All of lore.kernel.org
 help / color / mirror / Atom feed
From: Alejandro del Castillo <alejandro.delcastillo@ni.com>
To: Philip Tricca <flihp@twobit.us>
Cc: yocto@yoctoproject.org, Xin.Ouyang@windriver.com
Subject: Re: [meta-selinux][PATCH] busybox: explicitly set temporary shell scripts to be executable
Date: Mon, 7 Mar 2016 09:59:57 -0600	[thread overview]
Message-ID: <56DDA57D.3060209@ni.com> (raw)
In-Reply-To: <56D67D70.6010608@twobit.us>



On 03/01/2016 11:43 PM, Philip Tricca wrote:
> On 03/01/2016 02:58 PM, Alejandro del Castillo wrote:
>> I first started looking at busybox postinst (modified on meta-selinux) since I
>> am getting errors during "opkg upgrade buysbox". The errors that I am getting
>> are of the form:
>>
>> update-alternatives: Error: not linking /bin/gunzip to
>> /usr/lib/busybox/bin/gunzip since /bin/gunzip exists and is not a link
>>
>> Which only happen for utilities that don't have any alternative installed after
>> removing busybox, hence no symlink. Utilities that do have an alternative
>> installed are correctly processed, for example:
>>
>> update-alternatives: Linking /usr/bin/ar to /usr/bin/x86_64-nilrt-linux-ar
>>
>> The root problem is that meta-selinux on commit
>> 521ca9c9cf370840e9f8c808a7955aa5da7c356e replaced the temporary symlink creation
>> on postinst with sh wrappers, which breaks update-alternatives when it tries to
>> update symlinks (which are now sh wrappers).
>>
>> Is there a reason why the postinst needs to be different on the selinux version
>> of busybox? If the postinst is just creating temporary symlinks which will go
>> away once the update-alternatives block later on the script runs, do we need the
>> sh wrappers?
> 
> Thanks for the additional context Alejandro. I don't have enough history
> yet to pull answers to stuff like this out at will though. Hopefully
> someone with the right background will chime in. Either way I'll circle
> back once the easy stuff in the backlog is done and we'll try to sort
> something out.

Sounds good, I hope someone can provide background since it looks to me like the
postinst is wrong, but I don't have historic context.

Let me know if I can help in any way.

-- 
Cheers,

Alejandro


      reply	other threads:[~2016-03-07 16:00 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-02-19 22:45 [meta-selinux][PATCH] busybox: explicitly set temporary shell scripts to be executable Alejandro del Castillo
2016-03-01 22:58 ` Alejandro del Castillo
2016-03-02  5:43   ` Philip Tricca
2016-03-07 15:59     ` Alejandro del Castillo [this message]

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=56DDA57D.3060209@ni.com \
    --to=alejandro.delcastillo@ni.com \
    --cc=Xin.Ouyang@windriver.com \
    --cc=flihp@twobit.us \
    --cc=yocto@yoctoproject.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.