];
407			int i, n;
408	greg	2.1	/* load rays from stdin & process */
409			#ifdef getc_unlocked
410			flockfile(stdin); /* avoid lock/unlock overhead */
411			#endif
412	greg	2.6	while (getvec(orgdir[2ninq]) == 0 && getvec(orgdir[2ninq+1]) == 0) {
413	greg	2.18	const int zero_ray = orgdir[2*ninq+1][0] == 0.0 &&
414			(orgdir[2*ninq+1][1] == 0.0) &
415			(orgdir[2*ninq+1][2] == 0.0);
416			ninq += !zero_ray;
417			/* Zero ray cannot go in input queue */
418			if (zero_ray ? ninq : ninq >= qlimit \|\|
419	greg	2.6	lastray/accumulate != (lastray+ninq)/accumulate) {
420	greg	2.5	i = next_child_nq(0); /* manages output */
421	greg	2.6	n = ninq;
422	greg	2.18	if (accumulate > 1) /* need terminator? */
423	greg	2.6	memset(orgdir[2n++], 0, sizeof(FVECT)2);
424			n = sizeof(FVECT)2; /* send assignment */
425	greg	2.21	if (writebuf(kidpr[i].w, (char *)orgdir, n) != n)
426	greg	2.1	error(SYSTEM, "pipe write error");
427	greg	2.6	kida[i].r1 = lastray+1;
428			lastray += kida[i].nr = ninq; /* mark as busy */
429			if (lastray < lastdone) { /* RNUMBER wrapped? */
430			while (next_child_nq(1) >= 0)
431			;
432	greg	2.17	lastray -= ninq;
433			lastdone = lastray %= accumulate;
434	greg	2.6	}
435	greg	2.17	ninq = 0;
436	greg	2.1	}
437	greg	2.18	if (zero_ray) { /* put bogus record? */
438			if ((yres <= 0) \| (xres <= 1) &&
439			(lastray+1) % accumulate == 0) {
440			while (next_child_nq(1) >= 0)
441			; /* clear the queue */
442			lastdone = lastray = accumulate-1;
443			waitflush = 1; /* flush next */
444			}
445			put_zero_record(++lastray);
446			}
447	greg	2.1	if (raysleft && !--raysleft)
448			break; /* preemptive EOI */
449			}
450			while (next_child_nq(1) >= 0) /* empty results queue */
451			;
452	greg	2.10	if (account < accumulate) {
453			error(WARNING, "partial accumulation in final record");
454			free_binq(out_bq); /* XXX just ignore it */
455	greg	2.3	out_bq = NULL;
456	greg	2.1	}
457	greg	2.10	free_binq(NULL); /* clean up */
458			lu_done(&ofiletab);
459	greg	2.1	if (raysleft)
460			error(USER, "unexpected EOF on input");
461	greg	2.10	}
462
463
464			/* Wait for the next available child by monitoring "to" pipes */
465			static int
466			next_child_ready()
467			{
468			fd_set writeset, errset;
469	greg	2.11	int i, n;
470	greg	2.10
471			for (i = nchild; i--; ) /* see if there's one free first */
472			if (!kida[i].nr)
473			return(i);
474			/* prepare select() call */
475			FD_ZERO(&writeset); FD_ZERO(&errset);
476			n = 0;
477			for (i = nchild; i--; ) {
478	greg	2.21	FD_SET(kidpr[i].w, &writeset);
479			FD_SET(kidpr[i].r, &errset);
480			if (kidpr[i].w >= n)
481			n = kidpr[i].w + 1;
482			if (kidpr[i].r >= n)
483			n = kidpr[i].r + 1;
484	greg	2.10	}
485			errno = 0;
486			n = select(n, NULL, &writeset, &errset, NULL);
487			if (n < 0)
488			error(SYSTEM, "select() error in next_child_ready()");
489			n = -1; /* identify waiting child */
490			for (i = nchild; i--; ) {
491	greg	2.21	if (FD_ISSET(kidpr[i].r, &errset))
492	greg	2.10	error(USER, "rendering process died");
493	greg	2.21	if (FD_ISSET(kidpr[i].w, &writeset))
494	greg	2.10	kida[n = i].nr = 0;
495			}
496			return(n); /* first available child */
497			}
498
499
500			/* Modified parental loop for full accumulation mode (-c 0) */
501			void
502			feeder_loop()
503			{
504			static int ignore_warning_given = 0;
505			int ninq = 0;
506			FVECT orgdir[2*MAXIQ];
507			int i, n;
508			/* load rays from stdin & process */
509			#ifdef getc_unlocked
510			flockfile(stdin); /* avoid lock/unlock overhead */
511			#endif
512			while (getvec(orgdir[2ninq]) == 0 && getvec(orgdir[2ninq+1]) == 0) {
513			if (orgdir[2ninq+1][0] == 0.0 && / asking for flush? */
514			(orgdir[2*ninq+1][1] == 0.0) &
515			(orgdir[2*ninq+1][2] == 0.0)) {
516			if (!ignore_warning_given++)
517			error(WARNING,
518			"dummy ray(s) ignored during accumulation\n");
519			continue;
520			}
521			if (++ninq >= MAXIQ) {
522			i = next_child_ready(); /* get eager child */
523			n = sizeof(FVECT)2 ninq; /* give assignment */
524	greg	2.21	if (writebuf(kidpr[i].w, (char *)orgdir, n) != n)
525	greg	2.10	error(SYSTEM, "pipe write error");
526			kida[i].r1 = lastray+1;
527			lastray += kida[i].nr = ninq;
528			if (lastray < lastdone) /* RNUMBER wrapped? */
529			lastdone = lastray = 0;
530	greg	2.18	ninq = 0;
531	greg	2.10	}
532			if (raysleft && !--raysleft)
533			break; /* preemptive EOI */
534			}
535			if (ninq) { /* polish off input */
536			i = next_child_ready();
537			n = sizeof(FVECT)2 ninq;
538	greg	2.21	if (writebuf(kidpr[i].w, (char *)orgdir, n) != n)
539	greg	2.10	error(SYSTEM, "pipe write error");
540			kida[i].r1 = lastray+1;
541			lastray += kida[i].nr = ninq;
542			ninq = 0;
543			}
544	greg	2.13	memset(orgdir, 0, sizeof(FVECT)2); / get results */
545			for (i = nchild; i--; ) {
546	greg	2.21	writebuf(kidpr[i].w, (char )orgdir, sizeof(FVECT)2);
547	greg	2.10	queue_results(i);
548			}
549			if (recover) /* and from before? */
550			queue_modifiers();
551	greg	2.12	end_children(0); /* free up file descriptors */
552	greg	2.10	for (i = 0; i < nmods; i++)
553			mod_output(out_bq->mca[i]); /* output accumulated record */
554			end_record();
555			free_binq(out_bq); /* clean up */
556			out_bq = NULL;
557			free_binq(NULL);
558	greg	2.1	lu_done(&ofiletab);
559	greg	2.10	if (raysleft)
560			error(USER, "unexpected EOF on input");
561	greg	2.1	}