I'd just like to revisit the time glitch I'm seeing.
Aug 5 03:48:59 localhost user.info ptp4l: [395552.302] rms 231 max 907 freq
+1545 +/- 677 delay 4575 +/- 55
Aug 5 03:49:13 localhost user.info phc2sys: [395566.452] rms 205 max 542
freq +1544 +/- 6 delay 1176 +/- 7
Aug 5 03:52:13 localhost user.info phc2sys: [395746.480] rms 213 max 582
freq +1545 +/- 5 delay 1176 +/- 8
Aug 5 03:53:15 localhost user.info ptp4l: [395808.302] rms 216 max 859 freq
+1545 +/- 676 delay 4608 +/- 43
Aug 5 03:55:13 localhost user.info phc2sys: [395926.513] rms 179 max 742
freq +1545 +/- 5 delay 1176 +/- 7
Aug 5 03:57:31 localhost user.info ptp4l: [396064.302] rms 234 max 906 freq
+1544 +/- 710 delay 4636 +/- 29
Aug 5 03:58:13 localhost user.info phc2sys: [396106.550] rms 189 max 511
freq +1545 +/- 8 delay 1177 +/- 8
Aug 5 04:01:13 localhost user.info phc2sys: [396286.578] rms 200 max 497
freq +1545 +/- 8 delay 1176 +/- 7
Aug 5 04:01:47 localhost user.info ptp4l: [396320.302] rms 222 max 867 freq
+1546 +/- 648 delay 4598 +/- 47
Aug 5 04:04:13 localhost user.info phc2sys: [396466.614] rms 182 max 626
freq +1545 +/- 8 delay 1176 +/- 8
Aug 5 04:06:03 localhost user.info ptp4l: [396576.302] rms 237 max 890 freq
+1545 +/- 737 delay 4604 +/- 37
Aug 5 04:07:13 localhost user.info phc2sys: [396646.649] rms 204 max 535
freq +1545 +/- 5 delay 1176 +/- 8
Aug 5 04:10:13 localhost user.info phc2sys: [396826.682] rms 179 max 593
freq +1545 +/- 5 delay 1175 +/- 8
Aug 5 04:10:19 localhost user.info ptp4l: [396832.302] rms 229 max 874 freq
+1546 +/- 678 delay 4571 +/- 48
Aug 5 04:13:14 localhost user.info phc2sys: [397006.709] rms 79587509 max
1000000314 freq +43900783 +/- 58841702 delay 1178 +/- 9
Aug 5 04:14:35 localhost user.info ptp4l: [397088.213] rms 240 max 895 freq
+1546 +/- 751 delay 4595 +/- 58
Aug 5 04:16:14 localhost user.info phc2sys: [397186.748] rms 68362504 max
92771816 freq -85603435 +/- 45202655 delay 1176 +/- 8
Aug 5 04:18:51 localhost user.info ptp4l: [397344.409] rms 234 max 888 freq
+1543 +/- 790 delay 4583 +/- 49
Aug 5 04:19:14 localhost user.info phc2sys: [397366.776] rms 124563926 max
151649798 freq +100000000 +/- 0 delay 1176 +/- 8
Aug 5 04:22:14 localhost user.info phc2sys: [397546.810] rms 149505897 max
218734480 freq -54210744 +/- 80647416 delay 1174 +/- 8
Aug 5 04:23:07 localhost user.info ptp4l: [397600.124] rms 220 max 855 freq
+1546 +/- 571 delay 4622 +/- 52
Aug 5 04:25:14 localhost user.info phc2sys: [397726.840] rms 140195572 max
217904540 freq -57689713 +/- 78449823 delay 1179 +/- 8
Aug 5 04:27:23 localhost user.info ptp4l: [397856.582] rms 220 max 867 freq
+1545 +/- 715 delay 4590 +/- 44
Aug 5 04:28:14 localhost user.info phc2sys: [397906.868] rms 249960103 max
282920259 freq +100000000 +/- 0 delay 1176 +/- 8
Aug 5 04:31:14 localhost user.info phc2sys: [398086.894] rms 155857306 max
260561450 freq +12936785 +/- 96684047 delay 1173 +/- 8
Aug 5 04:31:39 localhost user.info ptp4l: [398111.997] rms 252 max 896 freq
+1546 +/- 822 delay 4601 +/- 40
Aug 5 04:34:14 localhost user.info phc2sys: [398266.927] rms 327011985 max
355888572 freq -100000000 +/- 0 delay 1176 +/- 8
Aug 5 04:35:55 localhost user.info ptp4l: [398368.340] rms 247 max 894 freq
+1546 +/- 783 delay 4601 +/- 35
Aug 5 04:37:14 localhost user.info phc2sys: [398446.958] rms 176666362 max
295681670 freq -10874317 +/- 97058373 delay 1180 +/- 8
Aug 5 04:40:11 localhost user.info ptp4l: [398624.679] rms 258 max 912 freq
+1544 +/- 773 delay 4575 +/- 58
Aug 5 04:40:14 localhost user.info phc2sys: [398627.000] rms 396056670 max
428470961 freq +100000000 +/- 0 delay 1175 +/- 7
Aug 5 04:42:58 localhost user.notice ptp4l: [398791.635] selected best master
clock 000cec.fffe.0a085d
Aug 5 04:43:14 localhost user.info phc2sys: [398807.026] rms 214608659 max
371819041 freq +48058105 +/- 85162373 delay 1172 +/- 7
Notice that there is no hardware MAC reset in this instance and the PHC remains
locked. The system clock appears to have glitched out by a second.
I've noticed that this type of glitch is always a second. It's as if the system
thinks it needs to insert a leap second!
I first thought that application code may be stamping on some configuration
registers, and that might be the case, but to have the glitch always be a
second and not some random time is a bit weird.
Thoughts?
Ian T.
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Linuxptp-users mailing list
Linuxptp-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-users
