feather-trace-tools.git/src/ft2csv.c, branch 2012.3 [ARCHIVE] ftcat, ft2csv, and friends. Now on GitHub. Remove two sources of outliers 2012-10-19T20:54:29+00:00 Bjoern Brandenburg bbb@mpi-sws.org 2012-09-10T07:09:50+00:00 23624a2fd62515146c0eace874583fb66586de4c - Don't enforce sequentiality constraint on SEND_RESCHED samples. - Refuse samples with negative time deltas (possible due to clock skew). 
- Don't enforce sequentiality constraint on SEND_RESCHED samples.
- Refuse samples with negative time deltas (possible due to clock skew).
Add filtering of IRQ-disturbed samples (on by default) 2012-10-19T20:54:29+00:00 Bjoern Brandenburg bbb@mpi-sws.org 2012-08-19T18:05:00+00:00 4531e293279cdc817f7ac37bff9472e74cb085ee Interrupts can cause massive outliers. Instead of trying to filter them afterwards, let's not even include such samples in the first place. 
Interrupts can cause massive outliers. Instead of trying to filter
them afterwards, let's not even include such samples in the first
place.
Support matching based on PID in ft2csv 2012-10-19T20:38:25+00:00 Bjoern Brandenburg bbb@mpi-sws.org 2012-08-15T14:14:52+00:00 44da17bf24f2e67ab71ecb552ebf7b460e64999e 






Remove hard-coded threshold
2012-08-01T07:48:48+00:00

Bjoern Brandenburg
bbb@mpi-sws.org

2012-02-13T17:27:18+00:00

fe67e69e273ea7d589488465e95ba2b5b6d0a9a6

The hard-coded threshold is a leftover from the beginnings.  It can be
surprising and is no longer needed: outlier filtering should happen
during post-processing, if at all.





The hard-coded threshold is a leftover from the beginnings.  It can be
surprising and is no longer needed: outlier filtering should happen
during post-processing, if at all.







Support filtering by CPU identity.
2011-02-21T15:53:51+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2011-02-21T15:53:51+00:00

bee50d638a562218b3a263439daa5eea5b6ee09b












allow binary output in ft2csv
2011-02-18T07:58:10+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2011-02-18T07:58:10+00:00

cff74703ad71f333491941d85baaa3f1835101fc

This can help with large data sets.





This can help with large data sets.







Remove endian swapping from ft2csv; don't skip over holes
2011-02-18T03:17:53+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2011-02-18T03:17:53+00:00

690b3db97d0f5e74ab8d14cb70f0601a4eb0cb79












add support for release latency; make ftdump a bit prettier
2011-02-17T23:58:18+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2011-02-17T23:58:18+00:00

26b6535eaa2750df407ec06a37d05aca6183e163












Parametrize CYCLES per US
2010-04-21T21:30:31+00:00

Andrea Bastoni
bastoni@cs.unc.edu

2010-04-21T21:30:31+00:00

e84fce96e96d8b5ee5e9531400b0735fb7e4fbb6












ft2csv: accept short names (without the trailing _START)
2009-05-05T19:09:00+00:00

Bjoern B. Brandenburg
bbb@koruna.cs.unc.edu

2009-05-05T19:09:00+00:00

2e76f83c8ab6ce634d3a91724bfde81df88c15aa

Also, adjust line breaks to fit within 80 chars.





Also, adjust line breaks to fit within 80 chars.