[Xenomai-core] [announce] Xenomai v2.5.5

[Xenomai-core] [announce] Xenomai v2.5.5

[Gilles Chanteperdrix]

Hi,

Xenomai 2.5.5, aka "Ghosts", is available:
http://download.gna.org/xenomai/stable/xenomai-2.5.5.tar.bz2

It contains bug fixes for:
- the x86 I-pipe issue, which caused freeze-on-boot issue, very 
reproducible with grub 2
- a ppc I-pipe issue, which involved unaligned access to floating point 
values
- the debian patch generation script, which caused invalid patches to be
generated for recent kernel releases
- a few compilation issues with recent compilers and or innfrequently 
used compilation options
- issues with using fork with process-private user-space mutexes

And improvements:
- of the user-space latency on all platforms by optimizing the RPI code 
and the relax code
- of the ARM support to code, in order to support SMP systems
- of the analogy stack by Alexis
- of the user-space latency on ARM in general and OMAP3 in particular, 
as a result of some work both on Xenomai timer code and on the I-pipe 
patch
- of the scheduler "resched" bit handling by simplifying it, following 
the correction of an incorrect warning

Note: Xenomai 2.5.5 is the last release to contain a patch for Linux 2.4
on x86 and to export kernel-mode services to non GPL modules.

Enjoy.

The complete shortlog follows:
Alexis Berlemont (9):
      analogy: minor change in analogy Kconfig text
      analogy: add comments in the buffer management code
      analogy: fix a bug in the DIO subdevice's mite configuration
      analogy: minor changes in cmd_bits
      analogy: comsetic change in mite.c
      analogy: [pcimio] rework the mite setup procedures
      analogy: [pcimio] fix a forgotten warning
      analogy: [pcimio] add the initialization of the ring for gpct subd
      analogy: fix compilation warnings with gcc-4.4.1

Gilles Chanteperdrix (24):
      fix private heap unmapping upon fork.
      Do not assume that XN_NO_HANDLE is NULL
      Set user-space current thread handle to XN_NO_HANDLE after fork
      fix comment
      Fix typo in edaf1e2e54343b6e4bf5cf6ece9175ec0ab21cad
      sem_heap: map an inaccessible heap upon fork.
      posix: add a magic to internal structures.
      arm: fix VFP context handling on SMP systems
      Merge commit 'rpm/for-upstream'
      arm: fix atomic operations
      arm: fix typo in 4a739c5ce29b50bee780bf620852a2fff6a71431
      autotools: regenerate
      arm: better calibrate the timer programming latency
      arith: add the nodiv_imuldiv_ceil routine
      arm: use the nodiv_imuldiv routine for timer programming.
      hal: export missing apc symbols
      nucleus: modify /proc/xenomai/latency
      nucleus: add a margin for early timer shots
      nucleus/sched: fix rescheduling bit test macros
      arm: add calibrated scheduling latencies
      arm: remove debug printk
      smi: add PCI ID for old kernels
      build: boostrap
      doc: regenerate

Jan Kiszka (8):
      RTDM: Protect xnshadow_ppd_get via nklock
      nucleus: Fix symbolic status output of root threads
      nucleus: Create watchdog as non-blockable timer
      native: Improve documentation of rt_task_join and rt_task_delete
      x86: Add PCI ID of Series 5/3400 Intel chipset to SMI workaround
      x86: Add PCI ID of 631xESB/632xESB/3100 Intel chipset to SMI workaround
      x86: Add PCI ID of Intel ICH9M-E chipset to SMI workaround
      x86: Add PCI ID of Intel ICH9M chipset to SMI workaround

Mike Frysinger (1):
      blackfin: clean up syscall assembly

Peter Soetens (1):
      rtcan: Mask the SJA_MOD register when probing for channels

Philippe Gerum (36):
      powerpc: upgrade I-pipe support to 2.6.34.4-powerpc-2.10-04
      nucleus: demote RPI boost upon linux-originated signal
      blackfin: upgrade I-pipe support to 2.6.35.2-blackfin-1.15-00
      nucleus: requeue blocked non-periodic timers properly
      x86: upgrade I-pipe support to 2.6.32.20-x86-2.7-02, 2.6.34.5-x86-2.7-03
      arm: force enable preemptible switch support in SMP mode
      arm: enable VFP support in SMP
      arm: use rthal_processor_id() over non-linux contexts
      powerpc: resync thread switch code with mainline >= 2.6.32
      x86: increase SMP calibration value
      nucleus/sched: move locking to resume_rpi/suspend_rpi
      hal/generic: inline APC scheduling code
      nucleus, posix: use fast APC scheduling call
      nucleus/shadow: shorten the uninterruptible path to secondary mode
      nucleus/sched: prevent remote wakeup from triggering a debug assertion
      powerpc: upgrade I-pipe support to 2.6.35.4-powerpc-2.11-00
      nucleus/sched: fix race in non-atomic suspend path
      nucleus/sched: raise self-resched condition when unlocking scheduler
      nucleus: fix build issue introduced by 4860668b6
      x86: upgrade I-pipe support to 2.6.32.20-x86-2.7-03, 2.6.34.5-x86-2.7-04
      nucleus: please C++ parsers
      nucleus: do not rely on implicit header inclusion
      rtdm: fix build for non-pervasive mode
      rtipc: fix build for non-pervasive mode
      nucleus/shadow: fix build for simulation mode
      native: fix build for simulation mode
      nucleus/bufd: fix documentation group tag
      psos: fix t_shadow() regression
      psos: introduce t_setreg/t_getreg syscalls
      arm: upgrade adeos patch to 2.6.33-1.18-00
      x86: upgrade adeos patch to 2.6.35.7-2.7-04
      powerpc: upgrade I-pipe support to 2.6.34.4-powerpc-2.10-05, 2.6.35.7-powerpc-2.11-01
      blackfin: upgrade I-pipe support to 2.6.34-blackfin-1.15-01, 2.6.35.7-blackfin-1.15-01
      nucleus: disable u_mode watchdog for MMU-less
      can: adapt to platform driver infrastructure changes for 2.6.35
      powerpc: upgrade I-pipe support to 2.6.35.7-powerpc-2.11-02

Stefan Kisdaroczi (3):
      debian: fix asm include directory linking in prepare-patch.sh
      debian: move debian/prepare-patch.sh to scripts/
      RTDM device profiles: Document open_rt, socket_rt and close_rt deprecation

Wolfgang Grandegger (1):
      nucleus: switch to ioctl_unlocked


-- 
                                                                Gilles.

Re: [Xenomai-core] [announce] Xenomai v2.5.5

[Stefan Kisdaroczi]

[-- Attachment #1: Type: text/plain, Size: 296 bytes --]

Hi Gilles,

On 06.10.2010 14:05, Gilles Chanteperdrix wrote:
> [...]
> - the debian patch generation script, which caused invalid patches to be
> generated for recent kernel releases
>   

The file scripts/prepare-patch.sh is missing in the
xenomai-2.5.5.tar.bz2 archive.

Stefan



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]

Re: [Xenomai-core] [announce] Xenomai v2.5.5

[Gilles Chanteperdrix]

Stefan Kisdaroczi wrote:
> Hi Gilles,
> 
> On 06.10.2010 14:05, Gilles Chanteperdrix wrote:
>> [...]
>> - the debian patch generation script, which caused invalid patches to be
>> generated for recent kernel releases
>>   
> 
> The file scripts/prepare-patch.sh is missing in the
> xenomai-2.5.5.tar.bz2 archive.

I do not see it in scripts/Makefile.am EXTRA_DIST. Did I miss some part
of your patch?

-- 
					    Gilles.

Re: [Xenomai-core] [announce] Xenomai v2.5.5

[Stefan Kisdaroczi]

[-- Attachment #1: Type: text/plain, Size: 1146 bytes --]

On 06.10.2010 16:24, Gilles Chanteperdrix wrote:
> Stefan Kisdaroczi wrote:
>   
>> Hi Gilles,
>>
>> On 06.10.2010 14:05, Gilles Chanteperdrix wrote:
>>     
>>> [...]
>>> - the debian patch generation script, which caused invalid patches to be
>>> generated for recent kernel releases
>>>   
>>>       
>> The file scripts/prepare-patch.sh is missing in the
>> xenomai-2.5.5.tar.bz2 archive.
>>     
> I do not see it in scripts/Makefile.am EXTRA_DIST. Did I miss some part
> of your patch?
>   

No. I missed that in the patch, I just moved the file from debian/ to
scripts/.

Sorry, but a stupid guy like me simply does not understand why a script
in a git tree needs to be put in a Makefile.am so that it gets put in a
archive from that tree.

If I do "git archive v2.5.5 | bzip2 > ..." it works. The only difference
between the archives is the inclusion of the 3 pdf's in the doc/nodist
dir and the 3 small scripts in scripts/maint directory (and
prepare-patch.sh). The pdf's could be put out of the tree, but simple
solutions and workflows are probably just for stupid people like me.

Stefan (annoyed)



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]

Re: [Xenomai-core] [announce] Xenomai v2.5.5

[Gilles Chanteperdrix]

Stefan Kisdaroczi wrote:
> On 06.10.2010 16:24, Gilles Chanteperdrix wrote:
>> Stefan Kisdaroczi wrote:
>>   
>>> Hi Gilles,
>>>
>>> On 06.10.2010 14:05, Gilles Chanteperdrix wrote:
>>>     
>>>> [...]
>>>> - the debian patch generation script, which caused invalid patches to be
>>>> generated for recent kernel releases
>>>>   
>>>>       
>>> The file scripts/prepare-patch.sh is missing in the
>>> xenomai-2.5.5.tar.bz2 archive.
>>>     
>> I do not see it in scripts/Makefile.am EXTRA_DIST. Did I miss some part
>> of your patch?
>>   
> 
> No. I missed that in the patch, I just moved the file from debian/ to
> scripts/.
> 
> Sorry, but a stupid guy like me simply does not understand why a script
> in a git tree needs to be put in a Makefile.am so that it gets put in a
> archive from that tree.
> 
> If I do "git archive v2.5.5 | bzip2 > ..." it works. The only difference
> between the archives is the inclusion of the 3 pdf's in the doc/nodist
> dir and the 3 small scripts in scripts/maint directory (and
> prepare-patch.sh). The pdf's could be put out of the tree, but simple
> solutions and workflows are probably just for stupid people like me.

We use "make distcheck" which is the standard way of making a package
based on the autotools, and has been the way we have made Xenomai
package since day 1.

"make distcheck" checks that the package compilation works.

It ensures that users of the Xenomai package can regenerate a new
package from the sources we provide.

And it does not look that much more complicated than the pipeline using
git archive.

Anyway, I will add it, but I will wait a bit to re-generate a 2.5.5.1 in
order to see if we have no other issue.

-- 
					    Gilles.

Re: [Xenomai-core] [announce] Xenomai v2.5.5

[Gilles Chanteperdrix]

Gilles Chanteperdrix wrote:
> Anyway, I will add it, but I will wait a bit to re-generate a 2.5.5.1 in
> order to see if we have no other issue.

Ok. No other issues, I uploaded 2.5.5.1.

-- 
                                                                Gilles.

Re: [Xenomai-core] [announce] Xenomai v2.5.5

[Stefan Kisdaroczi]

Am Sonntag 10 Oktober 2010, 23:09:19 schrieben Sie:
> Gilles Chanteperdrix wrote:
> > Anyway, I will add it, but I will wait a bit to re-generate a 2.5.5.1 in
> > order to see if we have no other issue.
> 
> Ok. No other issues, I uploaded 2.5.5.1.

Thank you. Stefan