[LinuxPPS] Request for confirmation:

[Bernhard Schiffner]

Hi all out there,

is there someone who can do the "carzy" test I did:

Connect /ACK-pin of the parallel port and the DCD-pin of the serial port to 
the same electrical signal and register the /proc/pps/0*/* readouts linuxpps 
serves.

(I used a Maxim 3232 to convert a CMOS-signal. This caused a measured delay of 
less then 1µs on the serial line.)

My results (typical case):

bernd at bs4:/media/hda4/lnusers/bernd/developing/pcb$ ls -1 /proc/pps/0*/*
/proc/pps/00/assert
/proc/pps/01/assert
/proc/pps/01/clear
bernd at bs4:/media/hda4/lnusers/bernd/developing/pcb$ 
1168261686.889302674 #4580465
1168261686.889332088 #179
1168261698.412515975 #143

This shows an until now unexplained delay of about 30µs.
It seems to be caused by context-| task-switching. Me and others find this 
time high.

Can someone confirm this please?

tia!

Bernhard

----------------------
My PC:
root at bs4:~# uname -a
Linux bs4 2.6.19.1-rt15 #3 PREEMPT Wed Dec 27 12:28:01 CET 2006 i686 GNU/Linux
(+linuxpps)
root at bs4:~# cat /proc/cpuinfo
processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 6
model           : 6
model name      : AMD Athlon(tm) XP 2000+
stepping        : 2
cpu MHz         : 1662.080
..
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 sep mtrr pge mca cmov pat 
pse36 mmx fxsr sse syscall mmxext 3dnowext 3dnow ts
bogomips        : 3326.61