#include "ir.h" #include "u_bits.h" #include /* qsort */ static const uchar loadszcls[] = { [Oloads8 - Oloads8] = 1|KI32<<4, [Oloadu8 - Oloads8] = 1|KI32<<4, [Oloads16 - Oloads8] = 2|KI32<<4, [Oloadu16 - Oloads8] = 2|KI32<<4, [Oloads32 - Oloads8] = 4|KI32<<4, [Oloadu32 - Oloads8] = 4|KI32<<4, [Oloadi64 - Oloads8] = 8|KI64<<4, [Oloadf32 - Oloads8] = 4|KF32<<4, [Oloadf64 - Oloads8] = 8|KF64<<4, }; static const uchar load2ext[] = { [Oloads8 - Oloads8] = Oexts8, [Oloadu8 - Oloads8] = Oextu8, [Oloads16 - Oloads8] = Oexts16, [Oloadu16 - Oloads8] = Oextu16, [Oloads32 - Oloads8] = Oexts32, [Oloadu32 - Oloads8] = Oextu32, [Oloadi64 - Oloads8] = Ocopy, }; static const uchar storesz[] = { [Ostorei8 - Ostorei8] = 1, [Ostorei16 - Ostorei8] = 2, [Ostorei32 - Ostorei8] = 4, [Ostorei64 - Ostorei8] = 8, [Ostoref32 - Ostorei8] = 4, [Ostoref64 - Ostorei8] = 8, }; #define loadsz(o) (loadszcls[(o) - Oloads8] & 0xF) #define loadcls(o) (loadszcls[(o) - Oloads8] >> 4) #define load2ext(o) (load2ext[(o) - Oloads8]) #define storesz(o) (storesz[(o) - Ostorei8]) /* Implements algorithm in 'Simple and Efficient Construction of Static Single Assignment' (Braun et al) */ typedef struct SSABuilder { Arena **arena; imap_of(Ref *) curdefs; /* map of var to (map of block to def of var) */ BitSet *sealed, /* set of sealed blocks */ *marked; /* blocks marked, for 'Marker Algorithm' in the paper */ int lastvisit; int nblk; } SSABuilder; static Ref readvar(SSABuilder *, int var, enum irclass, Block *); static Ref deltrivialphis(SSABuilder *sb, int var, Block *blk, Ref phiref) { assert(instrtab[phiref.i].op == Ophi); Ref *args = phitab.p[instrtab[phiref.i].l.i]; Ref same = {0}; for (int i = 0; i < blk->npred; ++i) { if (args[i].bits == same.bits || args[i].bits == phiref.bits) { continue; /* unique value or self-reference */ } if (same.bits != 0) { return phiref; /* non-trivial */ } same = args[i]; } if (same.bits == 0) same = UNDREF; /* the phi is unreachable or in the start block */ /* replace uses */ replcuses(phiref, same); Ref **pcurdefs = imap_get(&sb->curdefs, var); assert (pcurdefs); for (int i = blk->id; i < sb->nblk; ++i) { if ((*pcurdefs)[i].bits == phiref.bits) (*pcurdefs)[i] = same; } /* stops infinite recursion and marks phi for removal */ instrtab[phiref.i].op = Onop; /* recursively try to remove all phi users as they might have become trivial */ Redo: for (IRUse *use = instruse[phiref.i]; use; use = use->next) { if (use->u != USERJUMP && instrtab[use->u].op == Ophi && use->u != phiref.i) { Ref it = mkref(RTMP, use->u); Ref vphi2 = deltrivialphis(sb, var, use->blk, it); if (vphi2.bits != it.bits) { same = vphi2; /* deletion happened so phiref use may have changed */ goto Redo; } } } deluses(phiref.i); return same; } static Ref addphiargs(SSABuilder *sb, int var, enum irclass cls, Block *blk, Ref phiref) { Ref *args = phitab.p[instrtab[phiref.i].l.i]; for (int i = 0; i < blk->npred; ++i) { Block *pred = blkpred(blk, i); args[i] = readvar(sb, var, cls, pred); adduse(blk, phiref.i, args[i]); } instrtab[phiref.i].r.bits = 0; return deltrivialphis(sb, var, blk, phiref); } static void writevar(SSABuilder *sb, int var, Block *blk, Ref val) { Ref **pcurdefs; if (!(pcurdefs = imap_get(&sb->curdefs, var))) { pcurdefs = imap_set(&sb->curdefs, var, allocz(sb->arena, sb->nblk * sizeof(Ref), 0)); } if (val.t == RTMP) assert(instrtab[val.i].op != Onop); (*pcurdefs)[blk->id] = val; } enum { RPENDINGPHI = 7 }; static Ref readvarrec(SSABuilder *sb, int var, enum irclass cls, Block *blk) { Ref val; assert(blk->npred > 0); if (!bstest(sb->sealed, blk->id)) { /* unsealed block */ /* add pending phi */ val = insertphi(blk, cls); instrtab[val.i].r = mkref(RPENDINGPHI, var); } else if (blk->npred == 1) { /* trivial case when block has a single predecessor */ val = readvar(sb, var, cls, blkpred(blk, 0)); } else if (!bstest(sb->marked, blk->id)) { /* Marker Algorithm to optimize the common case where a control flow join doesn't * need a phi function for the variable: mark the block in case we recursively reach * it again, collect defs from predecessors, only create a phi if they differ. * This avoids creating temporary trivial phi functions in acyclic data flow */ bsset(sb->marked, blk->id); for (int i = 0; i < blk->npred; ++i) { Block *pred = blkpred(blk, i); Ref it = readvar(sb, var, cls, pred); if (!bstest(sb->marked, blk->id)) { /* recursion reached this blk again, use its phi */ Ref **pcurdefs = imap_get(&sb->curdefs, var); /* must have called writevar */ assert(*pcurdefs && (*pcurdefs)[blk->id].bits); return (*pcurdefs)[blk->id]; } if (i == 0) val = it; else if (it.bits != val.bits) /* found a different definition, must add phi */ goto AddPhi; } bsclr(sb->marked, blk->id); } else { /* reached block while recursing */ AddPhi: val = insertphi(blk, cls); writevar(sb, var, blk, val); val = addphiargs(sb, var, cls, blk, val); bsclr(sb->marked, blk->id); return val; } writevar(sb, var, blk, val); return val; } static Ref readvar(SSABuilder *sb, int var, enum irclass cls, Block *blk) { Ref **pcurdefs; if ((pcurdefs = imap_get(&sb->curdefs, var)) && (*pcurdefs)[blk->id].bits) return (*pcurdefs)[blk->id]; if (blk->npred == 0) /* entry block, var is read before being written to */ return UNDREF; return readvarrec(sb, var, cls, blk); } static bool trysealrec(SSABuilder *sb, Block *blk) { Recur: if (bstest(sb->sealed, blk->id)) return 1; if (blk->id > sb->lastvisit) return 0; markvisited(blk); for (int i = 0; i < blk->npred; ++i) { Block *p = blkpred(blk, i); if (wasvisited(p)) continue; if (p->id > sb->lastvisit) return 0; } bsset(sb->sealed, blk->id); for (int i = 0; i < blk->phi.n; ++i) { Instr *ins = &instrtab[blk->phi.p[i]]; if (ins->r.t == RPENDINGPHI) addphiargs(sb, ins->r.i, ins->cls, blk, mkref(RTMP, blk->phi.p[i])); } if (blk->s1 && !blk->s2) { blk = blk->s1; goto Recur; } else if (blk->s1 && blk->s2) { trysealrec(sb, blk->s1); blk = blk->s2; goto Recur; } return 1; } static void tryseal(SSABuilder *sb, Block *blk) { startbbvisit(); trysealrec(sb, blk); } static int blkfindins(Block *blk, int ins) { for (int i = 0; i < blk->phi.n; ++i) if (blk->phi.p[i] == ins) return -1; for (int i = 0; i < blk->ins.n; ++i) if (blk->ins.p[i] == ins) return i; assert(0 && "ins not in blk"); } static int rcmpuse(const void *a, const void *b) { /* postorder sort */ const IRUse *ua = *(IRUse **)a, *ub = *(IRUse **)b; Block *blk = ua->blk; if (ua->blk != ub->blk) return ub->blk->id - ua->blk->id; assert(ua->u != USERJUMP && ub->u != USERJUMP); return blkfindins(blk, ub->u) - blkfindins(blk, ua->u); } void mem2reg(Function *fn) { SSABuilder sb = { fn->passarena, .nblk = fn->nblk }; FREQUIRE(FNUSE); sb.sealed = allocz(sb.arena, BSSIZE(fn->nblk) * sizeof *sb.sealed, 0); sb.marked = allocz(sb.arena, BSSIZE(fn->nblk) * sizeof *sb.sealed, 0); sortrpo(fn); Block *blk = fn->entry; do { for (int i = 0; i < blk->ins.n; ++i) { enum irclass k = 0; int sz = 0; enum op ext = Ocopy; int var = blk->ins.p[i]; Instr *ins = &instrtab[var]; IRUse *use; /* find allocas only used in loads/stores of uniform size */ if (!oisalloca(ins->op) || !(use = instruse[var])) continue; IRUse *usesbuf[64]; vec_of(IRUse *) uses = VINIT(usesbuf, countof(usesbuf)); do { if (use->u == USERJUMP) goto Skip; Instr *m = &instrtab[use->u]; if (oisload(m->op) && (!sz || sz == loadsz(m->op))) { sz = loadsz(m->op); k = loadcls(m->op); if (sz < 4) ext = load2ext(m->op); } else if (oisstore(m->op) && m->l.bits == mkref(RTMP, var).bits && (!sz || sz == storesz(m->op))) { sz = storesz(m->op); } else goto Skip; vpush(&uses, use); } while ((use = use->next)); /* visit uses in rpo */ qsort(uses.p, uses.n, sizeof *uses.p, rcmpuse); for (int i = uses.n-1; i >= 0; --i) { use = uses.p[i]; ins = &instrtab[use->u]; if (oisstore(ins->op)) { writevar(&sb, var, use->blk, ins->r); *ins = mkinstr0(Onop,0); } else if (oisload(ins->op)) { Ref val = readvar(&sb, var, k = ins->cls, use->blk); adduse(use->blk, use->u, val); if (ext != Ocopy && isintcon(val)) { /* fold constant int extension */ val = irunop(NULL, ext, k, val); assert(val.bits); ext = Ocopy; } *ins = mkinstr1(ext, k, val); } } /* remove alloca */ delinstr(blk, i--); Skip: vfree(&uses); } assert(sb.lastvisit == blk->id); ++sb.lastvisit; tryseal(&sb, blk); } while ((blk = blk->lnext) != fn->entry); do { /* remove phis marked for deletion */ for (int i = 0; i < blk->phi.n; ++i) if (instrtab[blk->phi.p[i]].op == Onop) delphi(blk, i--); } while ((blk = blk->lnext) != fn->entry); imap_free(&sb.curdefs); if (ccopt.dbg.m) { bfmt(ccopt.dbgout, "<< After mem2reg >>\n"); irdump(fn); } } /* vim:set ts=3 sw=3 expandtab: */