cvs commit: src/sys/netinet sctp.h sctp_asconf.c sctp_asconf.h sctp_bsd_addr.c sctp_constants.h sctp_indata.c sctp_input.c sctp_lock_bsd.h sctp_os_bsd.h sctp_output.c sctp_pcb.c sctp_pcb.h sctp_peeloff.c sctp_sysctl.c sctp_sysctl.h sctp_timer.c sctp_uio.h ...

Randall Stewart rrs at cisco.com
Fri Jun 15 20:50:33 UTC 2007


John Baldwin wrote:

> 
> Also, I would use KTR_SCTP and use the same #ifdef above so it is 0 or 
> KTR_SUBSYS depending on whether or not you want it enabled.
> 

Changing the name around is cosmetic.. and no problem..

The actual ifdef with 0, though.. we have a seperate
mask.. since there are 21 or so things that can be logged/traced
and they all generate quite a bit of output.. which can be graphed
and charted.. along with the usual debugging lot..

Thus there is a seperate mask that turns these on/off..

So .. if for instance you enable KTR_SUBSYS (aka KTR_SCTP) then
nothing will come out until you enable the various logging
levels in SCTP.

The question of if an ifdef is needed.. I did a little
performance study yesterday... since originally all the tracepoints
were ifdef'd.. so I decided to see if there was any noticable
performance penalty for having the additional if(logging mask) CTR6()
and I could not measure the difference.. thus all but a few of
the ifdef's around these (and thus the options in the build) came
out..

R

-- 
Randall Stewart
NSSTG - Cisco Systems Inc.
803-345-0369 <or> 803-317-4952 (cell)


More information about the cvs-all mailing list