Pavel Machek | 23 Oct 16:51 2014

eldk-5.5.2 issues


Eldk-5.5 seems to miscompile u-boot on altera socfpga (arm board), so
I tried updating to eldk-5.5.2... but I ran into problems compiling

Does that work for you?

I did cleanall, still I'm getting "object directory ... does not
exist". I guess rm -rf /opt/eldk/downloads/ would solve that, but
perhaps there's solution not involving excessive force?

include/stddef.h is missing. Should more dependencies be added to
u-boot-mkimage recipe?

Eldk-5.5 compilation works ok for me.

Best regards,

| DEBUG: Executing shell function do_compile
| HOSTLDFLAGS=-L/opt/work/pavel/build-wagabuibui/tmp/sysroots/x86_64-linux/usr/lib
| -L/opt/work/pavel/build-wagabuibui/tmp/sysroots/x86_64-linux/lib
| -Wl,-rpath-link,/opt/work/pavel/build-wagabuibui/tmp/sysroots/x86_64-linux/usr/lib
| -Wl,-rpath-link,/opt/work/pavel/build-wagabuibui/tmp/sysroots/x86_64-linux/lib
| -Wl,-rpath,/opt/work/pavel/build-wagabuibui/tmp/sysroots/x86_64-linux/usr/lib
| -Wl,-rpath,/opt/work/pavel/build-wagabuibui/tmp/sysroots/x86_64-linux/lib
| -Wl,-O1 HOSTSTRIP=true CONFIG_MX28=y tools
| error: object directory
(Continue reading)

Anatolij Gustschin | 8 Oct 15:09 2014

[ELDK-5.3 PATCH] meta-eldk: bash: add updates and recent 'Shellshock' fixes

Some projects still use Danny based ELDK-5.3, so add bash updates.
Taken from Chris Tapp's meta-danny-maintenance layer and updated
with recent bash fixes.

Signed-off-by: Anatolij Gustschin <agust@...>
 meta-eldk/recipes-extended/bash/bash_4.2.bbappend |  140 +++++++++++++++++++++
 1 file changed, 140 insertions(+)
 create mode 100644 meta-eldk/recipes-extended/bash/bash_4.2.bbappend

diff --git a/meta-eldk/recipes-extended/bash/bash_4.2.bbappend b/meta-eldk/recipes-extended/bash/bash_4.2.bbappend
new file mode 100644
index 0000000..4523d25
--- /dev/null
+++ b/meta-eldk/recipes-extended/bash/bash_4.2.bbappend
 <at>  <at>  -0,0 +1,140  <at>  <at> 
+# Update the patch level of bash 4.2
+# Created: 2014-09-28 to fix 'Shellshock' vulnerabilities
+# Make sure the PR gets 'bumped' so that use of this append after the base recipe has
+# been built will result in a rebuild.
+PRINC := "${ <at> int(PRINC) + 1}"
+SRC_URI_append = " \
+    ${GNU_MIRROR}/bash/bash-4.2-patches/bash42-011;apply=yes;striplevel=0;name=patch011 \
+    ${GNU_MIRROR}/bash/bash-4.2-patches/bash42-012;apply=yes;striplevel=0;name=patch012 \
+    ${GNU_MIRROR}/bash/bash-4.2-patches/bash42-013;apply=yes;striplevel=0;name=patch013 \
+    ${GNU_MIRROR}/bash/bash-4.2-patches/bash42-014;apply=yes;striplevel=0;name=patch014 \
+    ${GNU_MIRROR}/bash/bash-4.2-patches/bash42-015;apply=yes;striplevel=0;name=patch015 \
(Continue reading)

Marek Vasut | 9 Aug 16:13 2014

Please update ELDK 5.5.x to 5.5.2

Hello all,

I am getting reports about occasional kernel crashes on ARMv7 [3]. I found out 
this is caused by GCC bug [1], resolved in Yocto 1.5.2 [2] and thus ELDK 5.5.2 . 
The git commit ID in Yocto and ELDK is 97c9163d97911dad0afcd0dae39e5c426e27e921 
. To prevent being affected by this bug, please update your ELDK 5.5.x to ELDK 
5.5.2 or newer.


[3] This is the backtrace caused by the bug. Please note that the PC is pointing 
to 'find_get_entry' function, which is miscompiled by the GCC 4.8.x without the 
Unable to handle kernel paging request at virtual
address ffffffff
pgd = be138000
[ffffffff] *pgd=4fffd821, *pte=00000000, *ppte=00000000
Internal error: Oops: 17 [#1] SMP ARM
Modules linked in:
CPU: 0 PID: 31055 Comm: ip Not tainted 3.15.0+ #8
task: be2f4440 ti: be3e4000 task.ti: be3e4000
PC is at find_get_entry+0x5c/0xf8
LR is at radix_tree_lookup_slot+0x1c/0x2c
pc : [<800a6380>]    lr : [<802a3580>]    psr: a0000113
sp : be3e5ce8  ip : 00000000  fp : be3e5d1c
r10: a0000113  r9 : be8c4c80  r8 : 00000002
r7 : bed84558  r6 : 00000001  r5 : 00000000  r4 : bed84428
(Continue reading)

Sergei Poselenov | 17 Jul 19:10 2014

How to add a package to meta-toolchain-qte?


I added a package (log4cxx) to meta/recipes-devtools to ELDK-5.3 tree.
Everything is fine, the package built and populated to the target image
using "IMAGE_INSTALL +=".

However, I've run into a problem of adding the main and "-dev" package
to the SDK image. I need "liblog4cxx10" and "liblog4cxx-dev" (the
names of .ipk's) to be added to sysroot of meta-toolchain-qte, so I
could link a sample program with a header in include/log4cxx/ and
respective library.

Could someone please advice?

Another related question is what is the correct way to re-run
the creation of SKD? I'm doing "bitbake -cpopulated_sdk -f
meta-toolchain-qte" but not sure it picks up the changes.

Thanks in advance.


Jörn Wüstenfeld | 17 Jul 07:01 2014

Missing cmake binaries in eldk 5.5.1


I have build the current release of eldk-5.5.1 to be able to include 
additional packages, not available in the standard images( e.g. ntp, 
to syncronize the clock with the NFS server).
My Problem is that cmake is mostly missing in the created images.
After instalation I find some files in the core-image-lsb-sdk...


But there is no cmake binary! Also in the toolchain thejre is nothing 
regarding cmake! Looking thru the logfiles, the build of cmake-native- 
was successfull. In the directory:


is only one file:


(Continue reading)

Can Aydin | 24 Jun 21:20 2014

Compiling for powerpc-e500v2 using ELDK 5.5.2


I've been using ELDK 4.2 85xxDP toolchain and wanted to try out compiling a
simple hello world with 5.5.2 using the procedure described in the manual.

The output is below, I'm getting an error message saying "not configured
for SPE ABI", could anyone let me know if I'm doing something wrong?

I installed the eldk-5.5.2-powerpc-e500v2.iso and am doing the following:


user <at> host:~/src/tmp$ .

user <at> host:~/src/tmp$ echo $CC

powerpc-linux-gnuspe-gcc -m32 -mcpu=8548 -mabi=spe -mspe

user <at> host:~/src/tmp$ $CC helloworld.c

helloworld.c:1:0: error: not configured for SPE ABI

 #include <stdio.h>


helloworld.c:1:0: error: not configured for SPE instruction set
(Continue reading)

Claus Molden | 23 Jun 17:40 2014

Installing Drivers on m28evk


my goal is to use a LTE modem via USB with the m28evk.

I have got driver for the LTE card but I am not able to install the
driver direct on the board.

I know that I have to crosscompile the dirvers (gobi serial and gob net
drivers) for the system  and copy the solution then to the board, or
like this.

Is there any description how can  I do the cross compilation.?

Do I need the eldk for that?

If you have a describtion think about I am newbe!

This is the first time for me trying this.

Thanks in advance



eldk mailing list
(Continue reading)

Peter Oruba | 23 Jun 09:52 2014

U-Boot version required to boot kernel 3.10 on PowerPC


I am having a hard time booting Linux kernel 3.10 (linux-yocto-3.10) on my
PowerPC system (MPC5200). The system used to run kernel 2.6.38, but
application development requires an upgrade to ELDK 5.5 which brings kernel
3.10. Neither the kernel compiled by the ELDK nor a separately compiled one
say a word on the console (assuming here the boot command line is correct,
because it used to work with 2.6).

I have early printk enabled, but the last thing I see on the console is

## Flattened Device Tree blob at 00100000
     Booting using the fdt blob at 0x100000
     Uncompressing Kernel Image ... OK

My U-Boot version is "U-Boot 2010.06-00073-g54841ab-dirty". Is it possible
that this version is too old?


Peter Oruba
eldk mailing list
Sonali A Dwarakanath | 17 Jun 21:29 2014

xenomai and adeos patch


I am applying patches to the linux-2.6.33-xlnx from xenomai-2.5.6 and
getting exact same unable to patch error.
Where am I going wrong?? Please help.

eldk mailing list
Larry Baker | 17 Jun 00:14 2014


Wolfgang, et al.,

A recent situation has awakened me to a a behavior of U-Boot that maybe should be changed.

I am experimenting with a BeagleBone Black (BBB) single-board computer.  Occasionally when I perform a
reboot, the BBB never finishes booting Linux.  When I happen to have a console cable connected, I see the
U-Boot prompt.  This explains why the BBB has not booted Linux.  But it is a mystery why U-Boot is waiting for a
command when I did not request interrupting the autoboot sequence.

I found many others reporting the same experience -- and no solution.  Until I found Andrew Glen's
10/28/2013 post at!topic/beagleboard/aXv6An1xfqI about
U-Boot hangs likely due to noise on the UART0 port.  The solution he applied was to rebuild U-Boot to require
specific text -- "uboot" -- to interrupt the autoboot sequence.  (BBB have no flash, so there is no
opportunity to permanently alter any U-Boot environment variables, if this is settable.)

It occurs to me that this might be a more common occurrence on any number of circuit boards.

I have used many ARM single-board computers that have U-Boot.  As far as I recall, they all prompt "Hit any key
to stop autoboot:".  It is probably not such a good idea that ANY character can interrupt the U-Boot
autoboot sequence.  Perhaps the default should be to require an unusual character, like F2 or DEL for BIOS. 
Or a multi-character sequence within a short period of time, like the "+++" to enter command mode on the old
Hayes modems.

As I said, this awakening only occurred to me because of my recent experience with boot failures.  My
realization is that U-Boot could probably do a little better job of protecting our (likely unattended,
embedded) systems with a minimum of change to the code.  I plan to make it my practice to #define
CONFIG_AUTOBOOT_DELAY_STR.  I'm thinking that should probably be the default.

Thank you for your valuable contributions (U-Boot and more).

(Continue reading)

Wolfgang Denk | 9 Jun 20:31 2014

Re: Antwort: powerpc kernel compile issue

Dear Albrecht,

In message <1402161002.3372.0 <at> deneb.(none)> you wrote:
> > > Try 'unset LDFLAGS'. This helps at least when compiling u-boot.
> Ouch.  Looks as if I didn't read the web page thoroughly enough... :-/  This
> does of course compile the kernel perfectly now!


> Thanks for your help,
> and sorry for the noise,

Actually I wonder not more people have problems with this.  I
consider this change of behaviours a very unfortunate one.  LDFLAGS
has been in use like we do in U-Boot and Linux for a very, very long
time, and it is a real pity that Yocto came up with an uncompatible

Best regards,

Wolfgang Denk


DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd@...
"A dirty mind is a joy forever."                       - Randy Kunkee
(Continue reading)