Openembedded Devel Discussions
 help / color / mirror / Atom feed
From: j <vwyodapink@gmail.com>
To: openembedded-devel@lists.openembedded.org
Subject: Re: [oe-core] ld segfault cant figure it out
Date: Wed, 18 Apr 2012 20:32:25 -0700	[thread overview]
Message-ID: <4F8F8749.8000803@gmail.com> (raw)
In-Reply-To: <CAMKF1spO2U+Va9pUnr9pABL9rw_Bc1GBZvrpggXNNfcUg7XrGA@mail.gmail.com>

On 04/18/2012 03:00 PM, Khem Raj wrote:
> On Tue, Apr 17, 2012 at 10:40 PM, j<vwyodapink@gmail.com>  wrote:
>> Sorry to sound so dumb but if there is no recipe failing and 2 separate
>> systems are segfaulting in similar manners and both result with not fully
>> functioning images/packages where else can I look for info? I posted all the
>> logs to one of the builds, and the QA log seems to be the only thing really
>> spitting out anything. I am open to trying and or sharing any and all
>> relevant info to help sort it out or get me on track to sorting it out.
> what I am hearing is that your build succeeds despite these ld
> segfaults. Am I right ?
> but then when you run the resulting images then the images dont
> function correctly
> on your target
First off thank you all for your time and information, and putting up 
with me learning and trying to get things working!!
Yes you are correct. Nothing in the images or packages fail, but the 
build machine itself segfaults. Then if I put the image on my BB-XM 
sdcard they boot and "run" but things are not right, such as directories 
that should not be there are, some do not build a fully functioning 
package, such as task-native-sdk.



> and you also mentioned that its expat package when being built that
> causes the ld
> segfaults. Are you sure about that ? You could somehow forge ulimit to unlimited
> that will dump
No I am not sure about it. As I am just trying to watch it build and 
look for the build machine to segfault and see what packages it is 
working on. Though after further messing with expat I think you are 
correct in doubting that is the one causing it.

I just did a clean build of perl with no rm_work enabled, and I am 
trying to dig through the logs and info. From your earlier 
advice/mentioning of libtool-cross I do notice some things while it is 
building according to the info. I will post some of what stood out to 
me, which may in turn mean nothing. If I can post anymore of the logs or 
anything please let me know. I did notice today setting things to build 
one at a time, that it does appear that its working on libtool-cross, 
but again I could be wrong.

>
> on ubuntu edit /etc/security/limits.conf on build host and then add
> *         soft    core    unlimited
>
> that should generate core files. After this search for core* files in build tree
> and see if there are any then debug them with gdb to diagnose it further
>
> Now I am not sure if those segfaults directly relate to malfunction of image
> if thats the case but its likely. You should also try to explore the reasons why
> the image fails to behave as expected.
I will do as suggested as well as do a full image build again and see if 
I can find out all that is not working. Thank you for the extra 
debugging info and suggestions. Are there any keywords to search through 
all the files for including the normal log files, it may help me track 
down things even further.

>
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

I really am not trying to be a pain in the a** and would like to get 
things sorted out and understood on my side. Honestly if the resulting 
images worked and did not get any extra folders in their paths I would 
not care about the resulting segfaults on the build machine.

Another side note, I am borrowing another build machine for a few days 
to see, this is a non nvidia machine just in case that is the bad 
variable in both machines.

libtool-cross info::
log.do_configure
/configure: WARNING: unrecognized options: --disable-silent-rules, 
--with-sysroot/

log.do_compile:
help2man is installed on the system, whereis help2man
help2man: /usr/bin/help2man /usr/lib/help2man 
/usr/share/man/man1/help2man.1.gz

/WARNING: `help2man' is missing on your system.  You should only need it if
      you modified a dependency of a manual page.  You may need the
      `Help2man' package in order for those modifications to take
      effect.  You can get `Help2man' from any GNU archive site.
PATH=".:$PATH"; export PATH; /bin/sh 
/home/gusterson/OpenEmbed/setup-scripts/build/tmp-angstrom_2010_x-eglibc/work/armv7a-angstrom-linux-gnueabi/libtool-cross-2.4.2-r2.1/libtool-2.4.2/libltdl/config/missing 
--run help2man --output=doc/libtoolize.1 libtoolize
WARNING: `help2man' is missing on your system.  You should only need it if
      you modified a dependency of a manual page.  You may need the
      `Help2man' package in order for those modifications to take
      effect.  You can get `Help2man' from any GNU archive site./

log.do_package
/DEBUG: Preparing tree 
/home/gusterson/OpenEmbed/setup-scripts/build/tmp-angstrom_2010_x-eglibc/work/armv7a-angstrom-linux-gnueabi/libtool-cross-2.4.2-r2.1/shlibs 
for packaging at 
/home/gusterson/OpenEmbed/setup-scripts/build/tmp-angstrom_2010_x-eglibc/work/armv7a-angstrom-linux-gnueabi/libtool-cross-2.4.2-r2.1/sstate-build-package/shlibs
sed: no input files/

log.do_package_write_ipk & log.do_populate_lic
/sed: no input files/

log.do_populate_sysroot
/DEBUG: No site_config directory, skipping do_siteconfig/

Again a big thank you!


  reply	other threads:[~2012-04-19  3:42 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-04-13 22:46 ld segfault cant figure it out j
2012-04-13 22:50 ` j
2012-04-16 22:11   ` [oe-core] " j
2012-04-16 22:28     ` Gary Thomas
2012-04-16 22:32       ` j
2012-04-16 22:41         ` Gary Thomas
2012-04-16 22:54           ` j
2012-04-16 22:34       ` j
2012-04-17  1:43         ` j
2012-04-17 15:41           ` J. L. 
2012-04-17 15:54             ` Gary Thomas
2012-04-17 16:11               ` J. L. 
2012-04-17 16:13               ` J. L. 
2012-04-17 19:09                 ` j
2012-04-17 19:34                   ` Mark Hatle
2012-04-17 19:49                     ` j
2012-04-17 20:17                       ` j
2012-04-17 21:16                         ` j
2012-04-18  0:37                           ` j
2012-04-18  5:30                             ` Khem Raj
2012-04-18  5:40                               ` j
2012-04-18 22:00                                 ` Khem Raj
2012-04-19  3:32                                   ` j [this message]
2012-04-19  6:24                                     ` Khem Raj
2012-04-20  0:39                                       ` j
2012-04-20 11:13                                         ` Gary Thomas
2012-04-20 16:17                                           ` j
2012-04-20 18:03                                           ` j
2012-04-21 21:47                                             ` [oe-core] 3 build machines all failing in the same manner - was:ld " j
2012-04-24 20:39                                               ` j
2012-04-24 20:44                                                 ` Phil Blundell
2012-04-24 20:57                                                   ` j
2012-04-24 20:57                                                   ` j
2012-04-24 21:03                                                     ` Phil Blundell
2012-04-24 21:15                                                       ` j
2012-04-27 17:16                                                         ` j

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=4F8F8749.8000803@gmail.com \
    --to=vwyodapink@gmail.com \
    --cc=openembedded-devel@lists.openembedded.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox