Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Ulf Samuelsson <ulf@atmel.com>
To: buildroot@busybox.net
Subject: [Buildroot] Several issues by a buildroot newbie...
Date: Thu, 4 Oct 2007 06:14:47 +0200	[thread overview]
Message-ID: <003d01c8063e$26edb230$b07ee255@atmel.com> (raw)
In-Reply-To: 4704186D.6070303@cavebear.com

----- Original Message ----- 
From: "Karl Auerbach" <karl@cavebear.com>
To: <buildroot@uclibc.org>
Sent: Thursday, October 04, 2007 12:32 AM
Subject: [Buildroot] Several issues by a buildroot newbie...


> I'm fairly new to buildroot so please excuse me if I inadvertently touch 
> something that has been raised previously:
> 
> 1. Undefined reference in uClibc:
>    In function `__libc_pthread_init':
>    libc_pthread_init.c:(.text+0xf): undefined reference to
>    `__libc_multiple_threads'
> 
>    I dug through the code and it appears that the missing variable is 
> used as a return value for __libc_pthread_init() but that that return 
> value appears to be squirreled away and perhaps never used.
> 
>   In any event, this is, I suspect the cause of the next item, below.
> 
> 2. Python segfaults when being loaded.  For various reasons I'm very 
> suspicions that the uClibc dynamic library loader is segfaulting.  I've 
> tried to build a static image of Python but I can't because of the 
> unresolved in the prior item.
> 
> I had a similar problem with sudo and all the openssh tools with the 
> stable version of uClibc.  The segfaulting of those, but not of Python, 
> went away with the snapshot version of uClibc.
> 
>   Python did compile and run quite nicely on my year old snapshot of 
> buildroot, but it's effectively dead now.
> 
> 3. The sudo sources aren't available via the URL given in the 
> package/sudo/sudo.mk.  The site, http://www.courtesan.com/sudo/dist, has 
> upgraded to a newer version and the older one to be downloaded appears 
> no longer on that site.
> 
> 4. The lockfile-progs.c (in package lockfile-progs) uses rindex() but 
> does not include the correct header file that defines it.  This can be 
> repaired by a patch that either changes rindex() to strrchr() or 
> including the correct header.  (I did the former.)
> 

uClibc supports rindex if either of 
* UCLIBC_SUSV3_LEGACY or 
* UCLIBC_SYSV3_LEGACY MACROS 
is defined. 
Previously UCLIBC_SUSV3_LEGACY_MACROS was defined,
but Bernhard wants to get rid of all legacy stuff, and his idea to achieve
this it is by intentionally breaking the build if you enable a package 
containing SUSV3_LEGACY stuff.

> --karl--
> 
> 
BR
Ulf Samuelsson

  parent reply	other threads:[~2007-10-04  4:14 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-10-03 22:32 [Buildroot] Several issues by a buildroot newbie Karl Auerbach
2007-10-03 23:55 ` [Buildroot] GDB for threads Leonid
2007-10-09 17:51   ` Rex Ashbaugh
2007-10-04  4:14 ` Ulf Samuelsson [this message]
2007-10-07 16:30   ` [Buildroot] Several issues by a buildroot newbie Bernhard Fischer
2007-10-07 18:04     ` Ulf Samuelsson
2007-10-07 16:35 ` Bernhard Fischer
2007-10-07 19:19   ` Karl Auerbach

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='003d01c8063e$26edb230$b07ee255@atmel.com' \
    --to=ulf@atmel.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