rubyx/lib/vm/block.rb

126 lines
4.1 KiB
Ruby

require_relative "values"
module Vm
# Think flowcharts: blocks are the boxes. The smallest unit of linear code
# Blocks must end in control instructions (jump/call/return).
# And the only valid argument for a jump is a Block
# Blocks form a linked list
# There are four ways for a block to get data (to work on)
# - hard coded constants (embedded in code)
# - memory move
# - values passed in (from previous blocks. ie local variables)
# See Value description on how to create code/instructions
# Codes then get assembled into bytes (after linking)
class Block < Code
def initialize(name , function , next_block = nil)
super()
@function = function
@name = name.to_sym
@next = next_block
@codes = []
end
attr_reader :name , :next , :codes , :function
def length
cods = @codes.inject(0) {| sum , item | sum + item.length}
cods += @next.length if @next
cods
end
def add_code(kode)
if kode.is_a? Hash
raise "Hack only for 1 element #{inspect} #{kode.inspect}" unless kode.length == 1
instruction , result = kode.first
instruction.assign result
kode = instruction
end
raise "alarm #{kode}" if kode.is_a? Word
raise "alarm #{kode}" unless kode.is_a? Code
@codes << kode
self
end
alias :<< :add_code
alias :a :add_code
def link_at pos , context
@position = pos
@codes.each do |code|
code.link_at(pos , context)
pos += code.length
end
if @next
@next.link_at pos , context
pos += @next.length
end
pos
end
def assemble(io)
@codes.each do |obj|
obj.assemble io
end
@next.assemble(io) if @next
end
# create a new linear block after this block. Linear means there is no brach needed from this one
# to the new one. Usually the new one just serves as jump address for a control statement
# In code generation (assembly) , new new_block is written after this one, ie zero runtime cost
def new_block name
new_b = Block.new( name , @function , @next )
@next = new_b
return new_b
end
# to use the assignment syntax (see method_missing) the scope must be set, so variables can be resolved
# The scope you set should be a binding (literally, the kernel.binding)
# The function return the block, so it can be chained into an assignment
# Example (coding a function ) and having variable int defined
# b = function.body.scope(binding)
# b.int = 5 will create a mov instruction to set the register that int points to
def scope where
@scope = where
self
end
# sugar to create instructions easily. Actually just got double sweet with two versions:
# 1 for any method that ends in = we evaluate the method name in the current scope (see scope())
# for the result we call assign with the right value. The resulting instruction is added to
# the block.
# Thus we emulate assignment,
# Example: block b
# b.variable = value looks like what it does, but actually generates
# an instruction for the block (mov or add)
#
# 2- any other method will be passed on to the RegisterMachine and the result added to the block
# With this trick we can write what looks like assembler,
# Example b.instance_eval
# mov( r1 , r2 )
# add( r1 , r2 , 4)
# end
# mov and add will be called on Machine and generate Inststuction that are then added
# to the block
def method_missing(meth, *args, &block)
var = meth.to_s[0 ... -1]
if( args.length == 1) and ( meth.to_s[-1] == "=" )
if @scope.local_variable_defined? var.to_sym
l_val = @scope.local_variable_get var.to_sym
return add_code l_val.assign(args[0])
else
return super
end
end
add_code RegisterMachine.instance.send(meth , *args)
end
end
end