headers
jlc | 1 Dec 2009 01:08
Picon

patch for opkg_download.c compile abort

If configure is run without options, the compile dies because of FALSE
and TRUE being redefined.  This
is true up to R412.

Here is the compiler output followed by the patch:

In file included from opkg_download.c:50:
../libbb/libbb.h:32:1: error: "FALSE" redefined
In file included from /usr/include/krb5.h:7,
                 from /usr/include/openssl/kssl.h:72,
                 from /usr/include/openssl/ssl.h:192,
                 from opkg_download.c:29:
/usr/include/krb5/krb5.h:187:1: error: this is the location of the
previous definition
In file included from opkg_download.c:50:
../libbb/libbb.h:33:1: error: "TRUE" redefined
In file included from /usr/include/krb5.h:7,
                 from /usr/include/openssl/kssl.h:72,
                 from /usr/include/openssl/ssl.h:192,
                 from opkg_download.c:29:
/usr/include/krb5/krb5.h:190:1: error: this is the location of the
previous definition
make[2]: *** [opkg_download.lo] Error 1
make[1]: *** [all] Error 2
make: *** [all-recursive] Error 1

The patch just moves the fixed includes ahead of the conditional
includes:

libopkg/opkg_download.c
(Continue reading)

Permalink | Reply | 1 comment |
headers
Camille Moncelier | 1 Dec 2009 09:04

Re: patch for opkg_download.c compile abort

I commited your fix in r413 (with extra #ifndef TRUE and #ifndef FALSE
in libbb.h)

On Tue, Dec 1, 2009 at 1:08 AM, jlc <jlcster@...> wrote:
> If configure is run without options, the compile dies because of FALSE
> and TRUE being redefined.  This
> is true up to R412.
>
> Here is the compiler output followed by the patch:
>
> In file included from opkg_download.c:50:
> ../libbb/libbb.h:32:1: error: "FALSE" redefined
> In file included from /usr/include/krb5.h:7,
>                 from /usr/include/openssl/kssl.h:72,
>                 from /usr/include/openssl/ssl.h:192,
>                 from opkg_download.c:29:
> /usr/include/krb5/krb5.h:187:1: error: this is the location of the
> previous definition
> In file included from opkg_download.c:50:
> ../libbb/libbb.h:33:1: error: "TRUE" redefined
> In file included from /usr/include/krb5.h:7,
>                 from /usr/include/openssl/kssl.h:72,
>                 from /usr/include/openssl/ssl.h:192,
>                 from opkg_download.c:29:
> /usr/include/krb5/krb5.h:190:1: error: this is the location of the
> previous definition
> make[2]: *** [opkg_download.lo] Error 1
> make[1]: *** [all] Error 2
> make: *** [all-recursive] Error 1
>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Graham Gower | 2 Dec 2009 00:23
Picon

Re: opkg help and documentation

Excellent, thanks.

I'm not knowledgeable with nroff nor autotools, so please bear with me
while I familiarise myself with them so that this can be smoothly
integrated.

-Graham

--

You received this message because you are subscribed to the Google Groups "opkg-devel" group.
To post to this group, send email to opkg-devel@...
To unsubscribe from this group, send email to opkg-devel+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/opkg-devel?hl=en.
Permalink | Reply | 2 comments |
headers
John Willis | 2 Dec 2009 12:46
Favicon

Re: tagging opkg 0.1.7

Hi All,

> OE is using r402/trunk now, so if people don't start bitching on the
> OE lists soon, it looks pretty good.

There seem to be a number of small issues that have surfaced with opkg
(in it's -native OE form) since the SRC bump that are showing up when
building the rootfs for images (in both reported cases, big images, >
200meg rootfs size and lots of packages).

These manifest themselves as opkg errors in the rootfs log (and the
respective package failing leading to the rootfs build failing).

The errors look something like below but the packages very from image
to image (in my case udev seems a constant failure across images but
others have reported that there 'package issues' list is different for
there environment).

"
Installing udev (141-r15.5) to root...
Downloading file:/oe/omap3evm-oe/tmp/deploy/glibc/ipk/omap3-pandora/
udev_141-r15.5_omap3-pandora.ipk
omap3-pandora.ipk
omap3-pandora.ipk
(offline root mode: not running udev.preinst)
omap3-pandora.ipk
(offline root mode: not running udev.preinst)
omap3-pandora.ipk
(offline root mode: not running udev.preinst)
opkg: Cannot create symlink from ./lib/udev/rules.d to '/etc/udev/
(Continue reading)

Permalink | Reply | 7 comments |
headers
Camille Moncelier | 2 Dec 2009 17:30

Re: opkg help and documentation

As I sait previously, I'm not a big fan of nroff syntax for creating man pages.

What do you think about makeing somthing like this:

http://www.google.com/codesearch/p?hl=en#N-lCuqhWHXY/trunk/tovid/Makefile.am&q=makefile.am%20txt2tags

It use txt2tags to make a man page from a wiki like syntax. The author added
a dist hook in the makefile to package the compiled man pages in order to
not depends on txt2tags when releasing.

IMHO, using a tool like txt2tags can make make page editing A LOT easier.

A sample txt2tags man page:
- http://www.google.com/codesearch/p?hl=en#S7R6oHNNbis/pub/distfiles/tovid-0.28.tar.gz|6tww3XDXzOU/tovid-0.28/docs/src/en/idvid.t2t

Compiled result:
- http://www.google.com/codesearch/p?hl=en#S7R6oHNNbis/pub/distfiles/tovid-0.28.tar.gz|6tww3XDXzOU/tovid-0.28/docs/man/idvid.1

On Wed, Dec 2, 2009 at 12:23 AM, Graham Gower <graham.gower@...> wrote:
> Excellent, thanks.
>
> I'm not knowledgeable with nroff nor autotools, so please bear with me
> while I familiarise myself with them so that this can be smoothly
> integrated.
>
> -Graham
>
> --
>
> You received this message because you are subscribed to the Google Groups "opkg-devel" group.
(Continue reading)

Permalink | Reply | 1 comment |
headers
Koen Kooi | 2 Dec 2009 18:38
Picon

Depends: field getting truncated during parsing with latest trunk (r413)

Hi,

When installing large meta-packages like 'kernel-modules' opkg will
fail with:

Collected errors:
 * ERROR: Cannot satisfy the following dependencies for task-
beagleboard-demo:
         *  kerne *

I've looked at the control files and cannot see any occurence of
'kerne' which leads me to believe something is getting truncated.

regards,

Koen

--

You received this message because you are subscribed to the Google Groups "opkg-devel" group.
To post to this group, send email to opkg-devel@...
To unsubscribe from this group, send email to opkg-devel+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/opkg-devel?hl=en.
Permalink | Reply | 3 comments |
headers
Graham Gower | 2 Dec 2009 21:48
Picon

Re: Re: tagging opkg 0.1.7

2009/12/2 John Willis <John.Willis@...>:
> Hi All,
>
>> OE is using r402/trunk now, so if people don't start bitching on the
>> OE lists soon, it looks pretty good.
>
> There seem to be a number of small issues that have surfaced with opkg
> (in it's -native OE form) since the SRC bump that are showing up when
> building the rootfs for images (in both reported cases, big images, >
> 200meg rootfs size and lots of packages).
>
> These manifest themselves as opkg errors in the rootfs log (and the
> respective package failing leading to the rootfs build failing).

Please see libbb/unarchive.c: extract_archive() and let me know if any
of the failures there which set "err = -1" should be downgraded to
warnings. Previously these conditions were not failure conditions only
because they could not be *by design*.

Maybe you want more sanity checks (e.g. Issue 28) instead of fewer?

-Graham

--

You received this message because you are subscribed to the Google Groups "opkg-devel" group.
To post to this group, send email to opkg-devel@...
To unsubscribe from this group, send email to opkg-devel+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/opkg-devel?hl=en.
(Continue reading)

Permalink | Reply | 6 comments |
headers
Graham Gower | 2 Dec 2009 21:49
Picon

Re: opkg help and documentation

2009/12/3 Camille Moncelier <moncelier@...>:
> As I sait previously, I'm not a big fan of nroff syntax for creating man pages.
>
> What do you think about makeing somthing like this:
>

The big problem is that it adds an extra build time dependency. nroff
is standard and everyone is using it. Maybe its not the best, but its
functional and all the tools already work.

-Graham

--

You received this message because you are subscribed to the Google Groups "opkg-devel" group.
To post to this group, send email to opkg-devel@...
To unsubscribe from this group, send email to opkg-devel+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/opkg-devel?hl=en.
Permalink | Reply | 0 comments |
headers
Graham Gower | 3 Dec 2009 00:27
Picon

Re: Depends: field getting truncated during parsing with latest trunk (r413)

There's probably a bug somewhere in the ugly interaction between
pkg_hash_fetch_unsatisfied_dependencies() and add_unresolved_dep() or
merge_unresolved(). These functions are abusing my sanity.

I'm test installing task-beagleboard-base an offline root with the
following in opkg.conf:

arch beagleboard 2
arch armv7a 3
src/gz bb http://www.angstrom-distribution.org/feeds/2008/ipk/glibc/armv7a/machine/beagleboard/

There is definitely a problem here... investigating.

-Graham

--

You received this message because you are subscribed to the Google Groups "opkg-devel" group.
To post to this group, send email to opkg-devel@...
To unsubscribe from this group, send email to opkg-devel+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/opkg-devel?hl=en.
Permalink | Reply | 2 comments |
headers
Graham Gower | 3 Dec 2009 04:05
Picon

Re: Depends: field getting truncated during parsing with latest trunk (r413)

Fixed in r438. Thanks!

-Graham

--

You received this message because you are subscribed to the Google Groups "opkg-devel" group.
To post to this group, send email to opkg-devel@...
To unsubscribe from this group, send email to opkg-devel+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/opkg-devel?hl=en.
Permalink | Reply | 1 comment |

Gmane