#2268 closed defect (fixed)
threadstartmultitasking.c has a FIXME on SMP path
Reported by: | Joel Sherrill | Owned by: | Sebastian Huber |
---|---|---|---|
Priority: | lowest | Milestone: | 4.11 |
Component: | score | Version: | 4.11 |
Severity: | critical | Keywords: | |
Cc: | Blocked By: | ||
Blocking: |
Description
If this comment is correct, then there should be a ticket to address this issue. If this problem still exists, then elaborate in this ticket and reference it in the code. The milestone may be wrong. Please adjust. I set it to 4.11 so we could evaluate it.
- The _CPU_Context_Restart_self() implementations usually assume that self
- context is executing. *
- FIXME: We have a race condition here in case another thread already
- performed scheduler operations and moved our heir thread to another
- processor. The time frame for this is likely too small to be practically
- relevant.
Change History (8)
comment:1 Changed on 02/13/15 at 20:57:44 by Sebastian Huber
Priority: | normal → lowest |
---|---|
Severity: | normal → critical |
comment:2 Changed on 02/13/15 at 21:00:03 by Joel Sherrill
OK. Can you update the comment to reference this ticket?
comment:3 Changed on 02/13/15 at 22:23:22 by Joel Sherrill
Component: | General → SMP |
---|
comment:4 Changed on 02/14/15 at 13:40:43 by Sebastian Huber
Status: | new → assigned |
---|
comment:5 Changed on 02/17/15 at 08:41:55 by Sebastian Huber <sebastian.huber@…>
comment:6 Changed on 02/19/15 at 08:22:47 by Sebastian Huber
Resolution: | → fixed |
---|---|
Status: | assigned → closed |
comment:7 Changed on 10/10/17 at 06:27:10 by Sebastian Huber
Component: | SMP → score |
---|
comment:8 Changed on 10/10/17 at 06:29:01 by Sebastian Huber
Component: | score → cpukit |
---|
Note: See
TracTickets for help on using
tickets.
Yes, this problem still exists. In order to solve it we must update the is executing indicator similar to the code in _CPU_Context_switch().