This is the mail archive of the
systemtap@sourceware.org
mailing list for the systemtap project.
Re: Another Newbie question about measuring time in a sys call
- From: Li Guanglei <guanglei at cn dot ibm dot com>
- To: David A Sperry <David_A_Sperry at raytheon dot com>
- Cc: systemtap at sources dot redhat dot com
- Date: Wed, 08 Feb 2006 09:51:52 +0800
- Subject: Re: Another Newbie question about measuring time in a sys call
- Organization: IBM CSTL
- References: <OF9A980AD4.6A1AABB6-ON8525710F.0006CB90-8525710F.000827AB@mck.us.ray.com>
David A Sperry ??:
I'm guessing the syscall tapsets may have changed and the associative array
may be getting confused between a call and it's return (lseek vs
lseek.return)
I greped in /usr/local/share/systemtap/tapsets/syscall.stp, and found
for the return of syscalls, in some places, it set name to
syscallname, while in other places it set name to syscall.return, e.g.:
probe kernel.syscall.chdir.return =
kernel.function("sys_chdir").return {
name = "chdir.return"
}
probe syscall.accept.return = kernel.function("sys_accept").return {
name = "accept"
returnp = 1 /*seem to use returnp=1 to mark this as a return probe*/
}
so we need to make the name consistent.