On Sep 9, 2009, at 12:47 PM, Joel Reymont wrote:
I'm trying to time a function within Firefoxwith a pid$target probe
[2]. e.g
dtrace -Zqw -x dynvarsize=64m -x evaltime=preinit -p 13954 -s menu-
construct.d -s sigcont.d
elapsed: 12.7942481ms
cpu : 7.7911194ms
count : 40 times
I'm also measuring Firefox startup time by running it arguments like
this:
... file:///Users/joelr/work/mozilla/startup/startup.html#`python -c
'import time; print int(time.time() * 1000);'`
startup.html [3] prints the time delta.
I'm seing 0.9s without tracing with menu-construct.d and 17.9s when
tracing with it.
Now the question... Should I be looking at the elapsed time above in
the context of 0.9s or 17.9s?
Thanks, Joel
Joel,
Given that you have only 40 hits, my guess is you should use the .9s
context.
As Firefox starts up, each dyld load is "paused" until a handshake
with dtrace
has taken place. This allows dtrace to place probes in the newly
loaded library before
any code has executed.
I think the extra startup time you are seeing is related to dtrace
processing
the newly loaded libraries.
If those libraries include debug info (for example, libraries you
might have built yourself)
it can take longer to process them.
James M
_______________________________________________
dtrace-discuss mailing list
dtrace-discuss@opensolaris.org