changelog

pashload是应用在linux下的网络带宽测试源程序精确度比较高 ///////////////////////////////////// Pathload is a tool for es

$Log: CHANGELOG,v $
Revision 1.17  2006/05/19 19:12:26  jain
Checking in version 1.3.2. Included bug fixes only

Revision 1.15  2004/04/17 01:01:05  jain
*** empty log message ***


Revision 2004/04/15 - updates since version 1.1.1
- Modified algorithm to detect interrupt coalescence
- Algorithm to estimate available bandwidth in presence of interrupt 
  coalescence.

Revision 2003/12/14 - bug fix since version 1.1.0
- Modified README.
- Make new thread detached in pathload receiver. This fixes a bug
  in older version, where receiver would hang after approx 20 fleet as not 
  enough memory was available to create a new thread and receiver would exit
  with a message "Receiver not responding".
- Fixed a bug in sender that would cause sender running in iterative mode to 
  exit with "Broken pipe" if receiver was killed.
- Simplified code to detect interrupt coalescence.
- Several other minor changes in the code

Revision 1.12 2003/04/09 04:31:13  jain
Modified the README


Revision 2003/04/06 05:16:00
Added option to have output in netlogger
format.

Revision 2003/04/04 01:21:22
implemented function to measure minimum sleep
time using select.

Revision 2003/04/03 21:06:57
Estimate gettimeofday() latency on sender since
some freebsd versions have gettimeofdaylatency equal 
to 5 usec.

Revision 2003/04/02 01:42:10
Modified the output mesages on pathload receiver.

Revision 2003/04/01 16:16:12
Added select so that sender doesnot block and
now waits a maximum of 50 sec for a control message
from receiver during a ongoing measurement.
Also reduced the aggregate threshold from .7 -> to .6.

Revision 2003/03/25 19:42:22
get_adr() additionally checks for interrupt coallescence.

Revision 2003/03/23 04:29:40
Modified the command line interface in pathload receiver.

Revision 2003/03/21 01:41:17
Removed bugs.

Revision 2003/03/19 15:27:36
Removed bugs.

Revision  2003/03/18 12:33:07  jain
Thread implementation in place of select.
select introduces additional latency while 
is a issue in gigabit path.

Revision 2003/02/09 04:36:47 jain
Trend detection after each stream
code working in low bandwidth paths.

Revision 1.10  2002/10/31 18:13:35  jain
CHANGELOG : testing with $Log: CHANGELOG,v $
CHANGELOG : testing with Revision 1.17  2006/05/19 19:12:26  jain
CHANGELOG : testing with Checking in version 1.3.2. Included bug fixes only
CHANGELOG : testing with
CHANGELOG : testing with Revision 1.15  2004/04/17 01:01:05  jain
CHANGELOG : testing with *** empty log message ***
CHANGELOG : testing with
CHANGELOG : testing with Revision 1.14  2004/01/14 01:49:32  jain
CHANGELOG : testing with *** empty log message ***
CHANGELOG : testing with
CHANGELOG : testing with Revision 1.13  2003/12/14 08:40:55  jain
CHANGELOG : testing with 1.1.1-test
CHANGELOG : testing with
CHANGELOG : testing with Revision 1.12  2003/04/09 04:31:13  jain
CHANGELOG : testing with Modified the README
CHANGELOG : testing with
CHANGELOG : testing with Revision 1.11  2003/04/09 00:42:35  jain
CHANGELOG : testing with *** empty log message ***
CHANGELOG : testing with

Revision 1.9  2002/10/31 17:49:08  jain
CHANGELOG : modified CHANGELOG

Revision 1.8  2002/10/31 17:47:27  jain
pathload_rcv_func.c : replaced blocking recvfrom by select.
pathload_snd_func.c : deley FINISH_STREAM by 200 usec.

10/22/02
In eliminate_rcvr_side_CS, changed
<  if ( ( rcvr_time_stamp[i+1] -  rcvr_time_stamp[i]  )  < rcv_latency )
to
>  if ( ( rcvr_time_stamp[i+1] -  rcvr_time_stamp[i]  )  <= (1.2 * rcv_latency) )

08/02/2002 - 1.0.3.2
Added code measure mss on sender and receiver. Also tried some testing on gigabit machines. Geting 100% losses.Dont know why ?
08/29/2002 - 1.0.3.3
new modifications : first time from gatech
08/29/2002 - 1.0.3.3
new modifications : first time from gatech

09/12/2002 
Changed min packet size to 300 bytes.
Changed min time interval to 150 usec.

07/25/02
Added code in pathload_snd_func to send
stream at lower rates ( < 1..17 mbps ).

07/25/02
Made grey_bw_resolution 5% of ADR.
Earlier it was 1.5 MBPS.
what if ADR could not be measures ?