2 % (c) The GRASP/AQUA Project, Glasgow University, 1993-1994
4 \section[StgMacros]{C macros used in GHC-generated \tr{.hc} files}
11 %************************************************************************
13 \subsection[StgMacros-abbrev]{Abbreviatory(?) and general macros}
15 %************************************************************************
19 /* for function declarations */
20 #define STGFUN(f) F_ f(STG_NO_ARGS)
21 #define STATICFUN(f) static F_ f(STG_NO_ARGS)
23 /* for functions/data that are really external to this module */
24 #define EXTFUN(f) extern F_ f(STG_NO_ARGS)
25 #define EXTDATA(d) extern W_ d[]
26 #define EXTDATA_RO(d) extern const W_ d[] /* read-only */
28 /* for fwd decls to functions/data somewhere else in this module */
29 /* (identical for the mo') */
30 #define INTFUN(f) static F_ f(STG_NO_ARGS)
31 #define INTDATA(d) extern W_ d[]
32 #define INTDATA_RO(d) extern const W_ d[] /* read-only */
34 /* short forms of most of the above */
36 #define FN_(f) F_ f(STG_NO_ARGS)
37 #define IFN_(f) static F_ f(STG_NO_ARGS)
38 #define EF_(f) extern F_ f(STG_NO_ARGS)
39 #define ED_(d) extern W_ d[]
40 #define ED_RO_(d) extern const W_ d[] /* read-only */
41 #define IF_(f) static F_ f(STG_NO_ARGS)
43 /* GCC is uncooperative about the next one: */
44 /* But, the "extern" prevents initialisation... ADR */
46 #define ID_(d) extern W_ d[]
47 #define ID_RO_(d) extern const W_ d[] /* read-only */
49 #define ID_(d) static W_ d[]
50 #define ID_RO_(d) static const W_ d[] /* read-only */
54 General things; note: general-but-``machine-dependent'' macros are
55 given in \tr{StgMachDeps.lh}.
57 I_ STG_MAX PROTO((I_, I_)); /* GCC -Wall loves prototypes */
61 STG_MAX(I_ a, I_ b) { return((a >= b) ? a : b); }
62 /* NB: the naive #define macro version of STG_MAX
63 can lead to exponential CPP explosion, if you
64 have very-nested STG_MAXes.
68 Macros to combine two short words into a single
69 word and split such a word back into two.
71 Dependent on machine word size :-)
74 #define COMBINE_WORDS(word,short1,short2) \
76 ((packed_shorts *)&(word))->wu.s1 = short1; \
77 ((packed_shorts *)&(word))->wu.s2 = short2; \
80 #define SPLIT_WORD(word,short1,short2) \
82 short1 = ((packed_shorts *)&(word))->wu.s1; \
83 short2 = ((packed_shorts *)&(word))->wu.s2; \
88 %************************************************************************
90 \subsection[StgMacros-gen-stg]{General STGish macros}
92 %************************************************************************
94 Common sizes of vector-return tables.
96 Claim: don't need fwd decls for return pts in \tr{VECTBL*}, because
97 the AbsC flattener ensures that things come out sufficiently
101 #ifdef __STG_REV_TBLS__
102 #define UNVECTBL(staticp,label,a) /* nothing */
104 #define UNVECTBL(staticp,label,a) \
106 staticp const W_ label[] = { \
113 #if defined(USE_SPLIT_MARKERS)
114 #define __STG_SPLIT_MARKER(n) FN_(CAT2(__stg_split_marker,n)){ }
116 #define __STG_SPLIT_MARKER(n) /* nothing */
120 %************************************************************************
122 \subsection[StgMacros-exceptions]{Exception-checking macros}
124 %************************************************************************
126 Argument-satisfaction check, stack(s) overflow check, heap overflow
129 The @SUBTRACT(upper, lower)@ macros return a positive result in words
130 indicating the amount by which upper is above lower on the stack.
133 #define SUBTRACT_A_STK( upper, lower ) AREL( (lower) - (upper) )
134 #define SUBTRACT_B_STK( upper, lower ) BREL( (lower) - (upper) )
137 %************************************************************************
139 \subsubsection[StgMacros-arg-satis]{Argument-satisfaction checks}
141 %************************************************************************
143 @ARGS_CHK(n)@ sees of there are @n@ words of args on the A/B stack.
144 If not, it jumps to @UpdatePAP@.
146 @ARGS_CHK@ args are pre-directionified.
147 Notice that we do the comparisons in the form (x < a+n), for
148 some constant n. This generates more efficient code (with GCC at least)
152 #define ARGS_CHK_A(n) \
153 if (SuA /*SUBTRACT_A_STK( SpA, SuA )*/ < (SpA+(n))) { \
157 #define ARGS_CHK_A_LOAD_NODE(n, closure_addr) \
158 if (SuA /*SUBTRACT_A_STK( SpA, SuA )*/ < (SpA+(n))) { \
159 Node = (P_) closure_addr; \
163 #define ARGS_CHK_B(n) \
164 if (SpB /*SUBTRACT_B_STK( SpB, SuB )*/ < (SuB-(n))) { \
169 #define ARGS_CHK_B_LOAD_NODE(n, closure_addr) \
170 if (SpB /*SUBTRACT_B_STK( SpB, SuB )*/ < (SuB-(n))) { \
171 Node = (P_) closure_addr; \
176 %************************************************************************
178 \subsubsection[StgMacros-stk-chks]{Stack-overflow check}
180 %************************************************************************
182 @STK_CHK(a,b)@ [misc args omitted...] checks that we can allocate @a@
183 words of A stack and @b@ words of B stack. If not, it calls
184 @StackOverflow@ (which dies).
186 (It will be different in the parallel case.)
188 NB: args @a@ and @b@ are pre-direction-ified!
190 I_ SqueezeUpdateFrames PROTO((P_, P_, P_));
191 int sanityChk_StkO (P_ stko); /* ToDo: move to a sane place */
193 #if ! defined(CONCURRENT)
195 extern void StackOverflow(STG_NO_ARGS) STG_NORETURN;
197 #if STACK_CHECK_BY_PAGE_FAULT
199 #define STACK_OVERFLOW(liveness,hda,hdb,spa,spb,rtype,reenter) \
200 /* use memory protection instead; still need ticky-ness */
204 #define STACK_OVERFLOW(liveness,hda,hdb,spa,spb,rtype,reenter) \
205 ULTRASAFESTGCALL0(void,(void *),StackOverflow)
207 #endif /* not using page-faulting */
211 I_ StackOverflow PROTO((W_, W_));
214 * On a uniprocessor, we do *NOT* context switch on a stack overflow
215 * (though we may GC). Therefore, we never have to reenter node.
218 #define STACK_OVERFLOW(liveness,hda,hdb,spa,spb,rtype,reenter) \
219 DO_STACKOVERFLOW((hda+hdb)<<2|((rtype)<<1)|(reenter),((spa)<<20)|((spb)<<8)|(liveness))
221 #define STACK_OVERFLOW_HEADROOM(args,y) ((args) >> 2)
222 #define STACK_OVERFLOW_PRIM_RETURN(args,y) ((args) & 2)
223 #define STACK_OVERFLOW_REENTER(args,y) ((args) & 1)
225 #define STACK_OVERFLOW_AWORDS(x,args) (((args) >> 20) & 0x0fff)
226 #define STACK_OVERFLOW_BWORDS(x,args) (((args) >> 8) & 0x0fff)
227 #define STACK_OVERFLOW_LIVENESS(x,args) ((args) & 0xff)
229 #endif /* CONCURRENT */
231 #define STK_CHK(liveness_mask,a_headroom,b_headroom,spa,spb,ret_type,reenter)\
233 DO_ASTK_HWM(); /* ticky-ticky profiling */ \
235 if (STKS_OVERFLOW_OP((a_headroom) + 1, (b_headroom) + 1)) { \
236 STACK_OVERFLOW(liveness_mask,a_headroom,b_headroom,spa,spb,ret_type,reenter);\
241 %************************************************************************
243 \subsubsection[StgMacros-heap-chks]{Heap-overflow checks}
245 %************************************************************************
247 Please see the general discussion/commentary about ``what really
248 happens in a GC,'' in \tr{SMinterface.lh}.
251 void PerformGC PROTO((W_));
252 void RealPerformGC PROTO((W_ liveness, W_ reqsize, W_ always_reenter_node, rtsBool do_full_collection));
253 void checkInCCallGC(STG_NO_ARGS);
256 void StgPerformGarbageCollection(STG_NO_ARGS);
261 #define OR_MSG_PENDING /* never */
263 #define HEAP_OVERFLOW(liveness,n,reenter) \
265 DO_GC((((W_)n)<<8)|(liveness)); \
268 #define REQSIZE_BITMASK ((1L << ((BITS_IN(W_) - 8 + 1))) - 1)
269 #define HEAP_OVERFLOW_REQSIZE(args) (((args) >> 8) & REQSIZE_BITMASK)
270 #define HEAP_OVERFLOW_REENTER(args) 0
271 #define HEAP_OVERFLOW_LIVENESS(args) ((args) & 0xff)
273 #else /* CONCURRENT */
275 void ReallyPerformThreadGC PROTO((W_, rtsBool));
277 #define HEAP_OVERFLOW(liveness,n,reenter) \
279 DO_GC((((W_)(n))<<9)|((reenter)<<8)|(liveness)); \
282 #define REQSIZE_BITMASK ((1L << ((BITS_IN(W_) - 9 + 1))) - 1)
283 #define HEAP_OVERFLOW_REQSIZE(args) (((args) >> 9) & REQSIZE_BITMASK)
284 #define HEAP_OVERFLOW_REENTER(args) (((args) >> 8) & 0x1)
285 #define HEAP_OVERFLOW_LIVENESS(args) ((args) & 0xff)
289 #define OR_MSG_PENDING /* never */
293 extern int PacketsWaiting; /*Probes for incoming messages*/
294 extern int heapChkCounter; /*Not currently used! We check for messages when*/
295 /*a thread is resheduled PWT*/
296 /* #define OR_MSG_PENDING || (--heapChkCounter == 0 && PacketsWaiting())*/
297 #define OR_MSG_PENDING /* never */
300 #endif /* CONCURRENT */
302 #if 0 /* alpha_TARGET_ARCH */
303 #define CACHE_LINE 4 /* words */
304 #define LINES_AHEAD 3
305 #define PRE_FETCH(n) \
308 j = ((STG_VOLATILE StgInt *) Hp)[LINES_AHEAD * CACHE_LINE]; \
310 #define EXTRA_HEAP_WORDS (CACHE_LINE * LINES_AHEAD)
312 #define PRE_FETCH(reg)
313 #define EXTRA_HEAP_WORDS 0
317 #define HEAP_CHK(liveness_mask,n,reenter) \
319 /* TICKY_PARANOIA(__FILE__, __LINE__); */ \
320 /* THREAD_CONTEXT_SWITCH(liveness_mask,reenter); */ \
321 ALLOC_HEAP(n); /* ticky profiling */ \
322 GRAN_ALLOC_HEAP(n,liveness_mask); /* Granularity Simulation */ \
323 if (((Hp = Hp + (n)) > HpLim)) { \
324 /* Old: STGCALL3_GC(PerformGC,liveness_mask,n,StgFalse); */\
325 HEAP_OVERFLOW(liveness_mask,n,StgFalse); \
330 #define HEAP_CHK(liveness_mask,n,reenter) \
332 /* TICKY_PARANOIA(__FILE__, __LINE__); */ \
334 ALLOC_HEAP(n); /* ticky profiling */ \
335 if (((Hp = Hp + (n)) > HpLim) OR_INTERVAL_EXPIRED OR_CONTEXT_SWITCH OR_MSG_PENDING) { \
336 HEAP_OVERFLOW(liveness_mask,n,reenter); \
344 #define HEAP_CHK_AND_RESTORE_N(liveness_mask,n,reenter) \
346 /* TICKY_PARANOIA(__FILE__, __LINE__); */ \
348 ALLOC_HEAP(n); /* ticky profiling */ \
349 if (((Hp = Hp + (n)) > HpLim) OR_INTERVAL_EXPIRED OR_CONTEXT_SWITCH OR_MSG_PENDING) { \
350 HEAP_OVERFLOW(liveness_mask,n,reenter); \
351 n = TSO_ARG1(CurrentTSO); \
356 #define HEAP_CHK_AND_RESTORE_N(liveness_mask,n,reenter) \
357 HEAP_CHK(liveness_mask,n,reenter)
364 %************************************************************************
366 \subsection[StgMacros-prim-ops]{Primitive operations}
368 %************************************************************************
370 One thing to be {\em very careful about} with these macros that assign
371 to results is that the assignment must come {\em last}. Some of the
372 other arguments may be in terms of addressing modes that get clobbered
373 by the assignment. (Dirty imperative programming RULES!)
375 The order here is roughly that in \tr{compiler/prelude/PrimOps.lhs}.
377 %************************************************************************
379 \subsubsection[StgMacros-compare-primops]{Primitive comparison ops on basic types}
381 %************************************************************************
383 We cast the chars in case one of them is a literal (so C things work right
384 even for 8-bit chars).
386 #define gtCharZh(r,a,b) r=(I_)((a)> (b))
387 #define geCharZh(r,a,b) r=(I_)((a)>=(b))
388 #define eqCharZh(r,a,b) r=(I_)((a)==(b))
389 #define neCharZh(r,a,b) r=(I_)((a)!=(b))
390 #define ltCharZh(r,a,b) r=(I_)((a)< (b))
391 #define leCharZh(r,a,b) r=(I_)((a)<=(b))
393 /* Int comparisons: >#, >=# etc */
394 #define ZgZh(r,a,b) r=(I_)((a) >(b))
395 #define ZgZeZh(r,a,b) r=(I_)((a)>=(b))
396 #define ZeZeZh(r,a,b) r=(I_)((a)==(b))
397 #define ZdZeZh(r,a,b) r=(I_)((a)!=(b))
398 #define ZlZh(r,a,b) r=(I_)((a) <(b))
399 #define ZlZeZh(r,a,b) r=(I_)((a)<=(b))
401 #define gtWordZh(r,a,b) r=(I_)((a) >(b))
402 #define geWordZh(r,a,b) r=(I_)((a)>=(b))
403 #define eqWordZh(r,a,b) r=(I_)((a)==(b))
404 #define neWordZh(r,a,b) r=(I_)((a)!=(b))
405 #define ltWordZh(r,a,b) r=(I_)((a) <(b))
406 #define leWordZh(r,a,b) r=(I_)((a)<=(b))
408 #define gtAddrZh(r,a,b) r=(I_)((a) >(b))
409 #define geAddrZh(r,a,b) r=(I_)((a)>=(b))
410 #define eqAddrZh(r,a,b) r=(I_)((a)==(b))
411 #define neAddrZh(r,a,b) r=(I_)((a)!=(b))
412 #define ltAddrZh(r,a,b) r=(I_)((a) <(b))
413 #define leAddrZh(r,a,b) r=(I_)((a)<=(b))
415 #define gtFloatZh(r,a,b) r=(I_)((a)> (b))
416 #define geFloatZh(r,a,b) r=(I_)((a)>=(b))
417 #define eqFloatZh(r,a,b) r=(I_)((a)==(b))
418 #define neFloatZh(r,a,b) r=(I_)((a)!=(b))
419 #define ltFloatZh(r,a,b) r=(I_)((a)< (b))
420 #define leFloatZh(r,a,b) r=(I_)((a)<=(b))
422 /* Double comparisons: >##, >=#@ etc */
423 #define ZgZhZh(r,a,b) r=(I_)((a) >(b))
424 #define ZgZeZhZh(r,a,b) r=(I_)((a)>=(b))
425 #define ZeZeZhZh(r,a,b) r=(I_)((a)==(b))
426 #define ZdZeZhZh(r,a,b) r=(I_)((a)!=(b))
427 #define ZlZhZh(r,a,b) r=(I_)((a) <(b))
428 #define ZlZeZhZh(r,a,b) r=(I_)((a)<=(b))
431 %************************************************************************
433 \subsubsection[StgMacros-char-primops]{Primitive @Char#@ ops (and @LitString#@ish things, too)}
435 %************************************************************************
437 We cast the chars in case one of them is a literal (so C things work right
438 even for 8-bit chars).
440 #define ordZh(r,a) r=(I_)((W_) (a))
441 #define chrZh(r,a) r=(StgChar)((W_)(a))
444 %************************************************************************
446 \subsubsection[StgMacros-int-primops]{Primitive @Int#@ ops}
448 %************************************************************************
451 I_ stg_div PROTO((I_ a, I_ b));
453 #define ZpZh(r,a,b) r=(a)+(b)
454 #define ZmZh(r,a,b) r=(a)-(b)
455 #define ZtZh(r,a,b) r=(a)*(b)
456 #define quotIntZh(r,a,b) r=(a)/(b)
457 #define ZdZh(r,a,b) r=ULTRASAFESTGCALL2(I_,(void *, I_, I_),stg_div,(a),(b))
458 #define remIntZh(r,a,b) r=(a)%(b)
459 #define negateIntZh(r,a) r=-(a)
462 %************************************************************************
464 \subsubsection[StgMacros-word-primops]{Primitive @Word#@ ops}
466 %************************************************************************
469 #define andZh(r,a,b) r=(a)&(b)
470 #define orZh(r,a,b) r=(a)|(b)
471 #define notZh(r,a) r=~(a)
473 #define shiftLZh(r,a,b) r=(a)<<(b)
474 #define shiftRAZh(r,a,b) r=(a)>>(b)
475 #define shiftRLZh(r,a,b) r=(a)>>(b)
476 #define iShiftLZh(r,a,b) r=(a)<<(b)
477 #define iShiftRAZh(r,a,b) r=(a)>>(b)
478 #define iShiftRLZh(r,a,b) r=(a)>>(b)
480 #define int2WordZh(r,a) r=(W_)(a)
481 #define word2IntZh(r,a) r=(I_)(a)
484 %************************************************************************
486 \subsubsection[StgMacros-addr-primops]{Primitive @Addr#@ ops}
488 %************************************************************************
491 #define int2AddrZh(r,a) r=(A_)(a)
492 #define addr2IntZh(r,a) r=(I_)(a)
495 %************************************************************************
497 \subsubsection[StgMacros-float-primops]{Primitive @Float#@ ops}
499 %************************************************************************
502 #define plusFloatZh(r,a,b) r=(a)+(b)
503 #define minusFloatZh(r,a,b) r=(a)-(b)
504 #define timesFloatZh(r,a,b) r=(a)*(b)
505 #define divideFloatZh(r,a,b) r=(a)/(b)
506 #define negateFloatZh(r,a) r=-(a)
508 #define int2FloatZh(r,a) r=(StgFloat)(a)
509 #define float2IntZh(r,a) r=(I_)(a)
511 #define expFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),exp,a)
512 #define logFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),log,a)
513 #define sqrtFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),sqrt,a)
514 #define sinFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),sin,a)
515 #define cosFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),cos,a)
516 #define tanFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),tan,a)
517 #define asinFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),asin,a)
518 #define acosFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),acos,a)
519 #define atanFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),atan,a)
520 #define sinhFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),sinh,a)
521 #define coshFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),cosh,a)
522 #define tanhFloatZh(r,a) r=(StgFloat) SAFESTGCALL1(StgDouble,(void *, StgDouble),tanh,a)
523 #define powerFloatZh(r,a,b) r=(StgFloat) SAFESTGCALL2(StgDouble,(void *, StgDouble,StgDouble),pow,a,b)
525 /* encoding/decoding given w/ Integer stuff */
528 %************************************************************************
530 \subsubsection[StgMacros-double-primops]{Primitive @Double#@ ops}
532 %************************************************************************
535 #define ZpZhZh(r,a,b) r=(a)+(b)
536 #define ZmZhZh(r,a,b) r=(a)-(b)
537 #define ZtZhZh(r,a,b) r=(a)*(b)
538 #define ZdZhZh(r,a,b) r=(a)/(b)
539 #define negateDoubleZh(r,a) r=-(a)
541 #define int2DoubleZh(r,a) r=(StgDouble)(a)
542 #define double2IntZh(r,a) r=(I_)(a)
544 #define float2DoubleZh(r,a) r=(StgDouble)(a)
545 #define double2FloatZh(r,a) r=(StgFloat)(a)
547 #define expDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),exp,a)
548 #define logDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),log,a)
549 #define sqrtDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),sqrt,a)
550 #define sinDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),sin,a)
551 #define cosDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),cos,a)
552 #define tanDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),tan,a)
553 #define asinDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),asin,a)
554 #define acosDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),acos,a)
555 #define atanDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),atan,a)
556 #define sinhDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),sinh,a)
557 #define coshDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),cosh,a)
558 #define tanhDoubleZh(r,a) r=(StgDouble) SAFESTGCALL1(StgDouble,(void *, StgDouble),tanh,a)
560 #define ZtZtZhZh(r,a,b) r=(StgDouble) SAFESTGCALL2(StgDouble,(void *, StgDouble,StgDouble),pow,a,b)
563 %************************************************************************
565 \subsubsection[StgMacros-integer-primops]{Primitive @Integer@-related ops (GMP stuff)}
567 %************************************************************************
569 Dirty macros we use for the real business.
571 INVARIANT: When one of these macros is called, the only live data is
572 tidily on the STG stacks or in the STG registers (the code generator
573 ensures this). If there are any pointer-arguments, they will be in
574 the first \tr{Ret*} registers (e.g., \tr{da} arg of \tr{gmpTake1Return1}).
576 OK, here are the real macros:
578 #define gmpTake1Return1(size_chk_macro, liveness, mpz_op, ar,sr,dr, aa,sa,da) \
581 I_ space = size_chk_macro(sa); \
583 /* Check that there will be enough heap & make Hp visible to GMP allocator */ \
584 GMP_HEAP_LOOKAHEAD(liveness,space); \
586 /* Now we can initialise (post possible GC) */ \
589 arg.d = (unsigned long int *) (BYTE_ARR_CTS(da)); \
591 SAFESTGCALL1(void,(void *, MP_INT *),mpz_init,&result); \
593 /* Perform the operation */ \
594 SAFESTGCALL2(void,(void *, MP_INT *, MP_INT *),mpz_op,&result,&arg); \
596 GMP_HEAP_HANDBACK(); /* restore Hp */ \
597 (ar) = result.alloc; \
598 (sr) = result.size; \
599 (dr) = (B_) (result.d - DATA_HS); \
600 /* pt to *beginning* of object (GMP has been monkeying around in the middle) */ \
604 #define gmpTake2Return1(size_chk_macro, liveness, mpz_op, ar,sr,dr, a1,s1,d1, a2,s2,d2)\
608 I_ space = size_chk_macro(s1,s2); \
610 /* Check that there will be enough heap & make Hp visible to GMP allocator */ \
611 GMP_HEAP_LOOKAHEAD(liveness,space); \
613 /* Now we can initialise (post possible GC) */ \
616 arg1.d = (unsigned long int *) (BYTE_ARR_CTS(d1)); \
619 arg2.d = (unsigned long int *) (BYTE_ARR_CTS(d2)); \
621 SAFESTGCALL1(void,(void *, MP_INT *),mpz_init,&result); \
623 /* Perform the operation */ \
624 SAFESTGCALL3(void,(void *, MP_INT *, MP_INT *, MP_INT *),mpz_op,&result,&arg1,&arg2); \
626 GMP_HEAP_HANDBACK(); /* restore Hp */ \
627 (ar) = result.alloc; \
628 (sr) = result.size; \
629 (dr) = (B_) (result.d - DATA_HS); \
630 /* pt to *beginning* of object (GMP has been monkeying around in the middle) */ \
633 #define gmpTake2Return2(size_chk_macro, liveness, mpz_op, ar1,sr1,dr1, ar2,sr2,dr2, a1,s1,d1, a2,s2,d2) \
638 I_ space = size_chk_macro(s1,s2); \
640 /* Check that there will be enough heap & make Hp visible to GMP allocator */ \
641 GMP_HEAP_LOOKAHEAD(liveness,space); \
643 /* Now we can initialise (post possible GC) */ \
646 arg1.d = (unsigned long int *) (BYTE_ARR_CTS(d1)); \
649 arg2.d = (unsigned long int *) (BYTE_ARR_CTS(d2)); \
651 SAFESTGCALL1(void,(void *, MP_INT *),mpz_init,&result1); \
652 SAFESTGCALL1(void,(void *, MP_INT *),mpz_init,&result2); \
654 /* Perform the operation */ \
655 SAFESTGCALL4(void,(void *, MP_INT *, MP_INT *, MP_INT *, MP_INT *),mpz_op,&result1,&result2,&arg1,&arg2); \
657 GMP_HEAP_HANDBACK(); /* restore Hp */ \
658 (ar1) = result1.alloc; \
659 (sr1) = result1.size; \
660 (dr1) = (B_) (result1.d - DATA_HS); \
661 (ar2) = result2.alloc; \
662 (sr2) = result2.size; \
663 (dr2) = (B_) (result2.d - DATA_HS); \
667 Some handy size-munging macros: sometimes gratuitously {\em conservative}.
668 The \tr{+16} is to allow for the initial allocation of \tr{MP_INT} results.
669 The \tr{__abs} stuff is because negative-ness of GMP things is encoded
672 #define __abs(a) (( (a) >= 0 ) ? (a) : (-(a)))
673 #define GMP_SIZE_ONE() (2 + DATA_HS + 16)
674 #define GMP_SAME_SIZE(a) (__abs(a) + DATA_HS + 16)
675 #define GMP_MAX_SIZE(a,b) ((__abs(a) > __abs(b) ? __abs(a) : __abs(b)) + 1 + DATA_HS + 16)
676 /* NB: the +1 is for the carry (or whatever) */
677 #define GMP_2MAX_SIZE(a,b) (2 * GMP_MAX_SIZE(a,b))
678 #define GMP_ADD_SIZES(a,b) (__abs(a) + __abs(b) + 1 + DATA_HS + 16)
679 /* the +1 may just be paranoia */
682 For the Integer/GMP stuff, we have macros that {\em look ahead} for
683 some space, but don't actually grab it.
685 If there are live pointers at the time of the lookahead, the caller
686 must make sure they are in \tr{Ret1}, \tr{Ret2}, ..., so they can be
687 handled normally. We achieve this by having the code generator {\em
688 always} pass args to may-invoke-GC primitives in registers, using the
689 normal pointers-first policy. This means that, if we do go to garbage
690 collection, everything is already in the Right Place.
692 Saving and restoring Hp register so the MP allocator can see them. If we are
693 performing liftime profiling need to save and restore HpLim as well so that
694 it can be bumped if allocation occurs.
696 The second argument to @GMP_HEAP_LOOKAHEAD@ must be an lvalue so that
697 it can be restored from @TSO_ARG1@ after a failed @HEAP_CHK@ in
701 #define GMP_HEAP_LOOKAHEAD(liveness,n) \
703 HEAP_CHK_AND_RESTORE_N(liveness,n,0); \
705 UN_ALLOC_HEAP(n); /* Undo ticky-ticky */ \
706 SAVE_Hp = Hp; /* Hand over the hp */ \
707 DEBUG_SetGMPAllocBudget(n) \
710 #define GMP_HEAP_HANDBACK() \
712 DEBUG_ResetGMPAllocBudget()
716 void *stgAllocForGMP PROTO((size_t size_in_bytes));
717 void *stgReallocForGMP PROTO((void *ptr, size_t old_size, size_t new_size));
718 void stgDeallocForGMP PROTO((void *ptr, size_t size));
721 extern StgInt DEBUG_GMPAllocBudget;
722 #define DEBUG_SetGMPAllocBudget(n) DEBUG_GMPAllocBudget = (n);
723 #define DEBUG_ResetGMPAllocBudget() DEBUG_GMPAllocBudget = 0;
725 #define DEBUG_SetGMPAllocBudget(n) /*nothing*/
726 #define DEBUG_ResetGMPAllocBudget() /*nothing*/
730 The real business (defining Integer primops):
732 #define negateIntegerZh(ar,sr,dr, liveness, aa,sa,da) \
733 gmpTake1Return1(GMP_SAME_SIZE, liveness, mpz_neg, ar,sr,dr, aa,sa,da)
735 #define plusIntegerZh(ar,sr,dr, liveness, a1,s1,d1, a2,s2,d2) \
736 gmpTake2Return1(GMP_MAX_SIZE, liveness, mpz_add, ar,sr,dr, a1,s1,d1, a2,s2,d2)
737 #define minusIntegerZh(ar,sr,dr, liveness, a1,s1,d1, a2,s2,d2) \
738 gmpTake2Return1(GMP_MAX_SIZE, liveness, mpz_sub, ar,sr,dr, a1,s1,d1, a2,s2,d2)
739 #define timesIntegerZh(ar,sr,dr, liveness, a1,s1,d1, a2,s2,d2) \
740 gmpTake2Return1(GMP_ADD_SIZES, liveness, mpz_mul, ar,sr,dr, a1,s1,d1, a2,s2,d2)
742 /* div, mod, quot, rem are defined w/ quotRem & divMod */
744 #define quotRemIntegerZh(ar1,sr1,dr1, ar2,sr2,dr2, liveness, a1,s1,d1, a2,s2,d2) \
745 gmpTake2Return2(GMP_2MAX_SIZE, liveness, mpz_divmod, ar1,sr1,dr1, ar2,sr2,dr2, a1,s1,d1, a2,s2,d2)
746 #define divModIntegerZh(ar1,sr1,dr1, ar2,sr2,dr2, liveness, a1,s1,d1, a2,s2,d2) \
747 gmpTake2Return2(GMP_2MAX_SIZE, liveness, mpz_mdivmod, ar1,sr1,dr1, ar2,sr2,dr2, a1,s1,d1, a2,s2,d2)
750 Comparison ops (@<@, @>=@, etc.) are defined in terms of the cmp
751 fellow (returns -ve, 0, or +ve).
753 #define cmpIntegerZh(r, hp, a1,s1,d1, a2,s2,d2) /* calls mpz_cmp */ \
756 /* Does not allocate memory */ \
760 arg1.d = (unsigned long int *) (BYTE_ARR_CTS(d1)); \
763 arg2.d = (unsigned long int *) (BYTE_ARR_CTS(d2)); \
765 (r) = SAFESTGCALL2(I_,(void *, MP_INT *, MP_INT *),mpz_cmp,&arg1,&arg2); \
772 #define integer2IntZh(r, hp, aa,sa,da) \
774 /* Does not allocate memory */ \
778 arg.d = (unsigned long int *) (BYTE_ARR_CTS(da)); \
780 (r) = SAFESTGCALL1(I_,(void *, MP_INT *),mpz_get_si,&arg); \
783 /* Since we're forced to know a little bit about MP_INT layout to do this with
784 pre-allocated heap, we just inline the whole of mpz_init_set_si here.
785 ** DIRE WARNING. if mpz_init_set_si changes, so does this! ***
788 #define int2IntegerZh(ar,sr,dr, hp, i) \
789 { StgInt val; /* to snaffle arg to avoid aliasing */ \
791 val = (i); /* snaffle... */ \
793 SET_DATA_HDR((hp),ArrayOfData_info,CCC,DATA_VHS+MIN_MP_INT_SIZE,0); \
795 if ((val) < 0) { (sr) = -1; (hp)[DATA_HS] = -(val); } \
796 else if ((val) > 0) { (sr) = 1; (hp)[DATA_HS] = (val); } \
797 else /* val==0 */ { (sr) = 0; } \
799 (dr) = (B_)(hp); /* dr is an StgByteArray */ \
802 #define word2IntegerZh(ar,sr,dr, hp, i) \
803 { StgWord val; /* to snaffle arg to avoid aliasing */ \
805 val = (i); /* snaffle... */ \
807 SET_DATA_HDR((hp),ArrayOfData_info,CCC,DATA_VHS+MIN_MP_INT_SIZE,0); \
809 if ((val) != 0) { (sr) = 1; (hp)[DATA_HS] = (val); } \
810 else /* val==0 */ { (sr) = 0; } \
812 (dr) = (B_)(hp); /* dr is an StgByteArray */ \
817 Then there are a few oddments to make life easier:
821 The "str" argument must be a literal C string.
823 addr2Integer( ..., "foo") OK!
826 addr2Integer( ..., x) NO! NO!
829 #define addr2IntegerZh(ar,sr,dr, liveness, str) \
831 /* taking the number of bytes/8 as the number of words of lookahead \
832 is plenty conservative */ \
833 I_ space = GMP_SAME_SIZE(sizeof(str) / 8 + 1); \
835 GMP_HEAP_LOOKAHEAD(liveness, space); \
837 /* Perform the operation */ \
838 if (SAFESTGCALL3(I_,(void *, MP_INT *, char *, int), mpz_init_set_str,&result,(str),/*base*/10)) \
841 GMP_HEAP_HANDBACK(); /* restore Hp */ \
842 (ar) = result.alloc; \
843 (sr) = result.size; \
844 (dr) = (B_) (result.d - DATA_HS); \
845 /* pt to *beginning* of object (GMP has been monkeying around in the middle) */ \
849 Encoding and decoding float-ish things is pretty Integer-ish. We use
850 these pretty magical support functions, essentially stolen from Lennart:
852 StgFloat __encodeFloat PROTO((MP_INT *, I_));
853 void __decodeFloat PROTO((MP_INT * /*result1*/,
857 StgDouble __encodeDouble PROTO((MP_INT *, I_));
858 void __decodeDouble PROTO((MP_INT * /*result1*/,
863 Some floating-point format info, made with the \tr{enquire} program
864 (version~4.3) [comes with gcc].
866 /* this should be done by CPU architecture, insofar as possible [WDP] */
868 #if sparc_TARGET_ARCH \
869 || alpha_TARGET_ARCH \
870 || hppa1_1_TARGET_ARCH \
871 || i386_TARGET_ARCH \
872 || m68k_TARGET_ARCH \
873 || mipsel_TARGET_ARCH \
874 || mipseb_TARGET_ARCH \
875 || powerpc_TARGET_ARCH
877 /* yes, it is IEEE floating point */
878 #include "ieee-flpt.h"
880 #if alpha_dec_osf1_TARGET \
881 || i386_TARGET_ARCH \
882 || mipsel_TARGET_ARCH
884 #undef BIGENDIAN /* little-endian weirdos... */
889 #else /* unknown floating-point format */
891 ******* ERROR *********** Any ideas about floating-point format?
893 #endif /* unknown floating-point */
897 #if alpha_dec_osf1_TARGET
898 #define encodeFloatZh(r, hp, aa,sa,da, expon) encodeDoubleZh(r, hp, aa,sa,da, expon)
900 #define encodeFloatZh(r, hp, aa,sa,da, expon) \
902 /* Does not allocate memory */ \
906 arg.d = (unsigned long int *) (BYTE_ARR_CTS(da)); \
908 r = SAFESTGCALL2(StgFloat,(void *, MP_INT *, I_), __encodeFloat,&arg,(expon)); \
912 #define encodeDoubleZh(r, hp, aa,sa,da, expon) \
914 /* Does not allocate memory */ \
918 arg.d = (unsigned long int *) (BYTE_ARR_CTS(da)); \
920 r = SAFESTGCALL2(StgDouble,(void *, MP_INT *, I_), __encodeDouble,&arg,(expon));\
923 #if alpha_dec_osf1_TARGET
924 #define decodeFloatZh(exponr, ar,sr,dr, hp, f) decodeDoubleZh(exponr, ar,sr,dr, hp, f)
926 #define decodeFloatZh(exponr, ar,sr,dr, hp, f) \
929 StgFloat arg = (f); \
931 /* Be prepared to tell Lennart-coded __decodeFloat */ \
932 /* where mantissa.d can be put (it does not care about the rest) */ \
933 SET_DATA_HDR(hp,ArrayOfData_info,CCC,DATA_VHS+MIN_MP_INT_SIZE,0); \
934 mantissa.d = (hp) + DATA_HS; \
936 /* Perform the operation */ \
937 SAFESTGCALL3(void,(void *, MP_INT *, I_ *, StgFloat),__decodeFloat,&mantissa,&exponent,arg); \
939 ar = mantissa.alloc; \
940 sr = mantissa.size; \
945 #define decodeDoubleZh(exponr, ar,sr,dr, hp, f) \
948 StgDouble arg = (f); \
950 /* Be prepared to tell Lennart-coded __decodeDouble */ \
951 /* where mantissa.d can be put (it does not care about the rest) */ \
952 SET_DATA_HDR(hp,ArrayOfData_info,CCC,DATA_VHS+MIN_MP_INT_SIZE,0); \
953 mantissa.d = (hp) + DATA_HS; \
955 /* Perform the operation */ \
956 SAFESTGCALL3(void,(void *, MP_INT *, I_ *, StgDouble),__decodeDouble,&mantissa,&exponent,arg); \
958 ar = mantissa.alloc; \
959 sr = mantissa.size; \
964 %************************************************************************
966 \subsubsection[StgMacros-mv-floats]{Moving floats and doubles around (e.g., to/from stacks)}
968 %************************************************************************
970 With GCC, we use magic non-standard inlining; for other compilers, we
971 just use functions (see also \tr{runtime/prims/PrimArith.lc}).
973 (The @OMIT_...@ is only used in compiling some of the RTS, none of
974 which uses these anyway.)
977 #if alpha_TARGET_ARCH \
978 || i386_TARGET_ARCH \
981 #define ASSIGN_FLT(dst, src) *(StgFloat *)(dst) = (src);
982 #define PK_FLT(src) (*(StgFloat *)(src))
984 #define ASSIGN_DBL(dst, src) *(StgDouble *)(dst) = (src);
985 #define PK_DBL(src) (*(StgDouble *)(src))
987 #else /* not m68k || alpha || i[34]86 */
989 /* Special handling for machines with troublesome alignment constraints */
991 #define FLOAT_ALIGNMENT_TROUBLES TRUE
993 #if ! defined(__GNUC__) || ! defined(__STG_GCC_REGS__)
995 void ASSIGN_DBL PROTO((W_ [], StgDouble));
996 StgDouble PK_DBL PROTO((W_ []));
997 void ASSIGN_FLT PROTO((W_ [], StgFloat));
998 StgFloat PK_FLT PROTO((W_ []));
1000 #else /* yes, its __GNUC__ && we really want them */
1002 #if sparc_TARGET_ARCH
1004 #define ASSIGN_FLT(dst, src) *(StgFloat *)(dst) = (src);
1005 #define PK_FLT(src) (*(StgFloat *)(src))
1007 #define ASSIGN_DBL(dst,src) \
1008 __asm__("st %2,%0\n\tst %R2,%1" : "=m" (((P_)(dst))[0]), \
1009 "=m" (((P_)(dst))[1]) : "f" (src));
1011 #define PK_DBL(src) \
1012 ( { register double d; \
1013 __asm__("ld %1,%0\n\tld %2,%R0" : "=f" (d) : \
1014 "m" (((P_)(src))[0]), "m" (((P_)(src))[1])); d; \
1019 /* (not very) forward prototype declarations */
1020 void ASSIGN_DBL PROTO((W_ [], StgDouble));
1021 StgDouble PK_DBL PROTO((W_ []));
1022 void ASSIGN_FLT PROTO((W_ [], StgFloat));
1023 StgFloat PK_FLT PROTO((W_ []));
1027 ASSIGN_DBL(W_ p_dest[], StgDouble src)
1031 p_dest[0] = y.du.dhi;
1032 p_dest[1] = y.du.dlo;
1035 /* GCC also works with this version, but it generates
1036 the same code as the previous one, and is not ANSI
1038 #define ASSIGN_DBL( p_dest, src ) \
1039 *p_dest = ((double_thing) src).du.dhi; \
1040 *(p_dest+1) = ((double_thing) src).du.dlo \
1048 y.du.dhi = p_src[0];
1049 y.du.dlo = p_src[1];
1055 ASSIGN_FLT(W_ p_dest[], StgFloat src)
1071 #endif /* ! sparc */
1073 #endif /* __GNUC__ */
1075 #endif /* not __m68k__ */
1078 %************************************************************************
1080 \subsubsection[StgMacros-array-primops]{Primitive arrays}
1082 %************************************************************************
1084 We regularly use this macro to fish the ``contents'' part
1085 out of a DATA or TUPLE closure, which is what is used for
1086 non-ptr and ptr arrays (respectively).
1088 BYTE_ARR_CTS returns a @C_ *@!
1090 We {\em ASSUME} we can use the same macro for both!!
1094 #define BYTE_ARR_CTS(a) \
1095 ({ ASSERT(INFO_PTR(a) == (W_) ArrayOfData_info); \
1096 ((C_ *) (((StgPtr) (a))+DATA_HS)); })
1097 #define PTRS_ARR_CTS(a) \
1098 ({ ASSERT((INFO_PTR(a) == (W_) ArrayOfPtrs_info) \
1099 || (INFO_PTR(a) == (W_) ImMutArrayOfPtrs_info));\
1100 ((a)+MUTUPLE_HS);} )
1102 #define BYTE_ARR_CTS(a) ((char *) (((StgPtr) (a))+DATA_HS))
1103 #define PTRS_ARR_CTS(a) ((a)+MUTUPLE_HS)
1107 extern I_ genSymZh(STG_NO_ARGS);
1108 extern I_ resetGenSymZh(STG_NO_ARGS);
1109 extern I_ incSeqWorldZh(STG_NO_ARGS);
1111 extern I_ byteArrayHasNUL__ PROTO((const char *, I_));
1114 OK, the easy ops first: (all except \tr{newArr*}:
1116 (OLD:) VERY IMPORTANT! The read/write/index primitive ops
1117 on @ByteArray#@s index the array using a {\em BYTE} offset, even
1118 if the thing begin gotten out is a multi-byte @Int#@, @Float#@ etc.
1119 This is because you might be trying to take apart a C struct, where
1120 the offset from the start of the struct isn't a multiple of the
1121 size of the thing you're getting. Hence the @(char *)@ casts.
1123 EVEN MORE IMPORTANT! The above is a lie. The offsets for BlahArrays
1124 are in Blahs. WDP 95/08
1126 In the case of messing with @StgAddrs@ (@A_@), which are really \tr{void *},
1127 we cast to @P_@, because you can't index off an uncast \tr{void *}.
1129 In the case of @Array#@ (which contain pointers), the offset is in units
1130 of one ptr (not bytes).
1133 #define sameMutableArrayZh(r,a,b) r=(I_)((a)==(b))
1134 #define sameMutableByteArrayZh(r,a,b) r=(I_)((B_)(a)==(B_)(b))
1136 #define readArrayZh(r,a,i) r=((PP_) PTRS_ARR_CTS(a))[(i)]
1138 #define readCharArrayZh(r,a,i) indexCharOffAddrZh(r,BYTE_ARR_CTS(a),i)
1139 #define readIntArrayZh(r,a,i) indexIntOffAddrZh(r,BYTE_ARR_CTS(a),i)
1140 #define readAddrArrayZh(r,a,i) indexAddrOffAddrZh(r,BYTE_ARR_CTS(a),i)
1141 #define readFloatArrayZh(r,a,i) indexFloatOffAddrZh(r,BYTE_ARR_CTS(a),i)
1142 #define readDoubleArrayZh(r,a,i) indexDoubleOffAddrZh(r,BYTE_ARR_CTS(a),i)
1144 /* result ("r") arg ignored in write macros! */
1145 #define writeArrayZh(a,i,v) ((PP_) PTRS_ARR_CTS(a))[(i)]=(v)
1147 #define writeCharArrayZh(a,i,v) ((C_ *)(BYTE_ARR_CTS(a)))[i] = (v)
1148 #define writeIntArrayZh(a,i,v) ((I_ *)(BYTE_ARR_CTS(a)))[i] = (v)
1149 #define writeAddrArrayZh(a,i,v) ((PP_)(BYTE_ARR_CTS(a)))[i] = (v)
1150 #define writeFloatArrayZh(a,i,v) \
1151 ASSIGN_FLT((P_) (((StgFloat *)(BYTE_ARR_CTS(a))) + i),v)
1152 #define writeDoubleArrayZh(a,i,v) \
1153 ASSIGN_DBL((P_) (((StgDouble *)(BYTE_ARR_CTS(a))) + i),v)
1155 #define indexArrayZh(r,a,i) r=((PP_) PTRS_ARR_CTS(a))[(i)]
1157 #define indexCharArrayZh(r,a,i) indexCharOffAddrZh(r,BYTE_ARR_CTS(a),i)
1158 #define indexIntArrayZh(r,a,i) indexIntOffAddrZh(r,BYTE_ARR_CTS(a),i)
1159 #define indexAddrArrayZh(r,a,i) indexAddrOffAddrZh(r,BYTE_ARR_CTS(a),i)
1160 #define indexFloatArrayZh(r,a,i) indexFloatOffAddrZh(r,BYTE_ARR_CTS(a),i)
1161 #define indexDoubleArrayZh(r,a,i) indexDoubleOffAddrZh(r,BYTE_ARR_CTS(a),i)
1163 #define indexCharOffAddrZh(r,a,i) r= ((C_ *)(a))[i]
1164 #define indexIntOffAddrZh(r,a,i) r= ((I_ *)(a))[i]
1165 #define indexAddrOffAddrZh(r,a,i) r= ((PP_)(a))[i]
1166 #define indexFloatOffAddrZh(r,a,i) r= PK_FLT((P_) (((StgFloat *)(a)) + i))
1167 #define indexDoubleOffAddrZh(r,a,i) r= PK_DBL((P_) (((StgDouble *)(a)) + i))
1169 /* Freezing arrays-of-ptrs requires changing an info table, for the
1170 benefit of the generational collector. It needs to scavenge mutable
1171 objects, even if they are in old space. When they become immutable,
1172 they can be removed from this scavenge list. */
1173 #define unsafeFreezeArrayZh(r,a) \
1177 FREEZE_MUT_HDR(result,ImMutArrayOfPtrs_info); \
1181 #define unsafeFreezeByteArrayZh(r,a) r=(B_)(a)
1184 Now the \tr{newArr*} ops:
1188 --------------------
1189 Will: ToDo: we need to find suitable places to put this comment, and the
1190 "in-general" one which follows.
1192 ************ Nota Bene. The "n" in this macro is guaranteed to
1193 be a register, *not* (say) Node[1]. That means that it is guaranteed
1194 to survive GC, provided only that the register is kept unaltered.
1195 This is important, because "n" is used after the HEAP_CHK.
1197 In general, *all* parameters to these primitive-op macros are always
1198 registers. (Will: For exactly *which* primitive-op macros is this guaranteed?
1199 Exactly those which can trigger GC?)
1200 ------------------------
1202 NOTE: the above may now be OLD (WDP 94/02/10)
1206 For char arrays, the size is in {\em BYTES}.
1209 #define newCharArrayZh(r,liveness,n) newByteArray(r,liveness,(n) * sizeof(C_))
1210 #define newIntArrayZh(r,liveness,n) newByteArray(r,liveness,(n) * sizeof(I_))
1211 #define newAddrArrayZh(r,liveness,n) newByteArray(r,liveness,(n) * sizeof(P_))
1212 #define newFloatArrayZh(r,liveness,n) newByteArray(r,liveness,(n) * sizeof(StgFloat))
1213 #define newDoubleArrayZh(r,liveness,n) newByteArray(r,liveness,(n) * sizeof(StgDouble))
1215 #define newByteArray(r,liveness,n) \
1220 HEAP_CHK(liveness,DATA_HS+BYTES_TO_STGWORDS(n),0); \
1221 size = BYTES_TO_STGWORDS(n); \
1222 ALLOC_PRIM(DATA_HS,size,0,DATA_HS+size) /* ticky ticky */; \
1223 CC_ALLOC(CCC,DATA_HS+size,ARR_K); \
1225 result = Hp-(DATA_HS+size)+1; \
1226 SET_DATA_HDR(result,ArrayOfData_info,CCC,DATA_VHS+size,0); \
1231 Arrays of pointers need to be initialised; uses \tr{TUPLES}!
1232 The initialisation value is guaranteed to be in a register,
1233 and will be indicated by the liveness mask, so it's ok to do
1234 a \tr{HEAP_CHK}, which may trigger GC.
1237 /* The new array initialization routine for the NCG */
1238 void newArrZh_init PROTO((P_ result, I_ n, P_ init));
1240 #define newArrayZh(r,liveness,n,init) \
1245 HEAP_CHK(liveness, MUTUPLE_HS+(n),0); \
1246 ALLOC_PRIM(MUTUPLE_HS,(n),0,MUTUPLE_HS+(n)) /* ticky ticky */; \
1247 CC_ALLOC(CCC,MUTUPLE_HS+(n),ARR_K); /* cc prof */ \
1249 result = Hp + 1 - (MUTUPLE_HS+(n)); \
1250 SET_MUTUPLE_HDR(result,ArrayOfPtrs_info,CCC,MUTUPLE_VHS+(n),0) \
1251 for (p = result+MUTUPLE_HS; p < (result+MUTUPLE_HS+(n)); p++) { \
1259 %************************************************************************
1261 \subsubsection[StgMacros-SynchVar-primops]{Synchronizing Variables PrimOps}
1263 %************************************************************************
1266 ED_(PrelBase_Z91Z93_closure);
1268 #define newSynchVarZh(r, hp) \
1270 ALLOC_PRIM(MUTUPLE_HS,3,0,MUTUPLE_HS+3) /* ticky ticky */; \
1271 CC_ALLOC(CCC,MUTUPLE_HS+3,ARR_K); /* cc prof */ \
1272 SET_SVAR_HDR(hp,EmptySVar_info,CCC); \
1273 SVAR_HEAD(hp) = SVAR_TAIL(hp) = SVAR_VALUE(hp) = PrelBase_Z91Z93_closure; \
1281 void Yield PROTO((W_));
1283 #define takeMVarZh(r, liveness, node) \
1285 while (INFO_PTR(node) != (W_) FullSVar_info) { \
1286 if (SVAR_HEAD(node) == PrelBase_Z91Z93_closure) \
1287 SVAR_HEAD(node) = CurrentTSO; \
1289 TSO_LINK(SVAR_TAIL(node)) = CurrentTSO; \
1290 TSO_LINK(CurrentTSO) = (P_) PrelBase_Z91Z93_closure; \
1291 SVAR_TAIL(node) = CurrentTSO; \
1292 DO_YIELD(liveness << 1); \
1294 SET_INFO_PTR(node, EmptySVar_info); \
1295 r = SVAR_VALUE(node); \
1296 SVAR_VALUE(node) = PrelBase_Z91Z93_closure; \
1301 #define takeMVarZh(r, liveness, node) \
1303 if (INFO_PTR(node) != (W_) FullSVar_info) { \
1304 /* Don't wrap the calls; we're done with STG land */\
1306 fprintf(stderr, "takeMVar#: MVar is empty.\n"); \
1307 EXIT(EXIT_FAILURE); \
1309 SET_INFO_PTR(node, EmptySVar_info); \
1310 r = SVAR_VALUE(node); \
1311 SVAR_VALUE(node) = PrelBase_Z91Z93_closure; \
1322 /* Only difference to the !GRAN def: RunnableThreadsHd has been replaced by */
1323 /* ThreadQueueHd i.e. the tso is added at the end of the thread queue on */
1324 /* the CurrentProc. This means we have an implicit context switch after */
1325 /* putMVar even if unfair scheduling is used in GranSim (default)! -- HWL */
1327 #define putMVarZh(node, value) \
1330 if (INFO_PTR(node) == (W_) FullSVar_info) { \
1331 /* Don't wrap the calls; we're done with STG land */\
1333 fprintf(stderr, "putMVar#: MVar already full.\n"); \
1334 EXIT(EXIT_FAILURE); \
1336 SET_INFO_PTR(node, FullSVar_info); \
1337 SVAR_VALUE(node) = value; \
1338 tso = SVAR_HEAD(node); \
1339 if (tso != (P_) PrelBase_Z91Z93_closure) { \
1341 STGCALL3(void,(void *, char *, P_, P_),QP_Event2,do_qp_prof > 1 ? "RA" : "RG",tso,CurrentTSO); \
1342 if (ThreadQueueHd == PrelBase_Z91Z93_closure) \
1343 ThreadQueueHd = tso; \
1345 TSO_LINK(ThreadQueueTl) = tso; \
1346 ThreadQueueTl = tso; \
1347 SVAR_HEAD(node) = TSO_LINK(tso); \
1348 TSO_LINK(tso) = (P_) PrelBase_Z91Z93_closure; \
1349 if(SVAR_HEAD(node) == (P_) PrelBase_Z91Z93_closure) \
1350 SVAR_TAIL(node) = (P_) PrelBase_Z91Z93_closure; \
1356 #define putMVarZh(node, value) \
1359 if (INFO_PTR(node) == (W_) FullSVar_info) { \
1360 /* Don't wrap the calls; we're done with STG land */\
1362 fprintf(stderr, "putMVar#: MVar already full.\n"); \
1363 EXIT(EXIT_FAILURE); \
1365 SET_INFO_PTR(node, FullSVar_info); \
1366 SVAR_VALUE(node) = value; \
1367 tso = SVAR_HEAD(node); \
1368 if (tso != (P_) PrelBase_Z91Z93_closure) { \
1370 STGCALL3(void,(void *, char *, P_, P_),QP_Event2,do_qp_prof > 1 ? "RA" : "RG",tso,CurrentTSO); \
1371 if (RunnableThreadsHd == PrelBase_Z91Z93_closure) \
1372 RunnableThreadsHd = tso; \
1374 TSO_LINK(RunnableThreadsTl) = tso; \
1375 RunnableThreadsTl = tso; \
1376 SVAR_HEAD(node) = TSO_LINK(tso); \
1377 TSO_LINK(tso) = (P_) PrelBase_Z91Z93_closure; \
1378 if(SVAR_HEAD(node) == (P_) PrelBase_Z91Z93_closure) \
1379 SVAR_TAIL(node) = (P_) PrelBase_Z91Z93_closure; \
1387 #define putMVarZh(node, value) \
1390 if (INFO_PTR(node) == (W_) FullSVar_info) { \
1391 /* Don't wrap the calls; we're done with STG land */\
1393 fprintf(stderr, "putMVar#: MVar already full.\n"); \
1394 EXIT(EXIT_FAILURE); \
1396 SET_INFO_PTR(node, FullSVar_info); \
1397 SVAR_VALUE(node) = value; \
1406 #define readIVarZh(r, liveness, node) \
1408 if (INFO_PTR(node) != (W_) ImMutArrayOfPtrs_info) { \
1409 if (SVAR_HEAD(node) == PrelBase_Z91Z93_closure) \
1410 SVAR_HEAD(node) = CurrentTSO; \
1412 TSO_LINK(SVAR_TAIL(node)) = CurrentTSO; \
1413 TSO_LINK(CurrentTSO) = (P_) PrelBase_Z91Z93_closure; \
1414 SVAR_TAIL(node) = CurrentTSO; \
1415 DO_YIELD(liveness << 1); \
1417 r = SVAR_VALUE(node); \
1422 #define readIVarZh(r, liveness, node) \
1424 if (INFO_PTR(node) != (W_) ImMutArrayOfPtrs_info) { \
1425 /* Don't wrap the calls; we're done with STG land */\
1427 fprintf(stderr, "readIVar#: IVar is empty.\n"); \
1428 EXIT(EXIT_FAILURE); \
1430 r = SVAR_VALUE(node); \
1441 /* Only difference to the !GRAN def: RunnableThreadsHd has been replaced by */
1442 /* ThreadQueueHd i.e. the tso is added at the end of the thread queue on */
1443 /* the CurrentProc. This means we have an implicit context switch after */
1444 /* writeIVar even if unfair scheduling is used in GranSim (default)! -- HWL */
1446 #define writeIVarZh(node, value) \
1449 if (INFO_PTR(node) == (W_) ImMutArrayOfPtrs_info) { \
1450 /* Don't wrap the calls; we're done with STG land */\
1452 fprintf(stderr, "writeIVar#: IVar already full.\n");\
1453 EXIT(EXIT_FAILURE); \
1455 tso = SVAR_HEAD(node); \
1456 if (tso != (P_) PrelBase_Z91Z93_closure) { \
1457 if (ThreadQueueHd == PrelBase_Z91Z93_closure) \
1458 ThreadQueueHd = tso; \
1460 TSO_LINK(ThreadQueueTl) = tso; \
1461 while(TSO_LINK(tso) != PrelBase_Z91Z93_closure) { \
1463 STGCALL3(void,(void *, char *, P_, P_),QP_Event2,do_qp_prof > 1 ? "RA" : "RG",tso,CurrentTSO); \
1464 tso = TSO_LINK(tso); \
1467 STGCALL3(void,(void *, char *, P_, P_),QP_Event2,do_qp_prof > 1 ? "RA" : "RG",tso,CurrentTSO); \
1468 ThreadQueueTl = tso; \
1470 /* Don't use freeze, since it's conditional on GC */ \
1471 SET_INFO_PTR(node, ImMutArrayOfPtrs_info); \
1472 MUTUPLE_CLOSURE_SIZE(node) = (MUTUPLE_VHS+1); \
1473 SVAR_VALUE(node) = value; \
1478 #define writeIVarZh(node, value) \
1481 if (INFO_PTR(node) == (W_) ImMutArrayOfPtrs_info) { \
1482 /* Don't wrap the calls; we're done with STG land */\
1484 fprintf(stderr, "writeIVar#: IVar already full.\n");\
1485 EXIT(EXIT_FAILURE); \
1487 tso = SVAR_HEAD(node); \
1488 if (tso != (P_) PrelBase_Z91Z93_closure) { \
1489 if (RunnableThreadsHd == PrelBase_Z91Z93_closure) \
1490 RunnableThreadsHd = tso; \
1492 TSO_LINK(RunnableThreadsTl) = tso; \
1493 while(TSO_LINK(tso) != PrelBase_Z91Z93_closure) { \
1495 STGCALL3(void,(void *, char *, P_, P_),QP_Event2,do_qp_prof > 1 ? "RA" : "RG",tso,CurrentTSO); \
1496 tso = TSO_LINK(tso); \
1499 STGCALL3(void,(void *, char *, P_, P_),QP_Event2,do_qp_prof > 1 ? "RA" : "RG",tso,CurrentTSO); \
1500 RunnableThreadsTl = tso; \
1502 /* Don't use freeze, since it's conditional on GC */ \
1503 SET_INFO_PTR(node, ImMutArrayOfPtrs_info); \
1504 MUTUPLE_CLOSURE_SIZE(node) = (MUTUPLE_VHS+1); \
1505 SVAR_VALUE(node) = value; \
1512 #define writeIVarZh(node, value) \
1515 if (INFO_PTR(node) == (W_) ImMutArrayOfPtrs_info) { \
1516 /* Don't wrap the calls; we're done with STG land */\
1518 fprintf(stderr, "writeIVar#: IVar already full.\n");\
1519 EXIT(EXIT_FAILURE); \
1521 /* Don't use freeze, since it's conditional on GC */ \
1522 SET_INFO_PTR(node, ImMutArrayOfPtrs_info); \
1523 MUTUPLE_CLOSURE_SIZE(node) = (MUTUPLE_VHS+1); \
1524 SVAR_VALUE(node) = value; \
1530 %************************************************************************
1532 \subsubsection[StgMacros-Wait-primops]{Delay/Wait PrimOps}
1534 %************************************************************************
1539 /* ToDo: for GRAN */
1541 #define delayZh(liveness, us) \
1543 if (WaitingThreadsTl == PrelBase_Z91Z93_closure) \
1544 WaitingThreadsHd = CurrentTSO; \
1546 TSO_LINK(WaitingThreadsTl) = CurrentTSO; \
1547 WaitingThreadsTl = CurrentTSO; \
1548 TSO_LINK(CurrentTSO) = PrelBase_Z91Z93_closure; \
1549 TSO_EVENT(CurrentTSO) = (W_) ((us) < 1 ? 1 : (us)); \
1550 DO_YIELD(liveness << 1); \
1555 #define delayZh(liveness, us) \
1558 fprintf(stderr, "delay#: unthreaded build.\n"); \
1559 EXIT(EXIT_FAILURE); \
1566 /* ToDo: something for GRAN */
1568 #define waitReadZh(liveness, fd) \
1570 if (WaitingThreadsTl == PrelBase_Z91Z93_closure) \
1571 WaitingThreadsHd = CurrentTSO; \
1573 TSO_LINK(WaitingThreadsTl) = CurrentTSO; \
1574 WaitingThreadsTl = CurrentTSO; \
1575 TSO_LINK(CurrentTSO) = PrelBase_Z91Z93_closure; \
1576 TSO_EVENT(CurrentTSO) = (W_) (-(fd)); \
1577 DO_YIELD(liveness << 1); \
1582 #define waitReadZh(liveness, fd) \
1585 fprintf(stderr, "waitRead#: unthreaded build.\n"); \
1586 EXIT(EXIT_FAILURE); \
1593 /* ToDo: something for GRAN */
1595 #ifdef HAVE_SYS_TYPES_H
1596 #include <sys/types.h>
1597 #endif HAVE_SYS_TYPES_H */
1599 #define waitWriteZh(liveness, fd) \
1601 if (WaitingThreadsTl == PrelBase_Z91Z93_closure) \
1602 WaitingThreadsHd = CurrentTSO; \
1604 TSO_LINK(WaitingThreadsTl) = CurrentTSO; \
1605 WaitingThreadsTl = CurrentTSO; \
1606 TSO_LINK(CurrentTSO) = PrelBase_Z91Z93_closure; \
1607 TSO_EVENT(CurrentTSO) = (W_) (-(fd+FD_SETSIZE)); \
1608 DO_YIELD(liveness << 1); \
1613 #define waitWriteZh(liveness, fd) \
1616 fprintf(stderr, "waitWrite#: unthreaded build.\n"); \
1617 EXIT(EXIT_FAILURE); \
1624 %************************************************************************
1626 \subsubsection[StgMacros-IO-primops]{Primitive I/O, error-handling primops}
1628 %************************************************************************
1631 extern P_ TopClosure;
1632 EXTFUN(ErrorIO_innards);
1633 EXTFUN(__std_entry_error__);
1635 #define errorIOZh(a) \
1636 do { TopClosure=(a); \
1637 (void) SAFESTGCALL1(I_,(void *, FILE *),fflush,stdout); \
1638 (void) SAFESTGCALL1(I_,(void *, FILE *),fflush,stderr); \
1639 JMP_(ErrorIO_innards); \
1642 #if !defined(CALLER_SAVES_SYSTEM)
1643 /* can use the macros */
1644 #define stg_getc(stream) getc((FILE *) (stream))
1645 #define stg_putc(c,stream) putc((c),((FILE *) (stream)))
1647 /* must not use the macros (they contain embedded calls to _filbuf/whatnot) */
1648 #define stg_getc(stream) SAFESTGCALL1(I_,(void *, FILE *),fgetc,(FILE *) (stream))
1649 #define stg_putc(c,stream) SAFESTGCALL2(I_,(void *, char, FILE *),fputc,(c),((FILE *) (stream)))
1652 int initialize_virtual_timer(int us);
1653 int install_segv_handler(STG_NO_ARGS);
1654 int install_vtalrm_handler(STG_NO_ARGS);
1655 void initUserSignals(STG_NO_ARGS);
1656 void blockUserSignals(STG_NO_ARGS);
1657 void unblockUserSignals(STG_NO_ARGS);
1658 IF_RTS(void blockVtAlrmSignal(STG_NO_ARGS);)
1659 IF_RTS(void unblockVtAlrmSignal(STG_NO_ARGS);)
1660 IF_RTS(void AwaitEvent(I_ delta);)
1662 #if defined(_POSIX_SOURCE) && !defined(nextstep3_TARGET_OS)
1663 /* For nextstep3_TARGET_OS comment see stgdefs.h. CaS */
1664 extern I_ sig_install PROTO((I_, I_, sigset_t *));
1665 #define stg_sig_ignore(s,m) SAFESTGCALL3(I_,(void *, I_, I_),sig_install,s,STG_SIG_IGN,(sigset_t *)m)
1666 #define stg_sig_default(s,m) SAFESTGCALL3(I_,(void *, I_, I_),sig_install,s,STG_SIG_DFL,(sigset_t *)m)
1667 #define stg_sig_catch(s,sp,m) SAFESTGCALL3(I_,(void *, I_, I_),sig_install,s,sp,(sigset_t *)m)
1669 extern I_ sig_install PROTO((I_, I_));
1670 #define stg_sig_ignore(s,m) SAFESTGCALL2(I_,(void *, I_, I_),sig_install,s,STG_SIG_IGN)
1671 #define stg_sig_default(s,m) SAFESTGCALL2(I_,(void *, I_, I_),sig_install,s,STG_SIG_DFL)
1672 #define stg_sig_catch(s,sp,m) SAFESTGCALL2(I_,(void *, I_, I_),sig_install,s,sp)
1675 #define STG_SIG_DFL (-1)
1676 #define STG_SIG_IGN (-2)
1677 #define STG_SIG_ERR (-3)
1679 StgInt getErrorHandler(STG_NO_ARGS);
1681 void raiseError PROTO((StgStablePtr handler));
1682 StgInt catchError PROTO((StgStablePtr newErrorHandler));
1684 void decrementErrorCount(STG_NO_ARGS);
1686 #define stg_catchError(sp) SAFESTGCALL1(I_,(void *, StgStablePtr),catchError,sp)
1687 #define stg_decrementErrorCount() SAFESTGCALL0(void,(void *),decrementErrorCount)
1690 %************************************************************************
1692 \subsubsection[StgMacros-stable-ptr]{Primitive ops for manipulating stable pointers}
1694 %************************************************************************
1697 The type of these should be:
1700 makeStablePointer# :: a -> State# _RealWorld -> StateAndStablePtr# _RealWorld a
1701 deRefStablePointer# :: StablePtr# a -> State# _RealWorld -> StateAndPtr _RealWorld a
1704 Since world-tokens are no longer explicitly passed around, the
1705 implementations have a few less arguments/results.
1707 The simpler one is @deRefStablePointer#@ (which is only a primop
1708 because it is more polymorphic than is allowed of a ccall).
1713 #define deRefStablePtrZh(ri,sp) \
1716 fprintf(stderr, "deRefStablePtr#: no stable pointer support.\n");\
1717 EXIT(EXIT_FAILURE); \
1722 extern StgPtr _deRefStablePointer PROTO((StgInt, StgPtr));
1724 #define deRefStablePtrZh(ri,sp) \
1725 ri = SAFESTGCALL2(I_,(void *, I_, P_),_deRefStablePointer,sp,StorageMgrInfo.StablePointerTable);
1728 Declarations for other stable pointer operations.
1731 void freeStablePointer PROTO((I_ stablePtr));
1733 void enterStablePtr PROTO((StgStablePtr, StgFunPtr));
1734 void performIO PROTO((StgStablePtr));
1735 I_ enterInt PROTO((StgStablePtr));
1736 I_ enterFloat PROTO((StgStablePtr));
1737 P_ deRefStablePointer PROTO((StgStablePtr));
1738 IF_RTS(I_ catchSoftHeapOverflow PROTO((StgStablePtr, I_));)
1739 IF_RTS(I_ getSoftHeapOverflowHandler(STG_NO_ARGS);)
1740 IF_RTS(extern StgStablePtr softHeapOverflowHandler;)
1741 IF_RTS(void shutdownHaskell(STG_NO_ARGS);)
1743 EXTFUN(stopPerformIODirectReturn);
1744 EXTFUN(startPerformIO);
1745 EXTFUN(stopEnterIntDirectReturn);
1746 EXTFUN(startEnterInt);
1747 EXTFUN(stopEnterFloatDirectReturn);
1748 EXTFUN(startEnterFloat);
1750 void enterStablePtr PROTO((StgStablePtr stableIndex, StgFunPtr startCode));
1754 IF_RTS(extern I_ ErrorIO_call_count;)
1757 Somewhat harder is @makeStablePointer#@ --- it is usually simple but
1758 if we're unlucky, it will have to allocate a new table and copy the
1759 old bit over. Since we might, very occasionally, have to call the
1760 garbage collector, this has to be a macro... sigh!
1762 NB @newSP@ is required because it is entirely possible that
1763 @stablePtr@ and @unstablePtr@ are aliases and so we can't do the
1764 assignment to @stablePtr@ until we've finished with @unstablePtr@.
1766 Another obscure piece of coding is the recalculation of the size of
1767 the table. We do this just in case Jim's threads decide they want to
1768 context switch---in which case any stack-allocated variables may get
1769 trashed. (If only there was a special heap check which didn't
1770 consider context switching...)
1775 /* Calculate SP Table size from number of pointers */
1776 #define SPTSizeFromNoPtrs( newNP ) (DYN_VHS + 1 + 2 * (newNP))
1778 /* Calculate number of pointers in new table from number in old table:
1779 any strictly increasing expression will do here */
1780 #define CalcNewNoSPtrs( i ) ((i)*2 + 100)
1782 void enlargeSPTable PROTO((P_, P_));
1784 #define makeStablePtrZh(stablePtr,liveness,unstablePtr) \
1786 EXTDATA_RO(StablePointerTable_info); \
1787 EXTDATA(UnusedSP); \
1788 StgStablePtr newSP; \
1790 if (SPT_EMPTY(StorageMgrInfo.StablePointerTable)) { /* free stack is empty */ \
1791 { /* Variables used before the heap check */ \
1792 I_ OldNoPtrs = SPT_NoPTRS( StorageMgrInfo.StablePointerTable ); \
1793 I_ NewNoPtrs = CalcNewNoSPtrs( OldNoPtrs ); \
1794 I_ NewSize = SPTSizeFromNoPtrs( NewNoPtrs ); \
1795 HEAP_CHK(liveness, _FHS+NewSize, 0); \
1797 { /* Variables used after the heap check - same values */ \
1798 I_ OldNoPtrs = SPT_NoPTRS( StorageMgrInfo.StablePointerTable ); \
1799 I_ NewNoPtrs = CalcNewNoSPtrs( OldNoPtrs ); \
1800 I_ NewSize = SPTSizeFromNoPtrs( NewNoPtrs ); \
1801 P_ SPTable = Hp + 1 - (_FHS + NewSize); \
1803 CC_ALLOC(CCC, _FHS+NewSize, SPT_K); /* cc prof */ \
1804 SET_DYN_HDR(SPTable,StablePointerTable_info,CCC,NewSize,NewNoPtrs);\
1805 SAFESTGCALL2(void, (void *, P_, P_), enlargeSPTable, SPTable, StorageMgrInfo.StablePointerTable); \
1806 StorageMgrInfo.StablePointerTable = SPTable; \
1810 newSP = SPT_POP(StorageMgrInfo.StablePointerTable); \
1811 SPT_SPTR(StorageMgrInfo.StablePointerTable, newSP) = unstablePtr; \
1812 CHECK_SPT_CLOSURE( StorageMgrInfo.StablePointerTable ); \
1813 stablePtr = newSP; \
1818 #define makeStablePtrZh(stablePtr,liveness,unstablePtr) \
1821 fprintf(stderr, "makeStablePtr#: no stable pointer support.\n");\
1822 EXIT(EXIT_FAILURE); \
1828 %************************************************************************
1830 \subsubsection[StgMacros-unsafePointerEquality]{Primitive `op' for breaking referential transparency}
1832 %************************************************************************
1834 The type of this is @reallyUnsafePtrEquality :: a -> a -> Int#@ so we
1835 can expect three parameters: the two arguments and a "register" to put
1838 Message to Will: This primop breaks referential transparency so badly
1839 you might want to leave it out. On the other hand, if you hide it
1840 away in an appropriate monad, it's perfectly safe. [ADR]
1842 Note that this primop is non-deterministic: different results can be
1843 obtained depending on just what the garbage collector (and code
1844 optimiser??) has done. However, we can guarantee that if two objects
1845 are pointer-equal, they have the same denotation --- the converse most
1846 certainly doesn't hold.
1848 ToDo ADR: The degree of non-determinism could be greatly reduced by
1849 following indirections.
1852 #define reallyUnsafePtrEqualityZh(r,a,b) r=((StgPtr)(a) == (StgPtr)(b))
1855 %************************************************************************
1857 \subsubsection[StgMacros-parallel-primop]{Primitive `op' for sparking (etc)}
1859 %************************************************************************
1861 Assuming local sparking in some form, we can now inline the spark request.
1863 We build a doubly-linked list in the heap, so that we can handle FIFO
1864 or LIFO scheduling as we please.
1866 Anything with tag >= 0 is in WHNF, so we discard it.
1871 ED_(PrelBase_Z91Z93_closure);
1875 #define parZh(r,node) \
1876 PARZh(r,node,1,0,0,0,0,0)
1878 #define parAtZh(r,node,where,identifier,gran_info,size_info,par_info,rest) \
1879 parATZh(r,node,where,identifier,gran_info,size_info,par_info,rest,1)
1881 #define parAtAbsZh(r,node,proc,identifier,gran_info,size_info,par_info,rest) \
1882 parATZh(r,node,proc,identifier,gran_info,size_info,par_info,rest,2)
1884 #define parAtRelZh(r,node,proc,identifier,gran_info,size_info,par_info,rest) \
1885 parATZh(r,node,proc,identifier,gran_info,size_info,par_info,rest,3)
1887 #define parAtForNowZh(r,node,where,identifier,gran_info,size_info,par_info,rest) \
1888 parATZh(r,node,where,identifier,gran_info,size_info,par_info,rest,0)
1890 #define parATZh(r,node,where,identifier,gran_info,size_info,par_info,rest,local) \
1893 if (SHOULD_SPARK(node)) { \
1896 result = NewSpark((P_)node,identifier,gran_info,size_info,par_info,local); \
1897 if (local==2) { /* special case for parAtAbs */ \
1898 GranSimSparkAtAbs(result,(I_)where,identifier);\
1899 } else if (local==3) { /* special case for parAtRel */ \
1900 GranSimSparkAtAbs(result,(I_)(CurrentProc+where),identifier); \
1902 GranSimSparkAt(result,where,identifier); \
1904 context_switch = 1; \
1906 RestoreAllStgRegs(); \
1907 } else if (do_qp_prof) { \
1908 I_ tid = threadId++; \
1909 SAFESTGCALL2(void,(I_, P_),QP_Event0,tid,node); \
1911 r = 1; /* return code for successful spark -- HWL */ \
1914 #define parLocalZh(r,node,identifier,gran_info,size_info,par_info,rest) \
1915 PARZh(r,node,rest,identifier,gran_info,size_info,par_info,1)
1917 #define parGlobalZh(r,node,identifier,gran_info,size_info,par_info,rest) \
1918 PARZh(r,node,rest,identifier,gran_info,size_info,par_info,0)
1922 #define PARZh(r,node,rest,identifier,gran_info,size_info,par_info,local) \
1924 if (SHOULD_SPARK(node)) { \
1927 result = NewSpark((P_)node,identifier,gran_info,size_info,par_info,local);\
1928 add_to_spark_queue(result); \
1929 GranSimSpark(local,(P_)node); \
1930 context_switch = 1; \
1932 RestoreAllStgRegs(); \
1933 } else if (do_qp_prof) { \
1934 I_ tid = threadId++; \
1935 SAFESTGCALL2(void,(I_, P_),QP_Event0,tid,node); \
1937 r = 1; /* return code for successful spark -- HWL */ \
1942 #define PARZh(r,node,rest,identifier,gran_info,size_info,par_info,local) \
1945 if (SHOULD_SPARK(node)) { \
1946 result = NewSpark((P_)node,identifier,gran_info,size_info,par_info,local);\
1947 ADD_TO_SPARK_QUEUE(result); \
1948 SAFESTGCALL2(void,(W_),GranSimSpark,local,(P_)node); \
1949 /* context_switch = 1; not needed any more -- HWL */ \
1950 } else if (do_qp_prof) { \
1951 I_ tid = threadId++; \
1952 SAFESTGCALL2(void,(I_, P_),QP_Event0,tid,node); \
1954 r = 1; /* return code for successful spark -- HWL */ \
1959 #define copyableZh(r,node) \
1960 /* copyable not yet implemented!! */
1962 #define noFollowZh(r,node) \
1963 /* noFollow not yet implemented!! */
1967 extern I_ required_thread_count;
1970 #define COUNT_SPARK TSO_GLOBALSPARKS(CurrentTSO)++
1976 Note that we must bump the required thread count NOW, rather
1977 than when the thread is actually created.
1980 #define forkZh(r,liveness,node) \
1982 while (PendingSparksTl[REQUIRED_POOL] == PendingSparksLim[REQUIRED_POOL]) \
1983 DO_YIELD((liveness << 1) | 1); \
1985 if (SHOULD_SPARK(node)) { \
1986 *PendingSparksTl[REQUIRED_POOL]++ = (P_)(node); \
1987 } else if (DO_QP_PROF) { \
1988 I_ tid = threadId++; \
1989 SAFESTGCALL2(void,(I_, P_),QP_Event0,tid,node); \
1991 required_thread_count++; \
1992 context_switch = 1; \
1993 r = 1; /* Should not be necessary */ \
1996 #define parZh(r,node) \
1999 if (SHOULD_SPARK(node) && \
2000 PendingSparksTl[ADVISORY_POOL] < PendingSparksLim[ADVISORY_POOL]) { \
2001 *PendingSparksTl[ADVISORY_POOL]++ = (P_)(node); \
2005 I_ tid = threadId++; \
2006 SAFESTGCALL2(void,(I_, P_),QP_Event0,tid,node); \
2009 r = 1; /* Should not be necessary */ \
2015 The following seq# code should only be used in unoptimized code.
2016 Be warned: it's a potential bug-farm.
2018 First we push two words on the B stack: the current value of RetReg
2019 (which may or may not be live), and a continuation snatched largely out
2020 of thin air (it's a point within this code block). Then we set RetReg
2021 to the special polymorphic return code for seq, load up Node with the
2022 closure to be evaluated, and we're off. When the eval returns to the
2023 polymorphic seq return point, the two words are popped off the B stack,
2024 RetReg is restored, and we jump to the continuation, completing the
2025 primop and going on our merry way.
2031 #define seqZh(r,liveness,node) \
2034 /* STK_CHK(liveness,0,2,0,0,0,0); */ \
2035 /* SpB -= BREL(2); */ \
2036 SpB[BREL(0)] = (W_) RetReg; \
2037 SpB[BREL(1)] = (W_) &&cont; \
2038 RetReg = (StgRetAddr) vtbl_seq; \
2041 InfoPtr = (D_)(INFO_PTR(Node)); \
2042 JMP_(ENTRY_CODE(InfoPtr)); \
2044 r = 1; /* Should be unnecessary */ \
2047 #endif /* CONCURRENT */
2050 %************************************************************************
2052 \subsubsection[StgMacros-foreign-objects]{Foreign Objects}
2054 %************************************************************************
2056 [Based on previous MallocPtr comments -- SOF]
2058 This macro is used to construct a ForeignObj on the heap.
2060 What this does is plug the pointer (which will be in a local
2061 variable) together with its finalising/free routine, into a fresh heap
2062 object and then sets a result (which will be a register) to point
2063 to the fresh heap object.
2065 To accommodate per-object finalisation, augment the macro with a
2066 finalisation routine argument. Nothing spectacular, just plug the
2067 pointer to the routine into the ForeignObj -- SOF 4/96
2069 Question: what's this "SET_ACTIVITY" stuff - should I be doing this
2070 too? (It's if you want to use the SPAT profiling tools to
2071 characterize program behavior by ``activity'' -- tail-calling,
2072 heap-checking, etc. -- see Ticky.lh. It is quite specialized.
2075 (Swapped first two arguments to make it come into line with what appears
2076 to be `standard' format, return register then liveness mask. -- SOF 4/96)
2081 StgInt eqForeignObj PROTO((StgForeignObj p1, StgForeignObj p2));
2083 #define makeForeignObjZh(r, liveness, mptr, finalise) \
2087 HEAP_CHK((W_)liveness, _FHS + ForeignObj_SIZE,0); \
2088 CC_ALLOC(CCC,_FHS + ForeignObj_SIZE,ForeignObj_K); /* cc prof */ \
2090 result = Hp + 1 - (_FHS + ForeignObj_SIZE); \
2091 SET_ForeignObj_HDR(result,ForeignObj_info,CCC,_FHS + ForeignObj_SIZE,0); \
2092 ForeignObj_CLOSURE_DATA(result) = (P_)mptr; \
2093 ForeignObj_CLOSURE_FINALISER(result) = (P_)finalise; \
2094 ForeignObj_CLOSURE_LINK(result) = StorageMgrInfo.ForeignObjList; \
2095 StorageMgrInfo.ForeignObjList = result; \
2098 /*fprintf(stderr,"DEBUG: ForeignObj(0x%x) = <0x%x, 0x%x, 0x%x, 0x%x>\n", \
2100 result[0],result[1], \
2101 result[2],result[3]);*/ \
2103 CHECK_ForeignObj_CLOSURE( result ); \
2104 VALIDATE_ForeignObjList( StorageMgrInfo.ForeignObjList ); \
2106 (r) = (P_) result; \
2109 #define writeForeignObjZh(res,datum) ((PP_) ForeignObj_CLOSURE_DATA(res)) = ((P_)datum)
2112 #define makeForeignObjZh(r, liveness, mptr, finalise) \
2115 fprintf(stderr, "makeForeignObj#: no foreign object support.\n");\
2116 EXIT(EXIT_FAILURE); \
2119 #define writeForeignObjZh(res,datum) \
2122 fprintf(stderr, "writeForeignObj#: no foreign object support.\n");\
2123 EXIT(EXIT_FAILURE); \
2130 End-of-file's multi-slurp protection:
2132 #endif /* ! STGMACROS_H */