fixed indents

[libfirm] / ir / be / bespillbelady.c

/**
 * Author:      Daniel Grund
 * Date:                20.09.2005
 * Copyright:   (c) Universitaet Karlsruhe
 * Licence:     This file protected by GPL -  GNU GENERAL PUBLIC LICENSE.
 *
 */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif

#ifdef HAVE_ALLOCA_H
#include <alloca.h>
#endif

#ifdef HAVE_MALLOC_H
#include <malloc.h>
#endif

#include "obst.h"
#include "set.h"
#include "pset.h"
#include "irprintf_t.h"
#include "irgraph.h"
#include "irnode.h"
#include "irmode.h"
#include "irgwalk.h"
#include "iredges_t.h"
#include "ircons_t.h"
#include "irprintf.h"

#include "beutil.h"
#include "bearch.h"
#include "bespillbelady.h"
#include "beuses_t.h"
#include "besched_t.h"
#include "beirgmod.h"
#include "belive_t.h"
#include "benode_t.h"
#include "bechordal_t.h"

#define DBG_SPILL   1
#define DBG_WSETS   2
#define DBG_FIX     4
#define DBG_DECIDE  8
#define DBG_START  16
#define DBG_SLOTS  32
#define DBG_TRACE  64
#define DEBUG_LVL 0 //(DBG_START | DBG_DECIDE | DBG_WSETS | DBG_FIX | DBG_SPILL)
DEBUG_ONLY(static firm_dbg_module_t *dbg = NULL;)

#define MIN(a,b) (((a)<(b))?(a):(b))

typedef struct _workset_t workset_t;

typedef struct _belady_env_t {
        struct obstack ob;
        const arch_env_t *arch;
        const arch_register_class_t *cls;
        int n_regs;                     /** number of regs in this reg-class */

        workset_t *ws;          /**< the main workset used while processing a block. ob-allocated */
        be_uses_t *uses;        /**< env for the next-use magic */
        ir_node *instr;         /**< current instruction */
        unsigned instr_nr;      /**< current instruction number (relative to block start) */
        pset *used;                     /**< holds the values used (so far) in the current BB */
        pset *copies;           /**< holds all copies placed due to phi-spilling */

        spill_env_t *senv;      /* see bespill.h */
        pset *reloads;          /**< all reload nodes placed */
} belady_env_t;

struct _workset_t {
        belady_env_t *bel;
        int len;                        /**< current length */
        loc_t vals[1];          /**< inlined array of the values/distances in this working set */
};

void workset_print(const workset_t *w)
{
        int i;

        for(i = 0; i < w->len; ++i) {
                ir_printf("%+F %d\n", w->vals[i].irn, w->vals[i].time);
        }
}

/**
 * Alloc a new workset on obstack @p ob with maximum size @p max
 */
static INLINE workset_t *new_workset(struct obstack *ob, belady_env_t *bel) {
        workset_t *res;
        size_t size = sizeof(*res) + (bel->n_regs-1)*sizeof(res->vals[0]);
        res = obstack_alloc(ob, size);
        memset(res, 0, size);
        res->bel = bel;
        return res;
}

/**
 * Alloc a new instance on obstack and make it equal to @param ws
 */
static INLINE workset_t *workset_clone(struct obstack *ob, workset_t *ws) {
        workset_t *res;
        size_t size = sizeof(*res) + (ws->bel->n_regs-1)*sizeof(res->vals[0]);
        res = obstack_alloc(ob, size);
        memcpy(res, ws, size);
        return res;
}

/**
 * Do NOT alloc anything. Make @param tgt equal to @param src.
 * returns @param tgt for convinience
 */
static INLINE workset_t *workset_copy(workset_t *tgt, workset_t *src) {
        size_t size = sizeof(*src) + (src->bel->n_regs-1)*sizeof(src->vals[0]);
        memcpy(tgt, src, size);
        return tgt;
}

/**
 * Overwrites the current content array of @param ws with the
 * @param count locations given at memory @param locs.
 * Set the length of @param ws to count.
 */
#define workset_bulk_fill(ws, count, locs) memcpy(&(ws)->vals[0], locs, ((ws)->len=count)*sizeof(locs[0]));


/**
 * Inserts the value @p val into the workset, iff it is not
 * already contained. The workset must not be full.
 */
static INLINE void workset_insert(workset_t *ws, ir_node *val) {
        int i;
        /* check for current regclass */
        if (arch_get_irn_reg_class(ws->bel->arch, val, -1) != ws->bel->cls) {
                DBG((dbg, DBG_DECIDE, "Dropped %+F\n", val));
                return;
        }

        /* check if val is already contained */
        for(i=0; i<ws->len; ++i)
                if (ws->vals[i].irn == val)
                        return;

        /* insert val */
        assert(ws->len < ws->bel->n_regs && "Workset already full!");
        ws->vals[ws->len++].irn = val;
}

/**
 * Inserts all values in array @p vals of length @p cnt
 * into the workset. There must be enough space for the
 * entries.
 */
static INLINE void workset_bulk_insert(workset_t *ws, int cnt, ir_node **vals) {
        int i, o;

        for(o=0; o<cnt; ++o) {
                ir_node *val = vals[o];
                DBG((dbg, DBG_TRACE, "Bulk insert %+F\n", val));
                /* check for current regclass */
                if (arch_get_irn_reg_class(ws->bel->arch, val, -1) != ws->bel->cls) {
                        DBG((dbg, DBG_TRACE, "Wrong reg class\n"));
                        goto no_insert;
                }

                /* check if val is already contained */
                for(i=0; i<ws->len; ++i)
                        if (ws->vals[i].irn == val) {
                                DBG((dbg, DBG_TRACE, "Already contained\n"));
                                goto no_insert;
                        }

                /* insert val */
                assert(ws->len < ws->bel->n_regs && "Workset does not have enough room!");
                ws->vals[ws->len++].irn = val;
                DBG((dbg, DBG_TRACE, "Inserted\n"));

no_insert:
                /*epsilon statement :)*/;
        }
}

/**
 * Removes all entries from this workset
 */
#define workset_clear(ws) (ws)->len = 0;

/**
 * Removes the value @p val from the workset if present.
 */
static INLINE void workset_remove(workset_t *ws, ir_node *val) {
        int i;
        for(i=0; i<ws->len; ++i)
                if (ws->vals[i].irn == val) {
                        ws->vals[i] = ws->vals[--ws->len];
                        return;
                }
}

static INLINE int workset_contains(const workset_t *ws, const ir_node *val) {
        int i;
        for(i=0; i<ws->len; ++i)
                if (ws->vals[i].irn == val)
                        return 1;
        return 0;
}

/**
 * Iterates over all values in the working set.
 * @p ws The workset to iterate
 * @p v  A variable to put the current value in
 * @p i  An integer for internal use
 */
#define workset_foreach(ws, v, i)       for(i=0; \
                                                                                v=(i < ws->len) ? ws->vals[i].irn : NULL, i < ws->len; \
                                                                                ++i)

#define workset_set_time(ws, i, t) (ws)->vals[i].time=t
#define workset_set_length(ws, length) (ws)->len = length
#define workset_get_length(ws) ((ws)->len)
#define workset_get_val(ws, i) ((ws)->vals[i].irn)
#define workset_sort(ws) qsort((ws)->vals, (ws)->len, sizeof((ws)->vals[0]), loc_compare);

typedef struct _block_info_t {
        workset_t *ws_start, *ws_end;
} block_info_t;


static INLINE void *new_block_info(struct obstack *ob) {
        block_info_t *res = obstack_alloc(ob, sizeof(*res));
        res->ws_start = NULL;
        res->ws_end = NULL;

        return res;
}

#define get_block_info(blk)                     ((block_info_t *)get_irn_link(blk))
#define set_block_info(blk, info)       set_irn_link(blk, info)

static int is_mem_phi(const ir_node *irn, void *data) {
        workset_t *sws;
        ir_node *blk = get_nodes_block(irn);

        DBG((dbg, DBG_SPILL, "Is %+F a mem-phi?\n", irn));
        sws = get_block_info(blk)->ws_start;
        DBG((dbg, DBG_SPILL, "  %d\n", !workset_contains(sws, irn)));
        return !workset_contains(sws, irn);
}

/**
 * @return The distance to the next use
 *         Or 0 if irn is an ignore node
 */

static INLINE unsigned get_distance(belady_env_t *bel, const ir_node *from, unsigned from_step, const ir_node *def, int skip_from_uses)
{
        arch_irn_flags_t fl = arch_irn_get_flags(bel->arch, def);
        unsigned dist = be_get_next_use(bel->uses, from, from_step, def, skip_from_uses);

        if(!USES_IS_INIFINITE(dist) && (fl & (arch_irn_flags_ignore | arch_irn_flags_dont_spill)) != 0)
                return 0;

        return dist;
}

/**
 * Performs the actions necessary to grant the request that:
 * - new_vals can be held in registers
 * - as few as possible other values are disposed
 * - the worst values get disposed
 *
 * @p is_usage indicates that the values in new_vals are used (not defined)
 * In this case reloads must be performed
 */
static void displace(belady_env_t *bel, workset_t *new_vals, int is_usage) {
        ir_node *val;
        int i, len, max_allowed, demand, iter;
        workset_t *ws = bel->ws;
        ir_node **to_insert = alloca(bel->n_regs * sizeof(*to_insert));

        /*
         * 1. Identify the number of needed slots and the values to reload
         */
        demand = 0;
        workset_foreach(new_vals, val, iter) {
                /* mark value as used */
                if (is_usage)
                        pset_insert_ptr(bel->used, val);

                if (!workset_contains(ws, val)) {
                        DBG((dbg, DBG_DECIDE, "    insert %+F\n", val));
                        to_insert[demand++] = val;
                        if (is_usage)
                                be_add_reload(bel->senv, val, bel->instr);
                } else
                        DBG((dbg, DBG_DECIDE, "    skip %+F\n", val));
        }
        DBG((dbg, DBG_DECIDE, "    demand = %d\n", demand));

        /*
         * 2. Make room for at least 'demand' slots
         */
        len = workset_get_length(ws);
        max_allowed = bel->n_regs - demand;

        /* Only make more free room if we do not have enough */
        if (len > max_allowed) {
                /* get current next-use distance */
                for (i=0; i<ws->len; ++i)
                        workset_set_time(ws, i, get_distance(bel, bel->instr, bel->instr_nr, workset_get_val(ws, i), !is_usage));

                /* sort entries by increasing nextuse-distance*/
                workset_sort(ws);

                /* Logic for not needed live-ins: If a value is disposed
                   before its first usage, remove it from start workset */
                for (i=max_allowed; i<ws->len; ++i) {
                        ir_node *irn = ws->vals[i].irn;

                        if (!pset_find_ptr(bel->used, irn)) {
                                ir_node *curr_bb = get_nodes_block(bel->instr);
                                workset_t *ws_start = get_block_info(curr_bb)->ws_start;
                                workset_remove(ws_start, irn);

                                DBG((dbg, DBG_DECIDE, "    dispose %+F dumb\n", irn));
                        } else
                                DBG((dbg, DBG_DECIDE, "    dispose %+F\n", irn));
                }

                /* kill the last 'demand' entries in the array */
                workset_set_length(ws, max_allowed);
        }

        /*
         * 3. Insert the new values into the workset
         */
        workset_bulk_insert(bel->ws, demand, to_insert);
}

static void belady(ir_node *blk, void *env);

/**
 * Collects all values live-in at block @p blk and all phi results in this block.
 * Then it adds the best values (at most n_regs) to the blocks start_workset.
 * The phis among the remaining values get spilled: Introduce psudo-copies of
 *  their args to break interference and make it possible to spill them to the
 *  same spill slot.
 */
static block_info_t *compute_block_start_info(ir_node *blk, void *env) {
        belady_env_t *bel = env;
        ir_node *irn, *first;
        irn_live_t *li;
        int i, count, ws_count;
        loc_t loc, *starters;
        ir_graph *irg = get_irn_irg(blk);
        struct obstack ob;
        block_info_t *res = get_block_info(blk);

        /* Have we seen this block before? */
        if (res)
                return res;

        /* Create the block info for this block. */
        res = new_block_info(&bel->ob);
        set_block_info(blk, res);


        /* Get all values living at the block start sorted by next use*/
        obstack_init(&ob);

        DBG((dbg, DBG_START, "Living at start of %+F:\n", blk));
        first = sched_first(blk);
        count = 0;
        sched_foreach(blk, irn)
                if (is_Phi(irn) && arch_get_irn_reg_class(bel->arch, irn, -1) == bel->cls) {
                        loc.irn = irn;
                        loc.time = get_distance(bel, first, 0, irn, 0);
                        obstack_grow(&ob, &loc, sizeof(loc));
                        DBG((dbg, DBG_START, "    %+F:\n", irn));
                        count++;
                } else
                        break;

        live_foreach(blk, li)
                if (live_is_in(li) && arch_get_irn_reg_class(bel->arch, li->irn, -1) == bel->cls) {
                        loc.irn = (ir_node *)li->irn;
                        loc.time = get_distance(bel, first, 0, li->irn, 0);
                        obstack_grow(&ob, &loc, sizeof(loc));
                        DBG((dbg, DBG_START, "    %+F:\n", irn));
                        count++;
                }

        starters = obstack_finish(&ob);
        qsort(starters, count, sizeof(starters[0]), loc_compare);


        /* If we have only one predecessor, we want the start_set of blk to be the end_set of pred */
        if (get_Block_n_cfgpreds(blk) == 1 && blk != get_irg_start_block(get_irn_irg(blk))) {
                ir_node *pred_blk       = get_Block_cfgpred_block(blk, 0);
                block_info_t *pred_info = get_block_info(pred_blk);

                /* if pred block has not been processed yet, do it now */
                if (! pred_info) {
                        belady(pred_blk, bel);
                        pred_info = get_block_info(pred_blk);
                }

                /* now we have an end_set of pred */
                assert(pred_info->ws_end && "The recursive call (above) is supposed to compute an end_set");
                res->ws_start = workset_clone(&bel->ob, pred_info->ws_end);

        } else

        /* Else we want the start_set to be the values used 'the closest' */
        {
                /* Copy the best ones from starters to start workset */
                ws_count = MIN(count, bel->n_regs);
                res->ws_start = new_workset(&bel->ob, bel);
                workset_bulk_fill(res->ws_start, ws_count, starters);
        }


        /* The phis of this block which are not in the start set have to be spilled later.
         * Therefore we add temporary copies in the pred_blocks so the spills can spill
         * into the same spill slot.
         * After spilling these copies get deleted. */
        for (i=workset_get_length(res->ws_start); i<count; ++i) {
                int o, max;

                irn = starters[i].irn;
                if (!is_Phi(irn) || get_nodes_block(irn) != blk)
                        continue;

                DBG((dbg, DBG_START, "For %+F:\n", irn));

                for (max=get_irn_arity(irn), o=0; o<max; ++o) {
                        ir_node *arg = get_irn_n(irn, o);
                        ir_node *pred_block = get_Block_cfgpred_block(get_nodes_block(irn), o);
                        ir_node *cpy = be_new_Copy(bel->cls, irg, pred_block, arg);
                        pset_insert_ptr(bel->copies, cpy);
                        DBG((dbg, DBG_START, "    place a %+F of %+F in %+F\n", cpy, arg, pred_block));
                        sched_add_before(pred_block, cpy);
                        set_irn_n(irn, o, cpy);
                }
        }

        obstack_free(&ob, NULL);
        return res;
}


/**
 * For the given block @p blk, decide for each values
 * whether it is used from a register or is reloaded
 * before the use.
 */
static void belady(ir_node *blk, void *env) {
        belady_env_t *bel = env;
        workset_t *new_vals;
        ir_node *irn;
        int iter;
        block_info_t *blk_info;

        /* Don't do a block twice */
        if (get_block_info(blk))
                return;

        /* get the starting workset for this block */
        blk_info = compute_block_start_info(blk, bel);

        DBG((dbg, DBG_DECIDE, "\n"));
        DBG((dbg, DBG_DECIDE, "Decide for %+F\n", blk));

        workset_copy(bel->ws, blk_info->ws_start);
        DBG((dbg, DBG_WSETS, "Start workset for %+F:\n", blk));
        workset_foreach(bel->ws, irn, iter)
                DBG((dbg, DBG_WSETS, "  %+F\n", irn));

        /* process the block from start to end */
        DBG((dbg, DBG_WSETS, "Processing...\n"));
        bel->used = pset_new_ptr(32);
        bel->instr_nr = 0;
        new_vals = new_workset(&bel->ob, bel);
        sched_foreach(blk, irn) {
                assert(workset_get_length(bel->ws) <= bel->n_regs && "Too much values in workset!");


                /* projs are handled with the tuple value.
                 * Phis are no real instr (see insert_starters())
                 * instr_nr does not increase */
                if (is_Proj(irn) || is_Phi(irn)) {
                        DBG((dbg, DBG_DECIDE, "  ...%+F skipped\n", irn));
                        continue;
                }
                DBG((dbg, DBG_DECIDE, "  ...%+F\n", irn));

                /* set instruction in the workset */
                bel->instr = irn;

                /* allocate all values _used_ by this instruction */
                workset_clear(new_vals);
                workset_bulk_insert(new_vals, get_irn_arity(irn)+1, get_irn_in(irn));
                displace(bel, new_vals, 1);

                /* allocate all values _defined_ by this instruction */
                workset_clear(new_vals);
                if (get_irn_mode(irn) == mode_T) { /* special handling for tuples and projs */
                        ir_node *proj;
                        for(proj=sched_next(irn); is_Proj(proj); proj=sched_next(proj))
                                workset_insert(new_vals, proj);
                } else {
                        workset_insert(new_vals, irn);
                }
                displace(bel, new_vals, 0);

                bel->instr_nr++;
        }
        del_pset(bel->used);

        /* Remember end-workset for this block */
        blk_info->ws_end = workset_clone(&bel->ob, bel->ws);
        DBG((dbg, DBG_WSETS, "End workset for %+F:\n", blk));
        workset_foreach(blk_info->ws_end, irn, iter)
                DBG((dbg, DBG_WSETS, "  %+F\n", irn));
}

/**
 * 'decide' is block-local and makes assumptions
 * about the set of live-ins. Thus we must adapt the
 * live-outs to the live-ins at each block-border.
 */
static void fix_block_borders(ir_node *blk, void *env) {
        workset_t *wsb;
        belady_env_t *bel = env;
        int i, max, iter, iter2;

        DBG((dbg, DBG_FIX, "\n"));
        DBG((dbg, DBG_FIX, "Fixing %+F\n", blk));

        wsb = get_block_info(blk)->ws_start;

        /* process all pred blocks */
        for (i=0, max=get_irn_arity(blk); i<max; ++i) {
                ir_node *irnb, *irnp, *pred = get_Block_cfgpred_block(blk, i);
                workset_t *wsp = get_block_info(pred)->ws_end;

                DBG((dbg, DBG_FIX, "  Pred %+F\n", pred));

                workset_foreach(wsb, irnb, iter) {
                        /* if irnb is a phi of the current block we reload
                         * the corresponding argument, else irnb itself */
                        if(is_Phi(irnb) && blk == get_nodes_block(irnb))
                                irnb = get_irn_n(irnb, i);

                        /* Unknowns are available everywhere */
                        if(get_irn_opcode(irnb) == iro_Unknown)
                                continue;

                        /* check if irnb is in a register at end of pred */
                        workset_foreach(wsp, irnp, iter2)
                                if (irnb == irnp)
                                        goto next_value;

                        /* irnb is in memory at the end of pred, so we have to reload it */
                        DBG((dbg, DBG_FIX, "    reload %+F\n", irnb));
                        be_add_reload_on_edge(bel->senv, irnb, blk, i);

next_value:
                        /*epsilon statement :)*/;
                }
        }
}

/**
 * Removes all used reloads from bel->reloads.
 * The remaining nodes in bel->reloads will be removed from the graph.
 */
static void rescue_used_reloads(ir_node *irn, void *env) {
        pset *rlds = (pset *)env;
        if (pset_find_ptr(rlds, irn))
                pset_remove_ptr(rlds, irn);
}

/**
 * Removes all copies introduced for phi-spills
 */
static void remove_copies(belady_env_t *bel) {
        ir_node *irn;

        for (irn = pset_first(bel->copies); irn; irn = pset_next(bel->copies)) {
                ir_node *src, *user;

                assert(be_is_Copy(irn));
                assert(get_irn_n_edges(irn) == 1 && "This is not a copy introduced in 'compute_block_start_info()'. Who created it?");

                user = get_irn_edge(get_irn_irg(irn), irn, 0)->src;

                src = be_get_Copy_op(irn);
                set_irn_n(user, 0, src);
        }
}

/**
 * Finds all unused reloads and remove them from the schedule
 * Also removes spills if they are not used anymore after removing reloads
 */
static void remove_unused_reloads(ir_graph *irg, belady_env_t *bel) {
        ir_node *irn;

        irg_walk_graph(irg, rescue_used_reloads, NULL, bel->reloads);
        for(irn = pset_first(bel->reloads); irn; irn = pset_next(bel->reloads)) {
                ir_node *spill;
                DBG((dbg, DBG_SPILL, "Removing %+F before %+F in %+F\n", irn, sched_next(irn), get_nodes_block(irn)));

                if (be_is_Reload(irn))
                        spill = get_irn_n(irn, be_pos_Reload_mem);

                /* remove reload */
                set_irn_n(irn, 0, new_Bad());
                sched_remove(irn);

                if (be_is_Reload(irn)) {
                        /* if spill not used anymore, remove it too
                         * test of regclass is necessary since spill may be a phi-M */
                        if (get_irn_n_edges(spill) == 0 && bel->cls == arch_get_irn_reg_class(bel->arch, spill, -1)) {
                                set_irn_n(spill, 0, new_Bad());
                                sched_remove(spill);
                        }
                }
        }
}

void be_spill_belady(const be_chordal_env_t *chordal_env) {
        belady_env_t bel;

        FIRM_DBG_REGISTER(dbg, "ir.be.spillbelady");

        /* init belady env */
        obstack_init(&bel.ob);
        bel.arch      = chordal_env->birg->main_env->arch_env;
        bel.cls       = chordal_env->cls;
        bel.n_regs    = arch_register_class_n_regs(bel.cls);
        bel.ws        = new_workset(&bel.ob, &bel);
        bel.uses      = be_begin_uses(chordal_env->irg, chordal_env->birg->main_env->arch_env, bel.cls);
        bel.senv      = be_new_spill_env(chordal_env, is_mem_phi, NULL);
        DEBUG_ONLY(be_set_spill_env_dbg_module(bel.senv, dbg);)
        bel.reloads   = pset_new_ptr_default();
        bel.copies    = pset_new_ptr_default();

        DBG((dbg, LEVEL_1, "running on register class: %s\n", bel.cls->name));

        /* do the work */
        be_clear_links(chordal_env->irg);
        irg_block_walk_graph(chordal_env->irg, NULL, belady, &bel);
        irg_block_walk_graph(chordal_env->irg, fix_block_borders, NULL, &bel);
        be_insert_spills_reloads(bel.senv, bel.reloads);
        remove_unused_reloads(chordal_env->irg, &bel);
        remove_copies(&bel);

        /* clean up */
        del_pset(bel.reloads);
        be_delete_spill_env(bel.senv);
        be_end_uses(bel.uses);
        obstack_free(&bel.ob, NULL);
}