On 25/06/15 18:18, ragunath wrote:
This patch includes changes for porting cpsw ethernet driver for Beaglebone
black from Freebsd.
It also includes smsc phy ported from FreeBSD. Corresponding nexus devices are
added.
For networking to work properly, two additional changes are needed on the RT
On 26/06/15 02:33, Saurabh Gadia wrote:
@joel: It essentially means finding out the highest priority thread
(or ceiling) of any remaining mutex held, right?
If you are talking this in context of nested mutex problem --->
We never go and find the highest priority thread waiting on held
mutex.
On 24/06/2015 3:04 pm, Gene Smith wrote:
> modified: wscript
> Re-generated from "./freebsd-to-rtems -m" based on
> waf_generator.py
Pushed.
Thanks.
Chris
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel
You can configure the test environment according to your needs but 3 mutex
are sufficient to cover all test cases and we can have multiple threads
using verify() of JPF.
Thanks,
Saurabh Gadia
On Thu, Jun 25, 2015 at 6:37 PM, Saurabh Gadia wrote:
> Hi all,
> For time being I have cleaned up the
Hi all,
For time being I have cleaned up the code. I have also attached the trace
files where our JPF-Model can reproduce the RTEMS unbounded priority
inheritance problem(TraceWupi). And after running it through solution model
we are able to demonstrate that there are no assertion errors and it run
O(n) is the least that we can have!!
Thanks,
Saurabh Gadia
On Thu, Jun 25, 2015 at 5:39 PM, Saurabh Gadia wrote:
> I mean not recursion. just looping till top of stack of held mutexes.
> check this out:
> https://github.com/saurabhgadia4/lock-model/blob/master/rtems/Mutex.java
> -->updateRecPr
Ok. should I will have complete trace files on drive. okkk.. I will upload
the trace and then mention about why we got error and how we solved the
problem of unbounded priorities.
Thanks,
Saurabh Gadia
On Thu, Jun 25, 2015 at 5:40 PM, Cyrille Artho
wrote:
> Yes, let's discuss this in more dep
I mean not recursion. just looping till top of stack of held mutexes.
check this out:
https://github.com/saurabhgadia4/lock-model/blob/master/rtems/Mutex.java
-->updateRecPriority() line 143
Thanks,
Saurabh Gadia
On Thu, Jun 25, 2015 at 5:35 PM, Joel Sherrill
wrote:
>
>
> On June 25, 2015 7:33
On June 25, 2015 7:33:24 PM CDT, Saurabh Gadia wrote:
>@joel: It essentially means finding out the highest priority thread (or
>ceiling) of any remaining mutex held, right?
>
>If you are talking this in context of nested mutex problem --->
>
>We never go and find the highest priority thread wai
@joel: It essentially means finding out the highest priority thread (or
ceiling) of any remaining mutex held, right?
If you are talking this in context of nested mutex problem --->
We never go and find the highest priority thread waiting on held mutex.
Inversely highest priority thread will recursi
On 25/06/2015 11:22 pm, Sebastian Huber wrote:
> On 25/06/15 15:21, Yurii Shevtsov wrote:
>> What kind of debugging info do you mean? What else can I get except
>> serial output? Can I make it verbose?
>
> I would use a debugger and not serial output.
>
This is not easy on a RPi. The JTAG signal
On 25/06/2015 11:17 pm, Sebastian Huber wrote:
> The problem is that empty structures have a different size in C and C++.
> ---
> cpukit/score/include/rtems/score/percpu.h | 17 ++-
> cpukit/score/include/rtems/score/smplock.h | 167
> +
> cpukit/score/src/profilings
On 26/06/2015 4:29 am, Joel Sherrill wrote:
> Stack protection must be disabled (--disable-stack-protection) must
> be disabled to avoid bugs in older toolsets. If this can be made
> conditional somehow, it is supposed to be needed on CentOS 5 and 6.
> There are at least two patches in Qemu Patchwo
On 25-06-2015 14:33, Gedare Bloom wrote:
Ask what other folks using the RPi are doing.
I do not know if anyone using the Pi with RTEMS currently has a debugger
working. Last year Alan posted in his blog [1] a configuration using a
FT2232H mini module and OpenOCD that connects to the Pi throug
Hello,
(sorry for the long email)
I am currently doing an implementation for a generic RTEMS API, and I
would like to inquire potential users about pin references provided by
the API to user applications.
Most platforms use a "global" pin number to refer to a GPIO pin in
documentation, usua
Stack protection must be disabled (--disable-stack-protection) must
be disabled to avoid bugs in older toolsets. If this can be made
conditional somehow, it is supposed to be needed on CentOS 5 and 6.
There are at least two patches in Qemu Patchworks which are supposed
to fix this but they do not f
Hi Gedare,
I will send them separately.
Thanks,
Ragunath
On Thu, Jun 25, 2015 at 10:29 PM, Gedare Bloom wrote:
> Can you split the commits so that the import from BSD is separate from
> your modifications to it?
>
> On Thu, Jun 25, 2015 at 12:18 PM, ragunath wrote:
>> This patch includes cha
On Thu, Jun 25, 2015 at 12:20 PM, ragunath wrote:
> This patch has two changes that are needed for networking to work in BBB.
> We disable cache as it is causing random values to be learned in the cpsw
> Address
> Lookup Engine(ALE) causing tx to fail. Vector enable is done after handler is
> ca
Can you split the commits so that the import from BSD is separate from
your modifications to it?
On Thu, Jun 25, 2015 at 12:18 PM, ragunath wrote:
> This patch includes changes for porting cpsw ethernet driver for Beaglebone
> black from Freebsd.
> It also includes smsc phy ported from FreeBSD.
This patch has two changes that are needed for networking to work in BBB.
We disable cache as it is causing random values to be learned in the cpsw
Address
Lookup Engine(ALE) causing tx to fail. Vector enable is done after handler is
called by the server task.
---
c/src/lib/libbsp/arm/beagle/i
This patch includes changes for porting cpsw ethernet driver for Beaglebone
black from Freebsd.
It also includes smsc phy ported from FreeBSD. Corresponding nexus devices are
added.
For networking to work properly, two additional changes are needed on the RTEMS
side.
---
Makefile
On 25/06/15 15:32, Gedare Bloom wrote:
This code is quite a pain to read with the interleaved ifdefs and I'm
not that happy about it, but I don't have any better suggestions right
now. The set of patches looks fine to commit.
The interleaved ifdefs are only in this file, the lock user don't s
Ask what other folks using the RPi are doing.
On Thu, Jun 25, 2015 at 9:24 AM, Yurii Shevtsov wrote:
> How to set a break point? Is there any other way of debugging except
> printfs and tracing?
>
> 2015-06-25 16:00 GMT+03:00 Sebastian Huber
> :
>> I would set a break point to nexus_probe(). In
This code is quite a pain to read with the interleaved ifdefs and I'm
not that happy about it, but I don't have any better suggestions right
now. The set of patches looks fine to commit.
On Thu, Jun 25, 2015 at 9:17 AM, Sebastian Huber
wrote:
> The problem is that empty structures have a differen
How to set a break point? Is there any other way of debugging except
printfs and tracing?
2015-06-25 16:00 GMT+03:00 Sebastian Huber :
> I would set a break point to nexus_probe(). In this loop
>
> SET_FOREACH(nd, nexus) {
> device_add_child(dev, nd->name, nd->unit);
> }
>
> your d
On 6/25/2015 8:22 AM, Sebastian Huber wrote:
On 25/06/15 15:20, Joel Sherrill wrote:
The comment doesn't appear to match the change.
The problem is that the extern "C" covers the #include
.
OK. I missed that. I obviously need more caffeine this
morning.
On 6/25/2015 8:17 AM, Sebastian
On 25/06/15 15:21, Yurii Shevtsov wrote:
What kind of debugging info do you mean? What else can I get except
serial output? Can I make it verbose?
I would use a debugger and not serial output.
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone
On 25/06/15 15:20, Joel Sherrill wrote:
The comment doesn't appear to match the change.
The problem is that the extern "C" covers the #include
.
On 6/25/2015 8:17 AM, Sebastian Huber wrote:
---
cpukit/score/include/rtems/score/heap.h | 3 +--
1 file changed, 1 insertion(+), 2 deletio
What kind of debugging info do you mean? What else can I get except
serial output? Can I make it verbose?
2015-06-25 16:00 GMT+03:00 Gedare Bloom :
> Can you add more debugging info?
>
> I glanced at your code, some comments, make sure you follow the
> recommendations in libbsd.txt about modifying
It puts the include before extern C which is appropriate
On Thu, Jun 25, 2015 at 9:20 AM, Joel Sherrill
wrote:
> The comment doesn't appear to match the change.
>
> On 6/25/2015 8:17 AM, Sebastian Huber wrote:
>>
>> ---
>> cpukit/score/include/rtems/score/heap.h | 3 +--
>> 1 file changed, 1 i
The comment doesn't appear to match the change.
On 6/25/2015 8:17 AM, Sebastian Huber wrote:
---
cpukit/score/include/rtems/score/heap.h | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/cpukit/score/include/rtems/score/heap.h
b/cpukit/score/include/rtems/score/heap.h
ind
The problem is that empty structures have a different size in C and C++.
---
cpukit/score/include/rtems/score/percpu.h | 17 ++-
cpukit/score/include/rtems/score/smplock.h | 167 +
cpukit/score/src/profilingsmplock.c| 18 ++--
cpukit/score/src/smp.c
Drop the include since this file exposes a lot of
implementation details.
---
cpukit/posix/src/pthreadself.c | 3 +--
cpukit/score/include/rtems/score/userextimpl.h | 1 +
cpukit/score/include/rtems/system.h| 2 +-
cpukit/score/src/corespinlockrelease.c | 2 +-
Drop the include since this file exposes a lot of
implementation details.
---
c/src/lib/libbsp/powerpc/mvme5500/vectors/exceptionhandler.c | 1 +
c/src/lib/libbsp/shared/clockdrv_shell.h | 1 +
c/src/lib/libbsp/sparc/shared/timer/tlib_ckinit.c| 1 +
c/src/lib/libcp
---
cpukit/score/Makefile.am | 1 +
cpukit/score/include/rtems/score/isr.h | 36 +---
cpukit/score/src/isrisinprogress.c | 51 ++
3 files changed, 53 insertions(+), 35 deletions(-)
create mode 100644 cpukit/score/src/isrisinpr
---
cpukit/score/include/rtems/score/heap.h | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/cpukit/score/include/rtems/score/heap.h
b/cpukit/score/include/rtems/score/heap.h
index ab627b6..5db22cf 100644
--- a/cpukit/score/include/rtems/score/heap.h
+++ b/cpukit/score/includ
Avoid Thread_Control typedef in . This helps to
get rid of the include in
which exposes a lot of implementation details.
---
cpukit/posix/include/rtems/posix/psignalimpl.h | 1 +
cpukit/rtems/include/rtems/rtems/signalimpl.h | 1 +
cpukit/score/include/rtems/score/percpu.h | 37
Drop the include since this file exposes a lot of
implementation details.
---
cpukit/libmisc/stackchk/check.c | 1 +
cpukit/libmisc/stackchk/stackchk.h | 2 +-
testsuites/libtests/stackchk/blow.c | 1 +
3 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/cpukit/libmisc/stackchk/che
Can you add more debugging info?
I glanced at your code, some comments, make sure you follow the
recommendations in libbsd.txt about modifying code taken from freebsd.
You also might find it more convenient to work on a branch instead of
a master, and to avoid merge commits.
Gedare
On Thu, Jun 2
I would set a break point to nexus_probe(). In this loop
SET_FOREACH(nd, nexus) {
device_add_child(dev, nd->name, nd->unit);
}
your device must get added. I would also set break points to the probe
and attach functions of your device.
On 25/06/15 14:50, Yurii Shevtsov wrote:
This is ping message, with small update: the problem is not on the
linking stage, driver is linked to testsuite (checked with objdump)
2015-06-21 17:57 GMT+03:00 Yurii Shevtsov :
> Hello)
> Now I have apps from libbsd testsuite running. But DWC OTG driver doesn't
> loads.
> I added this lines to
On 24-06-2015 15:03, Gedare Bloom wrote:
On Tue, Jun 23, 2015 at 6:44 PM, André Marques
wrote:
On 23-06-2015 17:16, Gedare Bloom wrote:
+ uint32_t bank;
+ int handled_count;
+ int rv;
+
+ gpio = (gpio_pin*) arg;
+
+ bank = gpio->bank_number;
Validate args for errors.
These tasks are
42 matches
Mail list logo
