[llvm-dev] RFC: Comprehensive Static Instrumentation

TB Schardl via llvm-dev llvm-dev at lists.llvm.org
Thu Jun 16 17:58:18 PDT 2016

Hey John,

Thanks for your comments.  I've CC'd the CSI mailing list with your
comments and responded inline.  Please let me know if you have further


On Thu, Jun 16, 2016 at 3:51 PM, John Criswell <jtcriswel at gmail.com> wrote:

> Dear TB,
> Comments inline below.
> On 6/16/16 11:01 AM, TB Schardl via llvm-dev wrote:
> Hey LLVM-dev,
> We propose to build the CSI framework to provide a comprehensive suite of
> compiler-inserted instrumentation hooks that dynamic-analysis tools can use
> to observe and investigate program runtime behavior.  Traditionally, tools
> based on compiler instrumentation would each separately modify the compiler
> to insert their own instrumentation.  In contrast, CSI inserts a standard
> collection of instrumentation hooks into the program-under-test.  Each
> CSI-tool is implemented as a library that defines relevant hooks, and the
> remaining hooks are "nulled" out and elided during link-time optimization
> (LTO), resulting in instrumented runtimes on par with custom
> instrumentation.  CSI allows many compiler-based tools to be written as
> simple libraries without modifying the compiler, greatly lowering the bar
> for
> developing dynamic-analysis tools.
> Can you clarify the scope of tools that you want to develop?  Are these
> profiling tools, security enforcement tools, debugging tools, etc?  The
> type of tools you want to build will dictate whether such a framework makes
> sense.

For the first version of CSI, we've looked at tools including memory
checkers, race detectors, performance profilers, call-graph generators,
cache analyzers, and code-coverage analyzers.  Our experience working with
and developing such tools has shown us that these tools benefit from
instrumentation at the IR level, which is where we have currently targeted
CSI.  One benefit of targeting the IR is that CSI tools are thus
source-language and machine-architecture independent.  For future versions
of CSI, we are interested in including instrumentation in the front end and
back end.

> ================
> Motivation
> ================
> Key to understanding and improving the behavior of any system is
> visibility -- the ability to know what is going on inside the system.
> Various dynamic-analysis tools, such as race detectors, memory checkers,
> cache simulators, call-graph generators, code-coverage analyzers, and
> performance profilers, rely on compiler instrumentation to gain visibility
> into the program behaviors during execution.  With this approach, the tool
> writer modifies the compiler to insert instrumentation code into the
> program-under-test so that it can execute behind the scene while the
> program-under-test runs.  This approach, however, means that the
> development of new tools requires compiler work, which many potential tool
> writers are ill equipped to do, and thus raises the bar for building new
> and innovative tools.
> The goal of the CSI framework is to provide comprehensive static
> instrumentation through the compiler, in order to simplify the task of
> building efficient and effective platform-independent tools.  The CSI
> framework allows the tool writer to easily develop analysis tools that
> require
> compiler instrumentation without needing to understand the compiler
> internals or modifying the compiler, which greatly lowers the bar for
> developing dynamic-analysis tools.
> ================
> Approach
> ================
> The CSI framework inserts instrumentation hooks at salient locations
> throughout the compiled code of a program-under-test, such as function
> entry and exit points, basic-block entry and exit points, before and after
> each memory operation, etc.  Tool writers can instrument a
> program-under-test simply by first writing a library that defines the
> semantics of relevant hooks
> and then statically linking their compiled library with the
> program-under-test.
> At first glance, this brute-force method of inserting hooks at every
> salient location in the program-under-test seems to be replete with
> overheads.  CSI overcomes these overheads through the use of
> link-time-optimization (LTO), which is now readily available in most major
> compilers, including GCC and LLVM.  Using LTO, instrumentation hooks that
> are not used by a particular tool can be elided, allowing the overheads of
> these hooks to be avoided when the
> instrumented program-under-test is run.
> The algorithms for optimizing away run-time hooks is not necessarily
> uniform.  For example, if a tool instruments loads and stores to collect
> the set of memory locations accessed by a program, then optimizing away a
> redundant check on a store is okay.  If the instrumentation is meant to
> enforce memory safety, then redundant checks can only be optimized away if
> there is no intervening call to free() between the two checks (which may
> require inter-procedural analysis to determine).  In such a case, you
> either need to be very pessimistic about the optimizations that you use, or
> you will get incorrect optimizations for certain classes of dynamic
> analyses.

The safety of performing certain optimizations on an instrumented program
depends on the tool, as you describe, and exists regardless of whether the
tool is implemented with CSI or custom compiler instrumentation.  For
example, if the compiler implements a conditional in the source code with a
conditional move instruction, then a code-coverage tool will struggle to
determine whether both branches of the conditional were executed in a given
run of the program.  By leveraging LTO as it currently exists in LLVM, CSI
makes standard compiler analyses and optimizations available to tool
writers to optimize their tools.  As with any tool writer that uses
compiler instrumentation, tool writers that use CSI must determine what
optimizations tool users can safely perform on instrumented programs.

For some tool-specific optimizations, tool-writers can leverage the
properties passed to instrumentation hooks, which store the results of
commonly used compiler analysis.  For example, although a tool might simply
instrument every load or store, the same tool might benefit from ignoring
certain loads or stores.  If the property bit is available in CSI, that
tool could enjoy this optimization by implementing a load hook like this:

__csi_before_load(const csi_id_t load_id, const void *addr,
                  const int32_t num_bytes, const uint64_t prop) {
  process_load(addr, num_bytes);

Because prop is a compile-time constant, LTO is capable of inlining this
instrumentation and constant-folding prop to elide the instrumentation when
the condition is true.  This optimization is not available to tool writers
if the bit is not set in the property, of course.  Our plan is to add
information to these properties gradually based on general demand.

CSI does not provide as much flexibility as a custom LLVM pass, which can
perform arbitrary compiler analysis and optimization.  CSI trades off this
flexibility to dramatically simplify the task of writing many dynamic
analysis tools.  This is the 90-10 tradeoff that we opt into, and we feel
that making life easy for many would-be tool writers is worth the effort,
even if we can't quite cover everyone's needs.  Moreover, we see value in
CSI as a framework for prototyping tools; once a tool writer has written a
correct tool, for instance, then she can decide whether the tool needs more
performance and can take on writing a custom LLVM pass.

> Furthermore, LTO can optimize a tool's instrumentation within a program
> using traditional compiler optimizations.  Our initial study indicates that
> the use of LTO does not unduly slow down the build time, and the LTO can
> indeed optimize away unused hooks.  One of our experiments with Apache HTTP
> server shows that, compiling with CSI and linking with the "null" CSI-tool
> (which consists solely of empty hooks) slows down the build time of the
> Apache HTTP server by less than 40%, and the resulting tool-instrumented
> executable is as fast as the original uninstrumented code.
> Is your intention to have a compiler flag that enables insertions of
> hooks, or are you planning on having a pass always add the hooks and having
> libLTO always remove them?  I assume the former, but you should probably
> clarify.

With CSI's current design, we've added a single compiler flag that inserts
all of these hooks.  Although CSI can leverage LTO to remove unused
instrumentation, we don't rely on LTO for correctness.

> ================
> CSI version 1
> ================
> The initial version of CSI supports a basic set of hooks that covers the
> following categories of program objects: functions, function exits
> (specifically, returns), basic blocks, call sites, loads, and stores.
> Don't forget that atomic instructions (e.g., compare-and-swap) and some of
> the intrinsics (e.g., llvm.memcpy()) also access memory.

Addressing these instructions is on our radar for future versions of CSI.

> We prioritized instrumenting these IR objects based on the need of seven
> example CSI tools, including a race detector, a cache-reuse analyzer, and a
> code-coverage analyzer.  We plan to evolve the CSI API over time to be more
> comprehensive, and we have designed the CSI API to be extensible, allowing
> new instrumentation to be added as needs grow.  We chose to initially
> implement a minimal "core" set of hooks, because we felt it was best to add
> new instrumentation on an as-needed basis in order to keep the interface
> simple.
> There are three salient features about the design of CSI.  First, CSI
> assigns each instrumented program object a unique integer identifier within
> one of the (currently) six program-object categories.  Within each
> category, the ID's are consecutively numbered from 0 up to the number of
> such objects minus 1.  The contiguous assignment of the ID's allows the
> tool writer to easily keep track of IR objects in the program and iterate
> through all objects in a category (whether the object is encountered during
> execution or not).  Second, CSI provides a platform-independent means to
> relate a given program object to locations in the source code.
> Specifically, CSI provides "front-end-data (FED)" tables, which provide
> file name and source lines for each program object given the object's ID.
> Third, each CSI hook takes in as a parameter a "property": a 64-bit
> unsigned integer that CSI uses to export the results of compiler analyses
> and other information known at compile time.  The use of properties allow
> the tool to rely on compiler analyses to optimize instrumentation and
> decrease overhead.  In particular, since the value of a property is known
> at compile time, LTO can constant-fold the conditional test around a
> property to elide unnecessary instrumentation.
> ================
> Future plan
> ================
> We plan to expand CSI in future versions by instrumenting additional
> program objects, such as atomic instructions, floating-point instructions,
> and exceptions.  We are also planning to provide additional static
> information to tool writers, both through information encoded in the
> properties passed to hooks and by other means.  In particular, we are also
> looking at mechanisms to present tool writers with more complex static
> information, such as how different program objects relate to each other,
> e.g., which basic blocks belong to a given function.
> As an aside, I'm not sure that I buy the idea that tool developers should
> be oblivious to the compiler internals.  If a tool developer doesn't
> understand what the compiler is doing, then she/he may not understand the
> results of the output.  For example, LLVM load/stores do not include stack
> spill slots, memory accesses incurred by calling conventions, etc.
> Depending on where the instrumentation passes are placed in the pass
> pipeline, instrumentation calls can be moved or removed (perhaps in
> undesirable ways for some dynamic analysis applications).

> I would also argue that a key design feature of LLVM is to make writing
> such passes simple, and I think LLVM accomplishes this.  If one understands
> how to build an efficient dynamic analysis, one can probably handle writing
> the compiler passes.

I personally think there's still a significant difference between
understanding what a compiler does conceptually (e.g., "What is a basic
block?") and being able to manipulate the codebase of a mainstream compiler
(e.g., "How does LLVM represent basic blocks, and how do I add instructions
to them?").  Although LLVM is certainly easier to work with than other
compiler codebases, CSI makes writing dynamic analysis tools even easier,
equivalent to writing a C library.  We've encountered several researchers
interested in writing dynamic analysis tools who are daunted by the task of
learning LLVM.

A common instrumentation infrastructure such as CSI has other benefits as
well.  Providing a single compiler hook to provide instrumentation for many
different tools reducers clutter in the compiler's codebase.  Furthermore,
tool-writers have an easier time distributing their tool if they don't have
to worry about getting users to use their custom compiler or to get their
changes upstreamed.

> Overall, I think having common instrumentation infrastructure is a good
> thing.  However, I'm not sure how common it can be across different
> applications of instrumentation.  As an example, most memory safety
> solutions have the same instrumentation and optimization needs (and
> constraints on optimization) regardless of how they implement the checks on
> loads and stores.  However, it's less clear to me that a race detector and
> a memory safety compiler can safely use the same optimizations.  You may
> find yourself implementing a common infrastructure with each tool
> implementing specialized optimizations to make each type of dynamic
> analysis really fast.
> Food for thought,
> John Criswell
> _______________________________________________
> LLVM Developers mailing listllvm-dev at lists.llvm.orghttp://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
> --
> John Criswell
> Assistant Professor
> Department of Computer Science, University of Rochesterhttp://www.cs.rochester.edu/u/criswell
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20160616/b61a1031/attachment.html>

More information about the llvm-dev mailing list