[Xenomai-core] Buildbot detects failure in recent build

2006-12-23 Thread Niklaus Giger
Hi Philippe

It seems that your revision 1978 or 1979 broke building PSOS+ on ppc and
PPC405. 
See e.g. http://ngiger.dyndns.org/buildbot/hcu3_q/builds/97/step-mk_kernel/1
which complains about
  CC [M]  kernel/xenomai/skins/psos+/syscall.o
kernel/xenomai/skins/psos+/syscall.c: In function '__rn_create':
kernel/xenomai/skins/psos+/syscall.c:1083: error: 'psosrn_t' undeclared (first 
use in this function)
kernel/xenomai/skins/psos+/syscall.c:1083: error: (Each undeclared identifier 
is reported only once
kernel/xenomai/skins/psos+/syscall.c:1083: error: for each function it appears 
in.)
kernel/xenomai/skins/psos+/syscall.c:1083: error: 'rn' undeclared (first use 
in this function)
kernel/xenomai/skins/psos+/syscall.c:1084: warning: ISO C90 forbids mixed 
declarations and code
kernel/xenomai/skins/psos+/syscall.c:1112: error: parse error before ')' token
kernel/xenomai/skins/psos+/syscall.c: In function '__rn_delete':
...

Best regards
-- 
Niklaus Giger

___
Xenomai-core mailing list
Xenomai-core@gna.org
https://mail.gna.org/listinfo/xenomai-core


Re: [Xenomai-core] Buildbot detects failure in recent build

2006-12-23 Thread Philippe Gerum
On Sat, 2006-12-23 at 19:31 +0100, Niklaus Giger wrote:
 Hi Philippe
 
 It seems that your revision 1978 or 1979 broke building PSOS+ on ppc and
 PPC405. 

Fixed, thanks. Btw, could you ask the buildbot to spam me gently each
time a commit wrecks the Xenomai boat? It would be easier for me than
tracking the WEB interface status. TIA,

-- 
Philippe.



___
Xenomai-core mailing list
Xenomai-core@gna.org
https://mail.gna.org/listinfo/xenomai-core


Re: [Xenomai-core] Buildbot detects failure in recent build

2006-12-23 Thread Niklaus Giger
Am Samstag, 23. Dezember 2006 19:53 schrieb Philippe Gerum:
 On Sat, 2006-12-23 at 19:31 +0100, Niklaus Giger wrote:
  Hi Philippe
 
  It seems that your revision 1978 or 1979 broke building PSOS+ on ppc and
  PPC405.

 Fixed, thanks. Btw, could you ask the buildbot to spam me gently each
 time a commit wrecks the Xenomai boat? It would be easier for me than
 tracking the WEB interface status. TIA,
Sorry for a (wrong positive) mail from my test installation. 

You should receive an email on build failures. I changed the buildbot setup so 
that failures while running the test on my CPU-board are only flagged as 
warnings, as they were often reporting failures in my setup.
As my PPC405 board is also used for may other development work, it is 
sometimes unavailable for the buildbot and the buildbot would also flag this 
as an error.
The sim is reporting error since quite a few weeks, but Gilles never fixed 
this error (which is probably just an expectation or a time value which has 
varied a little bit):
Some problems were reported in:
vxworks-t010823-2:t010823-2.c:163: Expected sequence: SEQ(Test2,2004); got 
SEQ(Test2,1999)
vxworks-t010823-2: t010823-2.c:166, test finished: 1 failures/ 33 tests
If you want I could also made a failure in this buildstep be considered only a 
warning.

If you receive too many emails just let me know and I will try to improve this 
situation.

Joyeux noël

Niklaus Giger

___
Xenomai-core mailing list
Xenomai-core@gna.org
https://mail.gna.org/listinfo/xenomai-core


Re: [Xenomai-core] Patches: README.install examples/common for cross-compiling

2006-12-23 Thread Niklaus Giger
Am Freitag, 22. Dezember 2006 08:45 schrieb Wolfgang Grandegger:
 Hi Niklaus,

 Niklaus Giger wrote:
  Hi
 
  After switching my development environment to a MacMini and using the
  ELDK 4.0 I discovered that my examples for building using a
  cross-compiler for my PPC405 target were not correct. (Maybe a few of my
  previous problems were caused by not correctly specifying the target
  ARCH.)
 
  Also I would like to adapt and document the examples/common/Makefile to
  the needs for cross-compiling (KSRC/XENOCONFIG/DESTDIR useage). Also I
  added a target install to copy the generated application to
  $(DESTDIR)/usr/xenomai.

 Hm, I don't have examples/common/Makefile in my SVN Xenomai tree. And
Added by revision 1970 only. Makefile came from Jan for my example 
hw_direct_io.
 /usr/xenomai should be the configurable.
Does the attached patch look better for you (using EXEC_PREFIX) ?

Best regards

-- 
Niklaus Giger
Index: examples/common/Makefile
===
--- examples/common/Makefile	(Revision 1980)
+++ examples/common/Makefile	(Arbeitskopie)
@@ -3,9 +3,9 @@
 ### List of applications to be build
 APPLICATIONS = hw_direct_io
 
-### Note: to override the search path for the xeno-config script, use make XENO=...
+### Note: to override the search path for the xeno-config script, use make XENOCONFIG=...
+### Note: if you installed it into a non standard place add DESTDIR=/path/to/installed/xenomai
 
-
 ### List of modules to be build
 MODULES =
 
@@ -14,6 +14,8 @@
 
 ### Note: to override the kernel source path, use make KSRC=...
 
+# default place to install the resulting binaries
+EXEC_PREFIX := $(DESTDIR)/usr/xenomai/bin
 
 
 ## USER SPACE BUILD (no change required normally) ##
@@ -24,7 +26,9 @@
 ### Sanity check
 ifeq ($(XENOCONFIG),)
 all::
-	@echo  Invoke make like this: \make XENO=/path/to/xeno-config\ 
+	@echo  Invoke make like this: \make XENOCONFIG=/path/to/xeno-config\ 
+	@echo  add KSRC=.. to override the default kernel source at /lib/modules/$(shell uname -r)/build 
+	@echo  add DESTDIR=/path/to/installed/xenomai if you installed it into a non standard place 
 	@echo
 endif
 
@@ -40,6 +44,9 @@
 clean::
 	$(RM) $(APPLICATIONS) *.o
 
+install::   $(APPLICATIONS)
+	cp $(APPLICATIONS) $(EXEC_PREFIX)/
+
 endif
 
 
@@ -80,3 +87,4 @@
 	$(RM) -R .tmp*
 
 endif
+
___
Xenomai-core mailing list
Xenomai-core@gna.org
https://mail.gna.org/listinfo/xenomai-core


Re: [Xenomai-core] Patches: README.install

2006-12-23 Thread Niklaus Giger

Am Donnerstag, 21. Dezember 2006 23:40 schrieb Wolfgang Denk:
 Dear Niklaus,

Dear Wolfgang

Thanks your for spotting my errors. I should have rechecked this patch before 
submitting it, as it was based on a outdated script. Please review the 
attached patch again which is based on a script which actually ran and 
produced a running kernel.

I would also thanks you for providing the ELDK. It was really a lot easier
to setup the ELDK once I switched from my PPC PowerBook to a MacMini. I can 
really recommend it.

The only addition to the ELDK I made to run xeno-test was to copy my 
Debian /usr/bin/which to the rootfs.

..

Best regards

-- 
Niklaus Giger
Index: README.INSTALL
===
--- README.INSTALL	(Revision 1980)
+++ README.INSTALL	(Arbeitskopie)
@@ -192,6 +192,12 @@
 from the cross-compilation tools prefix, you will have to manually
 pass the name of all compilation tools on configure command line. 
 
+If you want to avoid to build your own cross compiler, you might if find
+easier to use ELDK (see http://www.denx.de/wiki/DULG/ELDK). It includes
+the GNU cross development tools, such as the compilers, binutils, gdb, 
+etc., and a number of pre-built target tools and libraries necessary
+to provide some functionality on the target system.
+
 See sections 2.2, 2.3 and 2.4 for examples.
 
 1.4 Building the Linux kernel in the user-space support build tree
@@ -276,19 +282,22 @@
 -
 
 A typical cross-compilation setup, in order to build Xenomai for a
-PowerPC-405-based system:
+PowerPC-405-based system. Here we use an ELDK cross-compiler and a
+2.6.14 kernel.
 
-$ $xenomai_root/scripts/prepare-kernel.sh --arch=powerpc \
-  --adeos=$xenomai_root/ksrc/arch/powerpc/patches/adeos-ipipe-2.6.14-ppc-X.Y-ZZ.patch \
+$ $xenomai_root/scripts/prepare-kernel.sh --arch=ppc \
+  --adeos=$xenomai_root/ksrc/arch/powerpc/patches/adeos-ipipe-2.6.14-ppc-1.5-*.patch \
   --linux=$linux_tree
 $ cd $linux_tree
-$ mkdir ../build-powerpc-405-2.6.14
-$ make ARCH=ppc CROSS_COMPILE=powerpc-405-linux-gnu- O=../build-powerpc-405-2.6.14 xconfig/gconfig/menuconfig 
+$ build_root /path/to/eldk
+$ mkdir $build_root
+$ make ARCH=ppc CROSS_COMPILE=ppc_4xx- O=$build_root xconfig/gconfig/menuconfig
 # select the kernel and Xenomai options
-$ make ARCH=ppc CROSS_COMPILE=powerpc-405-linux-gnu- O=../build-powerpc-405-2.6.14 bzImage modules 
-# then install as needed
-$ mkdir $build_root  cd $build_root
-$ $xenomai_root/configure --build=i686-linux --host=powerpc-405-linux-gnu
+$ make ARCH=ppc CROSS_COMPILE=ppc_4xx- O=$build_root bzImage modules \
+  modules_install
+$ cd $build_root
+$ $xenomai_root/configure --build=i686-linux --host=ppc CC=ppc_4xx-gcc \
+  CXX=ppc_4xx-g++ LD=ppc_4xx-ld
 $ make install
 
 2.3 Building for the IPF
___
Xenomai-core mailing list
Xenomai-core@gna.org
https://mail.gna.org/listinfo/xenomai-core


Re: [Xenomai-core] Patches: README.install

2006-12-23 Thread Wolfgang Denk
Dear Niklaus,

in message [EMAIL PROTECTED] you wrote:
 
 Thanks your for spotting my errors. I should have rechecked this patch before 
 submitting it, as it was based on a outdated script. Please review the 
 attached patch again which is based on a script which actually ran and 
 produced a running kernel.
 
 I would also thanks you for providing the ELDK. It was really a lot easier
 to setup the ELDK once I switched from my PPC PowerBook to a MacMini. I can 
 really recommend it.

Thanks. I'm glad you like it.

 The only addition to the ELDK I made to run xeno-test was to copy my 
 Debian /usr/bin/which to the rootfs.

alias which='type -p'

  -
  
  A typical cross-compilation setup, in order to build Xenomai for a
 -PowerPC-405-based system:
 +PowerPC-405-based system. Here we use an ELDK cross-compiler and a
 +2.6.14 kernel.
  
 -$ $xenomai_root/scripts/prepare-kernel.sh --arch=powerpc \
 -  
 --adeos=$xenomai_root/ksrc/arch/powerpc/patches/adeos-ipipe-2.6.14-ppc-X.Y-ZZ.patch
  \
 +$ $xenomai_root/scripts/prepare-kernel.sh --arch=ppc \
 +  
 --adeos=$xenomai_root/ksrc/arch/powerpc/patches/adeos-ipipe-2.6.14-ppc-1.5-*.patch
  \
--linux=$linux_tree
  $ cd $linux_tree
 -$ mkdir ../build-powerpc-405-2.6.14
 -$ make ARCH=ppc CROSS_COMPILE=powerpc-405-linux-gnu- 
 O=../build-powerpc-405-2.6.14 xconfig/gconfig/menuconfig 
 +$ build_root /path/to/eldk

/path/to/eldk might be misleading - at least  it  needs  an  expla-
nation;  I  would interpret this as /opt/eldk, which is the default
installation directory of the ELDK, while I guess you  actually  mean
/opt/eldk/ppc_4xx ?

 +$ mkdir $build_root
 +$ make ARCH=ppc CROSS_COMPILE=ppc_4xx- O=$build_root 
 xconfig/gconfig/menuconfig
  # select the kernel and Xenomai options
 -$ make ARCH=ppc CROSS_COMPILE=powerpc-405-linux-gnu- 
 O=../build-powerpc-405-2.6.14 bzImage modules 
 -# then install as needed
 -$ mkdir $build_root  cd $build_root
 -$ $xenomai_root/configure --build=i686-linux --host=powerpc-405-linux-gnu
 +$ make ARCH=ppc CROSS_COMPILE=ppc_4xx- O=$build_root bzImage modules \
 +  modules_install

Be careful here. modules_install will try to run the native depmod,
which will not work. Recent builds (*) of the ELDK will allow to  add
a DEPMOD=/opt/eldk/usr/bin/depmod.pl argument to the make call.

[This extension was added in July, so you will have to wait for  ELDK
version 4.1 unless you build the ELDK from scratch from the sources.]


Thanks, and a Merry Christmas and a Happy New Year!

Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: [EMAIL PROTECTED]
READ THIS BEFORE OPENING PACKAGE: According to Certain Suggested Ver-
sions of the Grand Unified Theory, the Primary Particles Constituting
this Product May Decay to Nothingness Within the  Next  Four  Hundred
Million Years.

___
Xenomai-core mailing list
Xenomai-core@gna.org
https://mail.gna.org/listinfo/xenomai-core


Re: [Xenomai-core] Buildbot detects failure in recent build

2006-12-23 Thread Philippe Gerum
On Sat, 2006-12-23 at 21:58 +0100, Niklaus Giger wrote:
 Am Samstag, 23. Dezember 2006 19:53 schrieb Philippe Gerum:
  On Sat, 2006-12-23 at 19:31 +0100, Niklaus Giger wrote:
   Hi Philippe
  
   It seems that your revision 1978 or 1979 broke building PSOS+ on ppc and
   PPC405.
 
  Fixed, thanks. Btw, could you ask the buildbot to spam me gently each
  time a commit wrecks the Xenomai boat? It would be easier for me than
  tracking the WEB interface status. TIA,
 Sorry for a (wrong positive) mail from my test installation. 
 
 You should receive an email on build failures. I changed the buildbot setup 
 so 
 that failures while running the test on my CPU-board are only flagged as 
 warnings, as they were often reporting failures in my setup.
 As my PPC405 board is also used for may other development work, it is 
 sometimes unavailable for the buildbot and the buildbot would also flag this 
 as an error.
 The sim is reporting error since quite a few weeks, but Gilles never fixed 
 this error (which is probably just an expectation or a time value which has 
 varied a little bit):
 Some problems were reported in:
 vxworks-t010823-2:t010823-2.c:163: Expected sequence: SEQ(Test2,2004); got 
 SEQ(Test2,1999)
 vxworks-t010823-2: t010823-2.c:166, test finished: 1 failures/ 33 tests
 If you want I could also made a failure in this buildstep be considered only 
 a 
 warning.

No problem for now. If I ever get sick of these warnings, I will spam
Gilles in turn. Eh.

 
 If you receive too many emails just let me know and I will try to improve 
 this 
 situation.
 

Great, thanks. Let's see how this evolves; my mail filters will help me
handling the notifications appropriately.

 Joyeux noël
 
 Niklaus Giger
-- 
Philippe.



___
Xenomai-core mailing list
Xenomai-core@gna.org
https://mail.gna.org/listinfo/xenomai-core


Re: [Xenomai-core] Buildbot detects failure in recent build

2006-12-23 Thread Philippe Gerum
On Sat, 2006-12-23 at 21:58 +0100, Niklaus Giger wrote:

 Joyeux noël

Merci Niklaus, de même. Frõhliches Weihnachten und glũcklich neues Jahr!
(eh, sounds a bit like ARTE tv, isn't it?)

In other words, Merry Christmas and Happy New Year to all.

-- 
Philippe.



___
Xenomai-core mailing list
Xenomai-core@gna.org
https://mail.gna.org/listinfo/xenomai-core