1 Introduction
New: CIL now has a Source Forge page:
http://sourceforge.net/projects/cil.
CIL (C Intermediate Language) is a high-level representation
along with a set of tools that permit easy analysis and source-to-source
transformation of C programs.
CIL is both lower-level than abstract-syntax trees, by clarifying ambiguous
constructs and removing redundant ones, and also higher-level than typical
intermediate languages designed for compilation, by maintaining types and a
close relationship with the source program. The main advantage of CIL is that
it compiles all valid C programs into a few core constructs with a very clean
semantics. Also CIL has a syntax-directed type system that makes it easy to
analyze and manipulate C programs. Furthermore, the CIL front-end is able to
process not only ANSI-C programs but also those using Microsoft C or GNU C
extensions. If you do not use CIL and want instead to use just a C parser and
analyze programs expressed as abstract-syntax trees then your analysis will
have to handle a lot of ugly corners of the language (let alone the fact that
parsing C itself is not a trivial task). See Section 16 for some
examples of such extreme programs that CIL simplifies for you.
In essence, CIL is a highly-structured, “clean” subset of C. CIL features a
reduced number of syntactic and conceptual forms. For example, all looping
constructs are reduced to a single form, all function bodies are given
explicit return statements, syntactic sugar like "->" is
eliminated and function arguments with array types become pointers. (For an
extensive list of how CIL simplifies C programs, see Section 4.)
This reduces the number of cases that must be considered when manipulating a C
program. CIL also separates type declarations from code and flattens scopes
within function bodies. This structures the program in a manner more amenable
to rapid analysis and transformation. CIL computes the types of all program
expressions, and makes all type promotions and casts explicit. CIL supports
all GCC and MSVC extensions except for nested functions and complex numbers.
Finally, CIL organizes C's imperative features into expressions, instructions
and statements based on the presence and absence of side-effects and
control-flow. Every statement can be annotated with successor and predecessor
information. Thus CIL provides an integrated program representation that can
be used with routines that require an AST (e.g. type-based analyses and
pretty-printers), as well as with routines that require a CFG (e.g., dataflow
analyses). CIL also supports even lower-level representations (e.g.,
three-address code), see Section 8.
CIL comes accompanied by a number of Perl scripts that perform generally
useful operations on code:
-
A driver which behaves as either the gcc or
Microsoft VC compiler and can invoke the preprocessor followed by the CIL
application. The advantage of this script is that you can easily use CIL and
the analyses written for CIL with existing make files.
- A whole-program merger that you can use as a
replacement for your compiler and it learns all the files you compile when you
make a project and merges all of the preprocessed source files into a single
one. This makes it easy to do whole-program analysis.
- A patcher makes it easy to create modified
copies of the system include files. The CIL driver can then be told to use
these patched copies instead of the standard ones.
CIL has been tested very extensively. It is able to process the SPECINT95
benchmarks, the Linux kernel, GIMP and other open-source projects. All of
these programs are compiled to the simple CIL and then passed to gcc and
they still run! We consider the compilation of Linux a major feat especially
since Linux contains many of the ugly GCC extensions (see Section 16.2).
This adds to about 1,000,000 lines of code that we tested it on. It is also
able to process the few Microsoft NT device drivers that we have had access
to. CIL was tested against GCC's c-torture testsuite and (except for the tests
involving complex numbers and inner functions, which CIL does not currently
implement) CIL passes most of the tests. Specifically CIL fails 23 tests out
of the 904 c-torture tests that it should pass. GCC itself fails 19 tests. A
total of 1400 regression test cases are run automatically on each change to
the CIL sources.
CIL is relatively independent on the underlying machine and compiler. When
you build it CIL will configure itself according to the underlying compiler.
However, CIL has only been tested on Intel x86 using the gcc compiler on Linux
and cygwin and using the MS Visual C compiler. (See below for specific
versions of these compilers that we have used CIL for.)
The largest application we have used CIL for is
CCured, a compiler that compiles C code into
type-safe code by analyzing your pointer usage and inserting runtime checks in
the places that cannot be guaranteed statically to be type safe.
You can also use CIL to “compile” code that uses GCC extensions (e.g. the
Linux kernel) into standard C code.
CIL also comes accompanies by a growing library of extensions (see
Section 8). You can use these for your projects or as examples of
using CIL.
PDF versions of this manual and the
CIL API are available. However, we recommend the
HTML versions because the postprocessed code examples are easier to
view.
If you use CIL in your project, we would appreciate letting us know. If you
want to cite CIL in your research writings, please refer to the paper “CIL:
Intermediate Language and Tools for Analysis and Transformation of C
Programs” by George C. Necula, Scott McPeak, S.P. Rahul and Westley Weimer,
in “Proceedings of Conference on Compilier Construction”, 2002.
2 Installation
You will need OCaml release 3.08 or higher to build CIL. CIL has been tested
on Linux and on Windows (where it can behave at either Microsoft Visual C or
gcc).
If you want to use CIL on Windows then you must get a complete installation
of cygwin and the source-code OCaml distribution and compile it yourself
using the cygwin tools (as opposed to getting the Win32 native-code version of
OCaml). If you have not done this before then take a look
here. (Don't need to worry about cvs and
ssh unless you will need to use the master CVS repository for CIL.)
-
Download the CIL distribution (latest version is
distrib/cil-1.3.5.tar.gz). See the Section 20 for recent changes to the CIL distribution.
- Unzip and untar the source distribution. This will create a directory
called cil whose structure is explained below.
tar xvfz cil-1.3.5.tar.gz
- Enter the cil directory and run the configure script and then
GNU make to build the distribution. If you are on Windows, at least the
configure step must be run from within bash.
cd cil
./configure
make
make quicktest
- You should now find cilly.asm.exe in a
subdirectory of obj. The name of the subdirectory is either x86_WIN32
if you are using cygwin on Windows or x86_LINUX if you are using
Linux (although you should be using instead the Perl wrapper bin/cilly).
Note that we do not have an install make target and you should use Cil
from the development directory.
- If you decide to use CIL, please
send us a note. This will help recharge
our batteries after more than a year of development. And of course, do send us
your bug reports as well.
The configure script tries to find appropriate defaults for your system.
You can control its actions by passing the following arguments:
-
CC=foo Specifies the path for the gcc executable. By default
whichever version is in the PATH is used. If CC specifies the Microsoft
cl compiler, then that compiler will be set as the default one. Otherwise,
the gcc compiler will be the default.
CIL requires an underlying C compiler and preprocessor. CIL depends on the
underlying compiler and machine for the sizes and alignment of types.The
installation procedure for CIL queries the underlying compiler for
architecture and compiler dependent configuration parameters, such as the size
of a pointer or the particular alignment rules for structure fields. (This
means, of course, that you should re-run ./configure when you move CIL to
another machine.)
We have tested CIL on the following compilers:
-
On Windows, cl compiler version 12.00.8168 (MSVC 6),
13.00.9466 (MSVC .Net), and 13.10.3077 (MSVC .Net 2003). Run cl
with no arguments to get the compiler version.
- On Windows, using cygwin and gcc version 2.95.3, 3.0,
3.2, 3.3, and 3.4.
- On Linux, using gcc version 2.95.3, 3.0, 3.2, 3.3, and 4.0.
Others have successfully used CIL with Mac OS X (on both PowerPC and
x86), Solaris, and *BSD. If you make any changes to the build
system in order to run CIL on your platform, please send us a patch.
3 Distribution Contents
The file distrib/cil-1.3.5.tar.gz
contains the complete source CIL distribution,
consisting of the following files:
Filename |
Description |
Makefile.in |
configure source for the
Makefile that builds CIL |
configure |
The configure script |
configure.in |
The autoconf source for configure |
config.guess, config.sub, install-sh |
stuff required by
configure |
|
doc/ |
HTML documentation of the CIL API |
obj/ |
Directory that will contain the compiled
CIL modules and executables |
bin/cilly.in |
The configure source for a Perl script
that can be invoked with the
same arguments as either gcc or
Microsoft Visual C and will convert the
program to CIL, perform some simple
transformations, emit it and compile it as
usual. |
lib/CompilerStub.pm |
A Perl class that can be used to write code
that impersonates a compiler. cilly
uses it. |
lib/Merger.pm |
A subclass of CompilerStub.pm that can
be used to merge source files into a single
source file.cilly
uses it. |
bin/patcher.in |
A Perl script that applies specified patches
to standard include files. |
|
src/check.ml,mli |
Checks the well-formedness of a CIL file |
src/cil.ml,mli |
Definition of CIL abstract syntax and
utilities for manipulating it |
src/clist.ml,mli |
Utilities for efficiently managing lists
that need to be concatenated often |
src/errormsg.ml,mli |
Utilities for error reporting |
src/ext/heapify.ml |
A CIL transformation that moves array local
variables from the stack to the heap |
src/ext/logcalls.ml,mli |
A CIL transformation that logs every
function call |
src/ext/sfi.ml |
A CIL transformation that can log every
memory read and write |
src/frontc/clexer.mll |
The lexer |
src/frontc/cparser.mly |
The parser |
src/frontc/cabs.ml |
The abstract syntax |
src/frontc/cprint.ml |
The pretty printer for CABS |
src/frontc/cabs2cil.ml |
The elaborator to CIL |
src/main.ml |
The cilly application |
src/pretty.ml,mli |
Utilities for pretty printing |
src/rmtmps.ml,mli |
A CIL tranformation that removes unused
types, variables and inlined functions |
src/stats.ml,mli |
Utilities for maintaining timing statistics |
src/testcil.ml |
A random test of CIL (against the resident
C compiler) |
src/trace.ml,mli |
Utilities useful for printing debugging
information |
|
ocamlutil/ |
Miscellaneous libraries that are not
specific to CIL. |
ocamlutil/Makefile.ocaml |
A file that is included by Makefile |
ocamlutil/Makefile.ocaml.build |
A file that is included by Makefile |
ocamlutil/perfcount.c |
C code that links with src/stats.ml
and reads Intel performance
counters. |
|
obj/@ARCHOS@/feature_config.ml |
File generated by the Makefile
describing which extra “features”
to compile. See Section 5 |
obj/@ARCHOS@/machdep.ml |
File generated by the Makefile containing
information about your architecture,
such as the size of a pointer |
src/machdep.c |
C program that generates
machdep.ml files |
4 Compiling C to CIL
In this section we try to describe a few of the many transformations that are
applied to a C program to convert it to CIL. The module that implements this
conversion is about 5000 lines of OCaml code. In contrast a simple program
transformation that instruments all functions to keep a shadow stack of the
true return address (thus preventing stack smashing) is only 70 lines of code.
This example shows that the analysis is so much simpler because it has to
handle only a few simple C constructs and also because it can leverage on CIL
infrastructure such as visitors and pretty-printers.
In no particular order these are a few of the most significant ways in which
C programs are compiled into CIL:
-
CIL will eliminate all declarations for unused entities. This means that
just because your hello world program includes stdio.h it does not mean
that your analysis has to handle all the ugly stuff from stdio.h.
- Type specifiers are interpreted and normalized:
int long signed x;
signed long extern x;
long static int long y;
// Some code that uses these declaration, so that CIL does not remove them
int main() { return x + y; }
See the CIL output for this
code fragment
- Anonymous structure and union declarations are given a name.
struct { int x; } s;
See the CIL output for this
code fragment
- Nested structure tag definitions are pulled apart. This means that all
structure tag definitions can be found by a simple scan of the globals.
struct foo {
struct bar {
union baz {
int x1;
double x2;
} u1;
int y;
} s1;
int z;
} f;
See the CIL output for this
code fragment
- All structure, union, enumeration definitions and the type definitions
from inners scopes are moved to global scope (with appropriate renaming). This
facilitates moving around of the references to these entities.
int main() {
struct foo {
int x; } foo;
{
struct foo {
double d;
};
return foo.x;
}
}
See the CIL output for this
code fragment
- Prototypes are added for those functions that are called before being
defined. Furthermore, if a prototype exists but does not specify the type of
parameters that is fixed. But CIL will not be able to add prototypes for those
functions that are neither declared nor defined (but are used!).
int f(); // Prototype without arguments
int f(double x) {
return g(x);
}
int g(double x) {
return x;
}
See the CIL output for this
code fragment
- Array lengths are computed based on the initializers or by constant
folding.
int a1[] = {1,2,3};
int a2[sizeof(int) >= 4 ? 8 : 16];
See the CIL output for this
code fragment
- Enumeration tags are computed using constant folding:
int main() {
enum {
FIVE = 5,
SIX, SEVEN,
FOUR = FIVE - 1,
EIGHT = sizeof(double)
} x = FIVE;
return x;
}
See the CIL output for this
code fragment
- Initializers are normalized to include specific initialization for the
missing elements:
int a1[5] = {1,2,3};
struct foo { int x, y; } s1 = { 4 };
See the CIL output for this
code fragment
- Initializer designators are interpreted and eliminated. Subobjects are
properly marked with braces. CIL implements
the whole ISO C99 specification for initializer (neither GCC nor MSVC do) and
a few GCC extensions.
struct foo {
int x, y;
int a[5];
struct inner {
int z;
} inner;
} s = { 0, .inner.z = 3, .a[1 ... 2] = 5, 4, y : 8 };
See the CIL output for this
code fragment
- String initializers for arrays of characters are processed
char foo[] = "foo plus bar";
See the CIL output for this
code fragment
- String constants are concatenated
char *foo = "foo " " plus " " bar ";
See the CIL output for this
code fragment
- Initializers for local variables are turned into assignments. This is in
order to separate completely the declarative part of a function body from the
statements. This has the unfortunate effect that we have to drop the const
qualifier from local variables !
int x = 5;
struct foo { int f1, f2; } a [] = {1, 2, 3, 4, 5 };
See the CIL output for this
code fragment
- Local variables in inner scopes are pulled to function scope (with
appropriate renaming). Local scopes thus disappear. This makes it easy to find
and operate on all local variables in a function.
int x = 5;
int main() {
int x = 6;
{
int x = 7;
return x;
}
return x;
}
See the CIL output for this
code fragment
- Global declarations in local scopes are moved to global scope:
int x = 5;
int main() {
int x = 6;
{
static int x = 7;
return x;
}
return x;
}
See the CIL output for this
code fragment
- Return statements are added for functions that are missing them. If the
return type is not a base type then a return without a value is added.
The guaranteed presence of return statements makes it easy to implement a
transformation that inserts some code to be executed immediately before
returning from a function.
int foo() {
int x = 5;
}
See the CIL output for this
code fragment
- One of the most significant transformations is that expressions that
contain side-effects are separated into statements.
int x, f(int);
return (x ++ + f(x));
See the CIL output for this
code fragment
Internally, the x ++ statement is turned into an assignment which the
pretty-printer prints like the original. CIL has only three forms of basic
statements: assignments, function calls and inline assembly.
- Shortcut evaluation of boolean expressions and the ?: operator are
compiled into explicit conditionals:
int x;
int y = x ? 2 : 4;
int z = x || y;
// Here we duplicate the return statement
if(x && y) { return 0; } else { return 1; }
// To avoid excessive duplication, CIL uses goto's for
// statement that have more than 5 instructions
if(x && y || z) { x ++; y ++; z ++; x ++; y ++; return z; }
See the CIL output for this
code fragment
- GCC's conditional expression with missing operands are also compiled
into conditionals:
int f();;
return f() ? : 4;
See the CIL output for this
code fragment
- All forms of loops (while, for and do) are compiled
internally as a single while(1) looping construct with explicit break
statement for termination. For simple while loops the pretty printer is
able to print back the original:
int x, y;
for(int i = 0; i<5; i++) {
if(i == 5) continue;
if(i == 4) break;
i += 2;
}
while(x < 5) {
if(x == 3) continue;
x ++;
}
See the CIL output for this
code fragment
- GCC's block expressions are compiled away. (That's right there is an
infinite loop in this code.)
int x = 5, y = x;
int z = ({ x++; L: y -= x; y;});
return ({ goto L; 0; });
See the CIL output for this
code fragment
- CIL contains support for both MSVC and GCC inline assembly (both in one
internal construct)
- CIL compiles away the GCC extension that allows many kinds of constructs
to be used as lvalues:
int x, y, z;
return &(x ? y : z) - & (x ++, x);
See the CIL output for this
code fragment
- All types are computed and explicit casts are inserted for all
promotions and conversions that a compiler must insert:
- CIL will turn old-style function definition (without prototype) into
new-style definitions. This will make the compiler less forgiving when
checking function calls, and will catch for example cases when a function is
called with too few arguments. This happens in old-style code for the purpose
of implementing variable argument functions.
- Since CIL sees the source after preprocessing the code after CIL does
not contain the comments and the preprocessing directives.
- CIL will remove from the source file those type declarations, local
variables and inline functions that are not used in the file. This means that
your analysis does not have to see all the ugly stuff that comes from the
header files:
#include <stdio.h>
typedef int unused_type;
static char unused_static (void) { return 0; }
int main() {
int unused_local;
printf("Hello world\n"); // Only printf will be kept from stdio.h
}
See the CIL output for this
code fragment
5 How to Use CIL
There are two predominant ways to use CIL to write a program analysis or
transformation. The first is to phrase your analysis as a module that is
called by our existing driver. The second is to use CIL as a stand-alone
library. We highly recommend that you use cilly, our driver.
5.1 Using cilly, the CIL driver
The most common way to use CIL is to write an Ocaml module containing your
analysis and transformation, which you then link into our boilerplate
driver application called cilly. cilly is a Perl script that
processes and mimics GCC and MSVC command-line arguments and then
calls cilly.byte.exe or cilly.asm.exe (CIL's Ocaml executable).
An example of such module is logwrites.ml, a transformation that is
distributed with CIL and whose purpose is to instrument code to print the
addresses of memory locations being written. (We plan to release a
C-language interface to CIL so that you can write your analyses in C
instead of Ocaml.) See Section 8 for a survey of other example
modules.
Assuming that you have written /home/necula/logwrites.ml,
here is how you use it:
- Modify logwrites.ml so that it includes a CIL “feature
descriptor” like this:
let feature : featureDescr =
{ fd_name = "logwrites";
fd_enabled = ref false;
fd_description = "generation of code to log memory writes";
fd_extraopt = [];
fd_doit =
(function (f: file) ->
let lwVisitor = new logWriteVisitor in
visitCilFileSameGlobals lwVisitor f)
}
The fd_name field names the feature and its associated
command-line arguments. The fd_enabled field is a bool ref.
“fd_doit” will be invoked if !fd_enabled is true after
argument parsing, so initialize the ref cell to true if you want
this feature to be enabled by default.
When the user passes the --dologwrites
command-line option to cilly, the variable associated with the
fd_enabled flag is set and the fd_doit function is called
on the Cil.file that represents the merger (see Section 13) of
all C files listed as arguments.
- Invoke configure with the arguments
./configure EXTRASRCDIRS=/home/necula EXTRAFEATURES=logwrites
This step works if each feature is packaged into its own ML file, and the
name of the entry point in the file is feature.
An alternative way to specify the new features is to change the build files
yourself, as explained below. You'll need to use this method if a single
feature is split across multiple files.
-
Put logwrites.ml in the src or src/ext directory. This
will make sure that make can find it. If you want to put it in some
other directory, modify Makefile.in and add to SOURCEDIRS your
directory. Alternately, you can create a symlink from src or
src/ext to your file.
- Modify the Makefile.in and add your module to the
CILLY_MODULES or
CILLY_LIBRARY_MODULES variables. The order of the modules matters. Add
your modules somewhere after cil and before main.
- If you have any helper files for your module, add those to
the makefile in the same way. e.g.:
CILLY_MODULES = $(CILLY_LIBRARY_MODULES) \
myutilities1 myutilities2 logwrites \
main
Again, order is important: myutilities2.ml will be able to refer
to Myutilities1 but not Logwrites. If you have any ocamllex or ocamlyacc
files, add them to both CILLY_MODULES and either MLLS or
MLYS.
- Modify main.ml so that your new feature descriptor appears in
the global list of CIL features.
let features : C.featureDescr list =
[ Logcalls.feature;
Oneret.feature;
Heapify.feature1;
Heapify.feature2;
makeCFGFeature;
Partial.feature;
Simplemem.feature;
Logwrites.feature; (* add this line to include the logwrites feature! *)
]
@ Feature_config.features
Features are processed in the order they appear on this list. Put
your feature last on the list if you plan to run any of CIL's
built-in features (such as makeCFGfeature) before your own.
Standard code in cilly takes care of adding command-line arguments,
printing the description, and calling your function automatically.
Note: do not worry about introducing new bugs into CIL by adding a single
line to the feature list.
- Now you can invoke the cilly application on a preprocessed file, or
instead use the cilly driver which provides a convenient compiler-like
interface to cilly. See Section 7 for details using cilly.
Remember to enable your analysis by passing the right argument (e.g.,
--dologwrites).
5.2 Using CIL as a library
CIL can also be built as a library that is called from your stand-alone
application. Add cil/src, cil/src/frontc, cil/obj/x86_LINUX
(or cil/obj/x86_WIN32) to your Ocaml project -I include paths.
Building CIL will also build the library cil/obj/*/cil.cma (or
cil/obj/*/cil.cmxa). You can then link your application against that
library.
You can call the Frontc.parse: string -> unit -> Cil.file function with
the name of a file containing the output of the C preprocessor.
The Mergecil.merge: Cil.file list -> string -> Cil.file function merges
multiple files. You can then invoke your analysis function on the resulting
Cil.file data structure. You might want to call
Rmtmps.removeUnusedTemps first to clean up the prototypes and variables
that are not used. Then you can call the function Cil.dumpFile:
cilPrinter -> out_channel -> Cil.file -> unit to print the file to a
given output channel. A good cilPrinter to use is
defaultCilPrinter.
Check out src/main.ml and bin/cilly for other good ideas
about high-level file processing. Again, we highly recommend that you just
our cilly driver so that you can avoid spending time re-inventing the
wheel to provide drop-in support for standard makefiles.
Here is a concrete example of compiling and linking your project against
CIL. Imagine that your program analysis or transformation is contained in
the single file main.ml.
$ ocamlopt -c -I $(CIL)/obj/x86_LINUX/ main.ml
$ ocamlopt -ccopt -L$(CIL)/obj/x86_LINUX/ -o main unix.cmxa str.cmxa \
$(CIL)/obj/x86_LINUX/cil.cmxa main.cmx
The first line compiles your analysis, the second line links it against CIL
(as a library) and the Ocaml Unix library. For more information about
compiling and linking Ocaml programs, see the Ocaml home page
at http://caml.inria.fr/ocaml/.
In the next section we give an overview of the API that you can use
to write your analysis and transformation.
6 CIL API Documentation
The CIL API is documented in the file src/cil.mli. We also have an
online documentation extracted from cil.mli. We
index below the main types that are used to represent C programs in CIL:
6.1 Using the visitor
One of the most useful tools exported by the CIL API is an implementation of
the visitor pattern for CIL programs. The visiting engine scans depth-first
the structure of a CIL program and at each node is queries a user-provided
visitor structure whether it should do one of the following operations:
-
Ignore this node and all its descendants
- Descend into all of the children and when done rebuild the node if any
of the children have changed.
- Replace the subtree rooted at the node with another tree.
- Replace the subtree with another tree, then descend into the children
and rebuild the node if necessary and then invoke a user-specified function.
- In addition to all of the above actions then visitor can specify that
some instructions should be queued to be inserted before the current
instruction or statement being visited.
By writing visitors you can customize the program traversal and
transformation. One major limitation of the visiting engine is that it does
not propagate information from one node to another. Each visitor must use its
own private data to achieve this effect if necessary.
Each visitor is an object that is an instance of a class of type Cil.cilVisitor..
The most convenient way to obtain such classes is to specialize the
Cil.nopCilVisitor.class (which just traverses the tree doing
nothing). Any given specialization typically overrides only a few of the
methods. Take a look for example at the visitor defined in the module
logwrites.ml. Another, more elaborate example of a visitor is the
[copyFunctionVisitor] defined in cil.ml.
Once you have defined a visitor you can invoke it with one of the functions:
Some transformations may want to use visitors to insert additional
instructions before statements and instructions. To do so, pass a list of
instructions to the Cil.queueInstr method of the specialized
object. The instructions will automatically be inserted before that
instruction in the transformed code. The Cil.unqueueInstr method
should not normally be called by the user.
6.2 Interpreted Constructors and Deconstructors
Interpreted constructors and deconstructors are a facility for constructing
and deconstructing CIL constructs using a pattern with holes that can be
filled with a variety of kinds of elements. The pattern is a string that uses
the C syntax to represent C language elements. For example, the following
code:
Formatcil.cType "void * const (*)(int x)"
is an alternative way to construct the internal representation of the type of pointer to function
with an integer argument and a void * const as result:
TPtr(TFun(TVoid [Attr("const", [])],
[ ("x", TInt(IInt, []), []) ], false, []), [])
The advantage of the interpreted constructors is that you can use familiar C
syntax to construct CIL abstract-syntax trees.
You can construct this way types, lvalues, expressions, instructions and
statements. The pattern string can also contain a number of placeholders that
are replaced during construction with CIL items passed as additional argument
to the construction function. For example, the %e:id placeholder means
that the argument labeled “id” (expected to be of form Fe exp) will
supply the expression to replace the placeholder. For example, the following
code constructs an increment instruction at location loc:
Formatcil.cInstr "%v:x = %v:x + %e:something"
loc
[ ("something", Fe some_exp);
("x", Fv some_varinfo) ]
An alternative way to construct the same CIL instruction is:
Set((Var some_varinfo, NoOffset),
BinOp(PlusA, Lval (Var some_varinfo, NoOffset),
some_exp, intType),
loc)
See Cil.formatArg for a definition of the placeholders that are
understood.
A dual feature is the interpreted deconstructors. This can be used to test
whether a CIL construct has a certain form:
Formatcil.dType "void * const (*)(int x)" t
will test whether the actual argument t is indeed a function pointer of
the required type. If it is then the result is Some [] otherwise it is
None. Furthermore, for the purpose of the interpreted deconstructors
placeholders in patterns match anything of the right type. For example,
Formatcil.dType "void * (*)(%F:t)" t
will match any function pointer type, independent of the type and number of
the formals. If the match succeeds the result is Some [ FF forms ] where
forms is a list of names and types of the formals. Note that each member
in the resulting list corresponds positionally to a placeholder in the
pattern.
The interpreted constructors and deconstructors do not support the complete C
syntax, but only a substantial fragment chosen to simplify the parsing. The
following is the syntax that is supported:
Expressions:
E ::= %e:ID | %d:ID | %g:ID | n | L | ( E ) | Unop E | E Binop E
| sizeof E | sizeof ( T ) | alignof E | alignof ( T )
| & L | ( T ) E
Unary operators:
Unop ::= + | - | ~ | %u:ID
Binary operators:
Binop ::= + | - | * | / | << | >> | & | ``|'' | ^
| == | != | < | > | <= | >= | %b:ID
Lvalues:
L ::= %l:ID | %v:ID Offset | * E | (* E) Offset | E -> ident Offset
Offsets:
Offset ::= empty | %o:ID | . ident Offset | [ E ] Offset
Types:
T ::= Type_spec Attrs Decl
Type specifiers:
Type_spec ::= void | char | unsigned char | short | unsigned short
| int | unsigned int | long | unsigned long | %k:ID | float
| double | struct %c:ID | union %c:ID
Declarators:
Decl ::= * Attrs Decl | Direct_decl
Direct declarators:
Direct_decl ::= empty | ident | ( Attrs Decl )
| Direct_decl [ Exp_opt ]
| ( Attrs Decl )( Parameters )
Optional expressions
Exp_opt ::= empty | E | %eo:ID
Formal parameters
Parameters ::= empty | ... | %va:ID | %f:ID | T | T , Parameters
List of attributes
Attrs ::= empty | %A:ID | Attrib Attrs
Attributes
Attrib ::= const | restrict | volatile | __attribute__ ( ( GAttr ) )
GCC Attributes
GAttr ::= ident | ident ( AttrArg_List )
Lists of GCC Attribute arguments:
AttrArg_List ::= AttrArg | %P:ID | AttrArg , AttrArg_List
GCC Attribute arguments
AttrArg ::= %p:ID | ident | ident ( AttrArg_List )
Instructions
Instr ::= %i:ID ; | L = E ; | L Binop= E | Callres L ( Args )
Actual arguments
Args ::= empty | %E:ID | E | E , Args
Call destination
Callres ::= empty | L = | %lo:ID
Statements
Stmt ::= %s:ID | if ( E ) then Stmt ; | if ( E ) then Stmt else Stmt ;
| return Exp_opt | break ; | continue ; | { Stmt_list }
| while (E ) Stmt | Instr_list
Lists of statements
Stmt_list ::= empty | %S:ID | Stmt Stmt_list
| Type_spec Attrs Decl ; Stmt_list
| Type_spec Attrs Decl = E ; Stmt_list
| Type_spec Attrs Decl = L (Args) ; Stmt_list
List of instructions
Instr_list ::= Instr | %I:ID | Instr Instr_list
Notes regarding the syntax:
-
In the grammar description above non-terminals are written with
uppercase initial
- All of the patterns consist of the % character followed by one or
two letters, followed by “:” and an indentifier. For each such
pattern there is a corresponding constructor of the Cil.formatArg
type, whose name is the letter 'F' followed by the same one or two letters as
in the pattern. That constructor is used by the user code to pass a
Cil.formatArg actual argument to the interpreted constructor and by
the interpreted deconstructor to return what was matched for a pattern.
- If the pattern name is uppercase, it designates a list of the elements
designated by the corresponding lowercase pattern. E.g. %E designated lists
of expressions (as in the actual arguments of a call).
- The two-letter patterns whose second letter is “o” designate an
optional element. E.g. %eo designates an optional expression (as in the
length of an array).
- Unlike in calls to printf, the pattern %g is used for strings.
- The usual precedence and associativity rules as in C apply
- The pattern string can contain newlines and comments, using both the
/* ... */ style as well as the // one.
- When matching a “cast” pattern of the form ( T ) E, the
deconstructor will match even expressions that do not have the actual cast but
in that case the type is matched against the type of the expression. E.g. the
patters "(int)%e" will match any expression of type int whether it
has an explicit cast or not.
- The %k pattern is used to construct and deconstruct an integer type of
any kind.
- Notice that the syntax of types and declaration are the same (in order
to simplify the parser). This means that technically you can write a whole
declaration instead of a type in the cast. In this case the name that you
declare is ignored.
- In lists of formal parameters and lists of attributes, an empty list in
the pattern matches any formal parameters or attributes.
- When matching types, uses of named types are unrolled to expose a real
type before matching.
- The order of the attributes is ignored during matching. The the pattern
for a list of attributes contains %A then the resulting formatArg will be
bound to all attributes in the list. For example, the pattern "const
%A" matches any list of attributes that contains const and binds the
corresponding placeholder to the entire list of attributes, including
const.
- All instruction-patterns must be terminated by semicolon
- The autoincrement and autodecrement instructions are not supported. Also
not supported are complex expressions, the && and || shortcut
operators, and a number of other more complex instructions or statements. In
general, the patterns support only constructs that can be represented directly
in CIL.
- The pattern argument identifiers are not used during deconstruction.
Instead, the result contains a sequence of values in the same order as the
appearance of pattern arguments in the pattern.
- You can mix statements with declarations. For each declaration a new
temporary will be constructed (using a function you provive). You can then
refer to that temporary by name in the rest of the pattern.
- The %v: pattern specifier is optional.
The following function are defined in the Formatcil module for
constructing and deconstructing:
Below is an example using interpreted constructors. This example generates
the CIL representation of code that scans an array backwards and initializes
every even-index element with an expression:
Formatcil.cStmts
loc
"int idx = sizeof(array) / sizeof(array[0]) - 1;
while(idx >= 0) {
// Some statements to be run for all the elements of the array
%S:init
if(! (idx & 1))
array[idx] = %e:init_even;
/* Do not forget to decrement the index variable */
idx = idx - 1;
}"
(fun n t -> makeTempVar myfunc ~name:n t)
[ ("array", Fv myarray);
("init", FS [stmt1; stmt2; stmt3]);
("init_even", Fe init_expr_for_even_elements) ]
To write the same CIL statement directly in CIL would take much more effort.
Note that the pattern is parsed only once and the result (a function that
takes the arguments and constructs the statement) is memoized.
6.2.1 Performance considerations for interpreted constructors
Parsing the patterns is done with a LALR parser and it takes some time. To
improve performance the constructors and deconstructors memoize the parsed
patterns and will only compile a pattern once. Also all construction and
deconstruction functions can be applied partially to the pattern string to
produce a function that can be later used directly to construct or
deconstruct. This function appears to be about two times slower than if the
construction is done using the CIL constructors (without memoization the
process would be one order of magnitude slower.) However, the convenience of
interpreted constructor might make them a viable choice in many situations
when performance is not paramount (e.g. prototyping).
6.3 Printing and Debugging support
The Modules Pretty and Errormsg contain respectively
utilities for pretty printing and reporting errors and provide a convenient
printf-like interface.
Additionally, CIL defines for each major type a pretty-printing function that
you can use in conjunction with the Pretty interface. The
following are some of the pretty-printing functions:
You can even customize the pretty-printer by creating instances of
Cil.cilPrinter.. Typically such an instance extends
Cil.defaultCilPrinter. Once you have a customized pretty-printer you
can use the following printing functions:
CIL has certain internal consistency invariants. For example, all references
to a global variable must point to the same varinfo structure. This
ensures that one can rename the variable by changing the name in the
varinfo. These constraints are mentioned in the API documentation. There
is also a consistency checker in file src/check.ml. If you suspect that
your transformation is breaking these constraints then you can pass the
--check option to cilly and this will ensure that the consistency checker
is run after each transformation.
6.4 Attributes
In CIL you can attach attributes to types and to names (variables, functions
and fields). Attributes are represented using the type Cil.attribute.
An attribute consists of a name and a number of arguments (represented using
the type Cil.attrparam). Almost any expression can be used as an
attribute argument. Attributes are stored in lists sorted by the name of the
attribute. To maintain list ordering, use the functions
Cil.typeAttrs to retrieve the attributes of a type and the functions
Cil.addAttribute and Cil.addAttributes to add attributes.
Alternatively you can use Cil.typeAddAttributes to add an attribute to
a type (and return the new type).
GCC already has extensive support for attributes, and CIL extends this
support to user-defined attributes. A GCC attribute has the syntax:
gccattribute ::= __attribute__((attribute)) (Note the double parentheses)
Since GCC and MSVC both support various flavors of each attribute (with or
without leading or trailing _) we first strip ALL leading and trailing _
from the attribute name (but not the identified in [ACons] parameters in
Cil.attrparam). When we print attributes, for GCC we add two leading
and two trailing _; for MSVC we add just two leading _.
There is support in CIL so that you can control the printing of attributes
(see Cil.setCustomPrintAttribute and
Cil.setCustomPrintAttributeScope). This custom-printing support is now
used to print the "const" qualifier as "const" and not as
"__attribute__((const))".
The attributes are specified in declarations. This is unfortunate since the C
syntax for declarations is already quite complicated and after writing the
parser and elaborator for declarations I am convinced that few C programmers
understand it completely. Anyway, this seems to be the easiest way to support
attributes.
Name attributes must be specified at the very end of the declaration, just
before the = for the initializer or before the , the separates a
declaration in a group of declarations or just before the ; that
terminates the declaration. A name attribute for a function being defined can
be specified just before the brace that starts the function body.
For example (in the following examples A1,...,An are type attributes
and N is a name attribute (each of these uses the __attribute__ syntax):
int x N;
int x N, * y N = 0, z[] N;
extern void exit() N;
int fact(int x) N { ... }
Type attributes can be specified along with the type using the following
rules:
-
The type attributes for a base type (int, float, named type, reference
to struct or union or enum) must be specified immediately following the
type (actually it is Ok to mix attributes with the specification of the
type, in between unsigned and int for example).
For example:
int A1 x N; /* A1 applies to the type int. An example is an attribute
"even" restricting the type int to even values. */
struct foo A1 A2 x; // Both A1 and A2 apply to the struct foo type
- The type attributes for a pointer type must be specified immediately
after the * symbol.
/* A pointer (A1) to an int (A2) */
int A2 * A1 x;
/* A pointer (A1) to a pointer (A2) to a float (A3) */
float A3 * A2 * A1 x;
Note: The attributes for base types and for pointer types are a strict
extension of the ANSI C type qualifiers (const, volatile and restrict). In
fact CIL treats these qualifiers as attributes.
- The attributes for a function type or for an array type can be
specified using parenthesized declarators.
For example:
/* A function (A1) from int (A2) to float (A3) */
float A3 (A1 f)(int A2);
/* A pointer (A1) to a function (A2) that returns an int (A3) */
int A3 (A2 * A1 pfun)(void);
/* An array (A1) of int (A2) */
int A2 (A1 x0)[]
/* Array (A1) of pointers (A2) to functions (A3) that take an int (A4) and
* return a pointer (A5) to int (A6) */
int A6 * A5 (A3 * A2 (A1 x1)[5])(int A4);
/* A function (A4) that takes a float (A5) and returns a pointer (A6) to an
* int (A7) */
extern int A7 * A6 (A4 x2)(float A5 x);
/* A function (A1) that takes a int (A2) and that returns a pointer (A3) to
* a function (A4) that takes a float (A5) and returns a pointer (A6) to an
* int (A7) */
int A7 * A6 (A4 * A3 (A1 x3)(int A2 x))(float A5) {
return & x2;
}
Note: ANSI C does not allow the specification of type qualifiers for function
and array types, although it allows for the parenthesized declarator. With
just a bit of thought (looking at the first few examples above) I hope that
the placement of attributes for function and array types will seem intuitive.
This extension is not without problems however. If you want to refer just to
a type (in a cast for example) then you leave the name out. But this leads to
strange conflicts due to the parentheses that we introduce to scope the
attributes. Take for example the type of x0 from above. It should be written
as:
int A2 (A1 )[]
But this will lead most C parsers into deep confusion because the parentheses
around A1 will be confused for parentheses of a function designator. To push
this problem around (I don't know a solution) whenever we are about to print a
parenthesized declarator with no name but with attributes, we comment out the
attributes so you can see them (for whatever is worth) without confusing the
compiler. For example, here is how we would print the above type:
int A2 /*(A1 )*/[]
Handling of predefined GCC attributes
GCC already supports attributes in a lot of places in declarations. The only
place where we support attributes and GCC does not is right before the { that
starts a function body.
GCC classifies its attributes in attributes for functions, for variables and
for types, although the latter category is only usable in definition of struct
or union types and is not nearly as powerful as the CIL type attributes. We
have made an effort to reclassify GCC attributes as name and type attributes
(they only apply for function types). Here is what we came up with:
-
GCC name attributes:
section, constructor, destructor, unused, weak, no_instrument_function,
noreturn, alias, no_check_memory_usage, dllinport, dllexport, exception,
model
Note: the "noreturn" attribute would be more appropriately qualified as a
function type attribute. But we classify it as a name attribute to make
it easier to support a similarly named MSVC attribute.
- GCC function type attributes:
fconst (printed as "const"), format, regparm, stdcall,
cdecl, longcall
I was not able to completely decipher the position in which these attributes
must go. So, the CIL elaborator knows these names and applies the following
rules:
-
All of the name attributes that appear in the specifier part (i.e. at
the beginning) of a declaration are associated with all declared names.
- All of the name attributes that appear at the end of a declarator are
associated with the particular name being declared.
- More complicated is the handling of the function type attributes, since
there can be more than one function in a single declaration (a function
returning a pointer to a function). Lacking any real understanding of how
GCC handles this, I attach the function type attribute to the "nearest"
function. This means that if a pointer to a function is "nearby" the
attribute will be correctly associated with the function. In truth I pray
that nobody uses declarations as that of x3 above.
Handling of predefined MSVC attributes
MSVC has two kinds of attributes, declaration modifiers to be printed before
the storage specifier using the notation "__declspec(...)" and a few
function type attributes, printed almost as our CIL function type
attributes.
The following are the name attributes that are printed using
__declspec right before the storage designator of the declaration:
thread, naked, dllimport, dllexport, noreturn
The following are the function type attributes supported by MSVC:
fastcall, cdecl, stdcall
It is not worth going into the obscure details of where MSVC accepts these
type attributes. The parser thinks it knows these details and it pulls
these attributes from wherever they might be placed. The important thing
is that MSVC will accept if we print them according to the rules of the CIL
attributes !
7 The CIL Driver
We have packaged CIL as an application cilly that contains certain
example modules, such as logwrites.ml (a module
that instruments code to print the addresses of memory locations being
written). Normally, you write another module like that, add command-line
options and an invocation of your module in src/main.ml. Once you compile
CIL you will obtain the file obj/cilly.asm.exe.
We wrote a driver for this executable that makes it easy to invoke your
analysis on existing C code with very little manual intervention. This driver
is bin/cilly and is quite powerful. Note that the cilly script
is configured during installation with the path where CIL resides. This means
that you can move it to any place you want.
A simple use of the driver is:
bin/cilly --save-temps -D HAPPY_MOOD -I myincludes hello.c -o hello
--save-temps tells CIL to save the resulting output files in the
current directory. Otherwise, they'll be put in /tmp and deleted
automatically. Not that this is the only CIL-specific flag in the
list – the other flags use gcc's syntax.
This performs the following actions:
-
preprocessing using the -D and -I arguments with the resulting
file left in hello.i,
- the invocation of the cilly.asm application which parses hello.i
converts it to CIL and the pretty-prints it to hello.cil.c
- another round of preprocessing with the result placed in hello.cil.i
- the true compilation with the result in hello.cil.o
- a linking phase with the result in hello
Note that cilly behaves like the gcc compiler. This makes it
easy to use it with existing Makefiles:
make CC="bin/cilly" LD="bin/cilly"
cilly can also behave as the Microsoft Visual C compiler, if the first
argument is --mode=MSVC:
bin/cilly --mode=MSVC /D HAPPY_MOOD /I myincludes hello.c /Fe hello.exe
(This in turn will pass a --MSVC flag to the underlying cilly.asm
process which will make it understand the Microsoft Visual C extensions)
cilly can also behave as the archiver ar, if it is passed an
argument --mode=AR. Note that only the cr mode is supported (create a
new archive and replace all files in there). Therefore the previous version of
the archive is lost.
Furthermore, cilly allows you to pass some arguments on to the
underlying cilly.asm process. As a general rule all arguments that start
with -- and that cilly itself does not process, are passed on. For
example,
bin/cilly --dologwrites -D HAPPY_MOOD -I myincludes hello.c -o hello.exe
will produce a file hello.cil.c that prints all the memory addresses
written by the application.
The most powerful feature of cilly is that it can collect all the
sources in your project, merge them into one file and then apply CIL. This
makes it a breeze to do whole-program analysis and transformation. All you
have to do is to pass the --merge flag to cilly:
make CC="bin/cilly --save-temps --dologwrites --merge"
You can even leave some files untouched:
make CC="bin/cilly --save-temps --dologwrites --merge --leavealone=foo --leavealone=bar"
This will merge all the files except those with the basename foo and
bar. Those files will be compiled as usual and then linked in at the very
end.
The sequence of actions performed by cilly depends on whether merging
is turned on or not:
-
If merging is off
-
For every file file.c to compile
-
Preprocess the file with the given arguments to
produce file.i
- Invoke cilly.asm to produce a file.cil.c
- Preprocess to file.cil.i
- Invoke the underlying compiler to produce file.cil.o
- Link the resulting objects
- If merging is on
-
For every file file.c to compile
-
Preprocess the file with the given arguments to
produce file.i
- Save the preprocessed source as file.o
- When linking executable hello.exe, look at every object
file that must be linked and see if it actually
contains preprocessed source. Pass all those files to a
special merging application (described in
Section 13) to produce hello.exe_comb.c
- Invoke cilly.asm to produce a hello.exe_comb.cil.c
- Preprocess to hello.exe_comb.cil.i
- Invoke the underlying compiler to produce hello.exe_comb.cil.o
- Invoke the actual linker to produce hello.exe
Note that files that you specify with --leavealone are not merged and
never presented to CIL. They are compiled as usual and then are linked in at
the end.
And a final feature of cilly is that it can substitute copies of the
system's include files:
make CC="bin/cilly --includedir=myinclude"
This will force the preprocessor to use the file myinclude/xxx/stdio.h
(if it exists) whenever it encounters #include <stdio.h>. The xxx is
a string that identifies the compiler version you are using. This modified
include files should be produced with the patcher script (see
Section 14).
7.1 cilly Options
Among the options for the cilly you can put anything that can normally
go in the command line of the compiler that cilly is impersonating.
cilly will do its best to pass those options along to the appropriate
subprocess. In addition, the following options are supported (a complete and
up-to-date list can always be obtained by running cilly --help):
-
--mode=mode This must be the first argument if present. It makes
cilly behave as a given compiled. The following modes are recognized:
-
GNUCC - the GNU C Compiler. This is the default.
- MSVC - the Microsoft Visual C compiler. Of course, you should
pass only MSVC valid options in this case.
- AR - the archiver ar. Only the mode cr is supported and
the original version of the archive is lost.
- --help Prints a list of the options supported.
- --verbose Prints lots of messages about what is going on.
- --stages Less than --verbose but lets you see what cilly
is doing.
- --merge This tells cilly to first attempt to collect into one
source file all of the sources that make your application, and then to apply
cilly.asm on the resulting source. The sequence of actions in this case is
described above and the merger itself is described in Section 13.
- --leavealone=xxx. Do not merge and do not present to CIL the files
whose basename is "xxx". These files are compiled as usual and linked in at
the end.
- --includedir=xxx. Override the include files with those in the given
directory. The given directory is the same name that was given an an argument
to the patcher (see Section 14). In particular this means that
that directory contains subdirectories named based on the current compiler
version. The patcher creates those directories.
- --usecabs. Do not CIL, but instead just parse the source and print
its AST out. This should looked like the preprocessed file. This is useful
when you suspect that the conversion to CIL phase changes the meaning of the
program.
- --save-temps=xxx. Temporary files are preserved in the xxx
directory. For example, the output of CIL will be put in a file
named *.cil.c.
- --save-temps. Temporay files are preserved in the current directory.
7.2 cilly.asm Options
All of the options that start with -- and are not understood by
cilly are passed on to cilly.asm. cilly also passes along to
cilly.asm flags such as --MSVC that both need to know
about. The following options are supported:
General Options:
-
--version output version information and exit
- --verbose Print lots of random stuff. This is passed on from cilly
- --warnall Show all warnings.
- --debug=xxx turns on debugging flag xxx
- --nodebug=xxx turns off debugging flag xxx
- --flush Flush the output streams often (aids debugging).
- --check Run a consistency check over the CIL after every operation.
- --nocheck turns off consistency checking of CIL.
- --noPrintLn Don't output #line directives in the output.
- --commPrintLn Print #line directives in the output, but
put them in comments.
- --log=xxx Set the name of the log file. By default stderr is used
- --MSVC Enable MSVC compatibility. Default is GNU.
- --ignore-merge-conflicts ignore merging conflicts.
- --extrafiles=filename: the name of a file that contains
a list of additional files to process, separated by whitespace.
- --stats Print statistics about the running time of the
parser, conversion to CIL, etc. Also prints memory-usage
statistics. You can time parts of your own code as well. Calling
(Stats.time “label” func arg) will evaluate (func arg)
and remember how long this takes. If you call Stats.time
repeatedly with the same label, CIL will report the aggregate
time.
If available, CIL uses the x86 performance counters for these
stats. This is very precise, but results in “wall-clock time.”
To report only user-mode time, find the call to Stats.reset in
main.ml, and change it to Stats.reset false.
Lowering Options
- --noLowerConstants do not lower constant expressions.
- --noInsertImplicitCasts do not insert implicit casts.
- --forceRLArgEval Forces right to left evaluation of function arguments.
- --disallowDuplication Prevent small chunks of code from being duplicated.
- --keepunused Do not remove the unused variables and types.
- --rmUnusedInlines Delete any unused inline functions. This is the default in MSVC mode.
Output Options:
- --printCilAsIs Do not try to simplify the CIL when
printing. Without this flag, CIL will attempt to produce prettier
output by e.g. changing while(1) into more meaningful loops.
- --noWrap do not wrap long lines when printing
- --out=xxx the name of the output CIL file. cilly
sets this for you.
- --mergedout=xxx specify the name of the merged file
- --cabsonly=xxx CABS output file name
Selected features. See Section 8 for more information.
- --dologcalls. Insert code in the processed source to print the name of
functions as are called. Implemented in src/ext/logcalls.ml.
- --dologwrites. Insert code in the processed source to print the
address of all memory writes. Implemented in src/ext/logwrites.ml.
- --dooneRet. Make each function have at most one 'return'.
Implemented in src/ext/oneret.ml.
- --dostackGuard. Instrument function calls and returns to
maintain a separate stack for return addresses. Implemeted in
src/ext/heapify.ml.
- --domakeCFG. Make the program look more like a CFG. Implemented
in src/cil.ml.
- --dopartial. Do interprocedural partial evaluation and
constant folding. Implemented in src/ext/partial.ml.
- --dosimpleMem. Simplify all memory expressions. Implemented in
src/ext/simplemem.ml.
For an up-to-date list of available options, run cilly.asm --help.
8 Library of CIL Modules
We are developing a suite of modules that use CIL for program analyses and
transformations that we have found useful. You can use these modules directly
on your code, or generally as inspiration for writing similar modules. A
particularly big and complex application written on top of CIL is CCured
(../ccured/index.html).
8.1 Control-Flow Graphs
The Cil.stmt datatype includes fields for intraprocedural
control-flow information: the predecessor and successor statements of
the current statement. This information is not computed by default.
If you want to use the control-flow graph, or any of the extensions in
this section that require it, you have to explicitly ask CIL to
compute the CFG.
8.1.1 The CFG module (new in CIL 1.3.5)
The best way to compute the CFG is with the CFG module. Just invoke
Cfg.computeFileCFG on your file. The Cfg API
describes the rest of actions you can take with this module, including
computing the CFG for one function at a time, or printing the CFG in
dot form.
8.1.2 Simplified control flow
CIL can reduce high-level C control-flow constructs like switch and
continue to lower-level gotos. This completely eliminates some
possible classes of statements from the program and may make the result
easier to analyze (e.g., it simplifies data-flow analysis).
You can invoke this transformation on the command line with
--domakeCFG or programatically with Cil.prepareCFG.
After calling Cil.prepareCFG, you can use Cil.computeCFGInfo
to compute the CFG information and find the successor and predecessor
of each statement.
For a concrete example, you can see how cilly --domakeCFG
transforms the following code (note the fall-through in case 1):
int foo (int predicate) {
int x = 0;
switch (predicate) {
case 0: return 111;
case 1: x = x + 1;
case 2: return (x+3);
case 3: break;
default: return 222;
}
return 333;
}
See the CIL output for this
code fragment
8.2 Data flow analysis framework
The Dataflow module (click for the ocamldoc) contains a
parameterized framework for forward and backward data flow
analyses. You provide the transfer functions and this module does the
analysis. You must compute control-flow information (Section 8.1)
before invoking the Dataflow module.
8.3 Dominators
The module Dominators contains the computation of immediate
dominators. It uses the Dataflow module.
8.4 Points-to Analysis
The module ptranal.ml contains two interprocedural points-to
analyses for CIL: Olf and Golf. Olf is the default.
(Switching from olf.ml to golf.ml requires a change in
Ptranal and a recompiling cilly.)
The analyses have the following characteristics:
-
Not based on C types (inferred pointer relationships are sound
despite most kinds of C casts)
- One level of subtyping
- One level of context sensitivity (Golf only)
- Monomorphic type structures
- Field insensitive (fields of structs are conflated)
- Demand-driven (points-to queries are solved on demand)
- Handle function pointers
The analysis itself is factored into two components: Ptranal,
which walks over the CIL file and generates constraints, and Olf
or Golf, which solve the constraints. The analysis is invoked
with the function Ptranal.analyze_file: Cil.file ->
unit. This function builds the points-to graph for the CIL file
and stores it internally. There is currently no facility for clearing
internal state, so Ptranal.analyze_file should only be called
once.
The constructed points-to graph supports several kinds of queries,
including alias queries (may two expressions be aliased?) and
points-to queries (to what set of locations may an expression point?).
The main interface with the alias analysis is as follows:
-
Ptranal.may_alias: Cil.exp -> Cil.exp -> bool. If
true, the two expressions may have the same value.
- Ptranal.resolve_lval: Cil.lval -> (Cil.varinfo
list). Returns the list of variables to which the given
left-hand value may point.
- Ptranal.resolve_exp: Cil.exp -> (Cil.varinfo list).
Returns the list of variables to which the given expression may
point.
- Ptranal.resolve_funptr: Cil.exp -> (Cil.fundec
list). Returns the list of functions to which the given
expression may point.
The precision of the analysis can be customized by changing the values
of several flags:
-
Ptranal.no_sub: bool ref.
If true, subtyping is disabled. Associated commandline option:
--ptr_unify.
- Ptranal.analyze_mono: bool ref.
(Golf only) If true, context sensitivity is disabled and the
analysis is effectively monomorphic. Commandline option:
--ptr_mono.
- Ptranal.smart_aliases: bool ref.
(Golf only) If true, “smart” disambiguation of aliases is
enabled. Otherwise, aliases are computed by intersecting points-to
sets. This is an experimental feature.
- Ptranal.model_strings: bool ref.
Make the alias analysis model string constants by treating them as
pointers to chars. Commandline option: --ptr_model_strings
- Ptranal.conservative_undefineds: bool ref.
Make the most pessimistic assumptions about globals if an undefined
function is present. Such a function can write to every global
variable. Commandline option: --ptr_conservative
In practice, the best precision/efficiency tradeoff is achieved by
setting Ptranal.no_sub to false, Ptranal.analyze_mono to
true, and Ptranal.smart_aliases to false. These are the
default values of the flags.
There are also a few flags that can be used to inspect or serialize
the results of the analysis.
-
Ptranal.debug_may_aliases.
Print the may-alias relationship of each pair of expressions in the
program. Commandline option: --ptr_may_aliases.
- Ptranal.print_constraints: bool ref.
If true, the analysis will print each constraint as it is
generated.
- Ptranal.print_types: bool ref.
If true, the analysis will print the inferred type of each
variable in the program.
If Ptranal.analyze_mono and Ptranal.no_sub are both
true, this output is sufficient to reconstruct the points-to
graph. One nice feature is that there is a pretty printer for
recursive types, so the print routine does not loop.
- Ptranal.compute_results: bool ref.
If true, the analysis will print out the points-to set of each
variable in the program. This will essentially serialize the
points-to graph.
8.5 StackGuard
The module heapify.ml contains a transformation similar to the one
described in “StackGuard: Automatic Adaptive Detection and Prevention of
Buffer-Overflow Attacks”, Proceedings of the 7th USENIX Security
Conference. In essence it modifies the program to maintain a separate
stack for return addresses. Even if a buffer overrun attack occurs the
actual correct return address will be taken from the special stack.
Although it does work, this CIL module is provided mainly as an example of
how to perform a simple source-to-source program analysis and
transformation. As an optimization only functions that contain a dangerous
local array make use of the special return address stack.
For a concrete example, you can see how cilly --dostackGuard
transforms the following dangerous code:
int dangerous() {
char array[10];
scanf("%s",array); // possible buffer overrun!
}
int main () {
return dangerous();
}
See the CIL output for this
code fragment
8.6 Heapify
The module heapify.ml also contains a transformation that moves all
dangerous local arrays to the heap. This also prevents a number of buffer
overruns.
For a concrete example, you can see how cilly --doheapify
transforms the following dangerous code:
int dangerous() {
char array[10];
scanf("%s",array); // possible buffer overrun!
}
int main () {
return dangerous();
}
See the CIL output for this
code fragment
8.7 One Return
The module oneret.ml contains a transformation the ensures that all
function bodies have at most one return statement. This simplifies a number
of analyses by providing a canonical exit-point.
For a concrete example, you can see how cilly --dooneRet
transforms the following code:
int foo (int predicate) {
if (predicate <= 0) {
return 1;
} else {
if (predicate > 5)
return 2;
return 3;
}
}
See the CIL output for this
code fragment
8.8 Partial Evaluation and Constant Folding
The partial.ml module provides a simple interprocedural partial
evaluation and constant folding data-flow analysis and transformation. This
transformation requires the --domakeCFG option.
For a concrete example, you can see how cilly --domakeCFG --dopartial
transforms the following code (note the eliminated if branch and the
partial optimization of foo):
int foo(int x, int y) {
int unknown;
if (unknown)
return y+2;
return x+3;
}
int main () {
int a,b,c;
a = foo(5,7) + foo(6,7);
b = 4;
c = b * b;
if (b > c)
return b-c;
else
return b+c;
}
See the CIL output for this
code fragment
8.9 Reaching Definitions
The reachingdefs.ml module uses the dataflow framework and CFG
information to calculate the definitions that reach each
statement. After computing the CFG (Section 8.1) and calling
computeRDs on a
function declaration, ReachingDef.stmtStartData will contain a
mapping from statement IDs to data about which definitions reach each
statement. In particular, it is a mapping from statement IDs to a
triple the first two members of which are used internally. The third
member is a mapping from variable IDs to Sets of integer options. If
the set contains Some(i), then the definition of that variable
with ID i reaches that statement. If the set contains None,
then there is a path to that statement on which there is no definition
of that variable. Also, if the variable ID is unmapped at a
statement, then no definition of that variable reaches that statement.
To summarize, reachingdefs.ml has the following interface:
-
computeRDs – Computes reaching definitions. Requires that
CFG information has already been computed for each statement.
- ReachingDef.stmtStartData – contains reaching
definition data after computeRDs is called.
- ReachingDef.defIdStmtHash – Contains a mapping
from definition IDs to the ID of the statement in which
the definition occurs.
- getRDs – Takes a statement ID and returns
reaching definition data for that statement.
- instrRDs – Takes a list of instructions and the
definitions that reach the first instruction, and for
each instruction calculates the definitions that reach
either into or out of that instruction.
- rdVisitorClass – A subclass of nopCilVisitor that
can be extended such that the current reaching definition
data is available when expressions are visited through
the get_cur_iosh method of the class.
8.10 Available Expressions
The availexps.ml module uses the dataflow framework and CFG
information to calculate something similar to a traditional available
expressions analysis. After computeAEs is called following a CFG
calculation (Section 8.1), AvailableExps.stmtStartData will
contain a mapping
from statement IDs to data about what expressions are available at
that statement. The data for each statement is a mapping for each
variable ID to the whole expression available at that point(in the
traditional sense) which the variable was last defined to be. So,
this differs from a traditional available expressions analysis in that
only whole expressions from a variable definition are considered rather
than all expressions.
The interface is as follows:
-
computeAEs – Computes available expressions. Requires
that CFG information has already been comptued for each statement.
- AvailableExps.stmtStartData – Contains available
expressions data for each statement after computeAEs has been
called.
- getAEs – Takes a statement ID and returns
available expression data for that statement.
- instrAEs – Takes a list of instructions and
the availalbe expressions at the first instruction, and
for each instruction calculates the expressions available
on entering or exiting each instruction.
- aeVisitorClass – A subclass of nopCilVisitor that
can be extended such that the current available expressions
data is available when expressions are visited through the
get_cur_eh method of the class.
8.11 Liveness Analysis
The liveness.ml module uses the dataflow framework and
CFG information to calculate which variables are live at
each program point. After computeLiveness is called
following a CFG calculation (Section 8.1), LiveFlow.stmtStartData will
contain a mapping for each statement ID to a set of varinfos
for varialbes live at that program point.
The interface is as follows:
-
computeLiveness – Computes live variables. Requires
that CFG information has already been computed for each statement.
- LiveFlow.stmtStartData – Contains live variable data
for each statement after computeLiveness has been called.
Also included in this module is a command line interface that
will cause liveness data to be printed to standard out for
a particular function or label.
-
–doliveness – Instructs cilly to comptue liveness
information and to print on standard out the variables live
at the points specified by –live_func and live_label.
If both are ommitted, then nothing is printed.
- –live_func – The name of the function whose
liveness data is of interest. If –live_label is ommitted,
then data for each statement is printed.
- –live_label – The name of the label at which
the liveness data will be printed.
8.12 Dead Code Elimination
The module deadcodeelim.ml uses the reaching definitions
analysis to eliminate assignment instructions whose results
are not used. The interface is as follows:
-
elim_dead_code – Performs dead code elimination
on a function. Requires that CFG information has already
been computed (Section 8.1).
- dce – Performs dead code elimination on an
entire file. Requires that CFG information has already
been computed.
8.13 Simple Memory Operations
The simplemem.ml module allows CIL lvalues that contain memory
accesses to be even futher simplified via the introduction of
well-typed temporaries. After this transformation all lvalues involve
at most one memory reference.
For a concrete example, you can see how cilly --dosimpleMem
transforms the following code:
int main () {
int ***three;
int **two;
***three = **two;
}
See the CIL output for this
code fragment
8.14 Simple Three-Address Code
The simplify.ml module further reduces the complexity of program
expressions and gives you a form of three-address code. After this
transformation all expressions will adhere to the following grammar:
basic::=
Const _
Addrof(Var v, NoOffset)
StartOf(Var v, NoOffset)
Lval(Var v, off), where v is a variable whose address is not taken
and off contains only "basic"
exp::=
basic
Lval(Mem basic, NoOffset)
BinOp(bop, basic, basic)
UnOp(uop, basic)
CastE(t, basic)
lval ::=
Mem basic, NoOffset
Var v, off, where v is a variable whose address is not taken and off
contains only "basic"
In addition, all sizeof and alignof forms are turned into
constants. Accesses to arrays and variables whose address is taken are
turned into "Mem" accesses. All field and index computations are turned
into address arithmetic.
For a concrete example, you can see how cilly --dosimplify
transforms the following code:
int main() {
struct mystruct {
int a;
int b;
} m;
int local;
int arr[3];
int *ptr;
ptr = &local;
m.a = local + sizeof(m) + arr[2];
return m.a;
}
See the CIL output for this
code fragment
8.15 Converting C to C++
The module canonicalize.ml performs several transformations to correct
differences between C and C++, so that the output is (hopefully) valid
C++ code. This may be incomplete — certain fixes which are necessary
for some programs are not yet implemented.
Using the --doCanonicalize option with CIL will perform the
following changes to your program:
-
Any variables that use C++ keywords as identifiers are renamed.
- C allows global variables to have multiple declarations and
multiple (equivalent) definitions. This transformation removes
all but one declaration and all but one definition.
- __inline is #defined to inline, and __restrict
is #defined to nothing.
- C allows function pointers with no specified arguments to be used on
any argument list. To make C++ accept this code, we insert a cast
from the function pointer to a type that matches the arguments. Of
course, this does nothing to guarantee that the pointer actually has
that type.
- Makes casts from int to enum types explicit. (CIL changes enum
constants to int constants, but doesn't use a cast.)
9 Controlling CIL
In the process of converting a C file to CIL we drop the unused prototypes
and even inline function definitions. This results in much smaller files. If
you do not want this behavior then you must pass the --keepunused argument
to the CIL application.
Alternatively you can put the following pragma in the code (instructing CIL
to specifically keep the declarations and definitions of the function
func1 and variable var2, the definition of type foo and of
structure bar):
#pragma cilnoremove("func1", "var2", "type foo", "struct bar")
10 GCC Extensions
The CIL parser handles most of the gcc
extensions
and compiles them to CIL. The following extensions are not handled (note that
we are able to compile a large number of programs, including the Linux kernel,
without encountering these):
-
Nested function definitions.
- Constructing function calls.
- Naming an expression's type.
- Complex numbers
- Hex floats
- Subscripts on non-lvalue arrays.
- Forward function parameter declarations
The following extensions are handled, typically by compiling them away:
-
Attributes for functions, variables and types. In fact, we have a clear
specification (see Section 6.4) of how attributes are interpreted. The
specification extends that of gcc.
- Old-style function definitions and prototypes. These are translated to
new-style.
- Locally-declared labels. As part of the translation to CIL, we generate
new labels as needed.
- Labels as values and computed goto. This allows a program to take the
address of a label and to manipulate it as any value and also to perform a
computed goto. We compile this by assigning each label whose address is taken
a small integer that acts as its address. Every computed goto in the body
of the function is replaced with a switch statement. If you want to invoke
the label from another function, you are on your own (the gcc
documentation says the same.)
- Generalized lvalues. You can write code like (a, b) += 5 and it gets
translated to CIL.
- Conditionals with omitted operands. Things like x ? : y are
translated to CIL.
- Double word integers. The type long long and the LL suffix on
constants is understood. This is currently interpreted as 64-bit integers.
- Local arrays of variable length. These are converted to uses of
alloca, the array variable is replaced with a pointer to the allocated
array and the instances of sizeof(a) are adjusted to return the size of
the array and not the size of the pointer.
- Non-constant local initializers. Like all local initializers these are
compiled into assignments.
- Compound literals. These are also turned into assignments.
- Designated initializers. The CIL parser actually supports the full ISO
syntax for initializers, which is more than both gcc and MSVC. I
(George) think that this is the most complicated part of the C language and
whoever designed it should be banned from ever designing languages again.
- Case ranges. These are compiled into separate cases. There is no code
duplication, just a larger number of case statements.
- Transparent unions. This is a strange feature that allows you to define
a function whose formal argument has a (tranparent) union type, but the
argument is called as if it were the first element of the union. This is
compiled away by saying that the type of the formal argument is that of the
first field, and the first thing in the function body we copy the formal into
a union.
- Inline assembly-language. The full syntax is supported and it is carried
as such in CIL.
- Function names as strings. The identifiers __FUNCTION__ and
__PRETTY_FUNCTION__ are replaced with string literals.
- Keywords typeof, alignof, inline are supported.
11 CIL Limitations
There are several implementation details of CIL that might make it unusable
or less than ideal for certain tasks:
-
CIL operates after preprocessing. If you need to see comments, for
example, you cannot use CIL. But you can use attributes and pragmas instead.
And there is some support to help you patch the include files before they are
seen by the preprocessor. For example, this is how we turn some
#defines that we don't like into function calls.
- CIL does transform the code in a non-trivial way. This is done in order
to make most analyses easier. But if you want to see the code e1, e2++
exactly as it appears in the code, then you should not use CIL.
- CIL removes all local scopes and moves all variables to function
scope. It also separates a declaration with an initializer into a declaration
plus an assignment. The unfortunate effect of this transformation is that
local variables cannot have the const qualifier.
12 Known Bugs and Limitations
- In the new versions of glibc there is a function
__builtin_va_arg that takes a type as its second argument. CIL
handles that through a slight trick. As it parses the function it changes a
call like:
mytype x = __builtin_va_arg(marker, mytype)
into
mytype x;
__builtin_va_arg(marker, sizeof(mytype), &x);
The latter form is used internally in CIL. However, the CIL pretty printer
will try to emit the original code.
Similarly, __builtin_types_compatible_p(t1, t2), which takes
types as arguments, is represented internally as
__builtin_types_compatible_p(sizeof t1, sizeof t2), but the
sizeofs are removed when printing.
- The implementation of bitsSizeOf does not take into account the
packing pragmas. However it was tested to be accurate on cygwin/gcc-2.95.3,
Linux/gcc-2.95.3 and on Windows/MSVC.
- We do not support tri-graph sequences (ISO 5.2.1.1).
- GCC has a strange feature called “extern inline”. Such a function can
be defined twice: first with the “extern inline” specifier and the second
time without it. If optimizations are turned off then the “extern inline”
definition is considered a prototype (its body is ignored). If optimizations
are turned on then the extern inline function is inlined at all of its
occurrences from the point of its definition all the way to the point where the
(optional) second definition appears. No body is generated for an extern
inline function. A body is generated for the real definition and that one is
used in the rest of the file.
CIL will rename your extern inline function (and its uses) with the suffix
__extinline. This means that if you have two such definition, that do
different things and the optimizations are not on, then the CIL version might
compute a different answer !
Also, if you have multiple extern inline declarations then CIL will ignore
but the first one. This is not so bad because GCC itself would not like it.
- There are still a number of bugs in handling some obscure features of
GCC. For example, when you use variable-length arrays, CIL turns them into
calls to alloca. This means that they are deallocated when the function
returns and not when the local scope ends.
Variable-length arrays are not supported as fields of a struct or union.
- CIL cannot parse arbitrary #pragma directives. Their
syntax must follow gcc's attribute syntax to be understood. If you
need a pragma that does not follow gcc syntax, add that pragma's name
to no_parse_pragma in src/frontc/clexer.mll to indicate that
CIL should treat that pragma as a monolithic string rather than try
to parse its arguments.
CIL cannot parse a line containing an empty #pragma.
- CIL only parses #pragma directives at the "top level", this is,
outside of any enum, structure, union, or function definitions.
If your compiler uses pragmas in places other than the top-level,
you may have to preprocess the sources in a special way (sed, perl,
etc.) to remove pragmas from these locations.
- CIL cannot parse the following code (fixing this problem would require
extensive hacking of the LALR grammar):
int bar(int ()); // This prototype cannot be parsed
int bar(int x()); // If you add a name to the function, it works
int bar(int (*)()); // This also works (and it is more appropriate)
- CIL also cannot parse certain K&R old-style prototypes with missing
return type:
g(); // This cannot be parsed
int g(); // This is Ok
- CIL does not understand some obscure combinations of type specifiers
(“signed” and “unsigned” applied to typedefs that themselves contain a
sign specification; you could argue that this should not be allowed anyway):
typedef signed char __s8;
__s8 unsigned uchartest; // This is unsigned char for gcc
- The statement x = 3 + x ++ will perform the increment of x
before the assignment, while gcc delays the increment after the
assignment. It turned out that this behavior is much easier to implement
than gcc's one, and either way is correct (since the behavior is unspecified
in this case). Similarly, if you write x = x ++; then CIL will perform
the increment before the assignment, whereas GCC and MSVC will perform it
after the assignment.
13 Using the merger
There are many program analyses that are more effective when
done on the whole program.
The merger is a tool that combines all of the C source files in a project
into a single C file. There are two tasks that a merger must perform:
-
Detect what are all the sources that make a project and with what
compiler arguments they are compiled.
- Merge all of the source files into a single file.
For the first task the merger impersonates a compiler and a linker (both a
GCC and a Microsoft Visual C mode are supported) and it expects to be invoked
(from a build script or a Makefile) on all sources of the project. When
invoked to compile a source the merger just preprocesses the source and saves
the result using the name of the requested object file. By preprocessing at
this time the merger is able to take into account variations in the command
line arguments that affect preprocessing of different source files.
When the merger is invoked to link a number of object files it collects the
preprocessed sources that were stored with the names of the object files, and
invokes the merger proper. Note that arguments that affect the compilation or
linking must be the same for all source files.
For the second task, the merger essentially concatenates the preprocessed
sources with care to rename conflicting file-local declarations (we call this
process alpha-conversion of a file). The merger also attempts to remove
duplicate global declarations and definitions. Specifically the following
actions are taken:
-
File-scope names (static globals, names of types defined with
typedef, and structure/union/enumeration tags) are given new names if they
conflict with declarations from previously processed sources. The new name is
formed by appending the suffix ___n, where n is a unique integer
identifier. Then the new names are applied to their occurrences in the file.
- Non-static declarations and definitions of globals are never renamed.
But we try to remove duplicate ones. Equality of globals is detected by
comparing the printed form of the global (ignoring the line number directives)
after the body has been alpha-converted. This process is intended to remove
those declarations (e.g. function prototypes) that originate from the same
include file. Similarly, we try to eliminate duplicate definitions of
inline functions, since these occasionally appear in include files.
- The types of all global declarations with the same name from all files
are compared for type isomorphism. During this process, the merger detects all
those isomorphisms between structures and type definitions that are required for the merged program to be legal. Such structure tags and
typenames are coalesced and given the same name.
- Besides the structure tags and type names that are required to be
isomorphic, the merger also tries to coalesce definitions of structures and
types with the same name from different file. However, in this case the merger
will not give an error if such definitions are not isomorphic; it will just
use different names for them.
- In rare situations, it can happen that a file-local global in
encountered first and it is not renamed, only to discover later when
processing another file that there is an external symbol with the same name.
In this case, a second pass is made over the merged file to rename the
file-local symbol.
Here is an example of using the merger:
The contents of file1.c is:
struct foo; // Forward declaration
extern struct foo *global;
The contents of file2.c is:
struct bar {
int x;
struct bar *next;
};
extern struct bar *global;
struct foo {
int y;
};
extern struct foo another;
void main() {
}
There are several ways in which one might create an executable from these
files:
-
gcc file1.c file2.c -o a.out
gcc -c file1.c -o file1.o
gcc -c file2.c -o file2.o
ld file1.o file2.o -o a.out
gcc -c file1.c -o file1.o
gcc -c file2.c -o file2.o
ar r libfile2.a file2.o
gcc file1.o libfile2.a -o a.out
gcc -c file1.c -o file1.o
gcc -c file2.c -o file2.o
ar r libfile2.a file2.o
gcc file1.o -lfile2 -o a.out
In each of the cases above you must replace all occurrences of gcc and
ld with cilly --merge, and all occurrences of ar with cilly
--merge --mode=AR. It is very important that the --merge flag be used
throughout the build process. If you want to see the merged source file you
must also pass the --keepmerged flag to the linking phase.
The result of merging file1.c and file2.c is:
// from file1.c
struct foo; // Forward declaration
extern struct foo *global;
// from file2.c
struct foo {
int x;
struct foo *next;
};
struct foo___1 {
int y;
};
extern struct foo___1 another;
14 Using the patcher
Occasionally we have needed to modify slightly the standard include files.
So, we developed a simple mechanism that allows us to create modified copies
of the include files and use them instead of the standard ones. For this
purpose we specify a patch file and we run a program caller Patcher which
makes modified copies of include files and applies the patch.
The patcher is invoked as follows:
bin/patcher [options]
Options:
--help Prints this help message
--verbose Prints a lot of information about what is being done
--mode=xxx What tool to emulate:
GNUCC - GNU CC
MSVC - MS VC cl compiler
--dest=xxx The destination directory. Will make one if it does not exist
--patch=xxx Patch file (can be specified multiple times)
--ppargs=xxx An argument to be passed to the preprocessor (can be specified
multiple times)
--ufile=xxx A user-include file to be patched (treated as \#include "xxx")
--sfile=xxx A system-include file to be patched (treated as \#include <xxx>)
--clean Remove all files in the destination directory
--dumpversion Print the version name used for the current compiler
All of the other arguments are passed to the preprocessor. You should pass
enough arguments (e.g., include directories) so that the patcher can find the
right include files to be patched.
Based on the given mode and the current version of the compiler (which
the patcher can print when given the dumpversion argument) the patcher
will create a subdirectory of the dest directory (say /usr/home/necula/cil/include), such as:
/usr/home/necula/cil/include/gcc_2.95.3-5
In that file the patcher will copy the modified versions of the include files
specified with the ufile and sfile options. Each of these options can
be specified multiple times.
The patch file (specified with the patch option) has a format inspired by
the Unix patch tool. The file has the following grammar:
<<< flags
patterns
===
replacement
>>>
The flags are a comma separated, case-sensitive, sequence of keywords or
keyword = value. The following flags are supported:
-
file=foo.h - will only apply the patch on files whose name is
foo.h.
- optional - this means that it is Ok if the current patch does not
match any of the processed files.
- group=foo - will add this patch to the named group. If this is not
specified then a unique group is created to contain just the current patch.
When all files specified in the command line have been patched, an error
message is generated for all groups for whom no member patch was used. We use
this mechanism to receive notice when the patch triggers are out-dated with
respect to the new include files.
- system=sysname - will only consider this pattern on a given
operating system. The “sysname” is reported by the “$Ô” variable in
Perl, except that Windows is always considered to have sysname
“cygwin.” For Linux use “linux” (capitalization matters).
- ateof - In this case the patterns are ignored and the replacement
text is placed at the end of the patched file. Use the file flag if you
want to restrict the files in which this replacement is performed.
- atsof - The patterns are ignored and the replacement text is placed
at the start of the patched file. Uf the file flag to restrict the
application of this patch to a certain file.
- disabled - Use this flag if you want to disable the pattern.
The patterns can consist of several groups of lines separated by the |||
marker. Each of these group of lines is a multi-line pattern that if found in
the file will be replaced with the text given at the end of the block.
The matching is space-insensitive.
All of the markers <<<, |||, === and >>> must appear at the
beginning of a line but they can be followed by arbitrary text (which is
ignored).
The replacement text can contain the special keyword @__pattern__@,
which is substituted with the pattern that matched.
15 Debugging support
Most of the time we debug our code using the Errormsg module along with the
pretty printer. But if you want to use the Ocaml debugger here is an easy way
to do it. Say that you want to debug the invocation of cilly that arises out
of the following command:
cilly -c hello.c
You must follow the installation instructions
to install the Elist support files for ocaml and to extend your .emacs
appropriately. Then from within Emacs you do
ALT-X my-camldebug
This will ask you for the command to use for running the Ocaml debugger
(initially the default will be “ocamldebug” or the last command you
introduced). You use the following command:
cilly --ocamldebug -c hello.c
This will run cilly as usual and invoke the Ocaml debugger when the cilly
engine starts. The advantage of this way of invoking the debugger is that the
directory search paths are set automatically and the right set or arguments is
passed to the debugger.
16 Who Says C is Simple?
When I (George) started to write CIL I thought it was going to take two weeks.
Exactly a year has passed since then and I am still fixing bugs in it. This
gross underestimate was due to the fact that I thought parsing and making
sense of C is simple. You probably think the same. What I did not expect was
how many dark corners this language has, especially if you want to parse
real-world programs such as those written for GCC or if you are more ambitious
and you want to parse the Linux or Windows NT sources (both of these were
written without any respect for the standard and with the expectation that
compilers will be changed to accommodate the program).
The following examples were actually encountered either in real programs or
are taken from the ISO C99 standard or from the GCC's testcases. My first
reaction when I saw these was: Is this C?. The second one was : What the hell does it mean?.
If you are contemplating doing program analysis for C on abstract-syntax
trees then your analysis ought to be able to handle these things. Or, you can
use CIL and let CIL translate them into clean C code.
16.1 Standard C
- Why does the following code return 0 for most values of x? (This
should be easy.)
int x;
return x == (1 && x);
See the CIL output for this
code fragment
- Why does the following code return 0 and not -1? (Answer: because
sizeof is unsigned, thus the result of the subtraction is unsigned, thus
the shift is logical.)
return ((1 - sizeof(int)) >> 32);
See the CIL output for this
code fragment
- Scoping rules can be tricky. This function returns 5.
int x = 5;
int f() {
int x = 3;
{
extern int x;
return x;
}
}
See the CIL output for this
code fragment
- Functions and function pointers are implicitly converted to each other.
int (*pf)(void);
int f(void) {
pf = &f; // This looks ok
pf = ***f; // Dereference a function?
pf(); // Invoke a function pointer?
(****pf)(); // Looks strange but Ok
(***************f)(); // Also Ok
}
See the CIL output for this
code fragment
- Initializer with designators are one of the hardest parts about ISO C.
Neither MSVC or GCC implement them fully. GCC comes close though. What is the
final value of i.nested.y and i.nested.z? (Answer: 2 and respectively
6).
struct {
int x;
struct {
int y, z;
} nested;
} i = { .nested.y = 5, 6, .x = 1, 2 };
See the CIL output for this
code fragment
- This is from c-torture. This function returns 1.
typedef struct
{
char *key;
char *value;
} T1;
typedef struct
{
long type;
char *value;
} T3;
T1 a[] =
{
{
"",
((char *)&((T3) {1, (char *) 1}))
}
};
int main() {
T3 *pt3 = (T3*)a[0].value;
return pt3->value;
}
See the CIL output for this
code fragment
- Another one with constructed literals. This one is legal according to
the GCC documentation but somehow GCC chokes on (it works in CIL though). This
code returns 2.
return ((int []){1,2,3,4})[1];
See the CIL output for this
code fragment
- In the example below there is one copy of “bar” and two copies of
“pbar” (static prototypes at block scope have file scope, while for all
other types they have block scope).
int foo() {
static bar();
static (*pbar)() = bar;
}
static bar() {
return 1;
}
static (*pbar)() = 0;
See the CIL output for this
code fragment
- Two years after heavy use of CIL, by us and others, I discovered a bug
in the parser. The return value of the following function depends on what
precedence you give to casts and unary minus:
unsigned long foo() {
return (unsigned long) - 1 / 8;
}
See the CIL output for this
code fragment
The correct interpretation is ((unsigned long) - 1) / 8, which is a
relatively large number, as opposed to (unsigned long) (- 1 / 8), which
is 0.
16.2 GCC ugliness
- GCC has generalized lvalues. You can take the address of a lot of
strange things:
int x, y, z;
return &(x ? y : z) - & (x++, x);
See the CIL output for this
code fragment
- GCC lets you omit the second component of a conditional expression.
extern int f();
return f() ? : -1; // Returns the result of f unless it is 0
See the CIL output for this
code fragment
- Computed jumps can be tricky. CIL compiles them away in a fairly clean
way but you are on your own if you try to jump into another function this way.
static void *jtab[2]; // A jump table
static int doit(int x){
static int jtab_init = 0;
if(!jtab_init) { // Initialize the jump table
jtab[0] = &&lbl1;
jtab[1] = &&lbl2;
jtab_init = 1;
}
goto *jtab[x]; // Jump through the table
lbl1:
return 0;
lbl2:
return 1;
}
int main(void){
if (doit(0) != 0) exit(1);
if (doit(1) != 1) exit(1);
exit(0);
}
See the CIL output for this
code fragment
- A cute little example that we made up. What is the returned value?
(Answer: 1);
return ({goto L; 0;}) && ({L: 5;});
See the CIL output for this
code fragment
- extern inline is a strange feature of GNU C. Can you guess what the
following code computes?
extern inline foo(void) { return 1; }
int firstuse(void) { return foo(); }
// A second, incompatible definition of foo
int foo(void) { return 2; }
int main() {
return foo() + firstuse();
}
See the CIL output for this
code fragment
The answer depends on whether the optimizations are turned on. If they are
then the answer is 3 (the first definition is inlined at all occurrences until
the second definition). If the optimizations are off, then the first
definition is ignore (treated like a prototype) and the answer is 4.
CIL will misbehave on this example, if the optimizations are turned off (it
always returns 3).
- GCC allows you to cast an object of a type T into a union as long as the
union has a field of that type:
union u {
int i;
struct s {
int i1, i2;
} s;
};
union u x = (union u)6;
int main() {
struct s y = {1, 2};
union u z = (union u)y;
}
See the CIL output for this
code fragment
- GCC allows you to use the __mode__ attribute to specify the size
of the integer instead of the standard char, short and so on:
int __attribute__ ((__mode__ ( __QI__ ))) i8;
int __attribute__ ((__mode__ ( __HI__ ))) i16;
int __attribute__ ((__mode__ ( __SI__ ))) i32;
int __attribute__ ((__mode__ ( __DI__ ))) i64;
See the CIL output for this
code fragment
- The “alias” attribute on a function declaration tells the
linker to treat this declaration as another name for the specified
function. CIL will replace the declaration with a trampoline
function pointing to the specified target.
static int bar(int x, char y) {
return x + y;
}
//foo is considered another name for bar.
int foo(int x, char y) __attribute__((alias("bar")));
See the CIL output for this
code fragment
16.3 Microsoft VC ugliness
This compiler has few extensions, so there is not much to say here.
-
Why does the following code return 0 and not -1? (Answer: because of a
bug in Microsoft Visual C. It thinks that the shift is unsigned just because
the second operator is unsigned. CIL reproduces this bug when in MSVC mode.)
return -3 >> (8 * sizeof(int));
- Unnamed fields in a structure seem really strange at first. It seems
that Microsoft Visual C introduced this extension, then GCC picked it up (but
in the process implemented it wrongly: in GCC the field y overlaps with
x!).
struct {
int x;
struct {
int y, z;
struct {
int u, v;
};
};
} a;
return a.x + a.y + a.z + a.u + a.v;
See the CIL output for this
code fragment
17 Authors
The CIL parser was developed starting from Hugues Casse's frontc
front-end for C although all the files from the frontc distribution have
been changed very extensively. The intermediate language and the elaboration
stage are all written from scratch. The main author is
George Necula, with significant
contributions from Scott McPeak,
Westley Weimer,
Ben Liblit,
Matt Harren,
Raymond To and Aman Bhargava.
This work is based upon work supported in part by the National Science
Foundation under Grants No. 9875171, 0085949 and 0081588, and gifts from
Microsoft Research. Any opinions, findings, and conclusions or recommendations
expressed in this material are those of the author(s) and do not necessarily
reflect the views of the National Science Foundation or the other sponsors.
18 License
Copyright (c) 2001-2005,
-
George C. Necula <necula@cs.berkeley.edu>
- Scott McPeak <smcpeak@cs.berkeley.edu>
- Wes Weimer <weimer@cs.berkeley.edu>
- Ben Liblit <liblit@cs.wisc.edu>
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
1. Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
3. The names of the contributors may not be used to endorse or promote
products derived from this software without specific prior written
permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
POSSIBILITY OF SUCH DAMAGE.
19 Bug reports
We are certain that there are still some remaining bugs in CIL. If you find
one please file a bug report in our Source Forge space
http://sourceforge.net/projects/cil.
You can find there the latest announcements, a source distribution,
bug report submission instructions and a mailing list: cil-users[at
sign]lists.sourceforge.net. Please use this list to ask questions about CIL,
as it will ensure your message is viewed by a broad audience.
20 Changes
-
May 20, 2006: Released version 1.3.5
- May 19, 2006: Makefile.cil.in/Makefile.cil have
been renamed Makefile.in/Makefile. And maincil.ml has
been renamed main.ml.
- May 18, 2006: Added a new module Cfg to compute the
control-flow graph. Unlike the older Cil.computeCFGInfo,
the new version does not modify the code.
- May 18, 2006: Added several new analyses: reaching
definitions, available expressions, liveness analysis, and dead code
elimination. See Section 8.
- May 2, 2006: Added a flag --noInsertImplicitCasts.
When this flag is used, CIL code will only include casts inserted by
the programmer. Implicit coercions are not changed to explicit casts.
- April 16, 2006: Minor improvements to the --stats
flag (Section 7.2). We now use Pentium performance
counters by default, if your processor supports them.
- April 10, 2006: Extended machdep.c to support
microcontroller compilers where the struct alignment of integer
types does not match the size of the type. Thanks to Nathan
Cooprider for the patch.
- April 6, 2006: Fix for global initializers of unions when
the union field being initialized is not the first one, and for
missing initializers of unions when the first field is not the
largest field.
- April 6, 2006: Fix for bitfields in the SFI module.
- April 6, 2006: Various fixes for gcc attributes.
packed, section, and always_inline attributes are now
parsed correctly. Also fixed printing of attributes on enum types.
- March 30, 2006: Fix for rmtemps.ml, which deletes
unused inline functions. When in gcc mode CIL now leaves all
inline functions in place, since gcc treats these as externally
visible.
- March 15, 2006: Fix for typeof(e) when e has type
void.
- March 3, 2006: Assume inline assembly instructions can
fall through for the purposes of adding return statements. Thanks to
Nathan Cooprider for the patch.
- February 27, 2006: Fix for extern inline functions when
the output of CIL is fed back into CIL.
- January 30, 2006: Fix parsing of switch without braces.
- January 30, 2006: Allow `$' to appear in identifiers.
- January 13, 2006: Added support for gcc's alias attribute
on functions. See Section 16.2, item 8.
- December 9, 2005: Christoph Spiel fixed the Golf and
Olf modules so that Golf can be used with the points-to analysis.
He also added performance fixes and cleaned up the documentation.
- December 1, 2005: Major rewrite of the ext/callgraph module.
- December 1, 2005: Preserve enumeration constants in CIL. Default
is the old behavior to replace them with integers.
- November 30, 2005: Added support for many GCC __builtin
functions.
- November 30, 2005: Added the EXTRAFEATURES configure
option, making it easier to add Features to the build process.
- November 23, 2005: In MSVC mode do not remove any locals whose name
appears as a substring in an inline assembly.
- November 23, 2005: Do not add a return to functions that have the
noreturn attribute.
- November 22, 2005: Released version 1.3.4
- November 21, 2005: Performance and correctness fixes for
the Points-to Analysis module. Thanks to Christoph Spiel for the
patches.
- October 5, 2005: CIL now builds on SPARC/Solaris. Thanks
to Nick Petroni and Remco van Engelen for the patches.
- September 26, 2005: CIL no longer uses the `-I-' flag
by default when preprocessing with gcc.
- August 24, 2005: Added a command-line option
“--forceRLArgEval” that forces function arguments to be evaluated
right-to-left. This is the default behavior in unoptimized gcc and
MSVC, but the order of evaluation is undefined when using
optimizations, unless you apply this CIL transformation. This flag
does not affect the order of evaluation of e.g. binary operators,
which remains undefined. Thanks to Nathan Cooprider for the patch.
- August 9, 2005: Fixed merging when there are more than 20
input files.
- August 3, 2005: When merging, it is now an error to
declare the same global variable twice with different initializers.
- July 27, 2005: Fixed bug in transparent unions.
- July 27, 2005: Fixed bug in collectInitializer. Thanks to
Benjamin Monate for the patch.
- July 26, 2005: Better support for extended inline assembly
in gcc.
- July 26, 2005: Added many more gcc __builtin* functions
to CIL. Most are treated as Call instructions, but a few are
translated into expressions so that they can be used in global
initializers. For example, “__builtin_offsetof(t, field)” is
rewritten as “&((t*)0)->field”, the traditional way of calculating
an offset.
- July 18, 2005: Fixed bug in the constant folding of shifts
when the second argument was negative or too large.
- July 18, 2005: Fixed bug where casts were not always
inserted in function calls.
- June 10, 2005: Fixed bug in the code that makes implicit
returns explicit. We weren't handling switch blocks correctly.
- June 1, 2005: Released version 1.3.3
- May 31, 2005: Fixed handling of noreturn attribute for function
pointers.
- May 30, 2005: Fixed bugs in the handling of constructors in gcc.
- May 30, 2005: Fixed bugs in the generation of global variable IDs.
- May 27, 2005: Reimplemented the translation of function calls so
that we can intercept some builtins. This is important for the uses of
__builtin_constant_p in constants.
- May 27, 2005: Export the plainCilPrinter, for debugging.
- May 27, 2005: Fixed bug with printing of const attribute for
arrays.
- May 27, 2005: Fixed bug in generation of type signatures. Now they
should not contain expressions anymore, so you can use structural equality.
This used to lead to Out_of_Memory exceptions.
- May 27, 2005: Fixed bug in type comparisons using
TBuiltin_va_list.
- May 27, 2005: Improved the constant folding in array lengths and
case expressions.
- May 27, 2005: Added the __builtin_frame_address to the set
of gcc builtins.
- May 27, 2005: Added the CIL project to SourceForge.
- April 23, 2005: The cattr field was not visited.
- March 6, 2005: Debian packaging support
- February 16, 2005: Merger fixes.
- February 11, 2005: Fixed a bug in --dopartial. Thanks to
Nathan Cooprider for this fix.
- January 31, 2005: Make sure the input file is closed even if a
parsing error is encountered.
- January 11, 2005: Released version 1.3.2
- January 11, 2005: Fixed printing of integer constants whose
integer kind is shorter than an int.
- January 11, 2005: Added checks for negative size arrays and arrays
too big.
- January 10, 2005: Added support for GCC attribute “volatile” for
tunctions (as a synonim for noreturn).
- January 10, 2005: Improved the comparison of array sizes when
comparing array types.
- January 10, 2005: Fixed handling of shell metacharacters in the
cilly command lione.
- January 10, 2005: Fixed dropping of cast in initialization of
local variable with the result of a function call.
- January 10, 2005: Fixed some structural comparisons that were
broken in the Ocaml 3.08.
- January 10, 2005: Fixed the unrollType function to not forget
attributes.
- January 10, 2005: Better keeping track of locations of function
prototypes and definitions.
- January 10, 2005: Fixed bug with the expansion of enumeration
constants in attributes.
- October 18, 2004: Fixed a bug in cabsvisit.ml. CIl would wrap a
BLOCK around a single atom unnecessarily.
- August 7, 2004: Released version 1.3.1
- August 4, 2004: Fixed a bug in splitting of structs using
--dosimplify
- July 29, 2004: Minor changes to the type typeSig (type signatures)
to ensure that they do not contain types, so that you can do structural
comparison without danger of nontermination.
- July 28, 2004: Ocaml version 3.08 is required. Numerous small
changes while porting to Ocaml 3.08.
- July 7, 2004: Released version 1.2.6
- July 2, 2004: Character constants such as 'c' should
have type int, not char. Added a utility function
Cil.charConstToInt that sign-extends chars greater than 128, if needed.
- July 2, 2004: Fixed a bug that was casting values to int
before applying the logical negation operator !. This caused
problems for floats, and for integer types bigger than int.
- June 13, 2004: Added the field sallstmts to a function
description, to hold all statements in the function.
- June 13, 2004: Added new extensions for data flow analyses, and
for computing dominators.
- June 10, 2004: Force initialization of CIL at the start of
Cabs2cil.
- June 9, 2004: Added support for GCC __attribute_used__
- April 7, 2004: Released version 1.2.5
- April 7, 2004: Allow now to run ./configure CC=cl and set the MSVC
compiler to be the default. The MSVC driver will now select the default name
of the .exe file like the CL compiler.
- April 7, 2004: Fixed a bug in the driver. The temporary files are
deleted by the Perl script before the CL compiler gets to them?
- April 7, 2004: Added the - form of arguments to the MSVC driver.
- April 7, 2004: Added a few more GCC-specific string escapes, (, [,
{, %, E.
- April 7, 2004: Fixed bug with continuation lines in MSVC.
- April 6, 2004: Fixed embarassing bug in the parser: the precedence
of casts and unary operators was switched.
- April 5, 2004: Fixed a bug involving statements mixed between
declarations containing initializers. Now we make sure that the initializers
are run in the proper order with respect to the statements.
- April 5, 2004: Fixed a bug in the merger. The merger was keeping
separate alpha renaming talbes (namespaces) for variables and types. This
means that it might end up with a type and a variable named the same way, if
they come from different files, which breaks an important CIL invariant.
- March 11, 2004 : Fixed a bug in the Cil.copyFunction function. The
new local variables were not getting fresh IDs.
- March 5, 2004: Fixed a bug in the handling of static function
prototypes in a block scope. They used to be renamed. Now we just consider
them global.
- February 20, 2004: Released version 1.2.4
- February 15, 2004: Changed the parser to allow extra semicolons
after field declarations.
- February 14, 2004: Changed the Errormsg functions: error, unimp,
bug to not raise an exception. Instead they just set Errormsg.hadErrors.
- February 13, 2004: Change the parsing of attributes to recognize
enumeration constants.
- February 10, 2004: In some versions of gcc the identifier
_{thread is an identifier and in others it is a keyword. Added code
during configuration to detect which is the case.
- January 7, 2004: Released version 1.2.3
- January 7, 2004: Changed the alpha renamer to be less
conservative. It will remember all versions of a name that were seen and will
only create a new name if we have not seen one.
- December 30, 2003 : Extended the cilly command to understand
better linker command options -lfoo.
- December 5, 2003: Added markup commands to the pretty-printer
module. Also, changed the “@<” left-flush command into “@''.
- December 4, 2003: Wide string literals are now handled
directly by Cil (rather than being exploded into arrays). This is
apparently handy for Microsoft Device Driver APIs that use intrinsic
functions that require literal constant wide-string arguments.
- December 3, 2003: Added support for structured exception handling
extensions for the Microsoft compilers.
- December 1, 2003: Fixed a Makefile bug in the generation of the
Cil library (e.g., cil.cma) that was causing it to be unusable. Thanks
to KEvin Millikin for pointing out this bug.
- November 26, 2003: Added support for linkage specifications
(extern “C”).
- November 26, 2003: Added the ocamlutil directory to contain some
utilities shared with other projects.
- November 25, 2003: Released version 1.2.2
- November 24, 2003: Fixed a bug that allowed a static local to
conflict with a global with the same name that is declared later in the
file.
- November 24, 2003: Removed the --keep option of the cilly
driver and replaced it with --save-temps.
- November 24, 2003: Added printing of what CIL features are being
run.
- November 24, 2003: Fixed a bug that resulted in attributes being
dropped for integer types.
- November 11, 2003: Fixed a bug in the visitor for enumeration
definitions.
- October 24, 2003: Fixed a problem in the configuration script. It
was not recognizing the Ocaml version number for beta versions.
- October 15, 2003: Fixed a problem in version 1.2.1 that was
preventing compilation on OCaml 3.04.
- September 17, 2003: Released version 1.2.1.
- September 7, 2003: Redesigned the interface for choosing
#line directive printing styles. Cil.printLn and
Cil.printLnComment have been merged into Cil.lineDirectiveStyle.
- August 8, 2003: Do not silently pad out functions calls with
arguments to match the prototype.
- August 1, 2003: A variety of fixes suggested by Steve Chamberlain:
initializers for externs, prohibit float literals in enum, initializers for
unsized arrays were not working always, an overflow problem in Ocaml, changed
the processing of attributes before struct specifiers
- July 14, 2003: Add basic support for GCC's "__thread" storage
qualifier. If given, it will appear as a "thread" attribute at the top of the
type of the declared object. Treatment is very similar to "__declspec(...)"
in MSVC
- July 8, 2003: Fixed some of the __alignof computations. Fixed
bug in the designated initializers for arrays (Array.get error).
- July 8, 2003: Fixed infinite loop bug (Stack Overflow) in the
visitor for __alignof.
- July 8, 2003: Fixed bug in the conversion to CIL. A function or
array argument of
the GCC __typeof() was being converted to pointer type. Instead, it should
be left alone, just like for sizeof.
- July 7, 2003: New Escape module provides utility functions
for escaping characters and strings in accordance with C lexical
rules.
- July 2, 2003: Relax CIL's rules for when two enumeration types are
considered compatible. Previously CIL considered two enums to be compatible if
they were the same enum. Now we follow the C99 standard.
- June 28, 2003: In the Formatparse module, Eric Haugh found and
fixed a bug in the handling of lvalues of the form “lv->field.more”.
- June 28, 2003: Extended the handling of gcc command lines
arguments in the Perl scripts.
- June 23, 2003: In Rmtmps module, simplified the API for
customizing the root set. Clients may supply a predicate that
returns true for each root global. Modifying various
“referenced” fields directly is no longer supported.
- June 17, 2003: Reimplement internal utility routine
Cil.escape_char. Faster and better.
- June 14, 2003: Implemented support for __attribute__s
appearing between "struct" and the struct tag name (also for unions and
enums), since gcc supports this as documented in section 4.30 of the gcc
(2.95.3) manual
- May 30, 2003: Released the regression tests.
- May 28, 2003: Released version 1.1.2
- May 26, 2003: Add the simplify module that compiles CIL
expressions into simpler expressions, similar to those that appear in a
3-address intermediate language.
- May 26, 2003: Various fixes and improvements to the pointer
analysis modules.
- May 26, 2003: Added optional consistency checking for
transformations.
- May 25, 2003: Added configuration support for big endian machines.
Now Cil.little_endian can be used to test whether the machine is
little endian or not.
- May 22, 2003: Fixed a bug in the handling of inline functions. The
CIL merger used to turn these functions into “static”, which is incorrect.
- May 22, 2003: Expanded the CIL consistency checker to verify
undesired sharing relationships between data structures.
- May 22, 2003: Fixed bug in the oneret CIL module: it was
mishandling certain labeled return statements.
- May 5, 2003: Released version 1.0.11
- May 5, 2003: OS X (powerpc/darwin) support for CIL. Special
thanks to Jeff Foster, Andy Begel and Tim Leek.
- April 30, 2003: Better description of how to use CIL for your
analysis.
- April 28, 2003: Fixed a bug with --dooneRet and
--doheapify. Thanks, Manos Renieris.
- April 16, 2003: Reworked management of
temporary/intermediate output files in Perl driver scripts. Default
behavior is now to remove all such files. To keep intermediate
files, use one of the following existing flags:
-
--keepmerged for the single-file merge of all sources
- --keep=<dir> for various other CIL and
CCured output files
- --save-temps for various gcc intermediate files; MSVC
has no equivalent option
As part of this change, some intermediate files have changed their
names slightly so that new suffixes are always preceded by a
period. For example, CCured output that used to appear in
“foocured.c” now appears in “foo.cured.c”.
- April 7, 2003: Changed the representation of the Cil.GVar
global constructor. Now it is possible to update the initializer without
reconstructing the global (which in turn it would require reconstructing the
list of globals that make up a program). We did this because it is often
tempting to use Cil.visitCilFileSameGlobals and the Cil.GVar
was the only global that could not be updated in place.
- April 6, 2003: Reimplemented parts of the cilly.pl script to make
it more robust in the presence of complex compiler arguments.
- March 10, 2003: Released version 1.0.9
- March 10, 2003: Unified and documented a large number of CIL
Library Modules: oneret, simplemem, makecfg, heapify, stackguard, partial.
Also documented the main client interface for the pointer analysis.
- February 18, 2003: Fixed a bug in logwrites that was causing it
to produce invalid C code on writes to bitfields. Thanks, David Park.
- February 15, 2003: Released version 1.0.8
- February 15, 2003: PDF versions of the manual and API are
available for those who would like to print them out.
- February 14, 2003: CIL now comes bundled with alias analyses.
- February 11, 2003: Added support for adding/removing options from
./configure.
- February 3, 2003: Released version 1.0.7
- February 1, 2003: Some bug fixes in the handling of variable
argument functions in new versions of gcc And glibc.
- January 29, 2003: Added the logical AND and OR operators.
Exapanded the translation to CIL to handle more complicated initializers
(including those that contain logical operators).
- January 28, 2003: Released version 1.0.6
- January 28, 2003: Added support for the new handling of
variable-argument functions in new versions of glibc.
- January 19, 2003: Added support for declarations in interpreted
constructors. Relaxed the semantics of the patterns for variables.
- January 17, 2003: Added built-in prototypes for the gcc built-in
functions. Changed the pGlobal method in the printers to print the
carriage return as well.
- January 9, 2003: Reworked lexer and parser's strategy for
tracking source file names and line numbers to more closely match
typical native compiler behavior. The visible CIL interface is
unchanged.
- January 9, 2003: Changed the interface to the alpha convertor. Now
you can pass a list where it will record undo information that you can use to
revert the changes that it makes to the scope tables.
- January 6, 2003: Released version 1.0.5
- January 4, 2003: Changed the interface for the Formatcil module.
Now the placeholders in the pattern have names. Also expanded the
documentation of the Formatcil module.
Now the placeholders in the pattern have names.
- January 3, 2003: Extended the rmtmps module to also remove
unused labels that are generated in the conversion to CIL. This reduces the
number of warnings that you get from cgcc afterwards.
- December 17, 2002: Fixed a few bugs in CIL related to the
representation of string literals. The standard says that a string literal
is an array. In CIL, a string literal has type pointer to character. This is
Ok, except as an argument of sizeof. To support this exception, we have
added to CIL the expression constructor SizeOfStr. This allowed us to fix
bugs with computing sizeof("foo bar") and sizeof((char*)"foo bar")
(the former is 8 and the latter is 4).
- December 8, 2002: Fixed a few bugs in the lexer and parser
relating to hex and octal escapes in string literals. Also fixed
the dependencies between the lexer and parser.
- December 5, 2002: Fixed visitor bugs that were causing
some attributes not to be visited and some queued instructions to be
dropped.
- December 3, 2002: Added a transformation to catch stack
overflows. Fixed the heapify transformation.
- October 14, 2002: CIL is now available under the BSD license
(see the License section or the file LICENSE). Released version 1.0.4
- October 9, 2002: More FreeBSD configuration changes, support
for the GCC-ims __signed and __volatile. Thanks to Axel
Simon for pointing out these problems. Released version 1.0.3
- October 8, 2002: FreeBSD configuration and porting fixes.
Thanks to Axel Simon for pointing out these problems.
- September 10, 2002: Fixed bug in conversion to CIL. Now we drop
all “const” qualifiers from the types of locals, even from the fields of
local structures or elements of arrays.
- September 7, 2002: Extended visitor interface to distinguish visitng
offsets inside lvalues from offsets inside initializer lists.
- September 7, 2002: Released version 1.0.1
- September 6, 2002: Extended the patcher with the ateof flag.
- September 4, 2002: Fixed bug in the elaboration to CIL. In some
cases constant folding of || and && was computed wrong.
- September 3, 2002: Fixed the merger documentation.
- August 29, 2002: Released version 1.0.0.
- August 29, 2002: Started numbering versions with a major nubmer,
minor and revisions. Released version 1.0.0.
- August 25, 2002: Fixed the implementation of the unique
identifiers for global variables and composites. Now those identifiers are
globally unique.
- August 24, 2002: Added to the machine-dependent configuration the
sizeofvoid. It is 1 on gcc and 0 on MSVC. Extended the implementation of
Cil.bitsSizeOf to handle this (it was previously returning an error when
trying to compute the size of void).
- August 24, 2002: Changed the representation of structure and
unions to distinguish between undefined structures and those that are defined
to be empty (allowed on gcc). The sizeof operator is undefined for the former
and returns 0 for the latter.
- August 22, 2002: Apply a patch from Richard H. Y. to support
FreeBSD installations. Thanks, Richard!
- August 12, 2002: Fixed a bug in the translation of wide-character
strings. Now this translation matches that of the underlying compiler. Changed
the implementation of the compiler dependencies.
- May 25, 2002: Added interpreted constructors and destructors.
- May 17, 2002: Changed the representation of functions to move the
“inline” information to the varinfo. This way we can print the “inline”
even in declarations which is what gcc does.
- May 15, 2002: Changed the visitor for initializers to make two
tail-recursive passes (the second is a List.rev and only done if one of
the initializers change). This prevents Stack_Overflow for large
initializers. Also improved the processing of initializers when converting to
CIL.
- May 15, 2002: Changed the front-end to allow the use of MSVC
mode even on machines that do not have MSVC. The machine-dependent parameters
for GCC will be used in that case.
- May 11, 2002: Changed the representation of formals in function
types. Now the function type is purely functional.
- May 4, 2002: Added the function
Cil.visitCilFileSameGlobals and changed Cil.visitCilFile to be
tail recursive. This prevents stack overflow on huge files.
- February 28, 2002: Changed the significance of the
CompoundInit in Cil.init to allow for missing initializers at the
end of an array initializer. Added the API function
Cil.foldLeftCompoundAll.
This document was translated from LATEX by
HEVEA.