We are pleased to announce a new release of Valgrind, version 3.3.0,
available from http://www.valgrind.org.
Valgrind is an open-source suite of simulation based debugging and
profiling tools. With the tools that come with Valgrind, you can
automatically detect many memory management and threading bugs, which
avoids hours of frustrating bug-hunting, and makes your code more
stable. You can also perform detailed time and space profiling to
help speed up and slim down your programs.
3.3.0 is a feature release with many significant improvements and the
usual collection of bug fixes. This release supports X86/Linux,
AMD64/Linux, PPC32/Linux and PPC64/Linux. See the release notes below
for details.
Our thanks to all those who contribute to Valgrind's development.
This release represents a great deal of time, energy and effort on the
part of many people.
Happy (and productive) debugging and profiling,
-- The Valgrind Developers
Release 3.3.0 (7 December 2007)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3.3.0 is a feature release with many significant improvements and the
usual collection of bug fixes. This release supports X86/Linux,
AMD64/Linux, PPC32/Linux and PPC64/Linux. Support for recent distros
(using gcc 4.3, glibc 2.6 and 2.7) has been added.
The main excitement in 3.3.0 is new and improved tools. Helgrind
works again, Massif has been completely overhauled and much improved,
Cachegrind now does branch-misprediction profiling, and a new category
of experimental tools has been created, containing two new tools:
Omega and DRD. There are many other smaller improvements. In detail:
- Helgrind has been completely overhauled and works for the first time
since Valgrind 2.2.0. Supported functionality is: detection of
misuses of the POSIX PThreads API, detection of potential deadlocks
resulting from cyclic lock dependencies, and detection of data
races. Compared to the 2.2.0 Helgrind, the race detection algorithm
has some significant improvements aimed at reducing the false error
rate. Handling of various kinds of corner cases has been improved.
Efforts have been made to make the error messages easier to
understand. Extensive documentation is provided.
- Massif has been completely overhauled. Instead of measuring
space-time usage -- which wasn't always useful and many people found
confusing -- it now measures space usage at various points in the
execution, including the point of peak memory allocation. Its
output format has also changed: instead of producing PostScript
graphs and HTML text, it produces a single text output (via the new
'ms_print' script) that contains both a graph and the old textual
information, but in a more compact and readable form. Finally, the
new version should be more reliable than the old one, as it has been
tested more thoroughly.
- Cachegrind has been extended to do branch-misprediction profiling.
Both conditional and indirect branches are profiled. The default
behaviour of Cachegrind is unchanged. To use the new functionality,
give the option --branch-sim=yes.
- A new category of "experimental tools" has been created. Such tools
may not work as well as the standard tools, but are included because
some people will find them useful, and because exposure to a wider
user group provides tool authors with more end-user feedback. These
tools have a "exp-" prefix attached to their names to indicate their
experimental nature. Currently there are two experimental tools:
* exp-Omega: an instantaneous leak detector. See
exp-omega/docs/omega_introduction.txt.
* exp-DRD: a data race detector based on the happens-before
relation. See exp-drd/docs/README.txt.
- Scalability improvements for very large programs, particularly those
which have a million or more malloc'd blocks in use at once. These
improvements mostly affect Memcheck. Memcheck is also up to 10%
faster for all programs, with x86-linux seeing the largest
improvement.
- Works well on the latest Linux distros. Has been tested on Fedora
Core 8 (x86, amd64, ppc32, ppc64) and openSUSE 10.3. glibc 2.6 and
2.7 are supported. gcc-4.3 (in its current pre-release state) is
supported. At the same time, 3.3.0 retains support for older
distros.
- The documentation has been modestly reorganised with the aim of
making it easier to find information on common-usage scenarios.
Some advanced material has been moved into a new chapter in the main
manual, so as to unclutter the main flow, and other tidying up has
been done.
- There is experimental support for AIX 5.3, both 32-bit and 64-bit
processes. You need to be running a 64-bit kernel to use Valgrind
on a 64-bit executable.
- There have been some changes to command line options, which may
affect you:
* --log-file-exactly and
--log-file-qualifier options have been removed.
To make up for this --log-file option has been made more powerful.
It now accepts a %p format specifier, which is replaced with the
process ID, and a %q{FOO} format specifier, which is replaced with
the contents of the environment variable FOO.
* --child-silent-after-fork=yes|no [no]
Causes Valgrind to not show any debugging or logging output for
the child process resulting from a fork() call. This can make the
output less confusing (although more misleading) when dealing with
processes that create children.
* --cachegrind-out-file, --callgrind-out-file and --massif-out-file
These control the names of the output files produced by
Cachegrind, Callgrind and Massif. They accept the same %p and %q
format specifiers that --log-file accepts. --callgrind-out-file
replaces Callgrind's old --base option.
* Cachegrind's 'cg_annotate' script no longer uses the --<pid>
option to specify the output file. Instead, the first non-option
argument is taken to be the name of the output file, and any
subsequent non-option arguments are taken to be the names of
source files to be annotated.
* Cachegrind and Callgrind now use directory names where possible in
their output files. This means that the -I option to
'cg_annotate' and 'callgrind_annotate' should not be needed in
most cases. It also means they can correctly handle the case
where two source files in different directories have the same
name.
- Memcheck offers a new suppression kind: "Jump". This is for
suppressing jump-to-invalid-address errors. Previously you had to
use an "Addr1" suppression, which didn't make much sense.
- Memcheck has new flags --malloc-fill=<hexnum> and
--free-fill=<hexnum> which free malloc'd / free'd areas with the
specified byte. This can help shake out obscure memory corruption
problems. The definedness and addressibility of these areas is
unchanged -- only the contents are affected.
- The behaviour of Memcheck's client requests VALGRIND_GET_VBITS and
VALGRIND_SET_VBITS have changed slightly. They no longer issue
addressability errors -- if either array is partially unaddressable,
they just return 3 (as before). Also, SET_VBITS doesn't report
definedness errors if any of the V bits are undefined.
- The following Memcheck client requests have been removed:
VALGRIND_MAKE_NOACCESS
VALGRIND_MAKE_WRITABLE
VALGRIND_MAKE_READABLE
VALGRIND_CHECK_WRITABLE
VALGRIND_CHECK_READABLE
VALGRIND_CHECK_DEFINED
They were deprecated in 3.2.0, when equivalent but better-named client
requests were added. See the 3.2.0 release notes for more details.
- The behaviour of the tool Lackey has changed slightly. First, the output
from --trace-mem has been made more compact, to reduce the size of the
traces. Second, a new option --trace-superblocks has been added, which
shows the addresses of superblocks (code blocks) as they are executed.
- The following bugs have been fixed. Note that "n-i-bz" stands for
"not in bugzilla" -- that is, a bug that was reported to us but
never got a bugzilla entry. We encourage you to file bugs in
bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than
mailing the developers (or mailing lists) directly.
n-i-bz x86_linux_REDIR_FOR_index() broken
n-i-bz guest-amd64/toIR.c:2512 (dis_op2_E_G): Assertion `0' failed.
n-i-bz Support x86 INT insn (INT (0xCD) 0x40 - 0x43)
n-i-bz Add sys_utimensat system call for Linux x86 platform
79844 Helgrind complains about race condition which does not exist
82871 Massif output function names too short
89061 Massif: ms_main.c:485 (get_XCon): Assertion `xpt->max_chi...'
92615 Write output from Massif at crash
95483 massif feature request: include peak allocation in report
112163 MASSIF crashed with signal 7 (SIGBUS) after running 2 days
119404 problems running setuid executables (partial fix)
121629 add instruction-counting mode for timing
127371 java vm giving unhandled instruction bytes: 0x26 0x2E 0x64 0x65
129937 ==150380
129576 Massif loses track of memory, incorrect graphs
132132 massif --format=html output does not do html entity escaping
132950 Heap alloc/usage summary
133962 unhandled instruction bytes: 0xF2 0x4C 0xF 0x10
134990 use -fno-stack-protector if possible
136382 ==134990
137396 I would really like helgrind to work again...
137714 x86/amd64->IR: 0x66 0xF 0xF7 0xC6 (maskmovq, maskmovdq)
141631 Massif: percentages don't add up correctly
142706 massif numbers don't seem to add up
143062 massif crashes on app exit with signal 8 SIGFPE
144453 (get_XCon): Assertion 'xpt->max_children != 0' failed.
145559 valgrind aborts when malloc_stats is called
145609 valgrind aborts all runs with 'repeated section!'
145622 --db-attach broken again on x86-64
145837 ==149519
145887 PPC32: getitimer() system call is not supported
146252 ==150678
146456 (update_XCon): Assertion 'xpt->curr_space >= -space_delta'...
146701 ==134990
146781 Adding support for private futexes
147325 valgrind internal error on syscall (SYS_io_destroy, 0)
147498 amd64->IR: 0xF0 0xF 0xB0 0xF (lock cmpxchg %cl,(%rdi))
147545 Memcheck: mc_main.c:817 (get_sec_vbits8): Assertion 'n' failed.
147628 SALC opcode 0xd6 unimplemented
147825 crash on amd64-linux with gcc 4.2 and glibc 2.6 (CFI)
148174 Incorrect type of freed_list_volume causes assertion [...]
148447 x86_64 : new NOP codes: 66 66 66 66 2e 0f 1f
149182 PPC Trap instructions not implemented in valgrind
149504 Assertion hit on alloc_xpt->curr_space >= -space_delta
149519 ppc32: V aborts with SIGSEGV on execution of a signal handler
149892 ==137714
150044 SEGV during stack deregister
150380 dwarf/gcc interoperation (dwarf3 read problems)
150408 ==148447
150678 guest-amd64/toIR.c:3741 (dis_Grp5): Assertion `sz == 4' failed
151209 V unable to execute programs for users with UID > 2^16
151938 help on --db-command= misleading
152022 subw $0x28, %%sp causes assertion failure in memcheck
152357 inb and outb not recognized in 64-bit mode
152501 vex x86->IR: 0x27 0x66 0x89 0x45 (daa)
152818 vex x86->IR: 0xF3 0xAC 0xFC 0x9C (rep lodsb)
Developer-visible changes:
- The names of some functions and types within the Vex IR have
changed. Run 'svn log -r1689 VEX/pub/libvex_ir.h' for full details.
Any existing standalone tools will have to be updated to reflect
these changes. The new names should be clearer. The file
VEX/pub/libvex_ir.h is also much better commented.
- A number of new debugging command line options have been added.
These are mostly of use for debugging the symbol table and line
number readers:
--trace-symtab-patt=<patt> limit debuginfo tracing to obj name <patt>
--trace-cfi=no|yes show call-frame-info details? [no]
--debug-dump=syms mimic /usr/bin/readelf --syms
--debug-dump=line mimic /usr/bin/readelf --debug-dump=line
--debug-dump=frames mimic /usr/bin/readelf --debug-dump=frames
--sym-offsets=yes|no show syms in form 'name+offset' ? [no]
- Internally, the code base has been further factorised and
abstractified, particularly with respect to support for non-Linux
OSs.
(3.3.0.RC1: 2 Dec 2007, vex r1803, valgrind r7268).
(3.3.0.RC2: 5 Dec 2007, vex r1804, valgrind r7282).
(3.3.0.RC3: 9 Dec 2007, vex r1804, valgrind r7288).
(3.3.0: 10 Dec 2007, vex r1804, valgrind r7290).
|
