systemtap release 0.9.5

From: Josh Stone
Date: Fri Mar 27 2009 - 16:16:04 EST


The SystemTap team proudly announces release 0.9.5.

Kernel tracepoint support, standard tapset functions and probe now have
manual pages, typecasting supports module search paths, on-file flight
recorder support, probe each instruction (block) step of a process,
syscall wrappers support, security fixes, tested against released
2.6.29 kernels.

= Where to get it

http://sourceware.org/systemtap/ - our project page
http://sourceware.org/systemtap/ftp/releases/systemtap-0.9.5.tar.gz
http://koji.fedoraproject.org/koji/packageinfo?packageID=615
git commit 756c9462

= How to build it

See the README and NEWS files at
http://sourceware.org/git/?p=systemtap.git;a=tree;hb=756c9462
Further information at http://sourceware.org/systemtap/wiki/

= Systemtap frontend (stap) changes

- On-file flight recorder is supported. It allows stap to record huge
trace log on the disk and to run in background. Passing -F option
with -o option runs stap in background mode. In this mode, staprun is
detached from console, and stap itself shows staprun's pid and exits.
Specifying the max size and the max number of log files are also
available by passing -S option. This option has one or two arguments
seperated by a comma. The first argument is the max size of a log
file in MB. If the size of a log file exceeds it, stap switches to
the next log file automatically. The second is how many files are
kept on the disk. If the number of log files exceeds it, the oldest
log file is removed automatically. The second argument can be
omitted.

For example, this will record output on log files each of them is
smaller than 1024MB and keep last 3 logs, in background.
% stap -F -o /tmp/staplog -S 1024,3 script.stp

- In guru mode (-g), the kernel probing blacklist is disabled, leaving
only a subset - the kernel's own internal kprobe blacklist - to
attempt to filter out areas unsafe to probe. The differences may be
enough to probe more interrupt handlers.

- Variables unavailable in current context may be skipped by setting a
session level flag with command line option --skip-badvars now
available. This replaces any dwarf $variable expressions that could
not be resolved with literal numeric zeros, along with a warning
message.

- Both kernel markers and kernel tracepoint support argument listing
through stap -L 'kernel.mark("*")' or stap -L 'kernel.trace("*")'

- Users can use -DINTERRUPTIBLE=0 to prevent interrupt reentrancy in
their script, at the cost of a bit more overhead to toggle the
interrupt mask.

- Added reentrancy debugging. If stap is run with the arguments
"-t -DDEBUG_REENTRANCY", additional warnings will be printed for
every reentrancy event, including the probe points of the resident
and interloper probes.

For an overview of the stap frontend see man stap(1).

= Systemtap script language changes

- Kernel tracepoints are now supported for probing predefined kernel
events without any debuginfo. Tracepoints incur less overhead than
kprobes, and context parameters are available with full type
information. Any kernel 2.6.28 and later should have defined
tracepoints. Try the following to see what's available:
$ stap -L 'kernel.trace("*")'

- New probes process().insn and process().insn.block that allows
inspection of the process after each instruction or block of
instructions executed. So to count the total number of instructions
a process executes during a run do something like:
$ stap -e 'global steps; probe process("/bin/ls").insn {steps++}
probe end {printf("Total instructions: %d\n", steps);}' \
-c /bin/ls
This feature can slow down execution of a process somewhat.

- Typecasting with @cast now supports modules search paths, which is
useful in case there are multiple places where the type definition
may be found. For example:
@cast(sdev, "scsi_device", "kernel:scsi_mod")->sdev_state

For an overview of available probes, functions and variables see
man stapprobes(3stap), man stapfuncs(3stap), man stapvars(3stap)

= Documentation

- Systemtap probes and function man pages extracted from the tapsets
are now available under 3stap. To show the page for the probe
vm.pagefault or the stap function pexecname do:
$ man [3stap] vm.pagefault
$ man [3stap] pexecname

= Miscellaneous changes

- Default to --disable-pie for configure. Use --enable-pie to turn it
back on.

- Improved sdt.h compatibility and test suite for static dtrace
compatible user space markers.

- Some architectures now use syscall wrappers (HAVE_SYSCALL_WRAPPERS).
The syscall tapset has been enhanced to take care of the syscall
wrappers in this release.

- Security fix for CVE-2009-0784: stapusr module-path checking race.

= Code contributors for this release

Ananth N Mavinakayanahalli, Breno Leitao, Dan Horak, Dave Brolley,
David Smith, Don Domingo, Eugeniy Meshcheryakov, Frank Ch. Eigler,
Josh Stone, Lubomir Rintel, Mahesh J Salgaonkar, Mark Wielaard,
Masami Hiramatsu, Maynard Johnson, Rajan Arora, Roland McGrath,
Stan Cox, Tim Moore, Wenji Huang and William Cohen

Thanks to Mark Wielaard for assembling these notes.

= Examples of tested kernel versions

2.6.9-70 (el4/i386)
2.6.18 (el5/ia64/i686/x86_64-xen)
2.6.27.19 (f10/i686/x86_64)
2.6.29 (i586/x86_64)

= Know issues with this release

Some kernel crashes continue to be reported when a script probes
broad kernel function wildcards.

Some 2.6.28- and 2.6.29-era kernels contain bugs that can more
easily trigger crashes upon systemtap scripts. Upgrade to the
latest -stable or -rc if possible.

= Problems resolved for this release

4311 support probing without debuginfo via symbol tables
4862 -rt kernel compatibility for locks
4901 tcp_sendmsg arguments changed in 2.6.23-rc2
5634 support pointer type punning
5951 process lifetime/memorymap monitor
6899 offer option to avoid disabling interrupts during more probe...
6921 tracepoint support
6930 Flight Recorder on file
6950 buildok/twentyfive.stp fails with elfutils 0.137
7045 user-space probe x86 on x86-64 host
7053 automatic global printing of statistic needs to check @count>0
7071 optional $context variables
7092 uprobes probes causing OOM
9692 stap-{client,server}-connect fail to build on Debian
9718 utrace - ftrace demo consumer
9719 improve type mismatch messages
9770 stap -l 'process("bla").mark("*")' should print marker names
9788 permissions error in staprun
9821 Support strftime like format for output file
9866 save_stack_trace code in runtime stack* needs more than CONF...
9870 man pages from tapset should not go under /usr/share/man/man5
9875 building rpm fails in checking for unpackaged file
9876 Syscall tapset broken on architectures that use SYSCALL_WRAP...
9882 drop Core from Fedora name
9883 static markers in userspace libraries segfaults
9896 Systemtap source compilation failure on latest Ubuntu
9919 array overflow error points to wrong statement
9922 need to configure with --disable-pie on ubuntu
9927 2.6.29-rc7 - kernel crash with sharedbuf.exp stap test script
9930 produce more information about probe reentrancy
9932 @cast module search path
9935 New sdt.exp test fails for mark probes with more than 4 argu...
9936 server.exp hangs
9947 lockdep triggered on rawhide with tracepoints
9951 runtime error: error: '_v' may be used uninitialized in this...
9956 sdt.exp doesn't run against gcc 4.1.2 with -O2
9959 tracepoints should gloss over unrecognized parameter types
9960 tracepoints should support complex parameters by-value
9964 Can't find seperate debuginfo file for user space probing
9967 raise MAXSKIPPED for -DINTERRUPTIBLE!=0 case
9974 adapt to utrace renaming of utrace_attached_engine -> utrace...
9980 sdt.h doesn't compile against gcc 4.3.2 on i386
9989 utrace or task-finder crash in upstream utrace git tree
9993 2.6.29-rc8-next-20090323 tracepoint build-breaking changes
10000 buildok/seventeen.stp fails "integer constant is too large"
10001 32-bit map error "array subscript is above array bounds"
10006 CVE-2009-0784: staprun race condition, privilege escalation

= Test results on various systems

After running "sudo make installcheck" from the test suite, on a
suitably equipped machine (kernel debugging data and other stuff
installed), you should see 500-600 passes and a small handful of
failures.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/