Openembedded Core Discussions
 help / color / mirror / Atom feed
From: Tudor Florea <tudor.florea@enea.com>
To: Mark Hatle <mark.hatle@windriver.com>,
	<openembedded-core@lists.openembedded.org>
Subject: Re: rpm generation question
Date: Wed, 20 Apr 2016 03:15:38 +0300	[thread overview]
Message-ID: <5716CA2A.4050001@enea.com> (raw)
In-Reply-To: <571672AD.9060706@windriver.com>



On 19/04/2016 21:02, Mark Hatle wrote:
> On 4/19/16 12:05 PM, Tudor Florea wrote:
>> As part of our test infrastructure we're attempting to install ptest
>> packages (and execute the tests accordingly).
>> Attempting to install ptest for valgrind fails with this error:
>>
>> error: Can't install valgrind-ptest-3.11.0-r0.1@ppce500mc: no package
>> provides /this/is/a/bogus/interpreter/name
> 
> You can turn off the file level requires and provides on a per package basis.
> 
> SKIP_FILEDEPS_${PN}-ptest = '1'

Mark,
Thank you for the information provided.
Unfortunately this did not work as expected.
First issue: the dependency on bogus interpreter still exists with the
line above.
I was able to force remove this dependency by setting
MERGEPERFILEDEPS = "0"
on meta/classes/package_rpm.bbclass. Of course this is not a fix but
only a test. I'm not sure if there is something missing around.

The second issue: The rpm created this way can be installed using the
command:
rpm -ivH ./valgrind-ptest-3.11.0-r0.4.ppce500mc.rpm
but still fails when attempt to install using smart.

Is this second issue a bug or smart is supposed to work this way?


Thank you,
 Tudor.

> 
> Otherwise (for rpm packages) the system will attempt to discover and use the
> per-file interpreter and other dependencies.  Since ptest is only tests, and the
> dependencies here are specific to tests -- it should be safe to disable them.
> 
> (If that isn't the right approach for some reason, there are some other ways to
> do a file level provide, but they're significantly more complicated and rarely
> used.)
> 
> --Mark
> 
>>
>> smart install valgrind-ptest-3.11.0-r0.1@ppce500mc
>> Loading cache...
>> Updating cache...
>> ###############################################################################
>> [100%]
>>
>> Computing transaction...
>>
>>  error: Can't install valgrind-ptest-3.11.0-r0.1@ppce500mc: no package
>> provides /this/is/a/bogus/interpreter/name
>>
>>
>> This is most probably caused by the file shell_badinterp contained in
>> this package having the following contents:
>>
>> #! /this/is/a/bogus/interpreter/name
>>
>> true
>>
>> Does anyone have an idea how to get rid of this error?
>> More specific: How can I exclude dependency on
>> "/this/is/a/bogus/interpreter/name" for an rpm package?
>>
>> Thank you very much.
>>   Tudor.
>>
> 


  reply	other threads:[~2016-04-20  0:15 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-04-19 17:05 rpm generation question Tudor Florea
2016-04-19 18:02 ` Mark Hatle
2016-04-20  0:15   ` Tudor Florea [this message]
2016-04-20  1:08     ` Mark Hatle
2016-04-21  8:24       ` Tudor Florea

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=5716CA2A.4050001@enea.com \
    --to=tudor.florea@enea.com \
    --cc=mark.hatle@windriver.com \
    --cc=openembedded-core@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