Opened 10 years ago
Last modified 4 years ago
#10926 new bug
[input_server] -q sometimes doesn't relaunch it
| Reported by: | diver | Owned by: | korli |
|---|---|---|---|
| Priority: | normal | Milestone: | R1.1 |
| Component: | Servers/input_server | Version: | R1/Development |
| Keywords: | Cc: | ||
| Blocked By: | Blocking: | ||
| Platform: | All |
Description
hrev47334 running in vbox.
Issuing /system/servers/input_server -q sometimes (1 out of 5 attempts) doesn't bring input_server back. Interestingly, it quits only after cursor is moved. Shouldn't it do it right away?
Change History (3)
comment:1 by , 10 years ago
comment:2 by , 9 years ago
I can't restart input server at all (hrev50113).
It does quit on "/system/servers/input_server -q" but is never restarted. In the syslog I saw at one time:
KERN: Mouse device exiting, Interrupted system call
and now I always see:
KERN: Mouse device exiting, Bad semaphore ID KERN: Mouse device exiting, Operation canceled
"launch_roster stop x-vnd.Be-input_server" claims:
launch_roster: Stopping job "x-vnd.Be-input_server" failed: Name not found
Besides, wouldn't it make sense to auto-restart the input_server? Or is it supposed to already?
comment:3 by , 4 years ago
| Milestone: | R1 → R1.1 |
|---|
Note:
See TracTickets
for help on using tickets.
The sending of B_ARGV_RECEIVED has been fixed in hrev49064, the handling of "-q" in hrev49067.
The input_server does not wait for the cursor directly, it waits for the pointing device add-on to quit, which is blocked on a read. Maybe the thread should explicitly be interrupted in that case.