Slot coalescing + SROA passes + other improvements

.clang-format

@@ -1,149 +1,40 @@
----
-Language:        Cpp
-# BasedOnStyle:  LLVM
-AccessModifierOffset: -2
-AlignAfterOpenBracket: Align
-AlignConsecutiveMacros: false
-AlignConsecutiveAssignments: false
-AlignConsecutiveBitFields: false
-AlignConsecutiveDeclarations: false
-AlignEscapedNewlines: Right
-AlignOperands:   Align
-AlignTrailingComments: true
-AllowAllArgumentsOnNextLine: true
-AllowAllConstructorInitializersOnNextLine: true
-AllowAllParametersOfDeclarationOnNextLine: true
-AllowShortEnumsOnASingleLine: true
-AllowShortBlocksOnASingleLine: Never
-AllowShortCaseLabelsOnASingleLine: false
-AllowShortFunctionsOnASingleLine: All
-AllowShortLambdasOnASingleLine: All
-AllowShortIfStatementsOnASingleLine: Never
-AllowShortLoopsOnASingleLine: false
-AlwaysBreakAfterDefinitionReturnType: None
-AlwaysBreakAfterReturnType: None
-AlwaysBreakBeforeMultilineStrings: false
-AlwaysBreakTemplateDeclarations: MultiLine
-BinPackArguments: true
-BinPackParameters: true
+BasedOnStyle: LLVM
+IndentWidth: 2
+ContinuationIndentWidth: 2
+UseTab: Never
+TabWidth: 2
+
+# Function definitions like:
+# static int
+# name(...)
+# {
 BraceWrapping:
-  AfterCaseLabel:  false
-  AfterClass:      false
-  AfterControlStatement: Never
-  AfterEnum:       false
-  AfterFunction:   false
-  AfterNamespace:  false
-  AfterObjCDeclaration: false
-  AfterStruct:     false
-  AfterUnion:      false
-  AfterExternBlock: false
-  BeforeCatch:     false
-  BeforeElse:      false
-  BeforeLambdaBody: false
-  BeforeWhile:     false
-  IndentBraces:    false
-  SplitEmptyFunction: true
-  SplitEmptyRecord: true
-  SplitEmptyNamespace: true
-BreakBeforeBinaryOperators: None
+  AfterFunction: true
+  AfterStruct: true
+  AfterClass: true
+  AfterNamespace: true
+  AfterExternBlock: true
+  BeforeCatch: false
+  BeforeElse: false
+
+# K&R control-flow braces (same-line)
 BreakBeforeBraces: Attach
-BreakBeforeInheritanceComma: false
-BreakInheritanceList: BeforeColon
-BreakBeforeTernaryOperators: true
-BreakConstructorInitializersBeforeComma: false
-BreakConstructorInitializers: BeforeColon
-BreakAfterJavaFieldAnnotations: false
-BreakStringLiterals: true
-ColumnLimit:     80
-CommentPragmas:  '^ IWYU pragma:'
-CompactNamespaces: false
-ConstructorInitializerAllOnOneLineOrOnePerLine: false
-ConstructorInitializerIndentWidth: 4
-ContinuationIndentWidth: 4
-Cpp11BracedListStyle: true
-DeriveLineEnding: true
-DerivePointerAlignment: false
-DisableFormat:   false
-ExperimentalAutoDetectBinPacking: false
-FixNamespaceComments: true
-ForEachMacros:
-  - foreach
-  - Q_FOREACH
-  - BOOST_FOREACH
-IncludeBlocks:   Preserve
-IncludeCategories:
-  - Regex:           '^"(llvm|llvm-c|clang|clang-c)/'
-    Priority:        2
-    SortPriority:    0
-  - Regex:           '^(<|"(gtest|gmock|isl|json)/)'
-    Priority:        3
-    SortPriority:    0
-  - Regex:           '.*'
-    Priority:        1
-    SortPriority:    0
-IncludeIsMainRegex: '(Test)?$'
-IncludeIsMainSourceRegex: ''
-IndentCaseLabels: false
-IndentCaseBlocks: false
-IndentGotoLabels: true
-IndentPPDirectives: None
-IndentExternBlock: AfterExternBlock
-IndentWidth:     2
-IndentWrappedFunctionNames: false
-InsertTrailingCommas: None
-JavaScriptQuotes: Leave
-JavaScriptWrapImports: true
-KeepEmptyLinesAtTheStartOfBlocks: true
-MacroBlockBegin: ''
-MacroBlockEnd:   ''
-MaxEmptyLinesToKeep: 1
-NamespaceIndentation: None
-ObjCBinPackProtocolList: Auto
-ObjCBlockIndentWidth: 2
-ObjCBreakBeforeNestedBlockParam: true
-ObjCSpaceAfterProperty: false
-ObjCSpaceBeforeProtocolList: true
-PenaltyBreakAssignment: 2
-PenaltyBreakBeforeFirstCallParameter: 19
-PenaltyBreakComment: 300
-PenaltyBreakFirstLessLess: 120
-PenaltyBreakString: 1000
-PenaltyBreakTemplateDeclaration: 10
-PenaltyExcessCharacter: 1000000
-PenaltyReturnTypeOnItsOwnLine: 60
-PointerAlignment: Right
-ReflowComments:  true
-SortIncludes:    true
-SortUsingDeclarations: true
+
+# Spacing rules
+SpaceBeforeParens: ControlStatements
 SpaceAfterCStyleCast: false
-SpaceAfterLogicalNot: false
-SpaceAfterTemplateKeyword: true
-SpaceBeforeAssignmentOperators: true
 SpaceBeforeCpp11BracedList: false
-SpaceBeforeCtorInitializerColon: true
-SpaceBeforeInheritanceColon: true
-SpaceBeforeParens: ControlStatements
-SpaceBeforeRangeBasedForLoopColon: true
-SpaceInEmptyBlock: false
-SpaceInEmptyParentheses: false
-SpacesBeforeTrailingComments: 1
-SpacesInAngles:  false
-SpacesInConditionalStatement: false
-SpacesInContainerLiterals: true
-SpacesInCStyleCastParentheses: false
-SpacesInParentheses: false
-SpacesInSquareBrackets: false
-SpaceBeforeSquareBrackets: false
-Standard:        Latest
-StatementMacros:
-  - Q_UNUSED
-  - QT_REQUIRE_VERSION
-TabWidth:        8
-UseCRLF:         false
-UseTab:          Never
-WhitespaceSensitiveMacros:
-  - STRINGIZE
-  - PP_STRINGIZE
-  - BOOST_PP_STRINGIZE
-...
+SpaceBeforeAssignmentOperators: true
+
+# Pointer alignment
+PointerAlignment: Right
 
+# Do not wrap long lines unless needed
+ColumnLimit: 80
+
+# Do not cram short functions onto one line
+AllowShortFunctionsOnASingleLine: None
+
+# Keep return type on its own line if desired
+AlwaysBreakAfterDefinitionReturnType: true
+AlwaysBreakAfterReturnType: None

Makefile

@@ -1,6 +1,7 @@
 SRC := ./src/
 
-.PHONY: build clean install uninstall test doc deps indent status-clean check-style
+.PHONY: build	clean	install uninstall test doc deps
+.PHONY: ocaml-indent clang-indent indent status-clean check-style
 
 all: install
 
@@ -25,9 +26,14 @@ doc:
 deps:
 	$(MAKE) deps -C $(SRC)
 
-indent:
+ocaml-indent:
 	sh tools/ocp-indent-all.sh
 
+clang-indent:
+	sh tools/clang-format-all.sh
+
+indent: ocaml-indent clang-indent
+
 status-clean:
 	git diff --quiet --exit-code
 

src/Makefile

@@ -1,5 +1,10 @@
 .PHONY: clean install uninstall test doc deps bap-deps
 
+PROFILE = release
+ifeq ($(DEBUG),1)
+	PROFILE = dev
+endif
+
 BAP_DEPS = \
 	bitvec \
 	bitvec-binprot \
@@ -12,7 +17,7 @@ BAP_DEPS = \
 BAP_REPO := git+https://github.com/BinaryAnalysisPlatform/opam-repository\#testing
 
 all:
-	dune build --profile=release
+	dune build --profile=$(PROFILE)
 
 clean:
 	dune clean
@@ -24,7 +29,7 @@ uninstall:
 	dune uninstall
 
 test:
-	dune test --profile=release
+	dune test --profile=$(PROFILE)
 
 doc:
 	dune build @doc

src/bin/cli.ml

@@ -1,23 +1,6 @@
 open Cmdliner
 
-(* Why is this here, when we have `Cgen.Target.find`?
-
-   The targets should be getting declared at the toplevel in their
-   respective modules, but the evaluation of these modules is not
-   guaranteed to happen by the time we start parsing command-line
-   arguments.
-
-   It just so happens that, for the front-end executable, we should
-   know all of the out-of-the-box targets provided by the library,
-   so this feels like a tolerable compromise in the design of extending
-   the supported targets.
-*)
-let targets =
-  Core.Map.of_alist_exn (module Core.String) @@
-  Core.List.map ~f:(fun t -> Cgen.Target.name t, t) @@
-  Cgen.Machine.[
-    X86.Amd64_sysv.target;
-  ]
+let () = Cgen.Machine.force_initialization ()
 
 let bail () = exit Cmd.Exit.ok
 
@@ -100,7 +83,8 @@ let dump_no_comment =
 let man_targets =
   `S "TARGET" ::
   `Pre "Supported target platforms" :: begin
-    Core.Map.data targets |>
+    Cgen.Target.enum_targets () |>
+    Core.Sequence.to_list |>
     Core.List.map ~f:(fun t ->
         `P (Format.asprintf "%a" Cgen.Target.pp t))
   end
@@ -134,7 +118,14 @@ type t = {
   target : Cgen.Target.t;
 }
 
-let go f file output dump nc target =
+let log_env = Cmd.Env.info "CGEN_LOG"
+
+let setup_log level =
+  Logs.set_level level;
+  Logs.set_reporter @@ Logs_fmt.reporter ()
+
+let go f file output dump nc target log_level =
+  setup_log log_level;
   let file = match file with
     | "" -> Istdin
     | _ -> Ifile file in
@@ -144,12 +135,20 @@ let go f file output dump nc target =
   let dump = match dump_of_string_opt dump with
     | None -> fatal "invalid dump option: %s\n%!" dump ()
     | Some d -> d in
-  let target = match Core.Map.find targets target with
+  let target = match Cgen.Target.find target with
     | None -> fatal "invalid target: %s\n%!" target ()
     | Some t -> t in
   f {file; output; dump; nc; target}
 
-let t f = Term.(const (go f) $ file $ output $ dump $ dump_no_comment $ target)
+let t f =
+  let open Term in
+  const (go f) $
+  file $
+  output $
+  dump $
+  dump_no_comment $
+  target $
+  Logs_cli.level ~env:log_env ()
 
 let man = List.concat [
     man_dump;

src/bin/dune

@@ -2,7 +2,7 @@
  (name cgen_main)
  (public_name cgen)
  (package cgen)
- (libraries cgen cmdliner)
+ (libraries cgen cmdliner logs.cli logs.fmt)
  (modules cgen_main cli)
  (flags -w -58)
  (ocamlopt_flags -O2)

src/cgen.opam

@@ -17,7 +17,9 @@ depends: [
   "core" {>= "v0.15"}
   "core_kernel" {>= "v0.15"}
   "dune" {>= "3.15"}
+  "fmt"
   "graphlib"
+  "logs"
   "menhir"
   "monads"
   "ocamldiff"

src/dune-project

@@ -23,7 +23,9 @@
     (core (>= v0.15))
     (core_kernel (>= v0.15))
     dune
+    fmt
     graphlib
+    logs
     menhir
     monads
     ocamldiff