fisk 2.0.0 → 2.3.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: '0809155eef95d321b2e39b2c3a801f5ed987f74e409ba40285a30f530964c558'
-  data.tar.gz: 0b44765ae9dd4d4611c9db1af5a49bbe8659ce4dba3e6a7ac8887e17ebdb6e4a
+  metadata.gz: f39cf1256e6c2030ee2f0dee50d91995cb2efff76d457a5f4c9eb25ca543059b
+  data.tar.gz: c4b92832e08b09b3660efeea2f1c3a629053cc4651ce7f224404e3065a7c8640
 SHA512:
-  metadata.gz: 1835a32e1579e07b1a7cf101a85ea94dee3471cb2c5c64e66a385a59807b8c049e1218c8b8152d7aa562757d585d3d7700b435b5791fd42ddaccabf99a403973
-  data.tar.gz: 90a32029797fb6f0a101fcfc81e3352e69294ac2bd02a48f9f3e077dde900b4d3a917e413b4f65be599cdf3a0fd4f3b0a5419a043774d16292eae9821aaa6e82
+  metadata.gz: a8d6c2be54f415b76997a3ffd76ee1f690a7088c17df3a83efb9b6b2830804e6621d89c8b1aeb2b935f954d6efade6d8cc5480bf203078663d1e5a17a75f311c
+  data.tar.gz: e1368e7bab672530a501aea3e201cdd2f72f5083708102d631165b4468df56a842a9a36d8399a901d4c46d750a4c16f195059b2a7868fbc97a216f222f8bada9

data/.github/workflows/ci.yml

@@ -0,0 +1,34 @@
+name: CI
+
+on: [push, pull_request]
+
+jobs:
+  test:
+    runs-on: ${{ matrix.os }}-latest
+
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu, macos]
+
+    steps:
+      - uses: actions/checkout@v2
+      - name: Install libcapstone Linux
+        if: matrix.os == 'ubuntu'
+        run: |
+            sudo apt-get update -y
+            sudo apt-get install -y libcapstone-dev
+
+      - name: Install libcapstone macOS
+        if: matrix.os == 'macos'
+        run: |
+            brew update
+            brew install capstone
+
+      - name: Set up Ruby
+        uses: ruby/setup-ruby@477b21f02be01bcb8030d50f37cfec92bfa615b6
+        with:
+          ruby-version: 3.0
+          bundler-cache: true # 'bundle install' and cache
+      - name: Run tests
+        run: bundle exec rake

data/CODE_OF_CONDUCT.md

@@ -0,0 +1,78 @@
+
+# Contributor Covenant Code of Conduct
+
+## Our Pledge
+
+In the interest of fostering an open and welcoming environment, we as
+contributors and maintainers pledge to make participation in our project and
+our community a harassment-free experience for everyone, regardless of age, body
+size, disability, ethnicity, sex characteristics, gender identity and expression,
+level of experience, education, socio-economic status, nationality, personal
+appearance, race, religion, or sexual identity and orientation.
+
+## Our Standards
+
+Examples of behavior that contributes to creating a positive environment
+include:
+
+* Using welcoming and inclusive language
+* Being respectful of differing viewpoints and experiences
+* Gracefully accepting constructive criticism
+* Focusing on what is best for the community
+* Showing empathy towards other community members
+
+Examples of unacceptable behavior by participants include:
+
+* The use of sexualized language or imagery and unwelcome sexual attention or
+  advances
+* Trolling, insulting/derogatory comments, and personal or political attacks
+* Public or private harassment
+* Publishing others' private information, such as a physical or electronic
+  address, without explicit permission
+* Other conduct which could reasonably be considered inappropriate in a
+  professional setting
+
+## Our Responsibilities
+
+Project maintainers are responsible for clarifying the standards of acceptable
+behavior and are expected to take appropriate and fair corrective action in
+response to any instances of unacceptable behavior.
+
+Project maintainers have the right and responsibility to remove, edit, or
+reject comments, commits, code, wiki edits, issues, and other contributions
+that are not aligned to this Code of Conduct, or to ban temporarily or
+permanently any contributor for other behaviors that they deem inappropriate,
+threatening, offensive, or harmful.
+
+## Scope
+
+This Code of Conduct applies within all project spaces, and it also applies when
+an individual is representing the project or its community in public spaces.
+Examples of representing a project or community include using an official
+project e-mail address, posting via an official social media account, or acting
+as an appointed representative at an online or offline event. Representation of
+a project may be further defined and clarified by project maintainers.
+
+## Enforcement
+
+Instances of abusive, harassing, or otherwise unacceptable behavior may be
+reported by contacting the project team at aaron.patterson at gmail.com. All
+complaints will be reviewed and investigated and will result in a response that
+is deemed necessary and appropriate to the circumstances. The project team is
+obligated to maintain confidentiality with regard to the reporter of an incident.
+Further details of specific enforcement policies may be posted separately.
+
+Project maintainers who do not follow or enforce the Code of Conduct in good
+faith may face temporary or permanent repercussions as determined by other
+members of the project's leadership.
+
+## Attribution
+
+This Code of Conduct is adapted from the [Contributor Covenant][homepage], version 1.4,
+available at https://www.contributor-covenant.org/version/1/4/code-of-conduct.html
+
+[homepage]: https://www.contributor-covenant.org
+
+For answers to common questions about this code of conduct, see
+https://www.contributor-covenant.org/faq
+

data/Gemfile

@@ -1,4 +1,3 @@
 source "https://rubygems.org"
 
-gem "minitest"
-gem "crabstone"
+gemspec

data/README.md

@@ -1,4 +1,6 @@
-# Fisk - A Pure Ruby Assembler
+# Fisk - A Pure Ruby x86-64 Assembler
+
+Tired of writing Ruby in Ruby?  Now you can write assembly in Ruby with Fisk!
 
 This is a pure Ruby x86-64 assembler (I guess).  I'm not 100% sure if it counts
 as pure Ruby because it just reads an XML file and metaprograms most of it.
@@ -24,7 +26,60 @@ binary = fisk.asm do
 end
 ```
 
-Though, it's not very fun to assemble something unless you can execute it.  So
+Fisk uses Intel assembly syntax, so the first operand is the *destination*, and
+the second operand is the source.  So for example `mov rax, imm8(1)` means "put
+an immediate that is 8 bits wide with the value of 1 in the RAX register".
+
+## Sizes and Memory Operands
+
+In order to select the right x86 instruction to emit, Fisk needs to know the
+types of the operands.  Register types are already implied, but things like
+immediates and memory operands need to be specified.
+
+Below are a few examples.
+
+Putting a 32 bit immediate in the RAX register:
+
+```ruby
+mov rax, imm32(0xFFF)
+```
+
+Dereferencing the value in RAX and storing the value in the RAX register:
+
+```ruby
+mov rax, m64(rax)
+```
+
+Dereferencing the value 8 bytes away RAX and storing the value in the R9 register:
+
+```ruby
+mov r9, m64(rax, 8)
+```
+
+## Register Allocation
+
+Fisk supports simple register assignment.  You can make temporary registers,
+then have Fisk assign registers for you.  For example:
+
+```ruby
+fisk = Fisk.new
+
+# Make some temporary registers
+reg1 = fisk.register("temp1")
+reg2 = fisk.register("temp2")
+
+# XOR the two virtual registers
+fisk.xor reg1, reg2
+
+# Ask Fisk to assign registers from the pool of registers passed in
+fisk.assign_registers([fisk.r9, fisk.r10])
+```
+
+Currently, Fisk won't spill registers for you, it just raises an exception.
+
+## Executing Assembly
+
+Now, it's not very fun to assemble something unless you can execute it.  So
 here is an example of how to execute the above assembly.  This assembly code
 will send an interrupt and tell the debugger to stop.  So let's write the
 machine code to some executable memory, and call it from a Ruby program that we
@@ -116,3 +171,8 @@ Target 0: (ruby) stopped.
     frame #32: 0x00007fff20530621 libdyld.dylib`start + 1
     frame #33: 0x00007fff20530621 libdyld.dylib`start + 1
 ```
+
+Note that in order to produce a stack trace like the above, the Ruby binary must include the debugging symbols (otherwise, the `ruby` frames will not be displayed); this is accomplished by specifying the `--ggdb3` compiler flag:
+
+- if compiling Ruby from source, use `./configure optflags=-gddb3` in the build process
+- if using a version manager, refer to the help; example for RVM: `optflags="-ggdb3" rvm install 3.0.2 --disable-binary`

data/Rakefile

@@ -10,8 +10,9 @@ INSNS_DIR    = "lib/fisk/instructions"
 
 file XML_FILE do
   Dir.mkdir 'tmp' unless File.directory?("tmp")
-  cd "tmp"
-  sh "git clone https://github.com/Maratyszcza/Opcodes.git"
+  cd "tmp" do
+    sh "git clone https://github.com/Maratyszcza/Opcodes.git"
+  end
 end
 
 file INSNS_DIR do
@@ -33,3 +34,4 @@ Rake::TestTask.new do |t|
 end
 
 task :test => LIB_FILE
+task :default => :test

data/bin/build-machine.rb

@@ -2,13 +2,27 @@ require "nokogiri"
 require "stringio"
 require "erb"
 
+def modrm_method operands
+  op_types = operands.map { |x| x["type"] }
+  case op_types
+  in [/^m\d*$/, /^r\d+$/]
+    "add_modrm_mem_reg"
+  in [/^r\d+$/, /^m\d*$/]
+    "add_modrm_reg_mem"
+  in [/^r\d+$/, /^r\d+$/]
+    "add_modrm_reg_reg"
+  else
+    "add_modrm"
+  end
+end
+
 def parse_rex v
   if v
     case v
     when /^(\d+)$/
       v
     when /^#(\d+)$/
-      "(operands[#{$1.to_i}].value >> 3)"
+      "operands[#{$1.to_i}].rex_value"
     else
       raise NotImplementedError, v
     end
@@ -23,7 +37,7 @@ def parse_operand_value v
     when /^(\d+)$/
       v
     when /^#(\d+)$/
-      "operands[#{$1.to_i}].value"
+      "operands[#{$1.to_i}].op_value"
     else
       raise NotImplementedError, v
     end

data/bin/instruction.rb.erb

@@ -3,89 +3,63 @@
 class Fisk
   module Instructions
 <%- name = instruction["name"] -%>
-    # Instruction <%= name %>
-    forms = []
+<%- summary = instruction["summary"] -%>
+    # Instruction <%= name %>: <%= summary %>
+    <%= name %> = Instruction.new(<%= name.dump %>, [
   <%-
     instruction.children.each do |form|
       next unless form.name == "InstructionForm"
   -%>
-    operands = []
-    encodings = []
     <%-
       operands = form.children.reject(&:text?).find_all { |c| c.name == "Operand" }
+      encodings = form.children.reject(&:text?).find_all { |c| c.name == "Encoding" }
     -%>
     # <%= form["gas-name"] %>: <%= operands.map { |op| op["type"] }.join(", ") %>
-    <%-
-      form.children.each do |child|
-        next if child.text?
-        case child.name
-    -%>
-      <%- when "Operand" -%>
-    operands << OPERAND_TYPES[<%= unique_operands.index(%w{ type input output }.map { |n| child[n] }) %>]
-      <%-
-        when "Encoding"
-          size = 0
-      -%>
-    encodings << Class.new(Fisk::Encoding) {
-      def encode buffer, operands
+      Form.new([
+      <%- operands.each do |child| -%>
+        OPERAND_TYPES[<%= unique_operands.index(%w{ type input output }.map { |n| child[n] }) %>],
+      <%- end -%>
+      ].freeze, [
+      <%- encodings.each do |child| -%>
+        Class.new(Fisk::Encoding) {
+          def encode buffer, operands
       <%-
         child.children.each do |encoding|
           next if encoding.text?
           case encoding.name
           when "Opcode"
-            size += 1
       -%>
-        add_opcode buffer, 0x<%= sprintf("%02X", encoding["byte"].to_i(16)) %>, <%= parse_operand_value encoding["addend"] %>
-        <%-
-            when "Immediate"
-              size += encoding["size"].to_i
-        -%>
-        add_immediate buffer, <%= parse_operand_value encoding["value"] %>, <%= encoding["size"].to_i %>
-        <%-
-            when "CodeOffset"
-              size += encoding["size"].to_i
-        -%>
-        add_code_offset buffer, <%= parse_operand_value encoding["value"] %>, <%= encoding["size"].to_i %>
-        <%-
-            when "REX"
-              size += 1 unless encoding["mandatory"] == "false"
-        -%>
-        add_rex(buffer, operands,
+            add_opcode(buffer, 0x<%= sprintf("%02X", encoding["byte"].to_i(16)) %>, <%= parse_operand_value encoding["addend"] %>) +
+        <%- when "Immediate" -%>
+            add_immediate(buffer, <%= parse_operand_value encoding["value"] %>, <%= encoding["size"].to_i %>) +
+        <%- when "CodeOffset" -%>
+            add_code_offset(buffer, <%= parse_operand_value encoding["value"] %>, <%= encoding["size"].to_i %>) +
+        <%- when "REX" -%>
+            add_rex(buffer, operands,
               <%= !(encoding["mandatory"] == "false") %>,
               <%= parse_rex encoding["W"] %>,
               <%= parse_rex encoding["R"] %>,
               <%= parse_rex encoding["X"] %>,
-              <%= parse_rex encoding["B"] %>)
-        <%-
-            when "ModRM"
-              size += 1
-        -%>
-        add_modrm(buffer, operands,
+              <%= parse_rex encoding["B"] %>) +
+        <%- when "ModRM" -%>
+            <%= modrm_method(operands) %>(buffer,
               <%= encoding["mode"].to_i(2) %>,
               <%= parse_operand_value encoding["reg"] %>,
-              <%= parse_operand_value encoding["rm"] %>)
+              <%= parse_operand_value encoding["rm"] %>, operands) +
         <%- when "Prefix" -%>
-        add_prefix buffer, operands, 0x<%= sprintf("%02X", encoding["byte"].to_i(16)) %>, <%= encoding["mandatory"] %>
+            add_prefix(buffer, operands, 0x<%= sprintf("%02X", encoding["byte"].to_i(16)) %>, <%= encoding["mandatory"] %>) +
         <%- when "DataOffset" -%>
-        add_data_offset buffer, <%= parse_operand_value encoding["value"] %>, <%= encoding["size"].to_i %>
+            add_data_offset(buffer, <%= parse_operand_value encoding["value"] %>, <%= encoding["size"].to_i %>) +
         <%- else -%>
-        add_<%= encoding.name %> buffer, operands
+            add_<%= encoding.name %>(buffer, operands)
         <%- end -%>
       <%- end -%>
-      end
-
-      def bytesize; <%= size %>; end
-    }.new
-      <%- when "ISA" -%>
-      <%- when "ImplicitOperand" -%>
-      <%-
-          else
-            raise NotImplementedError, child.name
-      -%>
+            0
+          end
+        }.new.freeze,
       <%- end -%>
-    <%- end -%>
-    forms << Form.new(operands, encodings)
+      ].freeze).freeze,
   <%- end -%>
-    <%= name %> = Instruction.new(<%= name.dump %>, forms)
+    ].freeze).freeze
   end
 end

data/bin/instructions.rb.erb

@@ -6,14 +6,44 @@ require "fisk/encoding"
 
 class Fisk
   module Instructions
-    Operand     = Struct.new(:type, :input, :output)
+    Operand     = Struct.new(:type, :input, :output) do
+      def immediate?; false; end
+    end
+
+    class Immediate < Operand
+      attr_reader :size
+
+      def initialize type, input, output, size
+        super(type, input, output)
+        @size = size
+      end
+
+      def immediate?; true; end
+    end
+
     Form        = Struct.new(:operands, :encodings)
-    Instruction = Struct.new(:name, :forms)
+    Instruction = Struct.new(:name, :forms) do
+      def check_performance(operands)
+        case name
+        when Instructions::MOV.name
+          if operands[0].register? && operands[1].register? && operands[0].name == operands[1].name
+            return "MOV with same register (#{operands[0].name})"
+          end
+        end
+      end
+    end
 
     OPERAND_TYPES = [
-<%- unique_operands.each do |u_op| -%>
+<%-
+  unique_operands.each do |u_op|
+    if u_op[0] =~ /^imm(\d+)$/
+      size = $1.to_i
+-%>
+      Immediate.new(<%= u_op[0].dump %>, <%= u_op[1] || "nil" %>, <%= u_op[2] || "nil" %>, <%= size %>),
+<%- else -%>
       Operand.new(<%= u_op[0].dump %>, <%= u_op[1] || "nil" %>, <%= u_op[2] || "nil" %>),
-<% end %>
+<%- end -%>
+<%- end -%>
     ]
 
     def self.instruction_names

data/fisk.gemspec

@@ -1,6 +1,10 @@
+$: << File.expand_path("lib")
+
+require "fisk/version"
+
 Gem::Specification.new do |s|
   s.name        = "fisk"
-  s.version     = "2.0.0"
+  s.version     = Fisk::VERSION
   s.summary     = "Write assembly in Ruby!"
   s.description = "Tired of writing Ruby in Ruby? Now you can write assembly in Ruby!"
   s.authors     = ["Aaron Patterson"]
@@ -9,4 +13,9 @@ Gem::Specification.new do |s|
   s.test_files  = s.files.grep(%r{^test/})
   s.homepage    = "https://github.com/tenderlove/fisk"
   s.license     = "Apache-2.0"
+
+  s.add_development_dependency 'minitest', '~> 5.14'
+  s.add_development_dependency 'crabstone', '~> 4.0'
+  s.add_development_dependency 'rake', '~> 13.0'
+  s.add_development_dependency 'nokogiri', '~> 1.11'
 end

data/lib/fisk/basic_block.rb

@@ -0,0 +1,162 @@
+class Fisk
+  class CFG < Struct.new(:variables, :blocks)
+    class BasicBlock < Struct.new(:name, :insns, :fall, :jump, :def, :use, :in, :out, :pred, :start_point, :end_point)
+      def jump?
+        insns.last.jump?
+      end
+
+      def jump_name
+        insns.last.target
+      end
+
+      def succ
+        list = []
+        list << fall if fall
+        list << jump if jump
+        list
+      end
+
+      def live; self.def | self.use | self.out; end
+    end
+
+    def self.build instructions
+      blocks = []
+      bb_index = 0
+
+      chunks = instructions.chunk { |insn|
+        if insn.label?
+          # start a new block here
+          bb_index += 1
+        end
+
+        ret_idx = bb_index
+
+        if insn.jump?
+          # end the block here
+          bb_index += 1
+        end
+
+        ret_idx
+      }.to_a
+
+      chunk_name = ->(chunk_id, insns) {
+        if insns.first.label?
+          insns.first.name
+        else
+          :"bb_#{chunk_id}"
+        end
+      }
+
+      jump_targets = {}
+      wants_target = []
+      insn_idx = 0
+
+      chunks.each_with_index do |(chunk_id, insns), idx|
+        prev_block = blocks.last
+
+        name = chunk_name.(chunk_id, insns)
+
+        next_chunk_id, next_insns = chunks[idx + 1]
+
+        outgoing = []
+
+        if next_chunk_id
+          outgoing << chunk_name.(next_chunk_id, next_insns)
+        end
+
+        bb = BasicBlock.new(name, insns)
+        bb.pred = []
+        bb.start_point = insn_idx
+        insn_idx += insns.length
+        bb.end_point = insn_idx - 1
+
+        if bb.jump?
+          target_name = bb.jump_name
+
+          # if we need to jump backwards, the label should already be filled in.
+          # Otherwise we need to find the block later (we're jumping forward)
+          if jump_targets[target_name]
+            jt = jump_targets[target_name]
+            bb.jump = jt
+            jt.pred << bb
+          else
+            wants_target << bb
+          end
+        end
+
+        jump_targets[bb.name] = bb
+
+        if prev_block
+          prev_block.fall = bb
+          bb.pred << prev_block
+        end
+
+        blocks << bb
+      end
+
+      # fill in any forward jumps
+      wants_target.each { |bb|
+        jt = jump_targets[bb.jump_name]
+        bb.jump = jt
+        jt.pred << bb
+      }
+
+      variables = live_variables blocks
+
+      CFG.new variables, blocks
+    end
+
+    def self.live_variables blocks
+      seen = Set.new
+
+      blocks.each do |block|
+        used = Set.new
+        defd = Set.new
+
+        block.insns.each do |insn|
+          if insn.has_temp_registers?
+            insn.temp_registers.each do |reg|
+              defd << reg unless seen.include?(reg)
+              used << reg unless defd.include?(reg)
+              seen << reg
+            end
+          end
+        end
+
+        block.use = used.to_a
+        block.def = defd.to_a
+        block.in  = []
+        block.out = []
+      end
+
+      worklist = blocks.dup
+      while block = worklist.pop
+        block.out = block.succ.inject([]) { |out_registers, succ_block|
+          out_registers | succ_block.in
+        }
+        ins = block.in | block.use | (block.out - block.def)
+
+        if ins == block.in
+          # Done with this block!
+          block.live.each do |live_register|
+            live_register.end_point ||= block.end_point
+
+            if live_register.start_point > block.start_point
+              live_register.start_point = block.start_point
+            end
+
+            if live_register.end_point < block.end_point
+              live_register.end_point = block.end_point
+            end
+          end
+        else
+          block.in = ins
+          worklist.concat(block.pred)
+        end
+      end
+
+      seen
+    end
+  end
+
+end