Re: Release task request for help

[Niteesh G. S.]

On Tue, Apr 28, 2020 at 8:10 AM Chris Johns  wrote:

> On 28/4/20 11:51 am, Niteesh G. S. wrote:
> >
> https://github.com/gs-niteesh/rtems-release/commit/8666bcd66066512f91194660817cd42cbb8c76e9
> > If it is okay for you, then I will send the patch to the mailing list.
>
> The repos list is now just the package names and the branch script as it
> is expects the repo name, i.e. .git at the end.
>

How should I do this? Shall I iterate over all the repos in rtems_repo
variable
and append '.git' to the end? This will happen right after the declaration
of the
variable.

Also the rtems-releases repo needs to be added to the list of repos (and
> not packages).
>

Done.
We should avoid branching this repo, right?


> I think it would be a good idea to be able to run this script which
> means a way to switch the git clone from ssh to the git protocol. I
> recently added command line support to rtems-release so I suggest you
> borrow that code and add an option so you can switch the git protocol
> and test. I would also add a switch to push, i.e. by default a push is
> not done.
>

Should we take the URL from the user similar to how it is done in
rtems-release
or just ask if he wants to test? In this case, the user has to just pass -t
or something
to cmd line and it would download the sources from git://git.rtems.org/



> After we have the branch script the tag script needs to be updated.
>
> Chris
>
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: Release task request for help

[Chris Johns]

On 28/4/20 6:26 pm, Niteesh G. S. wrote:
On Tue, Apr 28, 2020 at 8:10 AM Chris Johns > wrote:


On 28/4/20 11:51 am, Niteesh G. S. wrote:
 >

https://github.com/gs-niteesh/rtems-release/commit/8666bcd66066512f91194660817cd42cbb8c76e9
 > If it is okay for you, then I will send the patch to the mailing
list.

The repos list is now just the package names and the branch script
as it
is expects the repo name, i.e. .git at the end.


How should I do this? Shall I iterate over all the repos in rtems_repo 
variable
and append '.git' to the end? This will happen right after the 
declaration of the

variable.


Yes this is fine.


Also the rtems-releases repo needs to be added to the list of repos
(and
not packages).

Done.
We should avoid branching this repo, right?


No lets branch it. We might as well be consistent in how we manage repos 
in the project. It makes procedures in the eng manual easier, i.e. just 
one procedure to cover them all.



I think it would be a good idea to be able to run this script which
means a way to switch the git clone from ssh to the git protocol. I
recently added command line support to rtems-release so I suggest you
borrow that code and add an option so you can switch the git protocol
and test. I would also add a switch to push, i.e. by default a push is
not done.

Should we take the URL from the user similar to how it is done in 
rtems-release
or just ask if he wants to test? In this case, the user has to just pass 
-t or something
to cmd line and it would download the sources from git://git.rtems.org/ 



Hmm lets start with an option to switch between the git and ssh 
protocols. I cannot see a use case where there is a completely different 
location for the repos. If that does arise it can be sorted out then.


Chris
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: [PATCH v3] updates #3889, Added test for timer_create() using CLOCK_MONOTONIC

[Eshan Dhawan]

Hello everyone,

I sent this patch to add test for POSIX timer_create for
clock_id argument as CLOCK_MONOTONIC.
Other details are in the commit message.
If you all could review it.

thanks
-Eshan


On Mon, Apr 20, 2020 at 12:00 AM Eshan dhawan 
wrote:

> >I will create a timer using timer_create() function,
> and passing CLOCK_MONOTONIC in clock_id argument.
>
> >The test will fail as there is no support for
> CLOCK_MONOTONIC in cpukit/posix/src/psxtimercreate.c.
>
> >The added code compiles successfully without any errors.
>
> Signed-off-by: Eshan dhawan 
> ---
>  testsuites/psxtests/psxtimer02/psxtimer.c | 8 ++--
>  testsuites/psxtests/psxtimer02/psxtimer02.scn | 3 ++-
>  2 files changed, 8 insertions(+), 3 deletions(-)
>
> diff --git a/testsuites/psxtests/psxtimer02/psxtimer.c
> b/testsuites/psxtests/psxtimer02/psxtimer.c
> index 9f79d33c42..e071f98857 100644
> --- a/testsuites/psxtests/psxtimer02/psxtimer.c
> +++ b/testsuites/psxtests/psxtimer02/psxtimer.c
> @@ -62,9 +62,9 @@ void *POSIX_Init (
>status = timer_create( CLOCK_REALTIME, &event, NULL );
>fatal_posix_service_status_errno( status, EINVAL, "bad timer id" );
>
> -  puts( "timer_create - OK" );
> +  puts( "timer_create (CLOCK_REALTIME) - OK" );
>status = timer_create( CLOCK_REALTIME, NULL, &timer );
> -  posix_service_failed( status, "timer_create OK" );
> +  posix_service_failed( status, "timer_create (CLOCK_REALTIME)" );
>
>puts( "timer_create - too many - EAGAIN" );
>status = timer_create( CLOCK_REALTIME, NULL, &timer1 );
> @@ -127,6 +127,10 @@ void *POSIX_Init (
>status = timer_delete( timer );
>fatal_posix_service_status_errno( status, EINVAL, "bad id" );
>
> +  puts( "timer_create (CLOCK_MONOTONIC) - OK" );
> +  status = timer_create( CLOCK_MONOTONIC, NULL, &timer );
> +  posix_service_failed( status, "timer_create (CLOCK_MONOTONIC)" );
> +
>TEST_END();
>rtems_test_exit (0);
>  }
> diff --git a/testsuites/psxtests/psxtimer02/psxtimer02.scn
> b/testsuites/psxtests/psxtimer02/psxtimer02.scn
> index e78425a32e..7429bcf291 100644
> --- a/testsuites/psxtests/psxtimer02/psxtimer02.scn
> +++ b/testsuites/psxtests/psxtimer02/psxtimer02.scn
> @@ -1,7 +1,7 @@
>  *** POSIX Timers Test 02 ***
>  timer_create - bad clock id - EINVAL
>  timer_create - bad timer id pointer - EINVAL
> -timer_create - OK
> +timer_create (CLOCK_REALTIME) - OK
>  timer_create - too many - EAGAIN
>  timer_delete - bad id - EINVAL
>  timer_getoverrun - bad id - EINVAL
> @@ -13,4 +13,5 @@ timer_settime - bad itimer value - negative nanosecond -
> EINVAL
>  timer_settime - bad clock value - EINVAL
>  timer_delete - OK
>  timer_delete - bad id - EINVAL
> +timer_create (CLOCK_MONOTONIC) - OK
>  *** END OF POSIX Timers Test 02 ***
> --
> 2.17.1
>
>
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: Release task request for help

[Niteesh G. S.]

This is my first time writing shell scripts. If something seems like
a hack or if there is a better of doing it. Please let me know.

https://github.com/gs-niteesh/rtems-release/commit/70cbb5418ba68992a910844c3d6450ef4db17579
https://github.com/gs-niteesh/rtems-release/commit/58ed0cb3fc1f33f12de8e550979371a51e322754

On Tue, Apr 28, 2020 at 3:51 PM Chris Johns  wrote:

> On 28/4/20 6:26 pm, Niteesh G. S. wrote:
> > On Tue, Apr 28, 2020 at 8:10 AM Chris Johns  > > wrote:
> >
> > On 28/4/20 11:51 am, Niteesh G. S. wrote:
> >  >
> >
> https://github.com/gs-niteesh/rtems-release/commit/8666bcd66066512f91194660817cd42cbb8c76e9
> >  > If it is okay for you, then I will send the patch to the mailing
> > list.
> >
> > The repos list is now just the package names and the branch script
> > as it
> > is expects the repo name, i.e. .git at the end.
> >
> >
> > How should I do this? Shall I iterate over all the repos in rtems_repo
> > variable
> > and append '.git' to the end? This will happen right after the
> > declaration of the
> > variable.
>
> Yes this is fine.
>
> > Also the rtems-releases repo needs to be added to the list of repos
> > (and
> > not packages).
> >
> > Done.
> > We should avoid branching this repo, right?
>
> No lets branch it. We might as well be consistent in how we manage repos
> in the project. It makes procedures in the eng manual easier, i.e. just
> one procedure to cover them all.
>
> > I think it would be a good idea to be able to run this script which
> > means a way to switch the git clone from ssh to the git protocol. I
> > recently added command line support to rtems-release so I suggest you
> > borrow that code and add an option so you can switch the git protocol
> > and test. I would also add a switch to push, i.e. by default a push
> is
> > not done.
> >
> > Should we take the URL from the user similar to how it is done in
> > rtems-release
> > or just ask if he wants to test? In this case, the user has to just pass
> > -t or something
> > to cmd line and it would download the sources from git://git.rtems.org/
> > 
>
> Hmm lets start with an option to switch between the git and ssh
> protocols. I cannot see a use case where there is a completely different
> location for the repos. If that does arise it can be sorted out then.
>
> Chris
>
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

RTEMS 5 Release status

[Chris Johns]

Hello,

We are down to 4 remaining tickets. Many thanks to everyone for all the 
hard work getting to this point. It is appreciated.


The remaining tickets need to be closed or moved before I can branch the 
repos. The milestone page is ...


https://devel.rtems.org/milestone/5.1

... so please have a look and help if you can.

Release Change Notes


Joel has asked for a high level list of changes to be added to the 
release information. He would like this to be a community effort.


We have the detailed release notes created from the Trac tickets for the 
release so this is suppose to be something high level.


Please create or add to a file called rtems-notes-5.txt as markdown in 
the releases repo ...


 https://git.rtems.org/rtems-release/

Either push the changes or send a patch to this list and I will merge 
them. I will add support to place this text in the generated README once 
it exists. If there is no file at the time of the release no support 
will be added.


Tickets
---

New:

https://devel.rtems.org/ticket/3962

I will not be involved in this ticket or driving it and will wait for it 
to be closed or moved before we branch. Please help.


Assigned:

https://devel.rtems.org/ticket/2962

This ticket has no owner. Please review this ticket and help out. I 
added to the comments that heapwalk with RTEMS_DEBUG fails. Is this 
important?


https://devel.rtems.org/ticket/3813

I am working on this and should have it resolved soon.

Reopened:

https://devel.rtems.org/ticket/3800

I do not know the status of this ticket or what should happen.

Thanks
Chris
___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel

Re: RTEMS 5 Release status

[Sebastian Huber]

On 29/04/2020 02:28, Chris Johns wrote:


Release Change Notes


Joel has asked for a high level list of changes to be added to the 
release information. He would like this to be a community effort.


We have the detailed release notes created from the Trac tickets for 
the release so this is suppose to be something high level.


Please create or add to a file called rtems-notes-5.txt as markdown in 
the releases repo ...


https://git.rtems.org/rtems-release/

Either push the changes or send a patch to this list and I will merge 
them. I will add support to place this text in the generated README 
once it exists. If there is no file at the time of the release no 
support will be added. 


I added the file based with some initial content

https://git.rtems.org/rtems-release/tree/rtems-notes-5.txt

on the 4.11 release notes

https://devel.rtems.org/wiki/Release/4.11

We should create the notes for RTEMS 6.1 right after the 5.1 release. I 
have trouble to remember all the things which changed in the RTEMS 5 
development cycle. It would be better to update the notes in parallel 
with the development.


___
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel