Return to xxxprim CVS log

Up to [gforth] / gforth

more stack caching stuff

    1: \ Gforth primitives
    2: 
    3: \ Copyright (C) 1995,1996,1997,1998,2000,2003 Free Software Foundation, Inc.
    4: 
    5: \ This file is part of Gforth.
    6: 
    7: \ Gforth is free software; you can redistribute it and/or
    8: \ modify it under the terms of the GNU General Public License
    9: \ as published by the Free Software Foundation; either version 2
   10: \ of the License, or (at your option) any later version.
   11: 
   12: \ This program is distributed in the hope that it will be useful,
   13: \ but WITHOUT ANY WARRANTY; without even the implied warranty of
   14: \ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   15: \ GNU General Public License for more details.
   16: 
   17: \ You should have received a copy of the GNU General Public License
   18: \ along with this program; if not, write to the Free Software
   19: \ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111, USA.
   20: 
   21: 
   22: \ WARNING: This file is processed by m4. Make sure your identifiers
   23: \ don't collide with m4's (e.g. by undefining them).
   24: \ 
   25: \ 
   26: \ 
   27: \ This file contains primitive specifications in the following format:
   28: \ 
   29: \ forth name	( stack effect )	category	[pronunciation]
   30: \ [""glossary entry""]
   31: \ C code
   32: \ [:
   33: \ Forth code]
   34: \ 
   35: \ Note: Fields in brackets are optional.  Word specifications have to
   36: \ be separated by at least one empty line
   37: \
   38: \ Both pronounciation and stack items (in the stack effect) must
   39: \ conform to the C identifier syntax or the C compiler will complain.
   40: \ If you don't have a pronounciation field, the Forth name is used,
   41: \ and has to conform to the C identifier syntax.
   42: \ 
   43: \ These specifications are automatically translated into C-code for the
   44: \ interpreter and into some other files. I hope that your C compiler has
   45: \ decent optimization, otherwise the automatically generated code will
   46: \ be somewhat slow. The Forth version of the code is included for manual
   47: \ compilers, so they will need to compile only the important words.
   48: \ 
   49: \ Note that stack pointer adjustment is performed according to stack
   50: \ effect by automatically generated code and NEXT is automatically
   51: \ appended to the C code. Also, you can use the names in the stack
   52: \ effect in the C code. Stack access is automatic. One exception: if
   53: \ your code does not fall through, the results are not stored into the
   54: \ stack. Use different names on both sides of the '--', if you change a
   55: \ value (some stores to the stack are optimized away).
   56: \
   57: \ For superinstructions the syntax is:
   58: \
   59: \ forth-name [/ c-name] = forth-name forth-name ...
   60: \
   61: \ 
   62: \ The stack variables have the following types:
   63: \ 
   64: \ name matches	type
   65: \ f.*		Bool
   66: \ c.*		Char
   67: \ [nw].*	Cell
   68: \ u.*		UCell
   69: \ d.*		DCell
   70: \ ud.*		UDCell
   71: \ r.*		Float
   72: \ a_.*		Cell *
   73: \ c_.*		Char *
   74: \ f_.*		Float *
   75: \ df_.*		DFloat *
   76: \ sf_.*		SFloat *
   77: \ xt.*		XT
   78: \ f83name.*	F83Name *
   79: 
   80: \E stack data-stack   sp Cell
   81: \E stack fp-stack     fp Float
   82: \E stack return-stack rp Cell
   83: \E
   84: \E get-current prefixes set-current
   85: \E 
   86: \E s" Bool"		single data-stack type-prefix f
   87: \E s" Char"		single data-stack type-prefix c
   88: \E s" Cell"		single data-stack type-prefix n
   89: \E s" Cell"		single data-stack type-prefix w
   90: \E s" UCell"		single data-stack type-prefix u
   91: \E s" DCell"		double data-stack type-prefix d
   92: \E s" UDCell"		double data-stack type-prefix ud
   93: \E s" Float"		single fp-stack   type-prefix r
   94: \E s" Cell *"		single data-stack type-prefix a_
   95: \E s" Char *"		single data-stack type-prefix c_
   96: \E s" Float *"		single data-stack type-prefix f_
   97: \E s" DFloat *"		single data-stack type-prefix df_
   98: \E s" SFloat *"		single data-stack type-prefix sf_
   99: \E s" Xt"		single data-stack type-prefix xt
  100: \E s" struct F83Name *"	single data-stack type-prefix f83name
  101: \E s" struct Longname *" single data-stack type-prefix longname
  102: \E 
  103: \E return-stack stack-prefix R:
  104: \E inst-stream  stack-prefix #
  105: \E 
  106: \E set-current
  107: \E store-optimization on
  108: \E ' noop tail-nextp2 ! \ now INST_TAIL just stores, but does not jump
  109: \E
  110: \E include-skipped-insts on \ static superinsts include cells for components
  111: \E                          \ useful for dynamic programming and
  112: \E                          \ superinsts across entry points
  113: 
  114: \ 
  115: \ 
  116: \ 
  117: \ In addition the following names can be used:
  118: \ ip	the instruction pointer
  119: \ sp	the data stack pointer
  120: \ rp	the parameter stack pointer
  121: \ lp	the locals stack pointer
  122: \ NEXT	executes NEXT
  123: \ cfa	
  124: \ NEXT1	executes NEXT1
  125: \ FLAG(x)	makes a Forth flag from a C flag
  126: \ 
  127: \ 
  128: \ 
  129: \ Percentages in comments are from Koopmans book: average/maximum use
  130: \ (taken from four, not very representative benchmarks)
  131: \ 
  132: \ 
  133: \ 
  134: \ To do:
  135: \ 
  136: \ throw execute, cfa and NEXT1 out?
  137: \ macroize *ip, ip++, *ip++ (pipelining)?
  138: 
  139: \ set up stack caching states
  140: 
  141: \E register IPTOS Cell
  142: \E register spTOS Cell
  143: \E register sp1 Cell
  144: \E register sp2 Cell
  145: \E register sp3 Cell
  146: 
  147: \E create IPregs IPTOS ,
  148: \E create regs sp2 , sp1 , spTOS ,
  149: 
  150: \E IPregs 1 0 stack-state IPss1
  151: \E regs           0 -1 stack-state ss0
  152: \E regs 2 cells + 1  0 stack-state ss1
  153: \E regs 1 cells + 2  1 stack-state ss2
  154: \E regs 0 cells + 3  2 stack-state ss3
  155: 
  156: \E state S0
  157: \E state S1
  158: \E state S2
  159: \E state S3
  160: 
  161: \E ss0 data-stack S0 set-ss
  162: \E ss1 data-stack S1 set-ss
  163: \E ss2 data-stack S2 set-ss
  164: \E ss3 data-stack S3 set-ss
  165: 
  166: \E IPss1 inst-stream S0 set-ss
  167: \E IPss1 inst-stream S1 set-ss
  168: \E IPss1 inst-stream S2 set-ss
  169: \E IPss1 inst-stream S3 set-ss
  170: 
  171: \E data-stack to cache-stack
  172: \E here 4 cache-states 2! s0 , s1 , s2 , s3 ,
  173: 
  174: \E S1 to state-in-default
  175: \E S1 to state-out-default
  176: \E state-in-default to state-in
  177: \E state-out-default to state-out
  178: 
  179: +	( n1 n2 -- n )		core	plus
  180: n = n1+n2;
  181: 
  182: lit	( #w -- w )		gforth
  183: :
  184:  r> dup @ swap cell+ >r ;
  185: 
  186: over ( n1 n2 -- n1 n2 n1 )
  187: 
  188: drop ( n -- )
  189: 
  190: ?branch ( #a_target f -- ) f83	question_branch
  191: if (f==0) {
  192:   SET_IP((Xt *)a_target);
  193:   INST_TAIL; NEXT_P2;
  194: }
  195: SUPER_CONTINUE;
  196: 
  197: noop ( -- )
  198: 
  199: \E prim-states drop
  200: \E prim-states over
  201: 
  202: \E S1 branch-states ?branch
  203: 
  204: \E gen-transitions noop
  205: