From: Claudius Heine <claudius.heine.ext@siemens.com>
To: Jan Kiszka <jan.kiszka@siemens.com>,
isar-users <isar-users@googlegroups.com>
Subject: Re: Sporadic build failure of next
Date: Tue, 20 Aug 2019 09:37:08 +0200 [thread overview]
Message-ID: <b62546ce-ffa6-1043-969f-41a1f49410e4@siemens.com> (raw)
In-Reply-To: <e170d881-9672-dba3-d048-a7a9a43d6a77@siemens.com>
On 20/08/2019 09.09, Jan Kiszka wrote:
> On 20.08.19 08:54, Claudius Heine wrote:
>> Hi Jan,
>>
>> On 20/08/2019 07.37, [ext] Jan Kiszka wrote:
>>> Hi,
>>>
>>> attached a build failure of 49895c62cdca. I ran that build 3 more
>>> times, but it didn't trigger an issue again. Looks like some false
>>> sharing, but I do not have an idea yet of what exactly.
>>
>> This is probably the important part of the log:
>>
>> | dpkg-shlibdeps: error: no dependency information found for
>> /lib/arm-linux-gnueabihf/libc.so.6 (used by
>> debian/libhello/usr/lib/arm-linux-gnueabihf/libhello.so.0.0.0)
>> | Hint: check if the library actually comes from a package.
>> | dh_shlibdeps: dpkg-shlibdeps -Tdebian/libhello.substvars
>> debian/libhello/usr/lib/arm-linux-gnueabihf/libhello.so.0.0.0 returned
>> exit code 2
>>
>> The manpage dpkg-shlibdeps(1) states:
>>
>> no dependency information found for library-file (used by
>> binary).
>> The library needed by binary has been found by
>> dpkg-shlibdeps in library-
>> file but dpkg-shlibdeps has been unable to find any
>> dependency information
>> for that library. To find out the dependency, it has
>> tried to map the
>> library to a Debian package with the help of dpkg -S
>> library-file. Then
>> it checked the corresponding shlibs and
>> symbols files in
>> /var/lib/dpkg/info/, and in the various
>> package's build trees
>> (debian/*/DEBIAN/).
>>
>> This failure can be caused by a bad or missing shlibs
>> or symbols file in
>> the package of the library. It might also happen if
>> the library is built
>> within the same source package and if the shlibs files
>> has not yet been
>> created (in which case you must fix debian/rules
>> to create the shlibs
>> before calling dpkg-shlibdeps). Bad RPATH can also
>> lead to the library
>> being found under a non-canonical
>> name (example:
>> /usr/lib/openoffice.org/../lib/libssl.so.0.9.8
>> instead of
>> /usr/lib/libssl.so.0.9.8) that's not associated
>> to any package,
>> dpkg-shlibdeps tries to work around this by trying
>> to fallback on a
>> canonical name (using realpath(3)) but it might
>> not always work. It's
>> always best to clean up the RPATH of the binary to
>> avoid problems.
>>
>> Calling dpkg-shlibdeps in verbose mode (-v) will
>> provide much more
>> information about where it tried to find the
>> dependency information. This
>> might be useful if you don't understand why it's giving
>> you this error.
>>
>> Maybe inspection of the work space on the ci server when this error
>> occurred will help.
>
> Unfortunately, that is history already (disposed docker builds). But it
> has to be something that destroyed the information because the later
> builds got past this point.
Maybe its a issue with the libhello package? What I found odd was the
path. libc6-armhf-cross contains the libc here:
libc6-armhf-cross: /usr/arm-linux-gnueabihf/lib/libc.so.6
So a simple search (dpkg -S) for '/lib/arm-linux-gnueabihf/libc.so.6'
will not give that result.
Maybe the rpath needs to be fixed? Maybe the rpath is sometimes fixed,
and sometimes not? race?
regards,
Claudius
--
DENX Software Engineering GmbH, Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-54 Fax: (+49)-8142-66989-80 Email: ch@denx.de
next prev parent reply other threads:[~2019-08-20 7:37 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-20 5:37 Jan Kiszka
2019-08-20 6:54 ` Claudius Heine
2019-08-20 7:09 ` Jan Kiszka
2019-08-20 7:37 ` Claudius Heine [this message]
2019-08-20 10:31 ` Baurzhan Ismagulov
2019-08-20 10:40 ` Jan Kiszka
2019-08-20 11:05 ` Baurzhan Ismagulov
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=b62546ce-ffa6-1043-969f-41a1f49410e4@siemens.com \
--to=claudius.heine.ext@siemens.com \
--cc=isar-users@googlegroups.com \
--cc=jan.kiszka@siemens.com \
/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