All of lore.kernel.org
 help / color / mirror / Atom feed
From: Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org>
To: Xenomai core <Xenomai-core@domain.hid>,
	Xenomai help <xenomai@xenomai.org>
Subject: [Xenomai-core] [announce] Xenomai v2.5.2
Date: Sun, 28 Mar 2010 22:42:26 +0200	[thread overview]
Message-ID: <4BAFBF32.8030507@domain.hid> (raw)

Hi,

here comes Xenomai v2.5.2, codenamed "Souls Of Distortion" available
from Xenomai download area:
http://download.gna.org/xenomai/stable/xenomai-2.5.2.tar.bz2

It contains important fixes for several bugs found in previous versions:
- silent memory corruption induced by the "u_mode" feature;
- bad return values of posix and native condition variables (also add 
some unit tests for condition variables, to allow tracking regressions);
- fix error handling in xnintr_detach().

Some improvements in several area:
- a new --with-testdir option allowing to install xenomai testsuite 
programs in a special directory, to avoid collision with other programs;
- a bunch of improvements from the analogy branch;
- made a libxenomai library factoring the code common between all skins,
with a change in the xeno-config script, making it simpler to use (you 
now pass the script the skin for which you are compiling, and the output 
ldflags contain the name of all libraries);
- get the select syscall more in-line with the behaviour dictated by the
POSIX spec.

And I am probably forgetting some other important changes, short log
follows:
Alexis Berlemont (35):
      analogy: fix a declaration issue, a4l_sizeof_subd was not in analogy.h
      analogy: add the function a4l_sync_dio
      analogy: clean the fake driver and add a DIO subdevice
      analogy: [fake] fix a bad access in the attach procedure
      analogy: fix a bug in a4l_sync_dio
      analogy: add the program insn_bits to test digital subdevices
      analogy: [pcimio] fix a bug in the management of attach options
      analogy: print a kernel error message if an instruction failed
      analogy: add a missing error message in insn_read
      analogy: add a missing error message in insn_write
      analogy: [fake] fix a bad access in the detach procedure
      analogy: [fake] minor change in debug messages
      analogy: remove a useless locking during the detach procedure
      analogy: remove a forgotten variable which became useless
      analogy: add a4l_flush_sync()
      analogy: read() and poll() return no more -ENOENT when acquisition is over
      analogy: reinitialize events flags before accepting commands
      analogy: [fake] send events only if data are sent to userspace
      analogy: adapt cmd_read after having fixed read and poll syscalls
      analogy: in insn_read, no more real-time mode and automatic subd selection
      analogy: in insn_write, no more real-time mode and automatic subd selection
      analogy: improve robustness of the detach procedure
      analogy: minor changes in comments
      analogy: fix a potential missing initialization of the subd descriptor
      analogy: make a4l_config_subd support open drain configuration
      analogy: [pcimio] fix many race conditions in DMA output transfers
      analogy: make __pre_abs_get more robust at ends of acquisitions
      analogy: fix a bug in cmd_read with real-time mode enabled
      analogy: remove rt_task_set_mode from test programs
      analogy: change the error code in case of context error (-EPERM -> -ENOSYS)
      analogy: [loop] add the asynchronous callbacks for the write subdevice
      analogy: in cmd_{read, write}, replace a4l_sys_* by a4l_async_*
      analogy: declare a4l_async_read and a4l_async_write in analogy.h
      analogy: add a description of TRIG_WAKE_EOS
      analogy: at close time, cancel any busy subdevice

Andreas Glatz (3):
      native: corrected spelling mistake in macro __xeno_flush_rq()
      nucleus: check IRQ object before attaching or detaching it
      native: removed redundant code in rt_intr_delete()

Daniele Nicolodi (2):
      analogy: [pcimio] enable subdevice buffer mmapping
      analogy: fix a bug in a4l_ioctl_bufinfo when idle

Gilles Chanteperdrix (34):
      maint: extract-doc copies plain-text files too.
      testsuite: fix compilation of latency with uclibc
      arm: fix inline assembly for older binutils releases.
      nucleus: wake a select'ing thread up when closing a fd.
      timeconv: revert 97b78d45bf4c5571fccd9675fae8bb008a341769
      wrap-link: try to work correctly with the --as-needed option
      Make libxeno_common a real shared library
      Add libxenomai and a new --skin option to xeno-config
      build: bootstrap
      Use pthread_once for initializations which should be called once.
      Merge commit 'jan'
      asm-generic: remove any trace of mlock_alert.h, only bind.h needs it
      libxenomai: make xnsysinfo static, its contents depend on the skin
      xeno-config: fix typo
      xeno-config: support --skin=foo option in addition to --skin foo
      posix: fix recursive condvar implementation
      doc: add a warning about -EINTR return value for rt_cond_wait*
      testsuite: fix mutex-torture timed mutex test to work on UP
      testsuite: add condition variable testsuite
      posix: fix pthread_cond_*wait return value overwriting
      doc: document -ESRCH as a possible return value of rt_cond_delete
      native: fix rt_cond_*wait return value overwriting
      common: add access in user-space to nkvdso
      common: do not let u_mode exceptional cases leak out of current.[ch]
      testsuite: add some margin to avoid false errors in mutex-torture
      testsuite: suppress the cancel_with_signals testcase from sigtest.
      testsuite: fix compilation warnings
      common: modify leak warning
      arm: upgrade I-pipe support to 2.6.28.10-arm-1.12-07, 2.6.29-arm-1.13-05 and 2.6.30-arm-1.15-01
      x86 syscalls: make __xn_get_eip a macro
      Merge commit 'rpm'
      common: use vdso to detect presence of the drop_u_mode syscall.
      build: update version stamp
      doc: regenerate

Jan Kiszka (22):
      RTDM: Split up application and driver debugging
      Mark all time conversion symbols weak
      RTDM+POSIX: Avoid leaking binding objects on errors
      RTDM: Bind deleted sem/event objects, but mark them pending
      RTDM: Consistently use xnsynch_test_flags to test for RTDM_SYNCH_DELETED
      RTDM: Avoid calling cleanup_instance with held spin lock
      RTDM: Add smp barrier to rtdm_context_unlock
      Push time conversion services into libxeno_common
      Fix remapping of global sem_heap on fork
      Avoid false error reports of xeno_handle_mlock_alert
      testsuite: Fix posix mutex test for non-auto-shadow setups
      testsuite: Add timed mutex test
      Native: Fix error signedness in rt_cont_wait_until
      Native: Fix return code of in-kernel rt_cond_wait[_until]
      POSIX: Fix signedness issue in EFAULT path of cond_wait_prologue
      POSIX: Fix signedness in mutex re-acquisition on cond wait
      sigtest: Fix hard lock-up on disabled auto-shadowing
      Add __xn_linux_mux_p for Linux syscall demultiplexing
      Allow for deregistering current mode user space address
      common: Introduce xeno_get_current_fast
      rtdk: Introduce assert_nrt_fast
      common: Restore memory corruption warning for non-leak case

Philippe Gerum (38):
      x86: upgrade I-pipe support to 2.6.32.7-x86-2.6-00
      wrap-link.sh: force object files to stage1
      wrap-link.sh: introduce dry run mode (-n)
      x86: upgrade I-pipe support to 2.6.32.7-x86-2.6-01
      scripts: add wrapper script to run standard Xenomai commands
      build: introduce --with-testdir
      testsuite/unit: make vdso test depend on fastsynch availability
      asm-generic: fix dist rule to include timeconv.h
      scripts: fix xeno-config to return proper POSIX cflags
      scripts: fix xeno-config for the deprecated --posix-cflags form
      posix: do not uselessly detach IRQ descriptor before deletion
      nucleus/intr: factor out consistency checks
      nucleus/intr: factor out attachment bit setting
      nucleus/intr: fix error code from xnintr_detach()
      powerpc: define calibration data for tqm8555
      timeconv: init time conversion helper with hrclock frequency
      powerpc: move to SYSINFO v2
      powerpc: upgrade I-pipe support to 2.6.33-powerpc-2.9-00
      testsuite: silence harmless warnings
      rtdk: silence harmless warnings
      blackfin: initialize clockfreq value
      arm: initialize clockfreq value
      x86: initialize clockfreq value
      native: introduce rt_task_same()
      testsuite/latency: fix false positive in mode switch detection
      powerpc: remove dandling reference
      blackfin: upgrade I-pipe support to 2.6.33-blackfin-1.13-00
      scripts/xeno-config: fix --skin=posix --ldflags call form
      native: introduce rt_task_same()
      testsuite/latency: fix false positive in mode switch detection
      powerpc: remove dandling reference
      blackfin: upgrade I-pipe support to 2.6.33-blackfin-1.13-00
      scripts/xeno-config: fix --skin=posix --ldflags call form
      nucleus: warn about denied syscalls
      Merge branch 'master' into for-upstream
      rtdm: tag syscalls as conforming
      nucleus: reset deferred cancellation bit properly
      rtipc/bufp: prevent I/O waiters from missing a pulse

Roland Stigge (1):
      x86 syscalls: fix __xn_get_eip

Stefan Kisdaroczi (6):
      debian: copyright: fix typo and add project url
      debian: libxenomai1: sync from debian.org
      debian: linux-patch-xenomai.README.Debian: sync from debian.org
      debian: Build-Depends: debhelper (>= 7)
      debian: changelog: sync from debian.org
      debian: fix package libxenomai1

Stefan Schaal (1):
      analogy: add a4l_config_subd() declaration in analogy.h

Regards.

-- 
					    Gilles.


             reply	other threads:[~2010-03-28 20:42 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-03-28 20:42 Gilles Chanteperdrix [this message]
2010-03-29 20:01 ` [Xenomai-core] [announce] Xenomai v2.5.2 Roland Stigge
2010-04-13 15:41   ` Philippe Gerum

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4BAFBF32.8030507@domain.hid \
    --to=gilles.chanteperdrix@xenomai.org \
    --cc=Xenomai-core@domain.hid \
    --cc=xenomai@xenomai.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.