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] |
Hi, I'm working on tracing JBD layer and upper filesystem layer, and would like to get fine-grained setting of probes. Concretely speaking, I need to probe more static functions and some statements of in a large function.
I encounter some problems when resorting to module(xxx).function(yyy) and moule(xxx).statement(yyy).
Problem 1: Probes on some static function fail to work while others do work. For instance, probes on entry/exit of get_transaction(@fs/jbd/trasaction.c) fail because of semantic errors (failed to retrieve location attibution for target virable, or no match for probe point). But, probes on some other static functions(such as ext3_add_nondir@fs/ext3/namei.c, add_dirent_to_buf@fs/ext3/namei.c) do work.
With nm I found there is difference in corresponding object files. In one object file, there is symbol of static function. However, there isn't in the other one. I wonder the behind mechanism which cause the difference, and get concerned the effect on capability of systemtap.
Problem 2: Statement probe fail to work because of semantic error (no match for probe point)
I use systemtap-20070707 and kernel 2.6.22.1 on Fedora 7. A test script is attached. Thank you for your help. -- sunzen <<freedom & enjoyment>>
Attachment:
test.stp
Description: Binary data
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |