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] | |
On Mon, Dec 06, 2010 at 12:50:04PM +0530, h patel wrote:
> 2) Is it possible to use for Systemtap without debuginfo ? e.g. Only
> with Kernel Markers support and User space traces enabled. ?If yes,
> can you direct me to related help page ?
> As of now, I am working the long way of keeping debuginfo for target
> on Host box itself, I can't afford debuginfo on target as it goes
> beyond max image size.
With cross instrumentation you don't need debuginfo on the target. You
compile the module on the host (which has the debuginfos) then on the
target you only need the systemtap-runtime package (no debuginfo,
no compiler). This is detailed in http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/5/html/SystemTap_Beginners_Guide/cross-compiling.html
Furthermore, if you are using only markers or tracepoints
(as in *.mark("foo") and *.trace("bar")), you do not need debuginfos
on the host either. This is explained in the stapprobes(3stap) man page.
> 3) This is with performance related. Pardon my kernel knowledge. How
> the dynamic filtering of traces actually works ? E.g. If I have 10000+
> markers in 5 diff user apps, but enables or uses only one app markers
> in my systemtap script, how much overhead those remaining markers in
> other apps are going to add ?
I am not very familiar with the internals of how this works myself but
the best way to be sure the performances are acceptable for your use
case is probably to test :)
According to linux/Documentation/trace/tracepoints.txt:
When a tracepoint is "off" it has no effect, except for adding
a tiny time penalty (checking a condition for a branch) and space
penalty (adding a few bytes for the function call at the end of
the instrumented function and adds a data structure in a separate
section).
Not sure exactly how markers differ performance-wise (if they do) but if
you are running 2.6.32.10, it is probably not relevant anyway.
Attachment:
signature.asc
Description: Digital signature
| Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
|---|---|---|
| Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |