Skyeye on Ubuntu 13.04 and Fedora Update

139 views
Skip to first unread message

Joel Sherrill

unread,
Apr 29, 2013, 12:14:45 PM4/29/13
to skyeye-s...@googlegroups.com
Hi

I decided to follow your lead and try this on Ubuntu. After the final make install, 
this is what was installed:

joel@JoelD830:~$ ls ~/skyeye/install-git/bin
mknandflashdump  skyeye_cli.pyc   skyeye_lib_path.py   skyeye_module.py
prof_convert     skyeye_init.py   skyeye_lib_path.pyc  skyeye_module.pyc
skyeye_cli.py    skyeye_init.pyc  skyeye_main.py       uart_instance
joel@JoelD830:~$ cat /etc/issue
Ubuntu 13.04 \n \l

When I invoked skyeye_main.py, it reported a version number of 1.3.2. 

More questions and some answers:

(1) What happened to the binary named skyeye? If skyeye_cli.py is a replacement,
then it is not executable.

(2) I know there is a branch for 1.3.5. Why is master showing 1.3.2? 
Am I on the right git branch?

The answers:

More info on libreadline.a. The top level Makefile.in in git is NOT generated from
the Makefile.am. The Makefile.in has this:

all:third-party/readline/libreadline.a third-party/bfd/.libs/libbfd.so third-party/libiberty/libiberty.a third-party/opcodes/.libs/libopcodes.so

which does not match what is in the Makefile.am

all:third-party/bfd/.libs/libbfd.so third-party/libiberty/libiberty.a third-party/opcodes/.libs/libopcodes.so

Below my signature is the complete output of git status. Quite a few files were regenerated.

Fedora has the same version of automake but a different version of autoconf. Ubuntu
has 2.69. Fedora 17 has 2.68.

After I copied the tree back from Ubuntu to Fedora with the generated files, the build
completed exactly the same.

I think you need to commit the files generated by autoconf/automake. It may also be necessary
to have a script to touch the generated files to ensure a user doesn't have to regenerate them.
GCC's contrib/.

It is not necessary to keep the autom4te.cache directories under source code control.

--joel
RTEMS

# modified:   Makefile.in
# modified:   aclocal.m4
# modified:   arch/arm/Makefile.in
# modified:   arch/bfin/Makefile.in
# modified:   arch/coldfire/Makefile.in
# modified:   arch/mips/Makefile.in
# modified:   arch/ppc/Makefile.in
# modified:   arch/sparc/Makefile.in
# modified:   arch/x86/Makefile.in
# modified:   autom4te.cache/output.3
# modified:   autom4te.cache/requests
# modified:   autom4te.cache/traces.3
# modified:   common/Makefile.in
# modified:   configure
# modified:   device/Makefile.in
# modified:   libltdl/autom4te.cache/requests
# modified:   libltdl/config-h.in
# modified:   libltdl/configure
# modified:   soc/arm/Makefile.in
# deleted:    third-party/opcodes/s390-mkopc
# modified:   utils/Makefile.in
#
# Untracked files:
#   (use "git add <file>..." to include in what will be committed)
#
# autom4te.cache/output.4
# autom4te.cache/traces.4
# libltdl/autom4te.cache/output.2
# libltdl/autom4te.cache/traces.2



Michael.Kang

unread,
May 2, 2013, 7:56:02 AM5/2/13
to skyeye-simulator
Sorry for my late response due to my labour day vacation. I have not worked on the labour day....

The master branch is not updated for too long time. I am trying to clean it and update to the stable and latest state
to keep up with our development recently.

You can try to use branch skyeye_1_3_5_rel or wait one day or two day for the done of master branch updated.
After a quick testing of rtems testcases in SkyEye against skyeye_1_3_5_rel, I found now only the testcase
of edb7312 can run hello.exe successfully.

Thanks your indication of the existing issue of automake/autoconf in SkyEye source tree. I will commit the
Makefile.in and other auto-generated file. And at the same time, exclude autom4te.cache directory under
source control. Probably I really need to spend more time on the system of autoconf/automake .
And skyeye_cli.py is temporary plan and now skyeye binary is return.

Any questions, please do not hesitate to send email to the list.

Thanks
MK







--
您收到此邮件是因为您订阅了 Google 网上论坛的“skyeye-simulator”论坛。
要退订此论坛并停止接收此论坛的电子邮件,请发送电子邮件到 skyeye-simulat...@googlegroups.com
要向此网上论坛发帖,请发送电子邮件至 skyeye-s...@googlegroups.com
通过以下网址访问此论坛:http://groups.google.com/group/skyeye-simulator?hl=zh-CN。
要查看更多选项,请访问 https://groups.google.com/groups/opt_out。
 
 



--

Joel Sherrill

unread,
May 2, 2013, 11:16:12 AM5/2/13
to skyeye-s...@googlegroups.com
On Thu, May 2, 2013 at 6:56 AM, Michael.Kang <blackf...@gmail.com> wrote:
Sorry for my late response due to my labour day vacation. I have not worked on the labour day....

No problem.
 
The master branch is not updated for too long time. I am trying to clean it and update to the stable and latest state
to keep up with our development recently.


There are so many git development models.  RTEMS uses master like a head and
branches for releases.
 
You can try to use branch skyeye_1_3_5_rel or wait one day or two day for the done of master branch updated.

No problem. I am hoping to get enough groundwork laid where another person
can step in and take over.
 
After a quick testing of rtems testcases in SkyEye against skyeye_1_3_5_rel, I found now only the testcase
of edb7312 can run hello.exe successfully.


Hmm.. I wonder how long ago each one ran actually ran successfully. :(

Any advice on a git bisect strategy? I planned to start by testing all RTEMS 
executables in the tests against the end of all release branches? But once I have
that, should I git bisect the branch that is first broken on?

And ...how do you run skyeye these days? It doesn't look familiar. :(

And would it help to know about the other test cases working or not working
as we go through the testing?
 
Thanks your indication of the existing issue of automake/autoconf in SkyEye source tree. I will commit the
Makefile.in and other auto-generated file. And at the same time, exclude autom4te.cache directory under
source control. Probably I really need to spend more time on the system of autoconf/automake .
And skyeye_cli.py is temporary plan and now skyeye binary is return.

OK.
 
Any questions, please do not hesitate to send email to the list.


Thanks. 
Reply all
Reply to author
Forward
0 new messages