[Buildroot] uboot-tools autobuild failures

[Buildroot] uboot-tools autobuild failures

[Matthew Weber]

It looks like all of these recent failures are because the python swig
build does a "$(if $(shell which swig 2> /dev/null),_libfdt.so)" to
determine if the host system has swig and if so, try to build the
python bindings.  Obviously not taking into account any prefix we'd
provide to the host/ in out build.

http://autobuild.buildroot.org/?reason=uboot-tools-2017.03

I don't directly have a need for those bindings, would a patch that
disables this swig detect be reasonable at this point?  Other option
would be to rewrite how they're doing it to search a prefix path vs
using the shell's which.  Then work upstreaming of that change.

-Matt

[Buildroot] uboot-tools autobuild failures

[Baruch Siach]

Hi Matthew,

On Wed, Apr 19, 2017 at 08:55:58PM -0500, Matthew Weber wrote:
> It looks like all of these recent failures are because the python swig
> build does a "$(if $(shell which swig 2> /dev/null),_libfdt.so)" to
> determine if the host system has swig and if so, try to build the
> python bindings.  Obviously not taking into account any prefix we'd
> provide to the host/ in out build.
> 
> http://autobuild.buildroot.org/?reason=uboot-tools-2017.03
> 
> I don't directly have a need for those bindings, would a patch that
> disables this swig detect be reasonable at this point?  Other option
> would be to rewrite how they're doing it to search a prefix path vs
> using the shell's which.  Then work upstreaming of that change.

This was discussed in the past with no conclusion.

http://lists.busybox.net/pipermail/buildroot/2017-March/187638.html
http://lists.busybox.net/pipermail/buildroot/2017-March/187681.html
http://lists.busybox.net/pipermail/buildroot/2017-March/187700.html

baruch

-- 
     http://baruch.siach.name/blog/                  ~. .~   Tk Open Systems
=}------------------------------------------------ooO--U--Ooo------------{=
   - baruch at tkos.co.il - tel: +972.52.368.4656, http://www.tkos.co.il -

[Buildroot] uboot-tools autobuild failures

[Arnout Vandecappelle]

On 20-04-17 03:55, Matthew Weber wrote:
> It looks like all of these recent failures are because the python swig
> build does a "$(if $(shell which swig 2> /dev/null),_libfdt.so)" to
> determine if the host system has swig and if so, try to build the
> python bindings.  Obviously not taking into account any prefix we'd
> provide to the host/ in out build.
> 
> http://autobuild.buildroot.org/?reason=uboot-tools-2017.03
> 
> I don't directly have a need for those bindings, would a patch that
> disables this swig detect be reasonable at this point?  Other option
> would be to rewrite how they're doing it to search a prefix path vs
> using the shell's which.  Then work upstreaming of that change.

 The problem with patching uboot-tools is that you usually also built uboot when
you build uboot-tools, and most likely uboot will also fail to build (at least
if it's a recent version). But we can't patch uboot because we don't know which
version it is... So we'd need to do horrible sed-based "patching" in uboot.mk...

 So the ideal solution would be to convince the U-Boot build system that swig
isn't available, without patching... But that would be difficult...

 How about just touching tools/_libfdt.so, so that it doesn't get rebuilt?

 Regards,
 Arnout

-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF

[Buildroot] uboot-tools autobuild failures

[Fabio Estevam]

Hi Matthew,

On Wed, Apr 19, 2017 at 10:55 PM, Matthew Weber
<matthew.weber@rockwellcollins.com> wrote:
> It looks like all of these recent failures are because the python swig
> build does a "$(if $(shell which swig 2> /dev/null),_libfdt.so)" to
> determine if the host system has swig and if so, try to build the
> python bindings.  Obviously not taking into account any prefix we'd
> provide to the host/ in out build.
>
> http://autobuild.buildroot.org/?reason=uboot-tools-2017.03
>
> I don't directly have a need for those bindings, would a patch that
> disables this swig detect be reasonable at this point?  Other option
> would be to rewrite how they're doing it to search a prefix path vs
> using the shell's which.  Then work upstreaming of that change.

U-Boot mainline has a fix for this already:
http://git.denx.de/?p=u-boot.git;a=commitdiff;h=9963890b8be1d208035945abc5ba9f77637b6542

My host PC does not have Python 3, so I cannot reproduce this error myself.

Could you please try the attached Buildroot patch?

Regards,

Fabio Estevam
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-libfdtfix.patch
Type: text/x-patch
Size: 2173 bytes
Desc: not available
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20170423/d6629d56/attachment.bin>

[Buildroot] uboot-tools autobuild failures

[Matthew Weber]

Fabio

On Sun, Apr 23, 2017 at 6:44 PM, Fabio Estevam <festevam@gmail.com> wrote:
> Hi Matthew,
>
> On Wed, Apr 19, 2017 at 10:55 PM, Matthew Weber
> <matthew.weber@rockwellcollins.com> wrote:
>> It looks like all of these recent failures are because the python swig
>> build does a "$(if $(shell which swig 2> /dev/null),_libfdt.so)" to
>> determine if the host system has swig and if so, try to build the
>> python bindings.  Obviously not taking into account any prefix we'd
>> provide to the host/ in out build.
>>
>> http://autobuild.buildroot.org/?reason=uboot-tools-2017.03
>>
>> I don't directly have a need for those bindings, would a patch that
>> disables this swig detect be reasonable at this point?  Other option
>> would be to rewrite how they're doing it to search a prefix path vs
>> using the shell's which.  Then work upstreaming of that change.
>
> U-Boot mainline has a fix for this already:
> http://git.denx.de/?p=u-boot.git;a=commitdiff;h=9963890b8be1d208035945abc5ba9f77637b6542
>
> My host PC does not have Python 3, so I cannot reproduce this error myself.
>
> Could you please try the attached Buildroot patch?

No luck, same error.  I did notice this seems to be impacted by -j level.

Matt

[Buildroot] uboot-tools autobuild failures

[Fabio Estevam]

On Tue, Apr 25, 2017 at 2:07 PM, Matthew Weber
<matthew.weber@rockwellcollins.com> wrote:

> No luck, same error.  I did notice this seems to be impacted by -j level.

Please report this error in the U-boot list. Thanks

[Buildroot] uboot-tools autobuild failures

[Matthew Weber]

All,

On Tue, Apr 25, 2017 at 12:09 PM, Fabio Estevam <festevam@gmail.com> wrote:
> On Tue, Apr 25, 2017 at 2:07 PM, Matthew Weber
> <matthew.weber@rockwellcollins.com> wrote:
>
>> No luck, same error.  I did notice this seems to be impacted by -j level.
>
> Please report this error in the U-boot list. Thanks

Done and I submitted a workaround for now until a clean fix can be
found upstream.
https://patchwork.ozlabs.org/patch/757382/