On 06/10/2015 11:32 AM, Jiri Svoboda wrote:
> Hi,
> 
> are you saying that a thread is not stoppable while it's blocked waiting
> for a futex? That would be obviously wrong. Threads must be stoppable
> while being blocked in any syscall that's blocking by design - otherwise
> you would get poor debugging experience.

Well, unlike sys_ipc_wait_for_call(), sys_futex_sleep() does not
surround the wait by a stoppable section, so
TASK->udebug.not_stoppable_count remains undecremented for this thread
while sleeping.

> As far as what happened when the sleep was forcefully interrupted - I
> don't really follow you. I'd propably need to refresh my memory and look
> through the code. We can discuss this on the meeting.

Ok, I will try to explain and also show in gdb at the meeting.

Jakub


_______________________________________________
HelenOS-devel mailing list
[email protected]
http://lists.modry.cz/listinfo/helenos-devel

Reply via email to