2. Writing a New Valgrind Tool

Tools must define various functions for instrumenting programs that are called by Valgrind's core. They are then linked against Valgrind's core to define a complete Valgrind tool which will be used when the --tool option is used to select it.

To write your own tool, you'll need the Valgrind source code. You'll need a check-out of the Subversion repository for the automake/autoconf build instructions to work. See the information about how to do check-out from the repository at the Valgrind website.

Valgrind uses GNU automake and autoconf for the creation of Makefiles and configuration. But don't worry, these instructions should be enough to get you started even if you know nothing about those tools.

In what follows, all filenames are relative to Valgrind's top-level directory valgrind/.

These steps don't have to be followed exactly -- you can choose different names for your source files, and use a different --prefix for ./configure.

Now that we've setup, built and tested the simplest possible tool, onto the interesting stuff...

A tool must define at least these four functions:

  pre_clo_init()
  post_clo_init()
  instrument()
  fini()

The names can be different to the above, but these are the usual names. The first one is registered using the macro VG_DETERMINE_INTERFACE_VERSION. The last three are registered using the VG_(basic_tool_funcs) function.

In addition, if a tool wants to use some of the optional services provided by the core, it may have to define other functions and tell the core about them.

Most of the initialisation should be done in pre_clo_init. Only use post_clo_init if a tool provides command line options and must do some initialisation after option processing takes place ("clo" stands for "command line options").

First of all, various "details" need to be set for a tool, using the functions VG_(details_*). Some are all compulsory, some aren't. Some are used when constructing the startup message, detail_bug_reports_to is used if VG_(tool_panic) is ever called, or a tool assertion fails. Others have other uses.

Second, various "needs" can be set for a tool, using the functions VG_(needs_*). They are mostly booleans, and can be left untouched (they default to False). They determine whether a tool can do various things such as: record, report and suppress errors; process command line options; wrap system calls; record extra information about heap blocks; etc.

For example, if a tool wants the core's help in recording and reporting errors, it must call VG_(needs_tool_errors) and provide definitions of eight functions for comparing errors, printing out errors, reading suppressions from a suppressions file, etc. While writing these functions requires some work, it's much less than doing error handling from scratch because the core is doing most of the work.

Third, the tool can indicate which events in core it wants to be notified about, using the functions VG_(track_*). These include things such as heap blocks being allocated, the stack pointer changing, a mutex being locked, etc. If a tool wants to know about this, it should provide a pointer to a function, which will be called when that event happens.

For example, if the tool want to be notified when a new heap block is allocated, it should call VG_(track_new_mem_heap) with an appropriate function pointer, and the assigned function will be called each time this happens.

More information about "details", "needs" and "trackable events" can be found in include/pub_tool_tooliface.h.

instrument is the interesting one. It allows you to instrument VEX IR, which is Valgrind's RISC-like intermediate language. VEX IR is described in the comments of the header file VEX/pub/libvex_ir.h.

The easiest way to instrument VEX IR is to insert calls to C functions when interesting things happen. See the tool "Lackey" (lackey/lk_main.c) for a simple example of this, or Cachegrind (cachegrind/cg_main.c) for a more complex example.

This is where you can present the final results, such as a summary of the information collected. Any log files should be written out at this point.

Please note that the core/tool split infrastructure is quite complex and not brilliantly documented. Here are some important points, but there are undoubtedly many others that I should note but haven't thought of.

The files include/pub_tool_*.h contain all the types, macros, functions, etc. that a tool should (hopefully) need, and are the only .h files a tool should need to #include. They have a reasonable amount of documentation in it that should hopefully be enough to get you going.

Note that you can't use anything from the C library (there are deep reasons for this, trust us). Valgrind provides an implementation of a reasonable subset of the C library, details of which are in pub_tool_libc*.h.

When writing a tool, in theory you shouldn't need to look at any of the code in Valgrind's core, but in practice it might be useful sometimes to help understand something.

The include/pub_tool_basics.h and VEX/pub/libvex_basictypes.h files have some basic types that are widely used.

Ultimately, the tools distributed (Memcheck, Cachegrind, Lackey, etc.) are probably the best documentation of all, for the moment.

The VG_ macro is used heavily. This just prepends a longer string in front of names to avoid potential namespace clashes. It is defined in include/pub_tool_basics.h.

There are some assorted notes about various aspects of the implementation in docs/internals/. Much of it isn't that relevant to tool-writers, however.

Writing and debugging tools is not trivial. Here are some suggestions for solving common problems.

If you are getting segmentation faults in C functions used by your tool, the usual GDB command:

  gdb <prog> core

usually gives the location of the segmentation fault.

If you want to debug C functions used by your tool, there are instructions on how to do so in the file README_DEVELOPERS.

If you are having problems with your VEX IR instrumentation, it's likely that GDB won't be able to help at all. In this case, Valgrind's --trace-flags option is invaluable for observing the results of instrumentation.

If you just want to know whether a program point has been reached, using the OINK macro (in include/pub_tool_libcprint.h) can be easier than using GDB.

The other debugging command line options can be useful too (run valgrind --help-debug for the list).

If your tool reports errors and you want to suppress some common ones, you can add suppressions to the suppression files. The relevant files are *.supp; the final suppression file is aggregated from these files by combining the relevant .supp files depending on the versions of linux, X and glibc on a system.

Suppression types have the form tool_name:suppression_name. The tool_name here is the name you specify for the tool during initialisation with VG_(details_name).

If you are feeling conscientious and want to write some documentation for your tool, please use XML as the rest of Valgrind does. The file docs/README has more details on getting the XML toolchain to work; this can be difficult, unfortunately.

To write the documentation, follow these steps (using foobar as the example tool name again):

Valgrind has some support for regression tests. If you want to write regression tests for your tool:

Lots of profiling tools have trouble running Valgrind. For example, trying to use gprof is hopeless.

Probably the best way to profile a tool is with OProfile on Linux.

You can also use Cachegrind on it. Read README_DEVELOPERS for details on running Valgrind under Valgrind; it's a bit fragile but can usually be made to work.

If you add any directories under foobar/, you will need to add an appropriate Makefile.am to it, and add a corresponding entry to the AC_OUTPUT list in configure.in.

If you add any scripts to your tool (see Cachegrind for an example) you need to add them to the bin_SCRIPTS variable in foobar/Makefile.am and possible also to the AC_OUTPUT list in configure.in.

The core/tool interface evolves over time, but it's pretty stable. We deliberately do not provide backward compatibility with old interfaces, because it is too difficult and too restrictive. We view this as a good thing -- if we had to be backward compatible with earlier versions, many improvements now in the system could not have been added.

Because tools are statically linked with the core, if a tool compiles successfully then it should be compatible with the core. We would not deliberately violate this property by, for example, changing the behaviour of a core function without changing its prototype.