0000381: EventTicklerProcess blocked by long delay

(0000376 - 163 - 187 - 323 - 323 - 323 - 323)
laza
10-14-04 03:22

The Archive seems to be too large. Here is a link to the image:

http://www.lazarevic.de/Squeak3.8a-5987.zip [^]

(Despite the name the image is updated upto 6297)

(0000382 - 431 - 494 - 494 - 494 - 494 - 494)
laza
10-14-04 11:09
edited on: 10-14-04 11:21

To see that there is something wrong in 3.8a evaluate this:

Duration milliSeconds: (Sensor eventTicklerProcess suspendedContext receiver resumptionTime - Time millisecondClockValue)

In a 3.7 image I get the expected random values <500ms.
In a saved 3.8a image I get values upto 30 minutes!
An image updated from 3.7 to 3.8a (without saving the image) gives something like that:

-0:01:10:01.905

edited on: 10-14-04 11:21

(0000419 - 77 - 89 - 89 - 89 - 89 - 89)
laza
10-16-04 23:36

The problem shows up after this update

6295RestartTimerInterruptWatcher-nk

(0000448 - 173 - 212 - 212 - 212 - 212 - 212)
andreas
10-18-04 09:55

Yup, it looks as if the event sensor delay was just active when the timer interrupt got restarted. Simply evaluating:

EventSensor install.

Should do the trick - ar.

(0000508 - 57 - 57 - 57 - 57 - 57 - 57)
tim
10-22-04 03:43

Didn't this get resolved? Or was I imagining the emails?

(0000514 - 161 - 167 - 167 - 346 - 346 - 346)
laza
10-22-04 11:27

The bug reported for 3.7 in 0000371 gets correctly fixed with update 6293.
In my bugnote there I confused the bug with the one I now reported here for 3.8a in 0000381

(0000717 - 34 - 37 - 37 - 37 - 37 - 37)
laza
11-20-04 01:13

Seems to be fixed in 3.8b(>=6461).

Notes
(0000376 - 163 - 187 - 323 - 323 - 323 - 323) laza 10-14-04 03:22	The Archive seems to be too large. Here is a link to the image: http://www.lazarevic.de/Squeak3.8a-5987.zip [^] (Despite the name the image is updated upto 6297)

(0000382 - 431 - 494 - 494 - 494 - 494 - 494) laza 10-14-04 11:09 edited on: 10-14-04 11:21	To see that there is something wrong in 3.8a evaluate this: Duration milliSeconds: (Sensor eventTicklerProcess suspendedContext receiver resumptionTime - Time millisecondClockValue) In a 3.7 image I get the expected random values <500ms. In a saved 3.8a image I get values upto 30 minutes! An image updated from 3.7 to 3.8a (without saving the image) gives something like that: -0:01:10:01.905 edited on: 10-14-04 11:21

(0000419 - 77 - 89 - 89 - 89 - 89 - 89) laza 10-16-04 23:36	The problem shows up after this update 6295RestartTimerInterruptWatcher-nk

(0000448 - 173 - 212 - 212 - 212 - 212 - 212) andreas 10-18-04 09:55	Yup, it looks as if the event sensor delay was just active when the timer interrupt got restarted. Simply evaluating: EventSensor install. Should do the trick - ar.

(0000508 - 57 - 57 - 57 - 57 - 57 - 57) tim 10-22-04 03:43	Didn't this get resolved? Or was I imagining the emails?

(0000514 - 161 - 167 - 167 - 346 - 346 - 346) laza 10-22-04 11:27	The bug reported for 3.7 in 0000371 gets correctly fixed with update 6293. In my bugnote there I confused the bug with the one I now reported here for 3.8a in 0000381

(0000717 - 34 - 37 - 37 - 37 - 37 - 37) laza 11-20-04 01:13	Seems to be fixed in 3.8b(>=6461).

Issue History
Date Modified	Username	Field	Change
10-14-04 03:16	laza	New Issue
10-14-04 03:22	laza	Note Added: 0000376
10-14-04 11:09	laza	Note Added: 0000382
10-14-04 11:21	laza	Note Added: 0000383
10-14-04 11:21	laza	Note Deleted: 0000383
10-14-04 11:21	laza	Note Edited: 0000382
10-16-04 23:36	laza	Note Added: 0000419
10-18-04 09:55	andreas	Note Added: 0000448
10-22-04 03:43	tim	Note Added: 0000508
10-22-04 11:27	laza	Note Added: 0000514
11-20-04 01:13	laza	Note Added: 0000717
11-20-04 01:13	laza	Status	new => resolved
11-20-04 01:13	laza	Resolution	open => fixed
11-20-04 01:13	laza	Assigned To	=> laza
11-20-04 01:13	laza	Status	resolved => closed

Relationships