On 2014-06-11 14:08, Joel Sherrill wrote:
On Jun 11, 2014 2:36 AM, Sebastian Huber
wrote:
>
> Hello Joel,
>
> can you please commit this as soon as possible since the pending content move
> makes it harder to add the helping feature to the scheduler necessary for
MrsP.
:
This should address the issues Gedare pointed out with the previous
patch. It also fixes a problem seen on real hardware.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu
On 2014-06-10 14:32, Gedare Bloom wrote:
On Tue, Jun 10, 2014 at 2:51 AM, Sebastian Huber
wrote:
>Hello,
>
>the check submission script
>
>http://git.rtems.org/rtems-testing/tree/merge-helpers/check_submission
>
>contains a test for timing values
>
> # Hopefull
ame identical to this frame (corrupt stack?)
See also:
https://sourceware.org/bugzilla/show_bug.cgi?id=16215
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-bra
On 2014-06-09 11:06, Daniel Hellstrom wrote:
On 06/05/2014 04:54 PM, Sebastian Huber wrote:
On 2014-06-04 20:48, Daniel Hellstrom wrote:
On 06/04/2014 11:51 AM, Sebastian Huber wrote:
On 2014-06-04 11:23, Daniel Hellstrom wrote:
Instead of calling the system call TA instruction directly it
On 2014-06-06 16:24, Sebastian Huber wrote:
See also
https://sourceware.org/ml/binutils/2014-06/msg00059.html
This patch is no longer necessary since Alan Modra relaxed the check in the
linker.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
On 2014-06-09 16:36, Martin Boretto wrote:
I am sending you the new git patch attached with the code style according to
uncrustify. I used the config file attached in the previous mail.
Thanks, patch applied.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim
->Object.Node ) ) {
I will send a patch that moves this NULL pointer check into the order function
once this is committed.
[...]
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : seb
[[ ${c} -eq 0 ]]; then
echo "Please run the timing tests and include the results."
fi
Is it really useful to add such timing values to the sources? They may be out
of date right with the next commit.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178
/
+lpc176x_gpio_interrupt_function function;
+} lpc176x_registered_interrupt_function;
So all in all is different to what we have, but we have so many styles in the
BSP area, it doesn't matter.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puch
See also
https://sourceware.org/ml/binutils/2014-06/msg00059.html
On Fri, Jun 06, 2014 at 11:01:10AM +0200, Sebastian Huber wrote:
> I performed a git bisect and found this:
>
> 93d1b056cb396d6468781fe0e40dd769891bed32 is the first bad commit
> commit 93d1b056cb396d6468781fe0e40d
00:00:00 2001
From: Martin Boretto
Date: Thu, 5 Jun 2014 13:04:37 -0300
Subject: [PATCH] Lpc176x bsp folder was added.
---
[...]
c/src/lib/libbsp/arm/lpc176x/times-tests.txt | 328
[...]
I will omit this file, since its to hard to keep it up to date.
--
Sebastian Huber, embedded
On 2014-06-04 20:48, Daniel Hellstrom wrote:
On 06/04/2014 11:51 AM, Sebastian Huber wrote:
On 2014-06-04 11:23, Daniel Hellstrom wrote:
Instead of calling the system call TA instruction directly it
is better paractise to isolate the trap implementation to the
system call functions.
The
On 2014-06-05 16:14, Daniel Gutson wrote:
Hi Sebastian,
On Thu, Jun 5, 2014 at 10:49 AM, Sebastian Huber
wrote:
>On 2014-06-05 15:20, Marcos Díaz wrote:
>>
>>At first we started developing this bsp inside that folder. Although
>>it has many similarities with the
e the default for ARM, since this can break existing
applications that rely on the 4k.
I think this new BSP specific setting makes it harder for users to know what is
going on. Since I don't have a better alternative, your patch is fine.
--
Sebastian Huber, embedded brains GmbH
Address :
C1700 directory, then we can
add your specialized BSP. My feeling is that this will generate more work in
the long run though.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian
general architecture
specifics that determine the typical stack usage.
Which problem does this new define address? Is this the test suite? In case
the tests pass with a stack size of 1024 on this BSP, then they will pass on
all ARM BSPs using the same instruction set.
--
Sebastian Huber
Hello Marcos,
is the LPC1768 so greatly different from the LPC1778 that you cannot integrate
it into the c/src/lib/libbsp/arm/lpc24xx family of BSPs with some additional
pre-processor defines?
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone
On 2014-06-04 12:55, Chris Johns wrote:
On 4/06/2014 4:28 pm, Sebastian Huber wrote:
On 2014-05-29 07:10, Chris Johns wrote:
Remove rtems_current_shell_env as this is dangerous because
the env can be NULL if used outside of a valid shell with the
POSIX key to an env set up.
Clean up the usage
() and bsp_reset()?
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key available on request.
Diese Nachricht ist keine
oop forever */ \
- } while (0)
+#define _CPU_Fatal_halt( _source, _error ) BSP_Fatal_halt( _source, _error )
I would name it _BSP_Fatal_halt(), similar to _BSP_Exception_frame_print().
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key available on request.
Diese Nachricht ist keine geschäftliche Mitteilung im
rtems_global_shell_env. I
underestimated its widespread use, so it was an error that we removed it. Now
it is a constant, so you can use it only for the initialization with default
values.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189
Rename _Scheduler_Update() to _Scheduler_Update_priority(). Add
parameter for the new thread priority to avoid direct usage of
Thread_Control::current_priority in the scheduler operation.
---
cpukit/score/include/rtems/score/scheduler.h | 14 ++
cpukit/score/include/rtems/scor
Replace _Scheduler_Allocate() with _Scheduler_Node_initialize(). Remove
the return status and thus the node initialization must be always
successful.
Rename _Scheduler_Free() to _Scheduler_Node_destroy().
---
cpukit/score/Makefile.am |7 ++-
cpukit/score/include/rte
On 2014-06-03 16:14, Gedare Bloom wrote:
On Tue, Jun 3, 2014 at 6:55 AM, Sebastian Huber
wrote:
>
>I think the bigger problem is that the priority inheritance affects only one
>level, see new test sptests/spsem03.
>
>
Is there a bug PR associated with this test? The behavior u
On 2014-06-02 15:32, Joel Sherrill wrote:
On Jun 2, 2014 9:28 AM, Sebastian Huber
wrote:
>
> On 2014-06-02 15:24, Joel Sherrill wrote:
> > Does this resource tracking replace the code for strict order mutex?
>
> No, the MrsP (and maybe later the Migratory Priority Inher
On 2014-06-03 01:58, Chris Johns wrote:
On 3/06/2014 3:37 am, Gedare Bloom wrote:
On Mon, Jun 2, 2014 at 10:48 AM, Sebastian Huber wrote:
Module:rtems
Branch:master
Commit:1fc2e960cea37e8d78e142c71faec18262f356d2
Changeset:
http://git.rtems.org/rtems/commit/?id
On 2014-06-02 15:47, Gedare Bloom wrote:
On Mon, Jun 2, 2014 at 4:35 AM, Sebastian Huber
wrote:
>On 2014-05-28 23:48, Gedare Bloom wrote:
>>>
>>>+/**
>>> >+ * @brief Returns true if this is the most recent resource of the node,
>>> >and
>>
I
am not sure if its worth to add this overhead. The implementation with the
resource count is analytically not correct, but very efficient. Was this
really a source for trouble in the last 20 years or so?
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim
On 2014-06-02 14:45, Joel Sherrill wrote:
On 6/2/2014 1:53 AM, Sebastian Huber wrote:
>On 2014-06-02 03:24, Joel Sherrill wrote:
>>On Jun 1, 2014 8:00 PM, Gedare Bloom wrote:
>> >
>> > On Sun, Jun 1, 2014 at 7:10 PM, Joel Sherrill wrote:
>> > > Mod
t loader. Maybe this is the point in which GPLv2
creates derivative work.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key avai
this object. So for the _Resource_Node_*()
functions this is a node. The term "recent" implies temporal order and thus is
well defined. If we use "first", "last", "top" or whatever then we would have
to explicitly specify the order which makes the name even lon
ur conventions too, eh?)
This change would affect code generation and would make it less efficient. The
_Scheduler_Is_priority_higher_than() calls a global function and thus owner
would have to be saved/restored around this call.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-
>>+ if ( lowest_scheduled &&
>>+ ( *order )( &thread->Object.Node, &lowest_scheduled->Object.Node ) ) {
>> Scheduler_SMP_Node *lowest_scheduled_node =
>> _Scheduler_SMP_Node_get( lowest_scheduled );
>>
>I think its possibl
-time for RTEMS should go back to the default of using
POSIX keys. That will fix it for SMP and eliminate some of this code.
Since the Ada self context is part of Thread_Control you can also use something
like __getreent().
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82
hp/Coding_Conventions
I included some of the recent discussion points we've had, and some
other points that come up often during code review.
Gedare
_______
rtems-devel mailing list
rtems-devel@rtems.org
http://www.rtems.org/mailman/listinfo/rtems-devel
vb/startup/idle-thread.c:30
30 mpc55xx_wait_for_interrupt();
(gdb)
Now when I continue I will hit the breakpoint in bsp_reset:
Maybe the debugger has a problem with the wait instruction?
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89
mProbe(void *addr, int write, int size, void
*pval);
> >
> We should generally prefer to declare functions in a header file, or
> make them static if not exported.
Got a suggestion? Did I miss an existing prototype?
I also think that adding prototypes to source files to silence thi
On 05/30/2014 04:43 PM, Joel Sherrill wrote:
On 5/30/2014 8:33 AM, Sebastian Huber wrote:
>On 05/29/2014 09:28 PM, Joel Sherrill wrote:
>>Hi
>>
>>The priority affinity algorithm appears to be behaving as
>>we expect from a decision making standpoint. However,
>>
thread
context and is maintained at the lowest context switch level. For
proper migration the scheduler must ensure that,
1. an heir thread other than the migrating thread exists on the source
processor, and
2. the migrating thread is the heir thread on the destination processor.
--
Sebastian
ocument that describes
the coding style in detail. It can also be used to judge code contributions.
A "avoidance of deep nesting by using early returns" rule should cover what
happens if it leads to duplicated clean up code.
--
Sebastian Huber, embedded brains GmbH
Address : Dorniers
Do not change the scheduler with this function. Documentation. Coding
style.
---
cpukit/rtems/include/rtems/rtems/tasks.h | 11 +
cpukit/score/include/rtems/score/schedulerimpl.h | 10 +
cpukit/score/src/schedulersetaffinity.c | 52 +++---
doc/user/
Drop scheduler parameter. Coding style.
---
cpukit/posix/src/pthreadgetaffinitynp.c |1 -
cpukit/rtems/src/taskgetaffinity.c |1 -
cpukit/score/include/rtems/score/schedulerimpl.h |7 +--
cpukit/score/src/schedulergetaffinity.c | 44 ++
On 2014-05-20 20:43, Sebastian Huber wrote:
On 05/20/2014 05:30 PM, Joel Sherrill wrote:
On 5/20/2014 10:17 AM, Sebastian Huber wrote:
>On 2014-05-20 16:58, Joel Sherrill wrote:
>>On 5/13/2014 1:16 AM, Sebastian Huber wrote:
>>>>On 2014-05-13 01:28, Joel Sher
---
cpukit/rtems/src/schedulergetprocessorset.c | 28 +++---
1 files changed, 12 insertions(+), 16 deletions(-)
diff --git a/cpukit/rtems/src/schedulergetprocessorset.c
b/cpukit/rtems/src/schedulergetprocessorset.c
index e459b17..016c368 100644
--- a/cpukit/rtems/src/schedu
Provide rtems_global_shell_env as a public read-only variable for
backward compatibility.
---
cpukit/libmisc/shell/shell.c |2 +-
cpukit/libmisc/shell/shell.h |2 ++
2 files changed, 3 insertions(+), 1 deletions(-)
diff --git a/cpukit/libmisc/shell/shell.c b/cpukit/libmisc/shell/shell.c
i
---
cpukit/posix/src/pthread.c|4 ++--
cpukit/rtems/src/tasksetpriority.c|2 +-
cpukit/score/include/rtems/score/threadimpl.h | 17 +
cpukit/score/src/coremutexsurrender.c |2 +-
cpukit/score/src/threadrestart.c |
---
cpukit/score/include/rtems/score/thread.h |8
cpukit/score/include/rtems/score/threadimpl.h | 10 ++
cpukit/score/src/threadinitialize.c |2 ++
3 files changed, 20 insertions(+), 0 deletions(-)
diff --git a/cpukit/score/include/rtems/score/thread.h
b/
A resource is something that has at most one owner at a time and may
have multiple rivals in case an owner is present. The owner and rivals
are impersonated via resource nodes. A resource is represented via the
resource control structure. The resource controls and nodes are
organized as trees.
This enables proper resource dependency tracking and as a side-effect
deadlock detection.
---
cpukit/score/include/rtems/score/mrsp.h | 14 +-
cpukit/score/include/rtems/score/mrspimpl.h | 127 +---
doc/user/sem.t | 11 +-
testsuites/smptests/smpmrsp01
backfiles and TAGS.
Could you use a private ~/.gitignore for this development environment specific
stuff?
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded
Perhaps this should be encapsulated by a function call, considering it
is copied in some other start hooks?
The real problem is that this code is copied at all. We have to clean this up.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +
optimization disabled. I think this is an
extreme case with respect to the stack usage.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP
-users.cs.york.ac.uk/~burns/
http://www-users.cs.york.ac.uk/~burns/MRSPpaper.pdf
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key
duler simulator throughout the code base.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key available on request.
Diese Nachrich
buf or only the lock/condition
variables?
I do not think we need more POSIX enable work arounds.
Chris
___
rtems-devel mailing list
rtems-devel@rtems.org
http://www.rtems.org/mailman/listinfo/rtems-devel
--
Sebastian Huber, embedded brains GmbH
Addres
On 2014-05-27 16:47, Gedare Bloom wrote:
This spcache01 test is getting quite large. Does it make sense to split it up?
No, since in this case you have to do the same cache type detection multiple
times.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim
driver solved? Otherwise we
must not apply this patch series.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key available on
ion. You get
a fatal error at run-time. A compile-time error is bad, since this would break
the test suite build.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@
Do not use the Per_CPU_Control::started in
_SMP_Start_multitasking_on_secondary_processor() since this field may be
not up to date when a secondary processor reads it. Use the read-only
scheduler assignment instead.
Add a new fatal error SMP_FATAL_MULTITASKING_START_ON_INVALID_PROCESSOR.
This pre
On 2014-05-23 16:02, Joel Sherrill wrote:
This looks ok.
If no one else comments, I will commit these two.
Its fine for me.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail
differently in the fatal exception handler.
Thanks, patch checked in.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key available on
v2: Fixes after review by Gedare.
v3: Do not allow nested access for MrsP semaphores.
___
rtems-devel mailing list
rtems-devel@rtems.org
http://www.rtems.org/mailman/listinfo/rtems-devel
Elevate the priority of the creating task to the ceiling priority in
case a semaphore is created as initially locked.
---
cpukit/score/src/coremutex.c |5 +++--
testsuites/sptests/sp65/init.c | 35 +++
testsuites/sptests/sp65/sp65.doc |2 ++
3 files
Elevate the priority of the creating task to the ceiling priority in
case a semaphore is created as initially locked.
---
cpukit/score/src/coremutex.c | 19 ---
testsuites/sptests/sp65/init.c | 35 +++
testsuites/sptests/sp65/sp65.doc |
Elevate the priority of the creating task to the ceiling priority in
case a semaphore is created as initially locked.
---
cpukit/score/src/coremutex.c | 14 +-
testsuites/sptests/sp65/init.c | 35 +++
testsuites/sptests/sp65/sp65.doc |2 ++
19:06, Joel Sherrill wrote:
On 5/21/2014 9:13 AM, Sebastian Huber wrote:
On 2014-05-21 16:00, Joel Sherrill wrote:
Hi
We have an SMP behavioral decision to make and it isn't getting
enough discussion.
With cluster scheduling, there are potentially multiple scheduler
instances associated
On 2014-05-21 17:14, Sebastian Huber wrote:
+static rtems_status_code _Semaphore_Set_priorities(
Ok, next time its really fixed to _Semaphore_Set_priority().
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax
On 2014-05-21 16:53, Gedare Bloom wrote:
On Wed, May 21, 2014 at 8:09 AM, Sebastian Huber
wrote:
Add basic support for the Multiprocessor Resource Sharing Protocol
(MrsP).
The Multiprocessor Resource Sharing Protocol (MrsP) is defined in A.
Burns and A.J. Wellings, A Schedulability
. If it does not, the mutex is not
initialized.
Yes, this is the error case. In the good case the priority is not changed to
the ceiling priority like for a normal rtems_semaphore_obtain() (see
_CORE_mutex_Seize_interrupt_trylock_body()).
--
Sebastian Huber, embedded brains GmbH
Address : D
itialize(
&the_mutex->Wait_queue,
_CORE_mutex_Is_fifo( the_mutex_attributes ) ?
THREAD_QUEUE_DISCIPLINE_FIFO : THREAD_QUEUE_DISCIPLINE_PRIORITY,
STATES_WAITING_FOR_MUTEX,
CORE_MUTEX_TIMEOUT
);
return CORE_MUTEX_STATUS_SUCCESSFUL;
}
Is this working as intended?
--
Seb
ken is the _Scheduler_priority_affinity_SMP_Set_affinity() function.
Please have a look at the attached patch which I already sent to a similar
thread.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41
On 2014-05-21 15:44, Gedare Bloom wrote:
On Wed, May 21, 2014 at 4:37 AM, Daniel Hellstrom wrote:
On 05/21/2014 10:01 AM, Sebastian Huber wrote:
Hello Gedare,
thanks for your review.
On 2014-05-20 16:58, Gedare Bloom wrote:
@@ -54,7 +58,15 @@ SYM(_CPU_Context_switch):
ld
On 2014-05-20 17:32, Joel Sherrill wrote:
You asked earlier. I think changing affinity should work find from an
ISR. It is
just set state/clear state.
Does changing affinity include a scheduler change for you?
Scheduler changes from interrupt context make it quite complicated.
--
Sebastian
Hello Joel,
please have a look at the attached patch.
On 2014-05-20 23:24, Joel Sherrill wrote:
On 5/19/2014 1:34 AM, Sebastian Huber wrote:
>On 2014-05-16 19:48, Joel Sherrill wrote:
>>Hi
>>
>>Questions first:
>>
>>+ The affinity mask must be non-empty. This
spect to the usage of assembler macros in
this file?
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key available on request.
---
cpukit/rtems/src/semobtain.c |6 --
1 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/cpukit/rtems/src/semobtain.c b/cpukit/rtems/src/semobtain.c
index 697dde8..c9433ca 100644
--- a/cpukit/rtems/src/semobtain.c
+++ b/cpukit/rtems/src/semobtain.c
@@ -41,18 +41,20 @@ rtems_s
On 05/20/2014 05:30 PM, Joel Sherrill wrote:
On 5/20/2014 10:17 AM, Sebastian Huber wrote:
>On 2014-05-20 16:58, Joel Sherrill wrote:
>>On 5/13/2014 1:16 AM, Sebastian Huber wrote:
>>>>On 2014-05-13 01:28, Joel Sherrill wrote:
>>>>>>Hi
:13 AM, Joel Sherrill wrote:
Hi
I was looking for a simple obvious method to do this
and I don't see it.
I backed out from the per cpu pointer to an index
but thought there had to be a better way.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Ge
On 2014-05-20 16:58, Joel Sherrill wrote:
On 5/13/2014 1:16 AM, Sebastian Huber wrote:
>On 2014-05-13 01:28, Joel Sherrill wrote:
>>Hi
>>
>>Both schedulerident.c and schedulergetprocessorset.c do not follow
>>RTEMS Coding Conventions on avoidance of deep nesting by
On 2014-05-20 08:18, Sebastian Huber wrote:
What happens if you end up with a combination of these
actions at the same time? restart, scheduler change, affinity change.
Currently a scheduler change must not be issued in interrupt context. Apart
from this everything works fine even if you
nities?
Ok, so you plan to add a scheduler that supports arbitrary thread processor
affinities and this scheduler will assign a processor to the highest priority
threads provided the affinity sets of all threads are well chosen
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr
ns at the same time? restart, scheduler change, affinity change.
Currently a scheduler change must not be issued in interrupt context. Apart
from this everything works fine even if you restart a thread during a scheduler
change since thread dispatching is disabled.
--
Sebastian Huber, embedded b
On 2014-05-19 16:37, Gedare Bloom wrote:
On Thu, May 15, 2014 at 8:06 AM, Sebastian Huber
wrote:
>There is now an example available that shows how threads are scheduled with
>the SMP scheduler support:
>
>http://www.rtems.org/onlinedocs/doxygen/cpukit/html/group__ScoreSched
n
_Thread_Set_state( the_thread, STATES_MIGRATING );
update affinity
_Thread_Clear_state( the_thread, STATES_MIGRATING );
Comments?
Yes, this is how I would do it. Should interrupt service routines be able to
change the scheduler or the affinity set?
--
Sebastian Huber, embedde
to the begin of the
structure to be next to the context pointer.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key available on
On 2014-05-16 19:48, Joel Sherrill wrote:
On 5/16/2014 7:09 AM, Sebastian Huber wrote:
>Hello Joel,
>
>I think this rtems_test_* stuff is a hack. What we really need in the long run
>is a proper test framework.
I don't disagree but we have to have confidence in what is don
On 2014-05-17 08:04, dream wrote:
In rtems\c\src\lib\libbsp\arm\shared\start\start.S file, I think "ldr sp,
=bsp_stack_main_end " can delete,because when system power or reset, in vector
table bsp_stack_main_end will copy to MSP.
What happens if you start it via a boot loader?
--
This started with an examination of leon3 test output.
At least sp39 can exit with a "END OF" message and
a failure message. That needs to be fixed and is pretty
easy.
Yes, tests printing "END OF" in case of a test failure are broken.
Any thoughts on all this?
--
Sebastian Hu
e open issues on the SMP wiki page are all quite complex. One exception is
the Termios framework, but this is quite a boring job:
http://www.rtems.org/wiki/index.php/SMP#Termios_Framework
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49
There is now an example available that shows how threads are scheduled with the
SMP scheduler support:
http://www.rtems.org/onlinedocs/doxygen/cpukit/html/group__ScoreSchedulerSMP.html
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89
On 2014-05-15 13:49, Sebastian Huber wrote:
On 2014-05-14 20:38, Joel Sherrill wrote:
Hi
All of the SPARC BSPs are OK with --enable-smp but at least
psim doesn't link when it is enabled. I haven't checked ARM
but I suspect that some default magic needs to be in place
in the Mak
attached patch. If you want SMP support for
every BSP, then we need a solution with less copy and paste.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedde
e core/node that
the priority was changed on so we can't assume that
the processor ID is easily available.
So there are two hard things I don't see an obvious answer to.
I think this path has a decision.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, G
twork stack itself.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key available on request.
Diese Nachricht ist keine geschäft
On 05/14/2014 05:40 PM, Joel Sherrill wrote:
On 5/14/2014 10:32 AM, Sebastian Huber wrote:
The function to change a thread priority was too complex. Simplify it
with a new scheduler operation. This increases the average case
performance due to the simplified logic. The interrupt disabled
On 2014-05-14 14:44, Sebastian Huber wrote:
With Git revision 3f3d22271129993b865c25057b784dce13ffa778 all tests except
- SP2038,
- SPINTRCRITICAL 10, and
- SMPLOAD 1.
complete successfully on NGMP in the SMP configuration.
Sorry, the revision is wrong, it is
patches in this series looks good. Thanks!
Haven't tested it though. Please merge it since we will temporary branch from
mainline the coming days for the SMP project.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-1
Use __bss_start available via %g2 to clear the BSS section. The usage
of _edata resulted in a copy of [_edata, __bss_start) from ROM to RAM
and then a clear to zero of this area.
Clear now only [__bss_start, _end).
---
c/src/lib/libbsp/sparc/shared/start/start.S |2 --
1 files changed, 0 ins
1 - 100 of 1175 matches
Mail list logo
