This is the mail archive of the systemtap@sourceware.org mailing list for the systemtap project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug runtime/2725] Attempting to probe all kernel functions hangs FC5


------- Additional Comments From wcohen at redhat dot com  2006-06-01 19:39 -------
Which architecture is this on? There are some differences between the i386 and
x86_64 in setting up the probes. The x86_64 need executable regions of memory
allocated. Depending on which architecture is being used there could be
different failure modes.

Is there any other information output? One trick is to just have a regular
console, no X so you can see any oops that come up. Are there any entries in
/var/log/messages related to this problem?

It could be that this experiment is bumping against some limit. You might take a
look at the output of the follow to see how many probes and the size of the
result .ko:

stap -v -p4 -k -e 'probe kernel.function("*")  { printf ("%s  %s  %d  %s  ",
"test-> ", execname(), pid(),  probefunc() ) }'

It is also possible that some function that should be blacklisted is getting
instrumented. You might try to expand
tests/testsuite/systemtap.stress/all_kernel_functions.exp to see if you can get
a narrow down the boundary between working and crashing.


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Attempting to probe all     |Attempting to probe all
                   |kernel functions hangs FC5  |kernel functions hangs FC5


http://sourceware.org/bugzilla/show_bug.cgi?id=2725

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]