dtrace discuss - Apr 2006 - Dtrace - internals

What''s the best way to understand the internals of Dtrace? Aside from
the source code, are there any other files or documents?
   
  In particular, I''m interested in the workings of "entry",
in other words, how does Dtrace hook into the entry points of functions (kernel
and user-land)? E.g., does it modify the code space at runtime?
   
  Thanks in advance.
   
  Chester
   

		
---------------------------------
Talk is cheap. Use Yahoo! Messenger to make PC-to-Phone calls.  Great rates
starting at 1¢/min.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/dtrace-discuss/attachments/20060427/3ff87fd7/attachment.html>

(forgot to CC dtrace-discuss on my response, so sending it out for others to
see)

On 4/27/06, C B <chesbochesbo at yahoo.com> wrote:
>
> What''s the best way to understand the internals of Dtrace? Aside
from the
> source code, are there any other files or documents?
There''s the USENIX paper:

http://www.sun.com/bigadmin/content/dtrace/dtrace_usenix.pdf

See Section 4.1, in particular, for a description of the FBT provider,
which does kernel instrumentation.

For userland instrumentation, see Adam''s blog entry on the PID
provider;  it links to a presentation:

http://blogs.sun.com/roller/page/ahl/20050301
> In particular, I''m interested in the workings of
"entry", in other words,
> how does Dtrace hook into the entry points of functions (kernel and
> user-land)? E.g., does it modify the code space at runtime?
Yes;  see the above links for details.

Cheers,
- jonathan
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/dtrace-discuss/attachments/20060428/4af1da36/attachment.html>