Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: "Stefan Fröberg" <stefan.froberg@petroprogram.com>
To: buildroot@busybox.net
Subject: [Buildroot] [PATCH] busybox: needs linux-pam dependency added if using custom busybox .config & busybox login applet
Date: Sun, 18 Nov 2012 20:09:29 +0200	[thread overview]
Message-ID: <50A92459.1070404@petroprogram.com> (raw)
In-Reply-To: <20121118184849.660aad23@skate>

18.11.2012 19:48, Thomas Petazzoni kirjoitti:
> Dear Stefan Fr?berg,
>
> On Sun, 18 Nov 2012 19:39:16 +0200, Stefan Fr?berg wrote:
>
>> And that CONFIG_LOGIN checking should be done after buybox tarball has
>> been extracted but by then, it is already too late for
>> determining dependencies ....
>> Aarrggggh!
> I think you understood what the problem is :)
>
>>> It is really difficult to have dependencies that are needed only when
>>> some specific Busybox features are enabled.
>> Tell me about it.
>> What IMHO is needed, is some function or way to get these kinds of
>> information.
>> Just before dependencies are determined.
>>
>> There should be KCONFIG_GET_OPT or something similar inside
>> package/pkg-utils.mk
>> (like there already is KCONFIG_ENABLE_OPT, KCONFIG_DISABLE_OPT and
>> KCONFIG_SET_OPT)
>> that would help to determine not only buildroot features but also linux
>> features and any
>> else package that use kconfig like stuff.
>>
>> It would also help in case of  various package conflicts (say if busybox
>> unzip applet and real unzip are mistakenly
>> being installed, then this could be detected beforehand)
> KCONFIG_GET_OPT cannot work: you could only call it in a Makefile rule
> (i.e inside BUILD_CMDS or CONFIGURE_CMDS or a hook or something like
> that). But by the time those are executed, it is way too late to
> declare a dependency. The dependencies (in BUSYBOX_DEPENDENCIES) must
> be detailed when make parses all the .mk files, not when the rules are
> executed.
>
> Unless I'm wrong, linux-pam is only needed if ENABLE_PAM is defined
> when building Busybox, correct? So, CONFIG_LOGIN builds just fine
> without PAM, doesn't it?
Yes that is correct, there is a CONFIG_PAM in BusyBox .config  and althought
I have never actually tried to build it without pam it should in theory
work.

>
> So maybe you should rather do:
>
> ifeq ($(BR2_PACKAGE_LINUX_PAM),y)
> BUSYBOX_DEPENDENCIES += linux-pam
> BUSYBOX_CFLAGS += -DENABLE_PAM
> endif
>
> No?

Hmmm...
I will try that and see.

Thanks!

Regards
Stefan
> Thomas

  reply	other threads:[~2012-11-18 18:09 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-11-18  0:15 [Buildroot] [PATCH] busybox: needs linux-pam dependency added if using custom busybox .config & busybox login applet Stefan Fröberg
2012-11-18  2:14 ` Danomi Manchego
2012-11-18 11:51   ` Stefan Fröberg
2012-11-18 12:11     ` Stefan Fröberg
2012-11-18 14:01       ` Stefan Fröberg
2012-11-18 16:40 ` Thomas Petazzoni
2012-11-18 17:39   ` Stefan Fröberg
2012-11-18 17:48     ` Thomas Petazzoni
2012-11-18 18:09       ` Stefan Fröberg [this message]
2012-11-18 18:28         ` Thomas Petazzoni

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=50A92459.1070404@petroprogram.com \
    --to=stefan.froberg@petroprogram.com \
    --cc=buildroot@busybox.net \
    /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