batctl: Correct mdev calculation in ping subcommand
PATCH v2 1/1 batctl: Correct mdev calculation in ping subcommand
Fixes the calculation of the mean absolute deviation (mdev) of the roundtrip times. mdev was previously calculated by computing the difference of maximum and minimum round trip times.
The calculation is now done using the same formula as in the iputils ping program.
---
Makefile +1 -1
@@ -28,7 +28,7 @@ CC = gcc CFLAGS += -pedantic -Wall -W -g3 -std=gnu99 -Os -fno-strict-aliasing EXTRA_CFLAGS = -DREVISION_VERSION=$(REVISION_VERSION) -LDFLAGS += +LDFLAGS += -lm SBINDIR = $(INSTALL_PREFIX)/usr/sbin
ping.c +13 -2
@@ -29,6 +29,7 @@#include <signal.h> #include <fcntl.h> #include <string.h> +#include <math.h> #include "main.h" #include "ping.h"
@@ -76,7 +77,7 @@unsigned int seq_counter =0 , packets_out =0 , packets_in =0 , packets_loss;char *dst_string, *mac_string, *rr_string;double time_delta; -float min =0.0 , max =0.0 , avg =0.0 ; +float min =0.0 , max =0.0 , avg =0.0 , mdev =0.0 ;uint8_t last_rr_cur =0 , last_rr[BAT_RR_LEN ][ETH_ALEN ];size_t packet_len;
@@ -260,6 +261,7 @@if (time_delta > max) max = time_delta; avg += time_delta; + mdev += time_delta * time_delta; packets_in++;break ;case DESTINATION_UNREACHABLE :
@@ -289,11 +291,20 @@else packets_loss = ((packets_out - packets_in) *100 ) / packets_out; +if (packets_in) { + avg /= packets_in; + mdev /= packets_in; + mdev =sqrt (mdev - avg * avg); + }else { + avg =0.0 ; + mdev =0.0 ; + } +printf ("--- %s ping statistics ---\n" , dst_string);printf ("%u packets transmitted, %u received, %u%% packet loss\n" , packets_out, packets_in, packets_loss);printf ("rtt min/avg/max/mdev = %.3f/%.3f/%.3f/%.3f ms\n" , - min, (packets_in ? (avg / packets_in) :0.000 ), max, (max - min)); + min, avg, max, mdev); ret =EXIT_SUCCESS ;
Quoted 8 lines · Daniel Seither
Fixes the calculation of the mean absolute deviation (mdev) of the roundtrip times. mdev was previously calculated by computing the difference of maximum and minimum round trip times.
The calculation is now done using the same formula as in the iputils ping program.
Signed-off-by Daniel Seither <[email protected]>
Thanks a lot for your contributions,
Sven
Quoted 11 lines · Sven Eckelmann
Quoted 8 lines · Daniel Seither
Fixes the calculation of the mean absolute deviation (mdev) of the roundtrip times. mdev was previously calculated by computing the difference of maximum and minimum round trip times.
The calculation is now done using the same formula as in the iputils ping program.
Signed-off-by Daniel Seither <[email protected]>Acked-by Sven Eckelmann <[email protected]>
Ok, have a small suggestion. Please check that mdev - ... is positive. Otherwise we would have something like (mdev - avg * avg == -0.000....., sqrt(...) == NaN):
PING 02:00:00:00:00:02 (02:00:00:00:00:02) 19(47) bytes of data 19 bytes from 02:00:00:00:00:02 icmp_seq=1 ttl=50 time=0.10 ms --- 02:00:00:00:00:02 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss
rtt min/avg/max/mdev = 0.101/0.101/0.101/nan ms
Best regards,
Sven
Am 20.05.2010 19:00, schrieb Sven Eckelmann:
Quoted 10 lines
Sven Eckelmann wrote:
Ok, have a small suggestion. Please check that mdev - ... is positive. Otherwise we would have something like (mdev - avg * avg == -0.000....., sqrt(...) == NaN):PING 02:00:00:00:00:02 (02:00:00:00:00:02) 19(47) bytes of data 19 bytes from 02:00:00:00:00:02 icmp_seq=1 ttl=50 time=0.10 ms --- 02:00:00:00:00:02 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss
rtt min/avg/max/mdev = 0.101/0.101/0.101/nan ms
There seems to be a problem with the floats... The value of mdev - avg^2 cannot become negative if computed with infinite precision. In this case, it seems to be slightly less than zero because of rounding errors. In iputils' ping, all calculations are done using integers which prevents this kind of problems.
I think there are two solutions for the problem you discovered:
1) check whether the difference is <= 0 (easy)
2) rewrite the time measurement to use integers (more code to change, but calculations will be exact)
Please check whether replacing the line calling sqrt by the following code fixes the problem for you (first solution):
mdev = mdev - avg * avg; if (mdev > 0.0) mdev = sqrt(mdev); else mdev = 0.0;
Or should we aim for the second solution?
- Daniel
Quoted 12 lines · Daniel Seither
Am 20.05.2010 19:00, schrieb Sven Eckelmann:
Quoted 10 lines
Sven Eckelmann wrote:
Ok, have a small suggestion. Please check that mdev - ... is positive. Otherwise we would have something like (mdev - avg * avg == -0.000....., sqrt(...) == NaN):PING 02:00:00:00:00:02 (02:00:00:00:00:02) 19(47) bytes of data 19 bytes from 02:00:00:00:00:02 icmp_seq=1 ttl=50 time=0.10 ms --- 02:00:00:00:00:02 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss
rtt min/avg/max/mdev = 0.101/0.101/0.101/nan ms
Quoted 5 lines
There seems to be a problem with the floats... The value of mdev - avg^2 cannot become negative if computed with infinite precision. In this case, it seems to be slightly less than zero because of rounding errors. In iputils' ping, all calculations are done using integers which prevents this kind of problems.
Correct. Because of that I recommended to check before calling sqrt.
Quoted 7 lines
I think there are two solutions for the problem you discovered:
1) check whether the difference is <= 0 (easy)
2) rewrite the time measurement to use integers (more code to change, but calculations will be exact)
Quoted 8 lines
Please check whether replacing the line calling sqrt by the following code fixes the problem for you (first solution):
mdev = mdev - avg * avg; if (mdev > 0.0) mdev = sqrt(mdev); else mdev = 0.0;
Already done that - that's why I recommended it :)
Quoted 1 line
Or should we aim for the second solution?
I think the first solution is good enough for now. Please send the patch again with that fix and it is ok for now. If you want you can rewrite the code later to use only integers.
Best regards,
Sven