[Buildroot] [autobuild.buildroot.net] Build results for 2015-05-21

[Buildroot] [autobuild.buildroot.net] Build results for 2015-05-21

[Thomas Petazzoni]

Build statistics for 2015-05-21
===============================

        success : 293
       failures : 8  
       timeouts : 3  
          TOTAL : 304

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

       host-erlang-rebar-2.5.1 | 3 
                ipmiutil-2.9.5 | 2 
                     php-5.6.9 | 1 
                 cdrkit-1.1.11 | 1 
       opentyrian-9c9f0ec3532b | 1 
              spidev_test-v3.0 | 1 
               qt5webkit-5.4.1 | 1 
                 mpg123-1.22.1 | 1 

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

     powerpc |                  cdrkit-1.1.11 | NOK | http://autobuild.buildroot.net/results/23af5e660dd5b1cc0dd274f8fa809e772245b028/
     powerpc |        host-erlang-rebar-2.5.1 | TIM | http://autobuild.buildroot.net/results/7faece26b986e7db42472f6db66c137038573555/
         arm |        host-erlang-rebar-2.5.1 | TIM | http://autobuild.buildroot.net/results/7d73337ad9b50cc04dbd3414d85d3d125ad4108d/
         arm |        host-erlang-rebar-2.5.1 | TIM | http://autobuild.buildroot.net/results/739c4f448568eb6ba5ccffe8a075e0f797c7109b/
        i686 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/04a1c48d484debd7894fd32997ed50d3a0110b93/
      x86_64 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/9accdfec87aa9922a4096b84a61a6e2dfd53561e/
         arc |                  mpg123-1.22.1 | NOK | http://autobuild.buildroot.net/results/25f8088ecb276ff4559a56fcc4f1b92d11f7deac/
        bfin |        opentyrian-9c9f0ec3532b | NOK | http://autobuild.buildroot.net/results/5d215b8895ac082002b38159beb94693a04a51bb/
        bfin |                      php-5.6.9 | NOK | http://autobuild.buildroot.net/results/b4c0e62a2ce718cf4bfad0ce328c853dc0de7f11/
         arm |                qt5webkit-5.4.1 | NOK | http://autobuild.buildroot.net/results/485daba6024d8f25d7488eae2e7a18db0f9ac663/
        bfin |               spidev_test-v3.0 | NOK | http://autobuild.buildroot.net/results/0c862f64aebf6bc454af78116fe146af5f20547c/


-- 
http://autobuild.buildroot.net

[Buildroot] [autobuild.buildroot.net] Build results for 2015-05-21

[Peter Korsgaard]

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

 > Build statistics for 2015-05-21
 > ===============================

 >         success : 293
 >        failures : 8  

Wee, looks good!

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

 >      powerpc |                  cdrkit-1.1.11 | NOK | http://autobuild.buildroot.net/results/23af5e660dd5b1cc0dd274f8fa809e772245b028/

Looks like a static linking issue. Unless somebody feels like digging
into the cmake setup I'll simply mark it as !BR2_STATIC_LIBS.

 >      powerpc |        host-erlang-rebar-2.5.1 | TIM | http://autobuild.buildroot.net/results/7faece26b986e7db42472f6db66c137038573555/
 >          arm |        host-erlang-rebar-2.5.1 | TIM | http://autobuild.buildroot.net/results/7d73337ad9b50cc04dbd3414d85d3d125ad4108d/
 >          arm |        host-erlang-rebar-2.5.1 | TIM | http://autobuild.buildroot.net/results/739c4f448568eb6ba5ccffe8a075e0f797c7109b/

The funky erlang issues on gcc75.


 >         i686 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/04a1c48d484debd7894fd32997ed50d3a0110b93/
 >       x86_64 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/9accdfec87aa9922a4096b84a61a6e2dfd53561e/

Gaah, more ipmiutil issues. I believe Yann is looking into this.


 >          arc |                  mpg123-1.22.1 | NOK | http://autobuild.buildroot.net/results/25f8088ecb276ff4559a56fcc4f1b92d11f7deac/

Fixed by http://git.buildroot.net/buildroot/commit/?id=dea306c78b5d4c571555c50512e212a725b594b9


 >         bfin |        opentyrian-9c9f0ec3532b | NOK | http://autobuild.buildroot.net/results/5d215b8895ac082002b38159beb94693a04a51bb/

Sounds like a -lstdc++ issue with static linking. I'll try to find time
to have a closer look.

 >         bfin |                      php-5.6.9 | NOK | http://autobuild.buildroot.net/results/b4c0e62a2ce718cf4bfad0ce328c853dc0de7f11/

Sounds like the php bump brough some extra architecture dependencies:

/ssd1/thomas/autobuild/instance-1/output/build/php-5.6.9/ext/pcre/pcrelib/pcre_jit_compile.c:65:2: error: #error Unsupported architecture

It apparently also breaks for arc:

http://autobuild.buildroot.net/results/641/6418b26140b16b6f04452a9b05431c392de04238/build-end.log

I doubt there's a lot of php users on any of those archs, so perhaps we
should just disable it there?

>          arm |                qt5webkit-5.4.1 | NOK | http://autobuild.buildroot.net/results/485daba6024d8f25d7488eae2e7a18db0f9ac663/

The package apparently has an optional dependency on leveldb, but we
don't explicitly handle that and it looks like it forgets to link
against it:

Running configure tests...

Checking for fontconfig... no
Checking for gccdepends... yes
Checking for glx... no
Checking for gnuld... yes
Checking for icu... yes
Checking for leveldb... no

..

Use ........................... tiled_backing_store cross_platform_context_menus libxml2 zlib gstreamer gstreamer010 native_fullscreen_video leveldb glib

..


/home/chroot/media/code/buildroot/autobuilder/instance-1/output/build/qt5webkit-5.4.1/Source/WebCore//.obj/platform/leveldb/LevelDBDatabase.o: In function `WebCore::LevelDBDatabase::openInMemory(WebCore::LevelDBComparator const*)':
LevelDBDatabase.cpp:(.text._ZN7WebCore15LevelDBDatabase12openInMemoryEPKNS_17LevelDBComparatorE+0x22): undefined reference to `leveldb::NewMemEnv(leveldb::Env*)'
collect2: error: ld returned 1 exit status
Makefile.api:111: recipe for target '../lib/libQt5WebKit.so.5.4.1' failed


>         bfin |               spidev_test-v3.0 | NOK | http://autobuild.buildroot.net/results/0c862f64aebf6bc454af78116fe146af5f20547c/

kernel.org cgit instance apparently had issues and our mirroring fails
with the funky filename.

The problem is that the ?id=3.0 part gets interpreted by the web server
as the query string (which it is), so it looks for spidev_test.c which
doesn't exist.

To fix it we would need to URL encode these arguments (like
spidev_test.c%3fid=v3.0) whenever we try to access the mirrors.

-- 
Bye, Peter Korsgaard

[Buildroot] [autobuild.buildroot.net] Build results for 2015-05-21

[Thomas Petazzoni]

Peter,

On Fri, 22 May 2015 09:20:09 +0200, Peter Korsgaard wrote:

>  > Build statistics for 2015-05-21
>  > ===============================
> 
>  >         success : 293
>  >        failures : 8  
> 
> Wee, looks good!

Yeah!

> > Detail of failures
>  > ===================
> 
>  >      powerpc |                  cdrkit-1.1.11 | NOK | http://autobuild.buildroot.net/results/23af5e660dd5b1cc0dd274f8fa809e772245b028/
> 
> Looks like a static linking issue. Unless somebody feels like digging
> into the cmake setup I'll simply mark it as !BR2_STATIC_LIBS.

No, it's not a CMake setup problem. It's once again a uClibc bug with
redefined symbols between libc.a and libpthread.a. For now, we should
add an exception to the autobuilders. Long term, we'll switch to
uClibc-ng, which has fixed those problems (and if not, with an upstream
that is responsive: we have reported such problems to uClibc since
years and they have never been fixed).

>  >      powerpc |        host-erlang-rebar-2.5.1 | TIM | http://autobuild.buildroot.net/results/7faece26b986e7db42472f6db66c137038573555/
>  >          arm |        host-erlang-rebar-2.5.1 | TIM | http://autobuild.buildroot.net/results/7d73337ad9b50cc04dbd3414d85d3d125ad4108d/
>  >          arm |        host-erlang-rebar-2.5.1 | TIM | http://autobuild.buildroot.net/results/739c4f448568eb6ba5ccffe8a075e0f797c7109b/
> 
> The funky erlang issues on gcc75.

Yeah, I don't know what's going on.

> 
>  >         i686 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/04a1c48d484debd7894fd32997ed50d3a0110b93/
>  >       x86_64 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/9accdfec87aa9922a4096b84a61a6e2dfd53561e/
> 
> Gaah, more ipmiutil issues. I believe Yann is looking into this.

Ok, let's Yann have some fun. This package is causing a lot of
troubles, even though it's just some random not so complicated utility.

>  >         bfin |        opentyrian-9c9f0ec3532b | NOK | http://autobuild.buildroot.net/results/5d215b8895ac082002b38159beb94693a04a51bb/
> 
> Sounds like a -lstdc++ issue with static linking. I'll try to find time
> to have a closer look.

Yes, that's it.

>  >         bfin |                      php-5.6.9 | NOK | http://autobuild.buildroot.net/results/b4c0e62a2ce718cf4bfad0ce328c853dc0de7f11/
> 
> Sounds like the php bump brough some extra architecture dependencies:
> 
> /ssd1/thomas/autobuild/instance-1/output/build/php-5.6.9/ext/pcre/pcrelib/pcre_jit_compile.c:65:2: error: #error Unsupported architecture
> 
> It apparently also breaks for arc:
> 
> http://autobuild.buildroot.net/results/641/6418b26140b16b6f04452a9b05431c392de04238/build-end.log
> 
> I doubt there's a lot of php users on any of those archs, so perhaps we
> should just disable it there?

It's only the new PCRE JIT stuff that isn't available on all
architectures. There is a --disable-pcregrep-jit option in pcre to
disable it, but I'm not sure how to convince PHP to pass this down to
PCRE.

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

[Buildroot] [autobuild.buildroot.net] Build results for 2015-05-21

[Yann E. MORIN]

Peter, Thomas, All,

On 2015-05-22 09:20 +0200, Peter Korsgaard spake thusly:
>  >         i686 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/04a1c48d484debd7894fd32997ed50d3a0110b93/
>  >       x86_64 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/9accdfec87aa9922a4096b84a61a6e2dfd53561e/
> Gaah, more ipmiutil issues. I believe Yann is looking into this.

ipmiutil's Makefile.am-s are, let's say it politely, not the easiest to
work with...

So, regarding the parallel build issue, I'm bailing out. I am able to
reproduce it in about 1 build out of ten, but then hacking Makefile.am
is just of my league (they are ugly...).

Now, I'm looking at those new build failures. I suspect:
     25e15fd package/ipmiutil: really force md2 off

which introduced -DALLOW_GNU in the CFLAGS, and that has much mpre
impact...

Investigation in progress...

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'

[Buildroot] [autobuild.buildroot.net] Build results for 2015-05-21

[Peter Korsgaard]

>>>>> "Yann" == Yann E MORIN <yann.morin.1998@free.fr> writes:

 > Peter, Thomas, All,
 > On 2015-05-22 09:20 +0200, Peter Korsgaard spake thusly:
 >> >         i686 | ipmiutil-2.9.5 | NOK |
 >> > http://autobuild.buildroot.net/results/04a1c48d484debd7894fd32997ed50d3a0110b93/
 >> >       x86_64 | ipmiutil-2.9.5 | NOK |
 >> > http://autobuild.buildroot.net/results/9accdfec87aa9922a4096b84a61a6e2dfd53561e/
 >> Gaah, more ipmiutil issues. I believe Yann is looking into this.

 > ipmiutil's Makefile.am-s are, let's say it politely, not the easiest to
 > work with...

 > So, regarding the parallel build issue, I'm bailing out. I am able to
 > reproduce it in about 1 build out of ten, but then hacking Makefile.am
 > is just of my league (they are ugly...).

Yeah. I don't have a problem with us using MAKE1 as it's a fairly small
package.

 > Now, I'm looking at those new build failures. I suspect:
 >      25e15fd package/ipmiutil: really force md2 off

 > which introduced -DALLOW_GNU in the CFLAGS, and that has much mpre
 > impact...

 > Investigation in progress...

Thanks!

-- 
Venlig hilsen,
Peter Korsgaard 

[Buildroot] [autobuild.buildroot.net] Build results for 2015-05-21

[Yann E. MORIN]

Peter, Thomas, All,

On 2015-05-22 09:20 +0200, Peter Korsgaard spake thusly:
> >>>>> "Thomas" == Thomas Petazzoni <thomas.petazzoni@free-electrons.com> writes:
>  > Build statistics for 2015-05-21
>  >         i686 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/04a1c48d484debd7894fd32997ed50d3a0110b93/
>  >       x86_64 |                 ipmiutil-2.9.5 | NOK | http://autobuild.buildroot.net/results/9accdfec87aa9922a4096b84a61a6e2dfd53561e/
> 
> Gaah, more ipmiutil issues. I believe Yann is looking into this.

The first issue above is again about an openssl issue, I'm investigating
now.

The second issue is about missing internal symbols:

    isensor2.o: In function `read_sdr_binfile':
    /home/test/autobuild/instance-0/output/build/ipmiutil-2.9.5/util/isensor.c:2990:
    undefined reference to `get_LastError'
or:
    libipmiutil.a(ipmicmd.o): In function `ipmi_open':
    ipmicmd.c:(.text+0x5be): undefined reference to `ipmi_open_lan'

I'm unable (after ~25 builds) to reproduce that second issue. So I'm
bailing out on that one... :-(

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'