Opened 15 years ago
Closed 15 years ago
#137 closed defect (fixed)
Only one udebug session is possible, subsequent attempts fail.
| Reported by: | Tomas Brambora | Owned by: | Jiri Svoboda |
|---|---|---|---|
| Priority: | major | Milestone: | 0.4.2 |
| Component: | helenos/unspecified | Version: | mainline |
| Keywords: | Cc: | ||
| Blocker for: | Depends on: | ||
| See also: |
Description
Once an udebug session has finished, TASK→kb.finished is set to true,
which prevents any subsequent attempts to start a new session.
I believe that's rather a bug than a desired feature.
Note:
See TracTickets
for help on using tickets.
Not exactly. TASK→kb.finished was not set to NULL at the end of debugging session. Actually it was the test in kbox_proc_phone_hungup() which was broken. (A nice example of C's 'perfect' type checking. I was comparing a pointer and a boolean false). This was always evaluated to true, which meant kb.thread did not get reset to NULL. (Come to think of it, I cannot remember anymore why I am only detaching the thread unless the task is already terminating…) In the end a kbox thread would not get created when another debugging session started, which caused udebug_begin() to never finish.
Fixed in changeset:head,122.