gperftools/INSTALL

Copyright 1994, 1995, 1996, 1999, 2000, 2001, 2002 Free Software
Foundation, Inc.

   This file is free documentation; the Free Software Foundation gives
unlimited permission to copy, distribute and modify it.


Perftools-Specific Install Notes
================================

*** NOTE FOR 64-BIT LINUX SYSTEMS

The glibc built-in stack-unwinder on 64-bit systems has some problems
with the perftools libraries.  (In particular, the cpu/heap profiler
may be in the middle of malloc, holding some malloc-related locks when
they invoke the stack unwinder.  The built-in stack unwinder may call
malloc recursively, which may require the thread to acquire a lock it
already holds: deadlock.)

For that reason, if you use a 64-bit system, we strongly recommend you
install libunwind before trying to configure or install google
perftools.  libunwind can be found at

   http://download.savannah.gnu.org/releases/libunwind/libunwind-0.99-beta.tar.gz

Even if you already have libunwind installed, you should check the
version.  Versions older than this will not work properly; too-new
versions introduce new code that does not work well with perftools
(because libunwind can call malloc, which will lead to deadlock).

CAUTION: if you install libunwind from the url above, be aware that
you may have trouble if you try to statically link your binary with
perftools: that is, if you link with 'gcc -static -lgcc_eh ...'.  This
is because both libunwind and libgcc implement the same C++ exception
handling APIs, but they implement them differently on some platforms.
This is not likely to be a problem on ia64, but may be on x86-64.

Also, if you link binaries statically, make sure that you add 
-Wl,--eh-frame-hdr to your linker options. This is required so that
libunwind can find the information generated by the compiler required
for stack unwinding.

Using -static is rare, though, so unless you know this will affect you
it probably won't.

If you cannot or do not wish to install libunwind, you can still try
to use the built-in stack unwinder.  The built-in stack unwinder
requires that your application, the tcmalloc library, and system
libraries like libc, all be compiled with a frame pointer.  This is
*not* the default for x86-64.

If you are on x86-64 system, know that you have a set of system
libraries with frame-pointers enabled, and compile all your
applications with -fno-omit-frame-pointer, then you can enable the
built-in perftools stack unwinder by passing the
--enable-frame-pointers flag to configure.

Even with the use of libunwind, there are still known problems with
stack unwinding on 64-bit systems, particularly x86-64.  See the
"64-BIT ISSUES" section in README.

If you encounter problems, try compiling perftools with './configure
--enable-frame-pointers'.  Note you will need to compile your
application with frame pointers (via 'gcc -fno-omit-frame-pointer
...') in this case.


*** TCMALLOC LARGE PAGES: TRADING TIME FOR SPACE

You can set a compiler directive that makes tcmalloc faster, at the
cost of using more space (due to internal fragmentation).

Internally, tcmalloc divides its memory into "pages."  The default
page size is chosen to minimize memory use by reducing fragmentation.
The cost is that keeping track of these pages can cost tcmalloc time.
We've added a new, experimental flag to tcmalloc that enables a larger
page size.  In general, this will increase the memory needs of
applications using tcmalloc.  However, in many cases it will speed up
the applications as well, particularly if they allocate and free a lot
of memory.  We've seen average speedups of 3-5% on Google
applications.

This feature is still very experimental; it's not even a configure
flag yet.  To build libtcmalloc with large pages, run

   ./configure <normal flags> CXXFLAGS=-DTCMALLOC_LARGE_PAGES

(or add -DTCMALLOC_LARGE_PAGES to your existing CXXFLAGS argument).


*** SMALL TCMALLOC CACHES: TRADING SPACE FOR TIME

You can set a compiler directive that makes tcmalloc use less memory
for overhead, at the cost of some time.

Internally, tcmalloc keeps information about some of its internal data
structures in a cache.  This speeds memory operations that need to
access this internal data.  We've added a new, experimental flag to
tcmalloc that reduces the size of this cache, decresaing the memory
needs of applications using tcmalloc.

This feature is still very experimental; it's not even a configure
flag yet.  To build libtcmalloc with smaller internal caches, run

   ./configure <normal flags> CXXFLAGS=-DTCMALLOC_SMALL_BUT_SLOW

(or add -DTCMALLOC_SMALL_BUT_SLOW to your existing CXXFLAGS argument).


*** NOTE FOR ___tls_get_addr ERROR

When compiling perftools on some old systems, like RedHat 8, you may
get an error like this:
    ___tls_get_addr: symbol not found

This means that you have a system where some parts are updated enough
to support Thread Local Storage, but others are not.  The perftools
configure script can't always detect this kind of case, leading to
that error.  To fix it, just comment out the line
   #define HAVE_TLS 1
in your config.h file before building.


*** TCMALLOC AND DLOPEN

To improve performance, we use the "initial exec" model of Thread
Local Storage in tcmalloc.  The price for this is the library will not
work correctly if it is loaded via dlopen().  This should not be a
problem, since loading a malloc-replacement library via dlopen is
asking for trouble in any case: some data will be allocated with one
malloc, some with another.  If, for some reason, you *do* need to use
dlopen on tcmalloc, the easiest way is to use a version of tcmalloc
with TLS turned off; see the ___tls_get_addr note above.


*** COMPILING ON NON-LINUX SYSTEMS

Perftools has been tested on the following systems:
   FreeBSD 6.0 (x86)
   Linux Fedora Core 3 (x86)
   Linux Fedora Core 4 (x86)
   Linux Fedora Core 5 (x86)
   Linux Fedora Core 6 (x86)
   Linux Ubuntu 6.06.1 (x86)
   Linux Ubuntu 6.06.1 (x86_64)
   Linux RedHat 9 (x86)
   Linux Debian 4.0 (PPC)
   Mac OS X 10.3.9 (Panther) (PowerPC)
   Mac OS X 10.4.8 (Tiger) (PowerPC)
   Mac OS X 10.4.8 (Tiger) (x86)
   Mac OS X 10.5 (Leopard) (x86)
   Solaris 10 (x86)
   Windows XP, Visual Studio 2003 (VC++ 7) (x86)
   Windows XP, Visual Studio 2005 (VC++ 8) (x86)
   Windows XP, MinGW 5.1.3 (x86)
   Windows XP, Cygwin 5.1 (x86)

It works in its full generality on the Linux systems
tested (though see 64-bit notes above).  Portions of perftools work on
the other systems.  The basic memory-allocation library,
tcmalloc_minimal, works on all systems.  The cpu-profiler also works
fairly widely.  However, the heap-profiler and heap-checker are not
yet as widely supported.  In general, the 'configure' script will
detect what OS you are building for, and only build the components
that work on that OS.

Note that tcmalloc_minimal is perfectly usable as a malloc/new
replacement, so it is possible to use tcmalloc on all the systems
above, by linking in libtcmalloc_minimal.

** FreeBSD:

   The following binaries build and run successfully (creating
   libtcmalloc_minimal.so and libprofile.so in the process):
      % ./configure
      % make tcmalloc_minimal_unittest tcmalloc_minimal_large_unittest \
             addressmap_unittest atomicops_unittest frag_unittest \
             low_level_alloc_unittest markidle_unittest memalign_unittest \
             packed_cache_test stacktrace_unittest system_alloc_unittest \
             thread_dealloc_unittest profiler_unittest.sh
      % ./tcmalloc_minimal_unittest    # to run this test
      % [etc]                          # to run other tests

   Three caveats: first, frag_unittest tries to allocate 400M of memory,
   and if you have less virtual memory on your system, the test may
   fail with a bad_alloc exception.

   Second, profiler_unittest.sh sometimes fails in the "fork" test.
   This is because stray SIGPROF signals from the parent process are
   making their way into the child process.  (This may be a kernel
   bug that only exists in older kernels.)  The profiling code itself
   is working fine.  This only affects programs that call fork(); for
   most programs, the cpu profiler is entirely safe to use.

   Third, perftools depends on /proc to get shared library
   information.  If you are running a FreeBSD system without proc,
   perftools will not be able to map addresses to functions.  Some
   unittests will fail as a result.

   Finally, the new test introduced in perftools-1.2,
   profile_handler_unittest, fails on FreeBSD.  It has something to do
   with how the itimer works.  The cpu profiler test passes, so I
   believe the functionality is correct and the issue is with the test
   somehow.  If anybody is an expert on itimers and SIGPROF in
   FreeBSD, and would like to debug this, I'd be glad to hear the
   results!

   libtcmalloc.so successfully builds, and the "advanced" tcmalloc
   functionality all works except for the leak-checker, which has
   Linux-specific code:
      % make heap-profiler_unittest.sh maybe_threads_unittest.sh \
             tcmalloc_unittest tcmalloc_both_unittest \
             tcmalloc_large_unittest              # THESE WORK
      % make -k heap-checker_unittest.sh \
                heap-checker-death_unittest.sh    # THESE DO NOT

   Note that unless you specify --enable-heap-checker explicitly,
   'make' will not build the heap-checker unittests on a FreeBSD
   system.

   I have not tested other *BSD systems, but they are probably similar.

** Mac OS X:

   I've tested OS X 10.5 [Leopard], OS X 10.4 [Tiger] and OS X 10.3
   [Panther] on both intel (x86) and PowerPC systems.  For Panther
   systems, perftools does not work at all: it depends on a header
   file, OSAtomic.h, which is new in 10.4.  (It's possible to get the
   code working for Panther/i386 without too much work; if you're
   interested in exploring this, drop an e-mail.)

   For the other seven systems, the binaries and libraries that
   successfully build are exactly the same as for FreeBSD.  See that
   section for a list of binaries and instructions on building them.

   In addition, it appears OS X regularly fails profiler_unittest.sh
   in the "thread" test (in addition to occassionally failing in the
   "fork" test).  It looks like OS X often delivers the profiling
   signal to the main thread, even when it's sleeping, rather than
   spawned threads that are doing actual work.  If anyone knows
   details of how OS X handles SIGPROF (via setitimer()) events with
   threads, and has insight into this problem, please send mail to
   google-perftools@googlegroups.com.

** Solaris 10 x86:

   I've only tested using the GNU C++ compiler, not the Sun C++
   compiler.  Using g++ requires setting the PATH appropriately when
   configuring.

   % PATH=${PATH}:/usr/sfw/bin/:/usr/ccs/bin ./configure
   % PATH=${PATH}:/usr/sfw/bin/:/usr/ccs/bin make [...]

   Again, the binaries and libraries that successfully build are
   exactly the same as for FreeBSD.  (However, while libprofiler.so can
   be used to generate profiles, pprof is not very successful at
   reading them -- necessary helper programs like nm don't seem
   to be installed by default on Solaris, or perhaps are only
   installed as part of the Sun C++ compiler package.)  See that
   section for a list of binaries, and instructions on building them.

** Windows:

   Work on Windows is rather preliminary: we haven't found a good way
   to get stack traces in release mode on windows (that is, when FPO
   is enabled), so the heap profiling may not be reliable in that
   case.  Also, heap-checking and CPU profiling do not yet work at
   all.  But as in other ports, the basic tcmalloc library
   functionality, overriding malloc and new and such (and even
   windows-specific functions like _aligned_malloc!), is working fine,
   at least with VC++ 7.1 (Visual Studio 2003) and VC++ 8.0
   (Visual Studio 2005), in both debug and release modes.  See
   README.windows for instructions on how to install on Windows using
   Visual Studio.

   Cygwin can compile some but not all of perftools.  Furthermore,
   there is a problem with exception-unwinding in cygwin (it can call
   malloc, which can call the exception-unwinding-setup code, which
   can lead to an infinite loop).  I've comitted a workaround to the
   exception unwinding problem, but it only works in debug mode and
   when statically linking in tcmalloc.  I hope to have a more proper
   fix in a later release.  To configure under cygwin, run

      ./configure --disable-shared CXXFLAGS=-g && make

   Most of cygwin will compile (cygwin doesn't allow weak symbols, so
   the heap-checker and a few other pieces of functionality will not
   compile).  'make' will compile those libraries and tests that can
   be compiled.  You can run 'make check' to make sure the basic
   functionality is working.  I've heard reports that some versions of
   cygwin fail calls to pthread_join() with EINVAL, causing several
   tests to fail.  If you have any insight into this, please mail
   google-perftools@googlegroups.com.

   This Windows functionality is also available using MinGW and Msys,
   In this case, you can use the regular './configure && make'
   process.  'make install' should also work.  The Makefile will limit
   itself to those libraries and binaries that work on windows.


Basic Installation
==================

   These are generic installation instructions.

   The `configure' shell script attempts to guess correct values for
various system-dependent variables used during compilation.  It uses
those values to create a `Makefile' in each directory of the package.
It may also create one or more `.h' files containing system-dependent
definitions.  Finally, it creates a shell script `config.status' that
you can run in the future to recreate the current configuration, and a
file `config.log' containing compiler output (useful mainly for
debugging `configure').

   It can also use an optional file (typically called `config.cache'
and enabled with `--cache-file=config.cache' or simply `-C') that saves
the results of its tests to speed up reconfiguring.  (Caching is
disabled by default to prevent problems with accidental use of stale
cache files.)

   If you need to do unusual things to compile the package, please try
to figure out how `configure' could check whether to do them, and mail
diffs or instructions to the address given in the `README' so they can
be considered for the next release.  If you are using the cache, and at
some point `config.cache' contains results you don't want to keep, you
may remove or edit it.

   The file `configure.ac' (or `configure.in') is used to create
`configure' by a program called `autoconf'.  You only need
`configure.ac' if you want to change it or regenerate `configure' using
a newer version of `autoconf'.

The simplest way to compile this package is:

  1. `cd' to the directory containing the package's source code and type
     `./configure' to configure the package for your system.  If you're
     using `csh' on an old version of System V, you might need to type
     `sh ./configure' instead to prevent `csh' from trying to execute
     `configure' itself.

     Running `configure' takes awhile.  While running, it prints some
     messages telling which features it is checking for.

  2. Type `make' to compile the package.

  3. Optionally, type `make check' to run any self-tests that come with
     the package.

  4. Type `make install' to install the programs and any data files and
     documentation.

  5. You can remove the program binaries and object files from the
     source code directory by typing `make clean'.  To also remove the
     files that `configure' created (so you can compile the package for
     a different kind of computer), type `make distclean'.  There is
     also a `make maintainer-clean' target, but that is intended mainly
     for the package's developers.  If you use it, you may have to get
     all sorts of other programs in order to regenerate files that came
     with the distribution.

Compilers and Options
=====================

   Some systems require unusual options for compilation or linking that
the `configure' script does not know about.  Run `./configure --help'
for details on some of the pertinent environment variables.

   You can give `configure' initial values for configuration parameters
by setting variables in the command line or in the environment.  Here
is an example:

     ./configure CC=c89 CFLAGS=-O2 LIBS=-lposix

   *Note Defining Variables::, for more details.

Compiling For Multiple Architectures
====================================

   You can compile the package for more than one kind of computer at the
same time, by placing the object files for each architecture in their
own directory.  To do this, you must use a version of `make' that
supports the `VPATH' variable, such as GNU `make'.  `cd' to the
directory where you want the object files and executables to go and run
the `configure' script.  `configure' automatically checks for the
source code in the directory that `configure' is in and in `..'.

   If you have to use a `make' that does not support the `VPATH'
variable, you have to compile the package for one architecture at a
time in the source code directory.  After you have installed the
package for one architecture, use `make distclean' before reconfiguring
for another architecture.

Installation Names
==================

   By default, `make install' will install the package's files in
`/usr/local/bin', `/usr/local/man', etc.  You can specify an
installation prefix other than `/usr/local' by giving `configure' the
option `--prefix=PATH'.

   You can specify separate installation prefixes for
architecture-specific files and architecture-independent files.  If you
give `configure' the option `--exec-prefix=PATH', the package will use
PATH as the prefix for installing programs and libraries.
Documentation and other data files will still use the regular prefix.

   In addition, if you use an unusual directory layout you can give
options like `--bindir=PATH' to specify different values for particular
kinds of files.  Run `configure --help' for a list of the directories
you can set and what kinds of files go in them.

   If the package supports it, you can cause programs to be installed
with an extra prefix or suffix on their names by giving `configure' the
option `--program-prefix=PREFIX' or `--program-suffix=SUFFIX'.

Optional Features
=================

   Some packages pay attention to `--enable-FEATURE' options to
`configure', where FEATURE indicates an optional part of the package.
They may also pay attention to `--with-PACKAGE' options, where PACKAGE
is something like `gnu-as' or `x' (for the X Window System).  The
`README' should mention any `--enable-' and `--with-' options that the
package recognizes.

   For packages that use the X Window System, `configure' can usually
find the X include and library files automatically, but if it doesn't,
you can use the `configure' options `--x-includes=DIR' and
`--x-libraries=DIR' to specify their locations.

Specifying the System Type
==========================

   There may be some features `configure' cannot figure out
automatically, but needs to determine by the type of machine the package
will run on.  Usually, assuming the package is built to be run on the
_same_ architectures, `configure' can figure that out, but if it prints
a message saying it cannot guess the machine type, give it the
`--build=TYPE' option.  TYPE can either be a short name for the system
type, such as `sun4', or a canonical name which has the form:

     CPU-COMPANY-SYSTEM

where SYSTEM can have one of these forms:

     OS KERNEL-OS

   See the file `config.sub' for the possible values of each field.  If
`config.sub' isn't included in this package, then this package doesn't
need to know the machine type.

   If you are _building_ compiler tools for cross-compiling, you should
use the `--target=TYPE' option to select the type of system they will
produce code for.

   If you want to _use_ a cross compiler, that generates code for a
platform different from the build platform, you should specify the
"host" platform (i.e., that on which the generated programs will
eventually be run) with `--host=TYPE'.

Sharing Defaults
================

   If you want to set default values for `configure' scripts to share,
you can create a site shell script called `config.site' that gives
default values for variables like `CC', `cache_file', and `prefix'.
`configure' looks for `PREFIX/share/config.site' if it exists, then
`PREFIX/etc/config.site' if it exists.  Or, you can set the
`CONFIG_SITE' environment variable to the location of the site script.
A warning: not all `configure' scripts look for a site script.

Defining Variables
==================

   Variables not defined in a site shell script can be set in the
environment passed to `configure'.  However, some packages may run
configure again during the build, and the customized values of these
variables may be lost.  In order to avoid this problem, you should set
them in the `configure' command line, using `VAR=value'.  For example:

     ./configure CC=/usr/local2/bin/gcc

will cause the specified gcc to be used as the C compiler (unless it is
overridden in the site shell script).

`configure' Invocation
======================

   `configure' recognizes the following options to control how it
operates.

`--help'
`-h'
     Print a summary of the options to `configure', and exit.

`--version'
`-V'
     Print the version of Autoconf used to generate the `configure'
     script, and exit.

`--cache-file=FILE'
     Enable the cache: use and save the results of the tests in FILE,
     traditionally `config.cache'.  FILE defaults to `/dev/null' to
     disable caching.

`--config-cache'
`-C'
     Alias for `--cache-file=config.cache'.

`--quiet'
`--silent'
`-q'
     Do not print messages saying which checks are being made.  To
     suppress all normal output, redirect it to `/dev/null' (any error
     messages will still be shown).

`--srcdir=DIR'
     Look for the package's source code in directory DIR.  Usually
     `configure' can determine that directory automatically.

`configure' also accepts some other, not widely useful, options.  Run
`configure --help' for more details.