[Buildroot] [autobuild.buildroot.net] Build results for 2015-05-04

[Buildroot] [autobuild.buildroot.net] Build results for 2015-05-04

[Thomas Petazzoni]

Build statistics for 2015-05-04
===============================

        success : 166
       failures : 38 
       timeouts : 8  
          TOTAL : 212

Classification of failures by reason
====================================

            python-pyqt-4.11.3 | 8 
                      qt-4.8.6 | 3 
                     php-5.6.8 | 2 
         lua-periphery-1.0.4-1 | 2 
      host-gdb-arc-2014.12-gdb | 2 
                host-gdb-7.8.2 | 2 
               host-mono-4.0.1 | 2 
                tinyxml2-2.2.0 | 2 
          google-breakpad-1373 | 2 
                  guile-2.0.11 | 2 
          sane-backends-1.0.24 | 2 
                libtirpc-0.2.4 | 1 
                  snmppp-3.3.4 | 1 
               openldap-2.4.40 | 1 
               xapp_xwud-1.0.4 | 1 
                  xerces-3.1.2 | 1 
               valgrind-3.10.1 | 1 
                  cc-tool-0.26 | 1 
                       unknown | 1 
                ipmiutil-2.9.5 | 1 
znc-b396cafdb249544164ed029... | 1 
                poppler-0.32.0 | 1 
                     at-3.1.13 | 1 
filemq-482797b8aa30fcc9ea13... | 1 
                    neard-0.14 | 1 
               libcap-ng-0.7.4 | 1 
       xproto_presentproto-1.0 | 1 
                     vlc-2.2.1 | 1 

Detail of failures
===================

         arm |                      at-3.1.13 | NOK | http://autobuild.buildroot.net/results/6f25e80e8aedec91323ef9b67576b550d7abee60/
         arm |                   cc-tool-0.26 | NOK | http://autobuild.buildroot.net/results/69aff0b9d0ac8fe08e1d2f7ffb691f6a8fc8693b/
      xtensa | filemq-482797b8aa30fcc9ea13... | NOK | http://autobuild.buildroot.net/results/f685d73952a2562e0f3e1c378ddd7885d6433272/
     aarch64 |           google-breakpad-1373 | NOK | http://autobuild.buildroot.net/results/e063085f7d34f8d7940614b26313b28f5eb2396e/
     aarch64 |           google-breakpad-1373 | NOK | http://autobuild.buildroot.net/results/24baf2a135a1bdb6db7bd022e3899482f7f02c10/
     aarch64 |                   guile-2.0.11 | NOK | http://autobuild.buildroot.net/results/85bd849e17c6e5ed6b5b56a4d4ad1a73f7de99c2/
     aarch64 |                   guile-2.0.11 | NOK | http://autobuild.buildroot.net/results/38ce95ef3bf3b8270c48ca8fe839c5313ac51621/
         arm |                 host-gdb-7.8.2 | NOK | http://autobuild.buildroot.net/results/f8037de36e9a0d3b22bea1e52f563b8abeb50500/
      xtensa |                 host-gdb-7.8.2 | NOK | http://autobuild.buildroot.net/results/37238ee36af62a668e132a09490a02cf774b8889/
         arc |       host-gdb-arc-2014.12-gdb | NOK | http://autobuild.buildroot.net/results/f23333bcff3e8d317013faca00351023bf92de88/
         arc |       host-gdb-arc-2014.12-gdb | NOK | http://autobuild.buildroot.net/results/58f0276eea85f35edf7a247a958b1f42340b0a66/
     powerpc |                host-mono-4.0.1 | NOK | http://autobuild.buildroot.net/results/7980bbd50ab63ea2bf482b8f58493ac39790a213/
         arm |                host-mono-4.0.1 | NOK | http://autobuild.buildroot.net/results/1eeb6b7f4c055968f1683a37cec2b2142923cd28/
      x86_64 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/cd2e617f8e2b00581ab5936029f85e62ed3259ba/
       nios2 |                libcap-ng-0.7.4 | NOK | http://autobuild.buildroot.net/results/c8f19279cd49d13b14ec0d9dc116f527325d9c5b/
         arm |                 libtirpc-0.2.4 | NOK | http://autobuild.buildroot.net/results/09a7de29e21501a094f22e9afe470524f5de8c92/
    mips64el |          lua-periphery-1.0.4-1 | NOK | http://autobuild.buildroot.net/results/48efae954e6b030fa8743a5728fb3f30a892f142/
        mips |          lua-periphery-1.0.4-1 | NOK | http://autobuild.buildroot.net/results/ed2150dd5be56056de299d2011cad340dc1e3e56/
         arc |                     neard-0.14 | NOK | http://autobuild.buildroot.net/results/0a466cde55c5e128a2e201924f80f0ec6b8b5c2a/
         arm |                openldap-2.4.40 | NOK | http://autobuild.buildroot.net/results/b41c043fb3b2fad1d9cea0a95b512fb4942b5b19/
        bfin |                      php-5.6.8 | NOK | http://autobuild.buildroot.net/results/633581d8f95efd3fc92ce9875e0608fe40d054bf/
        bfin |                      php-5.6.8 | NOK | http://autobuild.buildroot.net/results/bf12eb189ab35ce00a2212695d2dbf8b8a126529/
         arm |                 poppler-0.32.0 | NOK | http://autobuild.buildroot.net/results/3a70305be4a78af9404b0bd027dbcdd011ca01b3/
         arm |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/bd8796a3001458de5326f96a9110a95bac247225/
         arm |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/156e4afe32f7928880859a69999abcc3c9b304d0/
        sh4a |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/04e209c1b54ec4ccfc0da67a3f077151b65171a7/
         arm |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/2dcc2b044a1235cf9df24e5d55423c5e4ca7196a/
         arm |             python-pyqt-4.11.3 | TIM | http://autobuild.buildroot.net/results/d2f426d839bebebef3d8a41d1d010a0c6ac285a9/
         sh4 |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/de975f29185aa95826b416ee5a60d2f190a979a6/
         arm |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/891dc6ad46039740867a0b436281fc489cfb2772/
     powerpc |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/65fe75b6aa965641f447c41b82d8511bd6be1a86/
    mips64el |                       qt-4.8.6 | NOK | http://autobuild.buildroot.net/results/e0df30e8068cd784bed8e47f5cd13f1f70284027/
    mips64el |                       qt-4.8.6 | NOK | http://autobuild.buildroot.net/results/ccd0534d1dfcb1628f808da91850ec62091f2178/
     powerpc |                       qt-4.8.6 | TIM | http://autobuild.buildroot.net/results/349922ebab920c82cf087080f96c12310d6e66bb/
         arm |           sane-backends-1.0.24 | NOK | http://autobuild.buildroot.net/results/f2a43b62aa2a720c538fc60c80c78c8ed5b04dea/
      xtensa |           sane-backends-1.0.24 | NOK | http://autobuild.buildroot.net/results/31115fe8d88f52d77ed0f2da769eb8896a1b34a2/
     aarch64 |                   snmppp-3.3.4 | NOK | http://autobuild.buildroot.net/results/23d696e3ecbd192669e0de5eae6d339730c1aea2/
        bfin |                 tinyxml2-2.2.0 | NOK | http://autobuild.buildroot.net/results/86a4da3f9f4c9025f2f2393800c797b2af050808/
        bfin |                 tinyxml2-2.2.0 | NOK | http://autobuild.buildroot.net/results/438ef12933aaac24029e352fdadeedb76bbd1cbb/
         arm |                        unknown | TIM | http://autobuild.buildroot.net/results/c2f41853747d32f253ba251f8919e02d47f69719/
         arm |                valgrind-3.10.1 | TIM | http://autobuild.buildroot.net/results/f5aa99efe31e5b417ba27a0de0408eed10926375/
         arm |                      vlc-2.2.1 | NOK | http://autobuild.buildroot.net/results/3612b54b0bbc2a55ba6f0febb7be32595b523d96/
    mips64el |                xapp_xwud-1.0.4 | TIM | http://autobuild.buildroot.net/results/5c02fe80e6072ee8d1794c20a31850af71eb10eb/
    mips64el |                   xerces-3.1.2 | TIM | http://autobuild.buildroot.net/results/9c5cc4182b561804476bdcbda5b7dbb763110aa2/
         arc |        xproto_presentproto-1.0 | TIM | http://autobuild.buildroot.net/results/b6abf41c886ade1e88d2cce7c0250f58a966aa7c/
         arm | znc-b396cafdb249544164ed029... | TIM | http://autobuild.buildroot.net/results/990172b1d3b2e028adba947efad9544a5a705f78/


-- 
http://autobuild.buildroot.net

[Buildroot] Analysis of build failures

[Thomas Petazzoni]

Hello,

Simon, Pascal, Gustavo, Vicente, Max, Samuel, Bernd, look below, there
are some questions/tasks for you! :-)

On Tue,  5 May 2015 08:30:15 +0200 (CEST), Thomas Petazzoni wrote:

>          arm |                      at-3.1.13 | NOK | http://autobuild.buildroot.net/results/6f25e80e8aedec91323ef9b67576b550d7abee60/

musl build problem. Yann has sent a patch to disable at on musl, but I
would actually prefer to fix at.

>          arm |                   cc-tool-0.26 | NOK | http://autobuild.buildroot.net/results/69aff0b9d0ac8fe08e1d2f7ffb691f6a8fc8693b/

There a -L/lib64 hardcoded somewhere.

>       xtensa | filemq-482797b8aa30fcc9ea13... | NOK | http://autobuild.buildroot.net/results/f685d73952a2562e0f3e1c378ddd7885d6433272/

Two issues:

 * Linking C++ code with gcc
 * Missing -lpthread since we're linking statically

Simon, can you have a look?

>      aarch64 |           google-breakpad-1373 | NOK | http://autobuild.buildroot.net/results/e063085f7d34f8d7940614b26313b28f5eb2396e/
>      aarch64 |           google-breakpad-1373 | NOK | http://autobuild.buildroot.net/results/24baf2a135a1bdb6db7bd022e3899482f7f02c10/

./src/client/linux/minidump_writer/linux_dumper.h:93:23: error: field 'regs' has incomplete type 'google_breakpad::user_pt_regs'
   struct user_pt_regs regs;

Does google-breakpad really works on AArch64 ?

Pascal, you're the original submitter of the google breakpad support,
can you have a look ?

>      aarch64 |                   guile-2.0.11 | NOK | http://autobuild.buildroot.net/results/85bd849e17c6e5ed6b5b56a4d4ad1a73f7de99c2/
>      aarch64 |                   guile-2.0.11 | NOK | http://autobuild.buildroot.net/results/38ce95ef3bf3b8270c48ca8fe839c5313ac51621/

Should be fixed by http://patchwork.ozlabs.org/patch/467829/.

>          arm |                 host-gdb-7.8.2 | NOK | http://autobuild.buildroot.net/results/f8037de36e9a0d3b22bea1e52f563b8abeb50500/
>       xtensa |                 host-gdb-7.8.2 | NOK | http://autobuild.buildroot.net/results/37238ee36af62a668e132a09490a02cf774b8889/
>          arc |       host-gdb-arc-2014.12-gdb | NOK | http://autobuild.buildroot.net/results/f23333bcff3e8d317013faca00351023bf92de88/
>          arc |       host-gdb-arc-2014.12-gdb | NOK | http://autobuild.buildroot.net/results/58f0276eea85f35edf7a247a958b1f42340b0a66/

These should be fixed by http://patchwork.ozlabs.org/patch/460494/ or
http://patchwork.ozlabs.org/patch/449686/. Not sure which one is the
most appropriate solution.

>      powerpc |                host-mono-4.0.1 | NOK | http://autobuild.buildroot.net/results/7980bbd50ab63ea2bf482b8f58493ac39790a213/
>          arm |                host-mono-4.0.1 | NOK | http://autobuild.buildroot.net/results/1eeb6b7f4c055968f1683a37cec2b2142923cd28/

Smells like parallel installation issue, but not sure. Angelo tried to
investigate, but was not able to reach a conclusion. Anyone to look at
this?

>       x86_64 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/cd2e617f8e2b00581ab5936029f85e62ed3259ba/

md2.o: In function `md2_sum':
md2.c:(.text+0x18): undefined reference to `EVP_md2'

Gustavo, maybe? Has OpenSSL removed md2 support by default or something
like that?

>        nios2 |                libcap-ng-0.7.4 | NOK | http://autobuild.buildroot.net/results/c8f19279cd49d13b14ec0d9dc116f527325d9c5b/

Compiler issue, let's disable libcap-ng on nios.

>          arm |                 libtirpc-0.2.4 | NOK | http://autobuild.buildroot.net/results/09a7de29e21501a094f22e9afe470524f5de8c92/

musl build problem. Will not re-appear for now since I disabled the
musl toolchain.

>     mips64el |          lua-periphery-1.0.4-1 | NOK | http://autobuild.buildroot.net/results/48efae954e6b030fa8743a5728fb3f30a892f142/
>         mips |          lua-periphery-1.0.4-1 | NOK | http://autobuild.buildroot.net/results/ed2150dd5be56056de299d2011cad340dc1e3e56/

Directly accessing the struct termios fields is bad, because depending
on the architecture, the structure is a bit different. There are some
accessors to do this, they should be used. Vincente, can you have a
look, since you fixed a similar problem in a different package some
time ago?

>          arc |                     neard-0.14 | NOK | http://autobuild.buildroot.net/results/0a466cde55c5e128a2e201924f80f0ec6b8b5c2a/

Should be fixed by applying http://patchwork.ozlabs.org/patch/467858/.

>          arm |                openldap-2.4.40 | NOK | http://autobuild.buildroot.net/results/b41c043fb3b2fad1d9cea0a95b512fb4942b5b19/

Should be fixed by applying http://patchwork.ozlabs.org/patch/467444/.

>         bfin |                      php-5.6.8 | NOK | http://autobuild.buildroot.net/results/633581d8f95efd3fc92ce9875e0608fe40d054bf/
>         bfin |                      php-5.6.8 | NOK | http://autobuild.buildroot.net/results/bf12eb189ab35ce00a2212695d2dbf8b8a126529/

Missing -lpthread when linking statically.

>          arm |                 poppler-0.32.0 | NOK | http://autobuild.buildroot.net/results/3a70305be4a78af9404b0bd027dbcdd011ca01b3/
>          arm |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/bd8796a3001458de5326f96a9110a95bac247225/
>          arm |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/156e4afe32f7928880859a69999abcc3c9b304d0/
>         sh4a |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/04e209c1b54ec4ccfc0da67a3f077151b65171a7/
>          arm |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/2dcc2b044a1235cf9df24e5d55423c5e4ca7196a/
>          arm |             python-pyqt-4.11.3 | TIM | http://autobuild.buildroot.net/results/d2f426d839bebebef3d8a41d1d010a0c6ac285a9/
>          sh4 |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/de975f29185aa95826b416ee5a60d2f190a979a6/
>          arm |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/891dc6ad46039740867a0b436281fc489cfb2772/
>      powerpc |             python-pyqt-4.11.3 | NOK | http://autobuild.buildroot.net/results/65fe75b6aa965641f447c41b82d8511bd6be1a86/
>     mips64el |                       qt-4.8.6 | NOK | http://autobuild.buildroot.net/results/e0df30e8068cd784bed8e47f5cd13f1f70284027/
>     mips64el |                       qt-4.8.6 | NOK | http://autobuild.buildroot.net/results/ccd0534d1dfcb1628f808da91850ec62091f2178/
>      powerpc |                       qt-4.8.6 | TIM | http://autobuild.buildroot.net/results/349922ebab920c82cf087080f96c12310d6e66bb/

All these issues are caused by
http://git.buildroot.net/buildroot/commit/?id=7619aba496208102e098e24454371b9513ec2f90.
I suggest to revert this patch.

>          arm |           sane-backends-1.0.24 | NOK | http://autobuild.buildroot.net/results/f2a43b62aa2a720c538fc60c80c78c8ed5b04dea/

Missing pthread, already fixed by
http://git.buildroot.net/buildroot/commit/?id=0045808c1b271eef75e08fe02b4032f3f0e3dfc7.

>       xtensa |           sane-backends-1.0.24 | NOK | http://autobuild.buildroot.net/results/31115fe8d88f52d77ed0f2da769eb8896a1b34a2/

kvs40xx.h:265:2: error: #error __BYTE_ORDER not defined
 #error __BYTE_ORDER not defined

Max, can you have a look?

>      aarch64 |                   snmppp-3.3.4 | NOK | http://autobuild.buildroot.net/results/23d696e3ecbd192669e0de5eae6d339730c1aea2/

Try 'libtool --help' for more information.
libtool:   error: unrecognised option: '-DHAVE_CONFIG_H'

Weird.

>         bfin |                 tinyxml2-2.2.0 | NOK | http://autobuild.buildroot.net/results/86a4da3f9f4c9025f2f2393800c797b2af050808/
>         bfin |                 tinyxml2-2.2.0 | NOK | http://autobuild.buildroot.net/results/438ef12933aaac24029e352fdadeedb76bbd1cbb/

tinyxml is trying to build a shared library in all cases. Samuel, you
are our CMake guy, can you have a look to fix this?

>          arm |                        unknown | TIM | http://autobuild.buildroot.net/results/c2f41853747d32f253ba251f8919e02d47f69719/
>          arm |                valgrind-3.10.1 | TIM | http://autobuild.buildroot.net/results/f5aa99efe31e5b417ba27a0de0408eed10926375/

Ignore.

>          arm |                      vlc-2.2.1 | NOK | http://autobuild.buildroot.net/results/3612b54b0bbc2a55ba6f0febb7be32595b523d96/

  CC       libgles2_plugin_la-gl.lo
arm-none-linux-gnueabi-gcc: ERROR: unsafe header/library path used in cross-compilation: '/usr/include/directfb'
make[6]: *** [libdirectfb_plugin_la-directfb.lo] Error 1

Bernd, can you have a look?

>     mips64el |                xapp_xwud-1.0.4 | TIM | http://autobuild.buildroot.net/results/5c02fe80e6072ee8d1794c20a31850af71eb10eb/
>     mips64el |                   xerces-3.1.2 | TIM | http://autobuild.buildroot.net/results/9c5cc4182b561804476bdcbda5b7dbb763110aa2/
>          arc |        xproto_presentproto-1.0 | TIM | http://autobuild.buildroot.net/results/b6abf41c886ade1e88d2cce7c0250f58a966aa7c/
>          arm | znc-b396cafdb249544164ed029... | TIM | http://autobuild.buildroot.net/results/990172b1d3b2e028adba947efad9544a5a705f78/

Ignore, all these are timeouts.

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] Analysis of build failures

[Gustavo Zacarias]

On 05/05/2015 04:59 AM, Thomas Petazzoni wrote:
>>       x86_64 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/cd2e617f8e2b00581ab5936029f85e62ed3259ba/
> 
> md2.o: In function `md2_sum':
> md2.c:(.text+0x18): undefined reference to `EVP_md2'
> 
> Gustavo, maybe? Has OpenSSL removed md2 support by default or something
> like that?

Yes, md2 has been disabled for some time, the problem is that ipmiutil's
configure uses LIB_DIR to check for md2 availability in libcrypto.
But we can't override that otherwise libipmiutil doesn't get installed
in the correct place :-/
Quick solution is to force it off, ideally a well-done tests would be
better.
Regards.

[Buildroot] Analysis of build failures

[Thomas Petazzoni]

Gustavo,

On Tue, 05 May 2015 10:44:18 -0300, Gustavo Zacarias wrote:

> Yes, md2 has been disabled for some time, the problem is that ipmiutil's
> configure uses LIB_DIR to check for md2 availability in libcrypto.
> But we can't override that otherwise libipmiutil doesn't get installed
> in the correct place :-/
> Quick solution is to force it off, ideally a well-done tests would be
> better.

Ok, thanks. Can you send a patch to force it off?

Thanks,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] Analysis of build failures

[Thomas Petazzoni]

Hello,

On Tue, 05 May 2015 15:24:23 +0200, Pascal Huerst wrote:

> According to:
> 
> https://breakpad.appspot.com/2824002/
> 
> support has been added for release r1407. I tried to build that
> revision, but I got the same error as reported above. I'm not really
> familiar with the internals of breakpad, but I asked on the breakpad
> mailing list. Maybe someone there already knows that issue... I'll keep
> you updated.

Thanks. We'll wait a bit for them to give some feedback, and if it
doesn't come soon enough, we'll disable breakpad on aarch64 for now.

Thanks,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] Analysis of build failures

[Max Filippov]

Hi Thomas,

On Tue, May 5, 2015 at 10:59 AM, Thomas Petazzoni
<thomas.petazzoni@free-electrons.com> wrote:
>
>>       xtensa |           sane-backends-1.0.24 | NOK | http://autobuild.buildroot.net/results/31115fe8d88f52d77ed0f2da769eb8896a1b34a2/
>
> kvs40xx.h:265:2: error: #error __BYTE_ORDER not defined
>  #error __BYTE_ORDER not defined
>
> Max, can you have a look?

Apparently sane-backends configure script cannot correctly detect target system
endianness without hints. It gets a hint in form of
ac_cv_c_bigendian=yes/no, that
is set in accordance to BR2_ENDIAN.
But for xtensa we don't have BR2_ENDIAN and endianness comes from the
configuration overlay. Do you think we need to provide BR2_ENDIAN setting
for xtensa?

-- 
Thanks.
-- Max

[Buildroot] Analysis of build failures

[Thomas Petazzoni]

Dear Max Filippov,

On Tue, 5 May 2015 17:52:23 +0300, Max Filippov wrote:

> Apparently sane-backends configure script cannot correctly detect target system
> endianness without hints. It gets a hint in form of
> ac_cv_c_bigendian=yes/no, that
> is set in accordance to BR2_ENDIAN.
> But for xtensa we don't have BR2_ENDIAN and endianness comes from the
> configuration overlay. Do you think we need to provide BR2_ENDIAN setting
> for xtensa?

Yes, we should have BR2_ENDIAN defined for Xtensa. Precisely to avoid
such issues :)

Thanks for the investigation!

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] Analysis of build failures

[Peter Korsgaard]

>>>>> "Thomas" == Thomas Petazzoni <thomas.petazzoni@free-electrons.com> writes:

 > Hello,
 > Simon, Pascal, Gustavo, Vicente, Max, Samuel, Bernd, look below, there
 > are some questions/tasks for you! :-)

It would probably have been good to add them in To: then ;)


 >> arm | at-3.1.13 | NOK |
 >> http://autobuild.buildroot.net/results/6f25e80e8aedec91323ef9b67576b550d7abee60/

 > musl build problem. Yann has sent a patch to disable at on musl, but I
 > would actually prefer to fix at.

J?rg mentioned that OE has a patch:

http://patchwork.openembedded.org/patch/91893/


 >> x86_64 | ipmiutil-2.9.5 | NOK |
 >> http://autobuild.buildroot.net/results/cd2e617f8e2b00581ab5936029f85e62ed3259ba/

 > md2.o: In function `md2_sum':
 > md2.c:(.text+0x18): undefined reference to `EVP_md2'

 > Gustavo, maybe? Has OpenSSL removed md2 support by default or something
 > like that?

I believe it simply gets confused about something on the build host.


 >> powerpc |                       qt-4.8.6 | TIM | http://autobuild.buildroot.net/results/349922ebab920c82cf087080f96c12310d6e66bb/

 > All these issues are caused by
 > http://git.buildroot.net/buildroot/commit/?id=7619aba496208102e098e24454371b9513ec2f90.
 > I suggest to revert this patch.

I've reverted it now.


 >> arm | vlc-2.2.1 | NOK |
 >> http://autobuild.buildroot.net/results/3612b54b0bbc2a55ba6f0febb7be32595b523d96/

 >   CC       libgles2_plugin_la-gl.lo
 > arm-none-linux-gnueabi-gcc: ERROR: unsafe header/library path used in
 > cross-compilation: '/usr/include/directfb'
 > make[6]: *** [libdirectfb_plugin_la-directfb.lo] Error 1

 > Bernd, can you have a look?

I believe it is fixed by:

http://git.buildroot.net/buildroot/commit/?id=ac95a7a27bf319a69db41

-- 
Bye, Peter Korsgaard

[Buildroot] Analysis of build failures

[Thomas Petazzoni]

Dear Peter Korsgaard,

On Tue, 05 May 2015 17:22:32 +0200, Peter Korsgaard wrote:

>  > Hello,
>  > Simon, Pascal, Gustavo, Vicente, Max, Samuel, Bernd, look below, there
>  > are some questions/tasks for you! :-)
> 
> It would probably have been good to add them in To: then ;)

They were in Cc, but this stupid mailing list removes people from the
Cc: list when they are subscribed. My initial e-mail headers were:

From: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
To: buildroot at uclibc.org
Cc: Simon Dawson <spdawson@gmail.com>, Pascal Huerst <pascal.huerst@gmail.com>, Gustavo Zacarias <gustavo@zacarias.com.ar>, Vicente Olivert Riera <Vincent.Riera@imgtec.com>, Max Filippov <jcmvbkbc@gmail.com>, Samuel Martin <s.martin49@gmail.com>, Bernd Kuhls <berndkuhls@hotmail.com>

If the mailing list configuration could be changed to not do this
stupid trimming of the Cc addresses, it would be great.

>  >> arm | at-3.1.13 | NOK |
>  >> http://autobuild.buildroot.net/results/6f25e80e8aedec91323ef9b67576b550d7abee60/
> 
>  > musl build problem. Yann has sent a patch to disable at on musl, but I
>  > would actually prefer to fix at.
> 
> J?rg mentioned that OE has a patch:
> 
> http://patchwork.openembedded.org/patch/91893/

So let's use it :-)

>  >> x86_64 | ipmiutil-2.9.5 | NOK |
>  >> http://autobuild.buildroot.net/results/cd2e617f8e2b00581ab5936029f85e62ed3259ba/
> 
>  > md2.o: In function `md2_sum':
>  > md2.c:(.text+0x18): undefined reference to `EVP_md2'
> 
>  > Gustavo, maybe? Has OpenSSL removed md2 support by default or something
>  > like that?
> 
> I believe it simply gets confused about something on the build host.

Yes, Gustavo has already submitted a patch.

>  >> powerpc |                       qt-4.8.6 | TIM | http://autobuild.buildroot.net/results/349922ebab920c82cf087080f96c12310d6e66bb/
> 
>  > All these issues are caused by
>  > http://git.buildroot.net/buildroot/commit/?id=7619aba496208102e098e24454371b9513ec2f90.
>  > I suggest to revert this patch.
> 
> I've reverted it now.

Great. We'll see which Qt related failures remain after that.

>  >> arm | vlc-2.2.1 | NOK |
>  >> http://autobuild.buildroot.net/results/3612b54b0bbc2a55ba6f0febb7be32595b523d96/
> 
>  >   CC       libgles2_plugin_la-gl.lo
>  > arm-none-linux-gnueabi-gcc: ERROR: unsafe header/library path used in
>  > cross-compilation: '/usr/include/directfb'
>  > make[6]: *** [libdirectfb_plugin_la-directfb.lo] Error 1
> 
>  > Bernd, can you have a look?
> 
> I believe it is fixed by:
> 
> http://git.buildroot.net/buildroot/commit/?id=ac95a7a27bf319a69db41

Excellent, thanks.

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] Analysis of build failures

[Peter Korsgaard]

>>>>> "Thomas" == Thomas Petazzoni <thomas.petazzoni@free-electrons.com> writes:

 > Dear Peter Korsgaard,
 > On Tue, 05 May 2015 17:22:32 +0200, Peter Korsgaard wrote:

 >> > Hello,
 >> > Simon, Pascal, Gustavo, Vicente, Max, Samuel, Bernd, look below, there
 >> > are some questions/tasks for you! :-)
 >> 
 >> It would probably have been good to add them in To: then ;)

 > They were in Cc, but this stupid mailing list removes people from the
 > Cc: list when they are subscribed. My initial e-mail headers were:

Ahh, ok.

 > From: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
 > To: buildroot at uclibc.org
 > Cc: Simon Dawson <spdawson@gmail.com>, Pascal Huerst
 > <pascal.huerst@gmail.com>, Gustavo Zacarias <gustavo@zacarias.com.ar>,
 > Vicente Olivert Riera <Vincent.Riera@imgtec.com>, Max Filippov
 > <jcmvbkbc@gmail.com>, Samuel Martin <s.martin49@gmail.com>, Bernd
 > Kuhls <berndkuhls@hotmail.com>

 > If the mailing list configuration could be changed to not do this
 > stupid trimming of the Cc addresses, it would be great.

I had a look around, and according to this, it is because these users
have the nodupes option enabled:

https://bugs.launchpad.net/mailman/+bug/1216960

As described below, I have changed the default for new subscribers to
NOT enable this option, but I don't know how to do it in any bulk way
for the current subscribers (and if people want that):

https://www.ietf.org/mail-archive/web/ietf/current/msg37128.html


 >> >> arm | at-3.1.13 | NOK |
 >> >> http://autobuild.buildroot.net/results/6f25e80e8aedec91323ef9b67576b550d7abee60/
 >> 
 >> > musl build problem. Yann has sent a patch to disable at on musl, but I
 >> > would actually prefer to fix at.
 >> 
 >> J?rg mentioned that OE has a patch:
 >> 
 >> http://patchwork.openembedded.org/patch/91893/

 > So let's use it :-)

Yeah, will you send a patch? ;)

 >> > Gustavo, maybe? Has OpenSSL removed md2 support by default or something
 >> > like that?
 >> 
 >> I believe it simply gets confused about something on the build host.

 > Yes, Gustavo has already submitted a patch.

Yes, applied already.

-- 
Venlig hilsen,
Peter Korsgaard 

[Buildroot] Analysis of build failures

[Thomas Petazzoni]

Peter,

On Tue, 05 May 2015 17:50:06 +0200, Peter Korsgaard wrote:

>  > From: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
>  > To: buildroot at uclibc.org
>  > Cc: Simon Dawson <spdawson@gmail.com>, Pascal Huerst
>  > <pascal.huerst@gmail.com>, Gustavo Zacarias <gustavo@zacarias.com.ar>,
>  > Vicente Olivert Riera <Vincent.Riera@imgtec.com>, Max Filippov
>  > <jcmvbkbc@gmail.com>, Samuel Martin <s.martin49@gmail.com>, Bernd
>  > Kuhls <berndkuhls@hotmail.com>
> 
>  > If the mailing list configuration could be changed to not do this
>  > stupid trimming of the Cc addresses, it would be great.
> 
> I had a look around, and according to this, it is because these users
> have the nodupes option enabled:
> 
> https://bugs.launchpad.net/mailman/+bug/1216960
> 
> As described below, I have changed the default for new subscribers to
> NOT enable this option, but I don't know how to do it in any bulk way
> for the current subscribers (and if people want that):
> 
> https://www.ietf.org/mail-archive/web/ietf/current/msg37128.html

I suppose it cannot be changed from the Mailman admin interface?

>  >> J?rg mentioned that OE has a patch:
>  >> 
>  >> http://patchwork.openembedded.org/patch/91893/
> 
>  > So let's use it :-)
> 
> Yeah, will you send a patch? ;)

If nobody beats me at it, I'll try.

Thanks,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] Analysis of build failures

[Peter Korsgaard]

>>>>> "Thomas" == Thomas Petazzoni <thomas.petazzoni@free-electrons.com> writes:

Hi,

>> As described below, I have changed the default for new subscribers to
 >> NOT enable this option, but I don't know how to do it in any bulk way
 >> for the current subscribers (and if people want that):
 >> 
 >> https://www.ietf.org/mail-archive/web/ietf/current/msg37128.html

 > I suppose it cannot be changed from the Mailman admin interface?

It can be done by clicking seperately for each subscriber, but I'm not
going to do that ~1300 times, and somebody might not like to get
doubles.

-- 
Venlig hilsen,
Peter Korsgaard 

[Buildroot] Analysis of build failures

[Angelo Compagnucci]

Dear Thomas Petazzoni,

>>      powerpc |                host-mono-4.0.1 | NOK | http://autobuild.buildroot.net/results/7980bbd50ab63ea2bf482b8f58493ac39790a213/
>>          arm |                host-mono-4.0.1 | NOK | http://autobuild.buildroot.net/results/1eeb6b7f4c055968f1683a37cec2b2142923cd28/
>
> Smells like parallel installation issue, but not sure. Angelo tried to
> investigate, but was not able to reach a conclusion. Anyone to look at
> this?

Stating at today (2015-05-06) report, the error is gone. Honestly I
think there is something in autobuild that makes parallel installation
screws, cause I repeated at least a dozen of tests with different
configurations and I'm using it in a side project (several more
compilations) and never encountered this bug on my quad core machine.

Sincerely, Angelo

[Buildroot] Analysis of build failures

[Thomas Petazzoni]

Dear Angelo Compagnucci,

On Wed, 6 May 2015 09:34:32 +0200, Angelo Compagnucci wrote:

> Stating at today (2015-05-06) report, the error is gone. Honestly I
> think there is something in autobuild that makes parallel installation
> screws, cause I repeated at least a dozen of tests with different
> configurations and I'm using it in a side project (several more
> compilations) and never encountered this bug on my quad core machine.

No it isn't gone: it occurred again two times during this night,
http://autobuild.buildroot.org/?reason=host-mono-4.0.1.

(The 2015-05-06 autobuild e-mail reports failures that took place
between 2015-05-05 00:00 and 2015-05-06 00:00, European time).

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] Analysis of build failures

[Pascal Huerst]

Hey Thomas,

On 05.05.2015 15:48, Thomas Petazzoni wrote:
> Hello,
> 
> On Tue, 05 May 2015 15:24:23 +0200, Pascal Huerst wrote:
> 
>> According to:
>>
>> https://breakpad.appspot.com/2824002/
>>
>> support has been added for release r1407. I tried to build that
>> revision, but I got the same error as reported above. I'm not really
>> familiar with the internals of breakpad, but I asked on the breakpad
>> mailing list. Maybe someone there already knows that issue... I'll keep
>> you updated.
> 
> Thanks. We'll wait a bit for them to give some feedback, and if it
> doesn't come soon enough, we'll disable breakpad on aarch64 for now.

Looks like there is a patch in pipeline, but not applied, yet. I would
suggest to disable breakpad for AArch64 for now and re-enable it on a
next version bump.

pending patch:

https://breakpad.appspot.com/10734002/

and Mike's reply:

https://groups.google.com/forum/#!topic/google-breakpad-discuss/PpWkw79SVIs

regards
Pascal